I’m going to wait till more is known before saying more. But here is yet another — I won’t say “common” but I will say “too common” — shootings that perhaps should but won’t become big national news. There probably won’t be protests. There won’t be unrest.
But did police really break into the house of Jack Jacquez and shoot him? I don’t know what happened, but I do know there’s a lot here that doesn’t sound good.
[Update: Cut to the chase. You might just want to read my summary post.]
Related to the “not 21 times” previous post, I received a tweet from one of the authors: “Differences in our methodologies: you count Hispanic homicides as white… deflate the results.”
So back to running stats for me. But there’s a problem in that the UCR homicide data does a particularly poor job in counting hispanics. Most cities simply do not record hispanic data.
As a result, 56% of homicide data has nothing for “hispanic or not.” I would guess that most of this 56% is non-hispanic, since cities without many hispanics are less likely to care about counting hispanics, but we do not know. In general, you really shouldn’t use data when half is missing.
[The UCR would like police departments to do like the census: record race and then overlay hispanic-or-not on top of that. (If you’re a cop, this is probably how you record domestics.) But I don’t think any police department does this. So what the UCR seems to do, for the departments that list hispanic at all, is just call them all white hispanics.]
But if one does exclude hispanic whites from the count of whites over the past three years, one finds all of 9 young white males shot by police over the past three years. If one then uses non-hispanic white for the population denominator, I get a black-to-white ratio of 21:1 [replicated! And updated from the original post].
But what I will quibble about is the validity of that number. It means very little because there’s just not enough data.
I mean, one could look at just one year. The last available year, 2012, has a black-to-white ratio for teen males killed by police a less headline worthy 7:1 [13:1 if you exclude hispanic whites]. But you can’t just look at one year — or three. Put bluntly, police don’t kill enough teens each year to be statistically useful (which is good news, I suppose).
And since we can look at more years, we should. So if one wants to only look at 15-19 year-olds males shot by police, let’s look at the past 15 years. The most shocking result I discover is that a majority of “whites” killed by police are listed as hispanic. (109 versus 95. And overall there are 6.3 million non-hispanic whites and 2.1 million hispanic white males 15-19.)
The overall black-to-white ratio (15-19 year-old males) is 5.5:1. If one removes white hispanics from the sample (I’m not sure you should), the black-to-white killed-by-police ratio goes up 9:1. Though if one removes white hispanics for the overall homicide rate, the overall black-to-white homicide ratio in society goes from 9:1 to 15:1. All this gets a bit silly.
So let’s include everybody.
The overall racial disparity in homicides — and presumably other violent crimes as well (but they’re not counted as reliably) — is 6:1. The racial disparity among police-involved killings is about 4:1 (3.8:1, to be exact). Given the former, I don’t find the latter disturbing high (though I suppose reasonable people could disagree).
Here’s the thing. We should focus on bad police-involved shootings. And also we should focus on overly aggressive use of less-lethal force. These are issues of training, issues of a relaxing a paranoid “warrior” mindset. Sure, race matters, but if you want to improve policing, you need to move past the idea that police only do bad things to black people. This isn’t a black and white issue. It’s a police issue.
[It’s always good to put a disclaimer in any post related to police-involved shooting. The data, in general, is very limited. That said, some of the UCR data on police-involved homicides is good. While one cannot infer absolute numbers, looking at ratio of included data, such as race, presents much less of a problem, since one is looking a ratio within the data.
[Update: Also, some of the numbers have changed as I’ve updated and corrected and double-checked figures. Nothing substantively major. But you’re not going crazy if you think the actual headline used to 3 times and now it says 4 times (the actual number is 3.8. Using different population figures and/or just making a mistake, I first came up with 3.3).]
[Update: Cut to the chase. You might just want to read my summary post.]
One of my liberal de Blasio-loving not-so-fond-of-cops friend send me an email with the subject “you gotta check yo facts” and a link to ProPublica: “Young black males in recent years were at a far greater risk of being shot dead by police than their white counterparts – 21 times greater.”
“Well, that’s interesting,” I thought, “It also can’t be true.” Since I kind of know these numbers (and had discussed them with my friend). So I guess I do have to check my facts. I then wasted a half day running the numbers myself (when I could have been giving my undivided attention to the Orioles’ loss).
