Thursday, May 19, 2016

Known unknowns and Texas criminal-justice data

The Marshall Project this week had an article describing 13 things no one knows about the criminal justice system. Grits, of course, views such matters through a unique lens. I only care about the Texas state justice system and needn't concern myself with the the lack of apples-to-apples comparables in Louisiana or New York, etc.. So as I read this, I realized we actually have data for three or four of these gaps. On the others, we're as bereft as the rest of the country. Below, I've interspersed a Texas-centric analysis after each bulleted item listed by the Marshall Project.
Among the things we don’t know about our criminal justice system:
  • how many people have a criminal record?
Unknown for Texas. Would depend in part on how one defines a "criminal record." Class C offenses like traffic fines and municipal regulatory violations might be civil penalties in other states. Here, we regulate everything through criminal law, including businesses. (That's why, for example, Texas has 11 felonies on the books you can commit with an oyster; we regulate the oyster industry through criminal law, not civil regulations.)
  • how many people have served time in prison or jail?
Unknown for Texas. Maybe we should count? If you've served time in prison or jail, raise your hand ...
  • how many children are on some type of supervision or probation?
The Texas Juvenile Justice Department undoubtedly knows this number, though it's not published in precisely that way. According to their annual report, "There were 62,535 formal referrals to juvenile probation departments throughout the state in fiscal year 2015.  This represents  a 2% decrease from  the  previous  year’s 63,914 formal  referrals." Also the court system tracks juvenile case filings in some detail, though you have to do a bit of work combining data from different court types to get to aggregate statewide numbers.
  • how many juvenile offenders graduate to become adult offenders?
Hmmmm. Nope.
  • how often people reoffend after being released from prison?
We have this information for Texas; the most recent report was published in February 2015.
  • how many shootings there are in America?
America's on it's own for this one. But on Texas' behalf, let me suggest a tweak to the data request: Be sure to flag in how many of those shootings was the alleged victim a "sonofabitch" who needed shooting. (Ask Billy Joe Shaver what I'm talking about, if you don't know.) As Shannon Edmonds once reminded us, there's an unwritten rule in every Texas county courthouse that "It ain't against the law to kill a sonofabitch." Then you can subtract that number out and reach the estimate on non-sonofabitch shootings that people actually care about. (/snark)
  • how many police are investigated or prosecuted for misconduct?
This will be secret in civil-service cities, unresearched in the others. Information is siloed, buried in individual department record keeping systems which are not shared. OTOH, there's no good reason such information could not be published.
  • how many people in America own guns?
Texas has more than one million concealed carry permit holders, so that's a start. Short answer: Enough Texans own guns that gun-control based solutions are a practical non-starter here and anti-violence advocates must look to other strategies if they hope to succeed under the current Republican regime. Or was that not the answer you were hoping for?
  • how often police stop pedestrians or motorists?
Texas has this data, sort of. Each jurisdictions reports the number of traffic stops in their annual racial profiling reports and some but not all jurisdictions also report pedestrian stops. These reports are compiled by TCLE on their website but are not routinely aggregated. So it's possible to know those totals - the data exists - but it would be a fair amount of legwork and statewide information is not immediately available.
  • how many incidents of domestic violence are reported to police?
The number of "reported" DV incidents is unknowable. Cases may be dismissed, plea to other things, etc., so it's hard to identify them in the system. They're not necessarily flagged or tracked that way. For that matter, you could say that about a lot of things. For example: "We don't know how many injuries or deaths happen in bar fights." That's true for the same reason, because the data's not tracked that way. UPDATE: We do have information on DV arrests and convictions, a commenter pointed out. But I don't think we have the number for "reports" that don't result in arrests.
  • what percentage of those eligible for parole are granted release from prison?
Texas knows this number. As of Aug. 31, 2014, right at half of TDCJ prisoners were eligible for parole, according to the 2014 annual statistical report, p. 17. And the number of inmates whose cases are reviewed and the outcomes can be determined from Board of Pardons and Parole reporting. If you couldn't get there from those published reports, you could with a choice open records request or two.
  • how many corrections officers are disciplined or prosecuted for abusing prisoners?
Somebody at TDCJ Office of Inspector General probably keeps count, presumably on their fingers.
  • how many criminal cases are referred to prosecutors and how they decide which to pursue?
We know how many arrests there are which would roughly correspond to referrals. And we know the number of case filings by prosecutors, which have gone down in recent years after a decade of rising in defiance of falling crime trends. But, with the exception of a few consultant's reports aimed at advising counties on reducing jail overcrowding, few people ever dig more deeply into the process to judge prosecutors' pre-plea decision making.
Bottom line: We have more data on a few of these topics for Texas than apparently is available nationwide. But there remain huge swaths of landscape through which we're flying blind.


Anonymous said...

