Tuesday, January 03, 2017

If tough-on-crime works so well, what's up with murder spikes in Texas cities?

So mass incarceration is working because crime is down, right?

Then what to make of the fact that Austin, San AntonioDallas, Houston, and Arlington all experienced significant increases in murders this year, as various media outlets have reported (although official data won't be available for a few months as the final cases of the year are categorized)?

These increases occurred despite Texas incarcerating more people - including more violent criminals - than any other state in the country. They occurred despite vast probation and parole systems monitoring hundreds of thousands of offenders in the community. They happened at a moment when more Texas cops are employed at more Texas law enforcement agencies than at any time in history, and in an era when Texas prosecutors have been granted nearly unfettered discretion. And it occurred even though the largest faction on the Court of Criminal Appeals essentially finds for the government in almost every case.

If big-government tough on crime policies failed to prevent murder spikes in Texas - which can hardly be accused of blue-state leniency when it comes to filling up courtrooms, jails, and prisons - then maybe it's not worth the expense?

That said, as is always the case when dealing with crime stats, one must place these increases in context. In Austin and San Antonio, headline writers emphasized that murder totals were the highest in two decades, which is certainly true. However, Austin had just one fewer murder in 2010, so the number isn't that much of an outlier from recent years. And besides, both Austin's and San Antonio's populations have grown monumentally since the mid-'90s (as has Arlington's, which saw an eyepopping increase in murder numbers this year). So, even after these recent spikes, murder rates remain low by comparison and residents of those cities are safer today, by a longshot, than 20 years ago. That's worth emphasizing more than press reports have done so far.

The press and public focus on murders because a) they're among the most egregious crimes b) they often come with the most compelling stories, and c) the numbers for murders are among the most solid in the crime-data realm because the dead are easy to count. Other crimes are less easy to categorize: Is it a burglary if later investigators come to suspect, but could never prove, that an adult child living with their parents was the real thief? How about if mom and dad decide not to press charges or later claim they were wrong and nothing was stolen? There are no such ambiguities when it comes to murder victims.

OTOH, murders are rare events, their numbers are generally small compared to other crimes, and totals can fluctuate quite a bit year to year for no apparent reason. Even with large, single-year spikes, one learns over time to look for long-term trends, not short-term fluctuations, to understand what's really happening. All of these cities have experienced one-year increases in murders before over the past couple of decades, even though the overall trajectory for each has been downward.

And what's happening may be different in different cities. After all, there's nothing connecting murderers but the outcome of their actions. They're solely responsible for what they do and police realistically have few means to prevent them from doing it. The cops' job is to clean up after them.

In San Antonio, the chief insisted that, “What we’re seeing now is a lot of spontaneous murders,” adding, “It’s really difficult to put a reason on it.” Dallas has also seen a spike, the majority of which the department attributed to home invasions of drug houses. But then, two years ago Dallas was crowing that its murder rates were the lowest since 1930. In Austin, the Statesman reported the murder total was "highest in nearly 20 years," but didn't mention that the per capta rates are now much lower because of the population boom. (Good for AP, which added that tidbit to their version.) Percentage increases look bigger when starting numbers are extremely small.*

Murders are terrible things but to act without understanding rarely achieves the desired results. And overreaction can also have its costs. As Chief McManus in San Antonio warned, attempts at “'arresting the problem away' or 'overpolicing' could just lead to distrust." When rare events like murders do happen, there needs to be a modicum of community trust for police to find and make use of leads, witnesses and other such indispensable assets.

I led off this column by joking about whether the murder spike means mass incarceration "works." The jibe is intended in part at my fellow Texans who've belittled other jurisdictions like Chicago or Los Angeles when they've suffered violent-crime woes. The truth is, crime has very little to do with government enforcement policies and is broadly more responsive to other societal trends and cues.

IMO, cops, prosecutors and prisons didn't have much to do with the bulk of the crime decline of the last two decades, so if they keep doing what they do and crime goes back up, that won't surprise me either. But it should surprise, and concern, anyone who has believed the tough-on-crime hype surrounding the benefits of mass incarceration. Even in red-state Texas, incarcerating more people than anywhere else, we're not immune to national trends and can't pretend being tougher has achieved any better outcomes. It hasn't.

*Pro tip for news consumers: When you see the words "fastest growing" in a headline or news article, substitute "smallest" to understand what the reporter is really describing. Things grow fast when they start small and their increases are relatively large compared to a tiny denominator. H/T to John Pfaff for harping on that observation.