Now it’s always dangerous to say my numbers are right and theirs are wrong. But I trust my numbers, because I just ran them. And I’m good at this. And then I ran them again. I’d like to see their numbers because, well, I think they’re wrong. But clearly one of us is wrong. I hope it’s not me.
In the past three years (2010-2012) among those 15-19 year old, 54 blacks and 36 have been shot and killed by police. This is according to the UCR stats that are not perfect. But while the data here are not complete, they’re OK in many ways. And the black-white ratio should hold-up just fine.
If my data are wrong, please do correct me.
In the 15-19 population population, there are 8,728,271 white males. (Click through to: “Annual Estimates … by Sex, Age, Race, and Hispanic Origin”) There are 1,978,081 black males, 15-19 years-old (2010 census).
Per year, for the past 3 years, this is a police-involved homicide rate of 0.14 per 100,000 for whites and 0.99 for blacks. 0.91 divided by 0.14 is 6.5, not 21. For the past three years black males 15-19 are 6 or 7 times more likely than white males to be shot and killed by police, not 21 times.
From ProPublica:
The 1,217 deadly police shootings from 2010 to 2012 captured in the federal data show that blacks, age 15 to 19, were killed at a rate of 31.17 per million, while just 1.47 per million white males in that age range died at the hands of police.
Now even if one takes a 3-year rate per million (which is statistically odd for two italicized reasons), the rate for blacks is 30 (close to 31 but not replicated). Where I think the error lies is that the rate for whites is not 1.47 but rather 4.3. That’s a big difference.
My numbers are based on the years 2010-2012: 36 whites shot and killed. 8.7 million white males 15-19.
[Their 95% confidence interval is vast: “between 10 and 40 times greater risk.” This, leaving aside the wrong number, seems to me to be a gross misuderstanding of confidence interval. The overall number (the “n,” in stat terminology) of young people killed by police over the past three years is not large. But there’s a difference between a small “population” and a small “sample” size.
A confidence interval tells you the odds your sample reflects the total population. Say you ask 100 potential voters if they would vote for Obama. Four or 40% say yes. So what are the odds that Obama would win 40% of the vote? Well you don’t know for sure because you didn’t ask everybody. But based on those 100 you did ask, you can come up with a range, say 35-45 percent, at which you can say there is 19 in 20 chance that if we did ask everybody, it would be in this range. That’s a confidence interval.
Again, if I’m wrong here, correct me! It’s been 18 years since I took a statistics class in graduate school. And I wasn’t even good at it.
If you poll everybody — if you have an election — you don’t have a confidence interval. You have a result! Even with its flaws, the UCR is pretty complete. If blacks are X-times more likely to be killed, that’s that! There is not a sample but a population. You don’t have a confidence interval if you sample everybody in a population. You have a number. But it is a small population.
I also wonder why they only picked people shot and killed, rather than all persons killed. It’s a minor difference, but why make more work when you don’t have to? 99.2 percent of people killed by cops are killed with a gun.)]
Well conveniently you can just add more years to get a larger population. I don’t know why they didn’t. (Well, I suspect because it’s work. It’s a bit of a pain to download and select from each year’s UCR sample. But that is what researchers do. I mean, I just happen to have the last 15 years compiled and ready to use because, well, that’s what researchers do. On a Saturday night. While watching baseball.)
So instead of looking at the past three years, let’s increase the population by looking at the past 15 years. From 1998-2012, 210 white and 242 black male 15-19 year-olds have been shot and killed by police. This comes out to an annual rate of 0.16 (per 100,000) for white males and 0.82 for black males.
So over the past 15 years black male teens are 5.1 times more likely — five times more likely — than whites to be shot and killed by police. Five times; not 21.
Now maybe 21 and 7 and 5 are close enough for you. Or maybe you think 5 times more is 5 times too many. But what number would be OK? Given ration disparities in violent crime, one shouldn’t expect 1:1. One might expect police to be more likely to shoot and kill people who shoot and kill other people. (Remember that we’re using rates here, which take into account the population difference, that there are 7 whites for every black in America.)