Many of these issues can be resolved through software programming. However, you have so many different agencies compiling different types of data in different software programs and juggling between which data their state wants and which the feds want. Every LE agency could dump all of their data in a database, but who's going to weed through it to understand it? It amazes me that CJAD knows how many probationers there are in Texas when there are multiple CSCD records software programs in use that don't always cooperate well with CSTS. Will we ever really know? Not until there's a financial investment in getting this data. Then again, we all agree our bridges need to fixed but still can't agree to actually fix them.

Gritsforbreakfast said...

With respect, 8:19, we can resolve our data issues a LOT more cheaply than we can fix the bridges and roads! Those costs differ by many orders of magnitude. There's even an argument to be made that, over time, new strategies made possible by improved data collection and analysis could improve success rates and reduce costs. You cannot manage what you cannot measure, meaning all these data-free zones in the justice system are probably also poorly managed from a perspective of fiduciary duty to taxpayers, much less from the perspective of a criminal-justice reformer.

You'll hear people say we should cut prisons to spend money on roads, schools, etc., but that's not real. The realistic goal is that Texas can save enough money via decarceration to be able to afford to run what's left of the system in a somewhat more responsible and less dysfunctional way. Basically, that means reducing low-risk admissions, closing down dangerous and/or understaffed units and focusing incarceration resources and programming on fewer, on-average higher risk offenders. We're talking about shifting around hundreds of millions of dollars within the budget to avoid future increases, whereas many billions, with a capital "B," are needed to solve Texas' road-and-bridge problems. Ditto for schools.

Speaking broadly, criminal justice is basically the fourth biggest state expense behind transportation, schools, and healthcare, but it's a distant fourth. Those others are much bigger dogs, and they must eat.

Anonymous said...

how many incidents of domestic violence are reported to police?

Texas has it - DPS reports it in Chapter 5 of "Crime in Texas". The total number of Texas family violence incidents in 2014 was 185,817. This represented a 0.2 percent increase when compared to 2013. These incidents involved 201,051 victims (up 0.7 percent from 2013) and 195,511 offenders (up 0.4 percent from 013). which has an absurdly detailed breakdown of the domestic relationship - wife (15%), common law husband (1%), all the way to male roommate (2%).

I think data exist to calculate 'how many people have served time in prison,' though there is no way to know if these people are still living or living in Texas. Same with 'how many people have a criminal record' - the data exists, but to your point the question needs to be defined a lot more clearly. And, 'how many juvenile offenders graduate to become adult offenders?' is also doable if that is Texas juvenile offenders becoming Texas adult offenders. I feel like this question has been answered before as a one off report, but Dr Google isn't helping confirm it.

Gritsforbreakfast said...

Thanks, 9:01, good point. In my mind's eye, I was distinguishing "reports" of DV from arrests and convictions. We can know the latter two, but there's a category of reports which don't result in arrests that aren't captured in DPS data.

Anonymous said...

Ah - yeah Grits, I see your distinction now that I had a cup of coffee. Beyond digging through 911 call logs (talk about horrible data systems), I have no idea. I wonder what Marshall Project actually meant with that question?

Gritsforbreakfast said...

Exactly, OTOH, I may be overthinking it. You're right it's hard to tell exactly what they meant. A number of the questions ask about things for which we have some relevant data but not precisely the datapoint they reference.

Anonymous said...

An interesting occurrence within TDCJ walls is the frequency of sexual relationships between inmates and officers. Though it IS a crime,even when consentual, rarely are ANY officers prosecuted and even more rare, IF ever, when it's a female officer.

Soronel Haetir said...

As I understood it from Prof Berman flagging this at SLP the piece was complaining more that we don't know these things in a timely manner - that it doesn't do us a whole lot of good that we know what the answer were for 2013 or even 2014 but not now.

However, given the way pretty much everything related to criminal enforcement is always in motion I'm not at all sure that we can do a lot better. The how many juveniles graduate to adult criminal careers is a prime example of this, no matter how much time passes all we can say about any particular person is that they have not yet returned to criminal activity.

Gritsforbreakfast said...

@Soronel, I didn't read it that way. But regardless, I don't think the delayed data is too a great disability because institutional change can't be implemented quickly, anyway and it's better to set policy reacting to long-term trends than short-term blips. The main downside is it gives the media so much time to engage in false narratives and demagoguery before real data comes out, especially on crime rates, etc.. These institutions are large, cumbersome and move incredibly slowly. Patience is an uncommon and unappreciated virtue these days, but given how, when, and why data is collected, waiting 18 months to get national data from thousands of independent jurisdictions isn't unreasonable to ask. Far more worrisome, to me, anyway, are the areas where no data exists at all and we're essentially flying blind.

Anonymous said...

Read this:

The Innocence Project May Have Framed A Man For A Crime He Didn't Commit.