Anonymous said...

Texas may be incarcerating more people than anyone else, but the better question is whether we're keeping up with the population growth? When is the last time Texas opened new prison? Without having the stats readily available, my guess is that the number or prison beds in Texas since the mid 90's hasn't increased nearly at the same rate as the population growth in this state. Even if the number of murders per capita go DOWN, if the population grows enough simple math will dictate the number of murders committed and the number of murderers who need to be incarcerated will increase. In my opinion, without an analysis of the population growth, your tough on crime argument is a non-sequitur.

Incidentally, an equally plausible (if not more justified) claim would be that the increase in murders is more directly related to the well publicized reduction in the use of the death penalty in this state over the last few years. Let that one sink in for a few moments...

Gritsforbreakfast said...

That last one's not really plausible, 6:47, since murder rates are lower in non-DP states.

My analysis, btw, includes population growth, which is why I think the rates are more important than the totals. As to your timeline, Texas filled up its last prison in the mid-aughts, mainly because conviction per arrest increased while crime plummeted. The folks diverted out are all low-level nonviolent offenders, not murderers. The facts just don't support your alternative theories.

Anonymous said...

I'm not sure how useful the average murder rate for a large city is. Nobody lives, or commits crimes, on the average. They do it at a specific place. At least in Austin, we seem to be seeing an increasing geographic concentration of murders - some neighborhoods consistently have a lot, some neighborhoods pretty much never. A more relevant measure of how safe a city is might be the size of the gap, not the overall average.

We also continue to see sharp disparity in the demographics of the people involved. Again, a better measure of a city's safety might be whether subpopulations are just as safe as the average.


Anonymous said...

Most violent crimes are attributable to drugs one way or another, one degree or another and media accounts rarely give us the true picture as to why a crime occurred. True research takes time and money and sound bytes are what keeps the public enthralled. I'd like to see a study that separates murders into two groups---domestic partners and total strangers and see if trends emerge. Obtaining money for drugs and irrational behavior resulting from drug usage appear to me as the two driving forces for murder but that is just my theory. I wonder if drugs cause the robberies and burglaries that so often result in murder?

On a related note, once TDCJ and Austin figure out that ill, elderly and harmless inmates should be put on the Medicare and Medicaid dime instead of the state budget prison space and revenue can be better allocated to handling the youthful murderers and others that need to be behind a wall.

Some surveys indicate that the statistical chances for a gun owner to be murdered are quite low for "death by intruder" versus "death by domestic partner".

FBI homicide data table eight divides murders out by the weapon of choice and might help in a study of murder.

There is an interesting cross over point between murders and suicides since each is sometimes mistakenly concluded to be the other and this might be affecting studies that often have nebulous results. Insurance investigators have told me that somewhere between two thirds and three quarters of all firearm deaths are in fact suicides---and these usually happen after a person gets bad medical news. The twist here is that, since many insurance policies don't pay on a suicide some policyholders get very creative in the scenario they leave police to unravel. A suicide usually means the family doesn't cash in on the beloved departed.

Whither the actual murder totals?

Anonymous said...

Of the 3 assertions in what might otherwise seem to be the money quote in your commentary("After all, [1] there's nothing connecting murderers but the outcome of their actions. [2] They're solely responsible for what they do and [3] police realistically have few means to prevent them from doing it."), all are overstatements, in varying degrees. Perhaps most egregious is [1]. Often, while most murderers do act independently in some sense, modest analysis can still find much connecting the events: factually, gang wars and drug turf battles drive significant segments of homicide totals; and social-cultural factors can readily be identified that link murderers who, otherwise, are too easily seen as acting independently. In one sense they do, of course, but many social-cultural factors connect them and their choices. Mass failures in the modern American family structure (mostly, the lack thereof) and the failure of American societal institutions (schools, mainly) to replace the lost sense of responsibility to an outside force or develop an adequate sense of conscience (lost with the rejection of organized religion) combine with the "usual" factors (poverty, gun accessibility, etc.) to connect murderers. The American deficit in impulse control, compared to other nations' residents, is stark, troubling, and little discussed. That deficit, too, connects many murderers. Moreover, total shootings, not murders, is a more important statistic, since whether a shooting becomes a murder is quite often a matter of luck (good or bad) of the shooter and of the victim (e.g., poor vs. good aim; body position at time of bullet entry; availability/quality of medical care).