The homicide rate for black men 15-19 is 9 times the rate for white men. (From 2010 to 2012, looking at men 15-19, 2,382 blacks and 1,209 whites have been murdered by criminals. The homicide rate for these young white men is 4.6 per 100,000. For these young black men, the homicide rate is 40.7.)
So given the 9:1 racial disparity in the homicide rate among young men, what racial disparity would one expect in police-involved shootings? There’s no right answer to this question. But I don’t think it’s unreasonable for the racial disparity of those young men shot and killed by police to be reflective of the racial disparity in violence and homicides among young men. And in fact, the police-involved ratio, at 5:1 (not 21:1 or even 9:1), is much less.
[Updated to reflect population data from 2010 census rather than ACS estimate. It doesn’t change much. Also, see next post and my summary.]
I asked Jim, my Dominican-born Austin-raised San Francisco-living white friend, why he thought so many Californian cities were high on my PIHN list. He thought for a very few short seconds and answered, “because hispanics aren’t violent but police think they are.”
I love over-generalizations and stereotypes that could very well be true.
So I got black and hispanic percentages for my 40 cities and ran correlations to see if there was anything related with race, hispanic, the city’s homicide rate, the police-involved homicide rate, and PIHN.
More blacks in a city correlates with a higher homicide rate but not significantly with the rate of police-involved homicides. That last part is surprising.
A higher hispanic percentage in a city correlates with a lower homicide rate (which shouldn’t be surprisingly, unless you only listen to Fox News) and is also not related to the rate of police-involved shootings. OK.
Of course a high homicide rate correlates very much more police-involved shootings (that I knew, and is the whole reason behind this PIHN idea).
And black and hispanic percentages in cities both correlate with PIHN, and in opposite directions. More hispanics mean a higher PIHN. More blacks a lower PIHN. Another way to look at this is to say that hispanics live in less violent cities, but those cities do not see the expected correlated decrease in police-involved shootings.
Now this might be counterintuitive to some, but it makes sense if once thinks of all the flack police can get when they shoot a black person (even an armed person who shot at police). For better and for worse, perhaps cities with more blacks are better organized to complain about police-involved shootings. Sure, these protests piss off police, but they could also lead to better training, fewer police-involved shootings, and police less likely to pull the trigger.
How often do whites or hispanics complain after a questionable shooting? Not so much.
So could police be disproportionately killing hispanics? Seems possible… but turns out not really.
In trigger-happy Riverside, which is 52 percent hispanic and 6 percent black — if the data is accurate — hispanics are not overrepresented in police-involved shootings (68 over 15 years). Other than the massive number of police-involved homicides, nothing jumps out at me. When hispanic-or-not is listed (80 percent of the time), 36 percent of those killed by police are listed as hispanic. 13 of the 68 were black (disproportionately but not unexpectedly high).
In Mesa, which is 28 percent hispanic and has only 3 blacks (just kidding, Mesa is 3 percent black), police killed 40 people over 15 years. Only one of the 40 was black. When ethnicity was listed, about one-third of those killed were hispanic.
I also looked at San Diego and Dallas, and could find nothing that stood out. So this seems to be a bit of a dead end. It’s also entirely possible that hispanics are listed as non-hispanic for whatever reason. I don’t know.
Basically, if there’s any conclusion to be reached, it seems that in cities with a lot of Mexicans, whites are more likely to get shot and killed by police. This isn’t what I really expected. Though it’s not hard to imagine a lot of poor messed-up whites living in trailer parks in the desert, maybe I watched too much Breaking Bad.
Any ideas? (Especially ones that aren’t particularly statistically advanced.)
The winner, still by far, is Riverside, CA. But sneaking into second place is Mesa, Arizona, the only non-Californian city in the top 6.
Here’s the top 20 with the PIHN. All 20 are west of the Mississippi:
Riverside: 31
Mesa: 14
San Diego: 12
Sacramento: 9
Bakersfield: 8
Seattle: 8
Portland: 8
Albuquerque: 7
Fresno: 7
Tucson: 7
San Jose: 7
Long Beach: 6
Colorado Springs: 6
Oklahoma City: 6
Denver: 6
Phoenix: 5
Tulsa: 5
Austin: 5
San Antonio: 5
The first city east of the Mississippi is Philadelphia, which has a PIHN of 2.9. This mean that police in San Diego are 4 times more likely to kill somebody, taking the overall homicide rate into account.
