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

by Peter Moskos

July 1, 2017

Prelude to a post

Homicide is going up. It's been going up for two years. And yet educated people still act shocked.

I'm tired of refuting the homicide-increase deniers, but their arguments comes down to these collectively nonsensical points: A) homicide isn't up in every city; B) homicide is up a whole lot in some cities; C) the increased risk of homicide isn't spread equally among society but disproportionately concentrated among poor young black males with access to guns living in neighborhoods with historic and systemic issues of racism and segregation; and D) homicide is still lower than what it was when it was really high. To which I say A) statistically speaking, that's why we look at averages; B) indeed, that's a big problem, but it doesn't negate the general increase; C) no shit, Sherlock, same as it's ever been; D) ah, go fuck yourself!

You see, writing about this same old topic has made me cranky because I can't believe I still have to. And I'm disappointed that so-called progressives waste time building a denialist house of cards instead of rolling up their sleeves and doing something to prevent poor black men (disproportionately) from getting murdered. But for whatever reason, a few years back, many of the left ceded crime prevention to conservatives. Somehow I missed the meeting where we decided that the only important criminal justice issues were to be police misconduct and the use of lethal force against African Americans (well, that and Mass Incarceration). And when generally respectable institutions like the Brennan Center make false statements about murder -- repeatedly -- we've got a problem.

To wit:
  • Alarmingly, Chicago accounted for 55.1 percent of the total increase in urban murders.
  • A similar phenomenon occurred in 2015, when three cities — Baltimore, Chicago, and Washington, D.C. — accounted for more than half (53.5 percent) of the increase in murders.
Since 2014, violence has increased. And it’s increased a lot. But Chicago neither accounted for "55 percent of the murder increase last year" nor "55.1 percent of the total increase in urban murders"! To say so once might be a mistake. To say it a few times might reflect statistical idiocy. But to do so again and again? I don’t get it. If forced to confront this false statement, they'll probably end up saying, "it was poorly worded and we meant 55 percent of the total of the cities we looked at." [Update: yup.] But regardless, it makes no statistical sense. Talking about the percent of total change one city makes in a small sample is bullshit, statistically and morally. Because it's possible to pick a sample in which Chicago is 100 percent of the increase. I don't think they're idiots. But if not, are they trying to deceive? Or do they just get there by accident? If Chicago's increase of 254 accounted for 55 percent of the murder increase last year, that would mean a total increase of 208 murders outside Chicago last year, nationwide. The actually increase in murders in 2016 is probably 2,000 more than 2015. And 2015 was 1,500 higher than 2014.

Second, in 2015, Baltimore, Chicago, and DC accounted for nothing close to half of the increase in murders. The national increase (2014-2015) was around 1,500. 255 is 17 percent of 1,500, not 53.5 percent. So how do they come up with these numbers? I've figured it out. Put it this way, if your sample only included Baltimore, Chicago, and DC, you could say these cities accounted for 100 percent of the increase in murder. Add a few cities, and that's basically what they've done.

There's a method to what, when, and why they do what they do. They don't just pull number from thin air. They use faulty methods until they get a number they can replicate. And then they just put it in words, knowing nobody ever checks these things. Either that or the authors are complete statistical idiots, but I doubt that.

Baltimore just finished the first half of 2017 with 170 homicides, the most since 1992, when the city had 115,000 more residents.

An assistant city health commissioner who oversees anti-violence initiatives was jumped and robbed in downtown Baltimore on his way back to work after having a sandwich for lunch. In the hospital, skull fractured, he said, "I think we need to look into what is causing people to engage in this kind of behavior." No. Actually, we don't. Cause I'll tell you the cause: bad or absent parenting on top of 500-years of systemic racism combined with 20th-century government programs designed to segregate and limit the ability of blacks to succeed. I can speak the liberal shibboleth. I even believe the liberal shibboleth! So what? Now what? One can and should acknowledge history, but that won’t change it. And the greater point, at least when it comes to crime and violence, is that none of this is new. Somehow, despite social injustice and white supremacy, crime and violence had been going down for basically 25 years. The violence problem has gotten worse just in the past two years. Talking about historic social issues, as important as they are, is nothing more than a distraction to avoid dealing with today's issues of criminals and wrong-doers.

Crime wasn't supposed to go up, of course. Crime reduction, say some, is just part of the grand social justice and intersectionality equation. DOJ reports (on policing in Baltimore, for instance) focused exclusively on improving police, necessarily as that is, and ending racially disparate policing. They managed this without even talking about crime prevention and racially disparate rates of violence. This recent crime rise needn't and shouldn't have been politicized, but, as I warned, if the left won't even acknowledge an increase in violence (disproportionately among poor black men) we effectively cede any crime "solution" to the "Trumpian right." So now we get BS talk crime and terrorism, like somehow crime and terrorism is mostly due illegal immigrants and Muslim grandmothers. So yeah, I'm cranky in my middle age.

But the past two years, 2014-2016, has seen the largest two-increase in homicide, in, well, probably ever. And the response of otherwise smart people is either to A) scratch their head and go, gosh, gee, maybe it's poverty and guns and historic policies of racism. Except those haven't changed in the past two years. Or B) it's not a problem because, well, homicide is really up in Chicago? I don't even know how to counter that. If you care more about right-wing overreaction to murder than the lives of those murdered, you win. Don't care. But for people with a conscious that trumps ideology, read on.

Here are the cities I looked at: Albuquerque, Atlanta, Austin, Bakersfield, Baltimore, Boston, Charlotte, Chicago, Cleveland, Columbus, Dallas, Denver, Detroit, Durham, El Paso, Fort Worth, Fresno, Hampton Roads (Norfolk, Newport News, et al), Houston, Indianapolis, Jacksonville, Kansas City, Las Vegas, Long Beach, Los Angeles, Louisville, Memphis, Miami, Milwaukee, Minneapolis, Nashville, New Orleans, New York, Oakland, Oklahoma City, Omaha, Philadelphia, Phoenix, Pittsburgh, Portland, Raleigh, Richmond, Sacramento, San Antonio, San Diego, San Francisco, San Jose, Seattle, St Louis, Tucson, Tulsa, Washington, Wichita.

I got the homicide numbers best I could for each city going back as far as possible. It's a lot of grunt work (but actually a bit easier than it used to be, thanks to journalists keeping track).

For those cities, 2013 was the least violent year ever, with a collective 4,900 homicides. It could have gone lower; God did not ordain an urban homicide rate of 9.8 be the bottom below which no more lives could be saved. Generally, overall, homicide had been decreasing for 25 years. It could have continued to go down. But alas, people decided that police were the problem. And the problem to bad policing wasn’t better policing but less policing. How'd that turn out?

I'll push the data in the next post.


Mike Maltz said...

FYI http://www.samefacts.com/2017/02/crime-incarceration/police-stops/


andrewpwheeler said...

I will post a fuller response when I get a chance, but FYI I compiled all the UCR reported homicides by jurisdiction from 65 to 15, there is a link to the data in https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2977556.