Friday, December 29, 2017

Murder down in big Texas cities in 2017, violent crime up in some

As the year closes out, the Brennan Center projected murder and violent crime rates for 2017 based on reporting from the nation's 30 largest cities, including six in Texas. They found that, in aggregate, both had declined from 2016 levels. Looking closer at Texas' data, complicated the story. Here are the changes in murder rates for the Texas cities in the survey:
Houston: -20.5%
San Antonio: -13.5%
Dallas: -1.7%
Austin: -27%
Fort Worth: +0.1%
El Paso: +29.4%
The increase in El Paso's murder rate was from 2.9 to 3.8 per 100,000 residents and represented an increase of 6 murders over the 2016 total.

Murders, though, are rare events occurring in relatively small numbers, which is why it's never ideal to look at one or two years in isolation. Overall, murder rates peaked in 1991 and have declined steadily ever since, with a couple of one or two year upward hiccups, as occurred in 2015 and 2016. But 2017 renewed that downward trend with what Brennan projects will be the second lowest rate for murders overall since the 1990s.

Other violent crime rates, however, rose slightly in Texas' largest cities, despite a slight overall national decline. For the same cities, here are Brennan's projections for year-over-year changes in violent-crime rates from 2016 to 2017:
Houston: +8.8%
San Antonio: +4%
Dallas: +4/8%
Austin: -7.5%
Fort Worth: +5%
El Paso: -7.8%
Now, these are single-digit increases in four of six cities following two decades of decline, so take them with a grain of salt, but it does buck the national trend, which saw violent crime decline 0.6% overall.

When you look at the data city-by-city, however, that average masks wide variation. For example, though declines in Austin and El Paso look quite good, some large cities posted even larger reductions, including: Columbus, Ohio (-12.4%), Indianapolis (-17.9%) and Washington, D.C. (-27.5%). And some quite large cities like New York (-7.4%), Detroit (-9.6%), and Boston (-9.7%) witnessed crime declines equal to or greater than any Texas jurisdiction.

What to make of this jumble? Grits honestly can't say. In most cases, imputing causation to the yearly vicissitudes of crime data (much less any partisan implications), IMO amounts to a fool's errand. I don't believe the new Houston Police Chief, Art Acevedo, can take credit for the decline in murders in his town, for example, nor do I believe he is responsible for the increase in violent crime on his watch. Crime rates measure societal trends which are bigger than the interests or actions of the government, including the police.

In fact, about the only thing you can tell from anyone's interpretation of short-term crime data at this historical moment are the hopes, aspirations, and interests of the interpreter. Crime rates have been dropping for a long time, but nobody knows whether or not we've hit the bottom. Are increases in violent crime rates in those four cities a blip, or the beginning of a longer-term bounce that will sustain over multiple years? No one but a demagogue can do aught but guess, and anybody who does so without offering lots of caveats is most certainly promoting an agenda.


BarkGrowlBite said...

Scott, please let me suggest you take a deep breath before you believe the crime stats that cities float around.

Several years ago professors John Eterno and Eli Silverman did a study of New York police crime stats. They found that NYPD routinely manipulated those stats to make the city look safer.
They even had the gall to list forcible rapes - and this is no joke - as illegal entry.

Some years back, Houston PD listed several deaths as suicides even though the victims had bee shot multiple times. And just recently a Los Angeles police captain revealed that LAPD, like NYPD was manipulating crime stats to make the City of Angels (?) look better.

And then there are those rosy Texas parole stats. There is no way Texas parole failures can be anywhere near as low as claimed when the rest of the states have failure rates of 50-60 percent.

Benjamin Disraeli may not have said it, but there is no denying that "There are three kinds of lies: lies, damned lies, and statistics."

Happy New Year!

Gritsforbreakfast said...

@BGB, first, UCR are reported crimes. The misclassified ones are a few around the edges, but broadly they do reflect societal trends, and the only alternative metric (the crime victim survey) has broadly reflected the same trends (though they JUST changed the survey's methodology to make it incomparable to past years, which is making me really grumpy). The problem is, you can't offer alternative data, only skepticism which cannot be answered since you have pre-decided to disbelieve all data. One can't defeat a circular argument whose foundation is "I won't believe any data which contradicts my view."

Re: parole. The reasons parole failures in Texas are low are two-fold: 1) We incarcerate vast numbers of low-risk people who would not have committed more prison-worthy offenses even if they had not been incarcerated. If we incarcerated people at the same rates as CA or NY, for example, we'd have parole failures just as high because they'd be mostly high-risk paroleees. That's the biggest difference. But also, 2) parole, unlike probation, has aggressively taken advantage of Intermediate Sanctions Facilities funded starting in 2007, particularly for offenders with addiction problems when they relapse, and that helped drive revocation numbers down.

It borders on philistinism to reflexively disbelieve the only available data because they fly in the face of your assumptions, which themselves are not supported by any documentable facts. Richard Feynman famously said, "The first principle is that you must not fool yourself – and you are the easiest person to fool." When you discount data because it disagrees with your opinion, that's exactly the outcome risked. Far better, IMO, to accept reality - including the best available data, acknowledging that it's imperfect - then let the chips fall where they may. It makes life more interesting when you do not require that the world always agree with you.

Happy new year to you, too, BGB! I hope you have an interesting one.

BarkGrowlBite said...

Thanks for your best wishes.

You gave me a bad rap when you said I disbelieve all data. But I certainly am skeptical of the UCR data submitted by NYPD, Houston PD and Los Angeles PD.

The manipulation of crime stats by NYPD was far from "around the edges." Professors Eterno and Silverman found those manipulations to be systemic for NYPD.

Last November, LAPD Capt. Lillian Carranza blew the whistle on the department for "underreporting" crime. That does not appear to be just "around the edges." (She must be getting ready to retire because her career with LAPD is kaput.)

According to the Board of Pardons and Parole statistical report for FY 2014, only 10.67 percent of all Texas parolees were returned to prison for either new convictions or technical parole revocations. Even if only "low risk" inmates are paroled in Texas, that low number is highly suspect. If true, then most of the parolees must have died while on parole.


Anonymous said...

So what do you have to back up your beliefs besides anecdotes and a career stoked on confirmation bias?