I compiled and ran the numbers for 40 cities for which I believe the UCR data on justified police-involved homicides seems valid for the past 15 years. By “seems” I mean me looking over the numbers to make sure there’s an entry for every year and that the overall number in close to what one might expect, based on population and crime. Once I supplemented missing data with other data (New York City), and once I just averaged from fewer years (San Antonio).
Cities which I think lack valid data include Boston, Charlotte, Detroit, El Paso, Fort Worth, Honolulu, Jacksonville, Louisville, Virginia Beach, Omaha, Arlington, Raleigh, Miami, Washington DC, and Wichita. But except for those, I compiled numbers for every city larger than 350,000 (and a few smaller ones, too).
But when the PIHN gets below two, I start to suspect some of the data is missing. But who knows? Maybe I’m not giving credit where credit is due.
There’s also the possibility that the PIHN adjusts too much for violence. It does, in effect, punish cities for being safe. But police officers in “safe” cities might be quicker to shoot, since they’re less used to danger. Certainly cities with low homicide rates rank high on the PIHN scale. But not always. Sacramento has a high homicide rate and a high PIHN. New York has a low homicide rate and a low PIHN. But it might be more interesting to make a scale which eliminates any correlation between PIHN and a city’s homicide rate. But I also suspect, based on experience, that police in high-crime areas deserve more credit than they get for not shooting. Some of the bad shootings I’ve seen recently… I can’t imagine a cop in Baltimore being so damn scared for no good reason.
There are fewer than 15 homicides a year in Riverside. Given that, it seems hard to believe that police kill almost five a year.
The city of Riverside, California appears to be, by far, the city in which police are most likely to commit justifiable homicide. I listed a rough rank order of cities in my previous post. Riverside is almost 50 percent higher than the next highest cities, St. Louis and Baltimore. (Even more so if one takes into account Riverside’s population gains over the past decade.)
Riverside police kill an average of 4.5 people a year. This is very high for a city with about 300,000 people. New York police kill about 13 people per year. But NYC has 8 million friggin’ people!
Other cities with a lot of police-involved homicides, like St. Louis and Baltimore, have a lot of crime. Not Riverside. Over the past decade (2003-2012) there have been 14.6 homicides per year in Riverside. This is on par with about the national average of 5 per 100,000. St. Louis, by comparison, about the same size as Riverside, sees about 126 murders annually. Baltimore, twice as large, has averaged 248 murders. Baltimore and St. Louis have a lot of murders. Since there are more murderers, one would expect police to shoot more of them.
But Riverside?
I’ve invented an acronym called PIHN. It stands for “Police-Involved Homicide Number.” I’ve also decided it’s pronounced “pin.”
PIHN takes a city’s violence into account and assumes a direct relationship between homicides in a city and police-involved shootings in that city. A higher PIHN means that there are more police-involved homicides for a given level of violence (presumably a poorly trained more trigger-happy police department). A low PIHN means fewer police-involved homicides (a better trained and less trigger-happy police department).
I applied PIHN to the 10 cities with the highest rate of justifiable police-involved homicides in America and also to the 10 largest American cities. First the cities in which police kill a lot of people, per capita.
Notice the cities ranked 2, 3, 4 (St. Louis, Baltimore, and Newark) in police-involved homicides drop way down if one takes the homicide rate into consideration.
Here are the 10 biggest cities. New York, even with a low crime rate, has a low PIHN. Not surprising to me, because the NYPD is very restrained in shooting (despite what you may read). And there’s a general clustering between 2 and 4 for the top five.
(Note the scale on this figure is half of the other one)
San Diego is interesting because it doesn’t even rank in the top 25 for the overall rate of police-involved homicides. But San Diego is a safe city, overall. Given the low number of homicides in San Diego, the high number of police-involved homicides — a PIHN close to 12 — is unexpected and striking.
Among the 20 cities I looked at, there’s a cluster of PIHNs between 2 and 3: Philadelphia, Atlanta, Chicago, St. Louis, Newark, New York, Baltimore, Dallas, and New Orleans. There’s variance, to be sure, but they’re all kind of in the same ballpark.
And then you go west of the Mississippi and the PIHNs skyrocket, particularly in California. Tulsa’s PIHN is 5.4. Sacramento is 8.9. And Riverside, California? Riverside’s PIHN is 31.1! That’s crazy high. Here are the highest PIHNs:
(Keep in mind this top-ten list comes only from the 20 cities I calculated, which are the 10 largest cities and the 10 cities with the highest police-involved homicide rate, based on URC data.)
Riverside, California. What is up with Riverside?
[You too can calculate PIHN! Divide the police-involved homicide rate or number by the average overall homicide rate or number. I used 2003-2012 homicide data from city-data). And then multiply by 100 to get a PIHN greater than 1. Or I’ll do it for you if you do some of the grunt work. Go to city-data, enter the city of your choice, scroll down to crime, add up the homicides numbers from 2003 to 2012.]
[Data for San Antonio police-involved homicides are averaged from just the past six year, since they didn’t report to the UCR before then.]
I’ve invented a statistic (and acronym) called PIHN (pronounced “pin”). It stands for “Police-Involved Homicide Number.”
PIHN looks at police-involved homicides but takes a city’s violence into account. PIHN assumes a (very questionable) direct relationship between homicides in a city and the number police-involved homicides one might expect.
A high PIHN means that there are more police-involved homicides for a given level of violence (and perhaps a poorly trained and/or more trigger-happy police department). A low PIHN means fewer police-involved homicides (perhaps a better trained and less trigger-happy police department).
You too can calculate PIHN! Divide the police-involved homicide rate or number by the average overall homicide rate or number. I used 2003-2012 homicide data from city-data. And then multiply by 100 to get a PIHN greater than 1.
The data on police-involved shootings are notoriously bad (that’s a link to Jon Stewart worth clicking on!). And yet, at least we kind of know which data are missing. That makes the data not as bad as you might think. At least when it comes to police-involved justifiable homicides (for shootings, we don’t know. But if you multiply homicides by two or three, you’ll probably be close enough).
Now I’ve compiled the UCR data on justifiable police homicides from 1998 to 2012. And the data are not complete. Some cities send data. Some don’t. Others send a few years… and then decide they have better things to do.
But of the 70 biggest cities in the US, only about 15 jumped out to me as horribly lacking in data. And if you take the 45 cities for which the data is probably good enough, which cities have the highest rate of police-involved homicides?
New York isn’t close to the top. (The UCR data from New York actually are not good — but the NYPD actually provides excellent data on shootings, but for some reason they just don’t bother reporting to the UCR.) But if one were to compile the data from the UCR and the NYPD (as I did) one would find an average of 12.2 annual police-involved shooting deaths. Let’s round up to 13, for the few cases of non-shooting police-involved homicides. That’s an annual rate of about 0.16, which is low. Very low. It puts New York 50th on the list of the 55 largest cites in the US (for which there is probably good data). New York’s rate is one-sixth the rate in Baltimore City, for instance.
My lovely Baltimore is number three on the list, in case you were wondering (I was). Baltimore City has seen 88 police-involved homicides over 15 years. Baltimore’s annual rate of police-involved homicide is under 1 per 100,000. These killings are justified, I should point out. I can vouch for a few of them 2nd-hand. At least one reader of this blog can vouch 1st-hand.
Number two in the USA is St. Louis (city, not county). 52 justifiable homicides over 15 years. The rate is just over 1 per 100,000.
Number one in the US — the city with the highest rate of police-involved homicides — is Riverside! Riverside? Where the hell is Riverside? Is that even a city? I had to look it up. Yes, it is a city. Also a county. In California. But since other cities in Riverside County report their own data, I have to assume that “Riverside” is just the city of Riverside.
According the UCR, there have been 68 justifiable police-involved homicides in Riverside over 15 years. With a (growing) population of 316,000, this is a crazy high police-involved homicide rate of 1.43. This rate of police-involved homicides in Riverside is higher than the overall homicide rate in 43 countries! Anyway…
The top 25 cities for justifiable police-involved homicides from 1998-2012 — and this is only based on data police provide to the UCR — but if you do rank high on this list, you probably do provide accurate numbers. It’s also only for the most populous 70 cities in America. Homicide numbers are total for 15 years. The rate is annual, per 100,000 residents. And there may be errors.
Also keep in mind this doesn’t take the city’s crime rate into account. Sure, Baltimore cops shoot a lot of people. But a lot of criminals in Baltimore need to be shot! Baltimorians shoot each other even more. So while the BPD does indeed kill citizens twice as often as say, police in Chicago, there are more situations, per capita, in Baltimore (and Newark and St. Louis) where police need to shoot. So more noteworthy are lower-crime cities that have a lot of police-involved shootings. I don’t know what to say about Riverside or Sacramento or Las Vegas, Nevada. And why are six of the cities in California?
Justifiable Police-Involved Homicides, UCR data, 1998-2012
The tenth and perhaps last in a series from Sgt. Adam Plantinga’s excellent 400 Things Cops Know: Street-Smart Lessons from a Veteran Patrolman:
Once the fleeing vehicle has finally come to a halt, your training dictates that you then conduct a high risk stop. You park your squad car in a position of tactical advantage and order the occupants out in a systematic, cautious fashion. But cops being cops, that’s not often how it goes. Instead of the high risk stop, you get the felony rush. Or the blue swarm, or the polyester pile, all different terms to describe cops, guns drawn, who run directly at the target car, sometimes in each other’s crossfire, sometimes jumping up on the hood of the fleeing vehicle, in order to yank the occupants out through any available open window, their adrenaline so high they can’t wait, like a kid tearing open her Christmas presents on December 24th. These same cops have the tendency to fade away once the excitement is over, and only the lengthy police reports loom. Everyone likes to go to the party but no one wants to clean up.
I might to do a few more of these. But then that’s it. There are still 390 left in the book, and I’m not going to go through all of them. So buy the damn book!
This is the second of two postson basic math.
Use rates when you want compare something in groups of different sizes.
Say New York City has 400 homicides a year. Say Baltimore City has 300 homicides. Is New York more dangerous than Baltimore because New York has more homicides. No. Because New York is much larger. But the homicide numbers don’t tell us that. Rates take different population sizes into account.
A rate in criminal justice is how often something happens per 100,000 people. (Rates don’t have to be per 100,000, but in criminal justice statistics, they almost always are.)
If Baltimore had 300 homicides and a population of 1,000,000 people (in reality both numbers are smaller, but I want to keep the math easy), the rate tells us how many homicides there are per 100,000 people. 100,000 is one-tenth of one million. So the homicide rate will be one-tenth the homicide number. You should be able to do that in your head, but on a calculator, divide 100,000 by 1,000,000. You get 0.1. So to convert Baltimore’s homicide numbers to a homicide rate, you multiply the homicide numbers by 0.1 (the same as dividing by 10). Baltimore’s homicide rate (per 100,000) would be 30.
New York City is larger. Much larger. About eight million people. In figuring out the homicide rate, we’re asking a hypothetical question about how many homicides New York would have if it had a population of 100,000. Then we can compare it Baltimore’s rate.
To do this in your head, if the numbers are nice are round, figure how many times 100,000 goes into the population 8,000,000. The answer is 80. And since we’re saying New York has 400 homicides a year, we would divide the number of homicides 400 by 80, which gives us a homicide rate of 5.
Same thing a different way, on your calculator. Take 100,000 and divide by 8,000,000. This gives you 0.0125. Multiply 0.0125 by the number of homicides, 400. This gives you a homicide rate or 5.
New York City has a homicide rate of 5; Baltimore’s homicide rate is 30, or 6 times higher than New York’s, even though New York has more murders.
And here’s one way to check your work. The rate is per 100,000. So if the population is less than 100,000, the rate will be greater than the number (a town with 50,000 people and 2 homicides has a homicide rate of 4 per 100,000). If the population is greater than 100,000, the rate will be less than the number (a town of 200,000 people has 8 homicides, the homicide rate is 4 per 100,000).