Tuesday, May 24, 2016

Visitation, data, metadata, predictions, and responding to drug war revisionism

Here are a few items which both merit Grits readers' attention and enable me to clear my browser tabs:

Dallas PD can't fill cadet class
Poor Dallas PD can't catch a break. In the wake of media demagoguery about a phony "crime wave," now they can't fill an academy class, reported WFAA. Two have been canceled and a third one may be because they can't get even ten qualified applicants to apply. The story puts the blame on low wages at DPD - that other area departments pay more, Fort Worth is poaching officers, etc.. I wonder if it's that simple?

Thoughts on visitation
Some of the content of this TDCJ 2014 report on visitation looks familiar but the format does not and it doesn't appear I've linked to it before. Though groups like the Texas Inmate Family Association and the Texas Criminal Justice Coalition have engaged the topic to some degree, there's fertile ground for advocacy work on visitation that lies fallow in Texas mainly for lack of resources. It's a gap in current movement work that a) families really care about and b) matters a lot when it comes to retaining family and community ties during incarceration that facilitate successful reentry. There were a few changes made in reaction to the above-linked survey, but much more to be done. Making visitation productive and non-traumatic for children in particular is important, as Dr. Melinda Tasca at Sam Houston State has ably documented. Recently, I've been in a couple of interesting discussions with our friends at the Prison Justice League about what would be required to make prison and jail visitation more child friendly, perhaps starting at the women's units. (The ones I've been to may require physical remodeling to make them welcoming to kids.) Visitation mostly get treated as an afterthought by both the agency and even many reformers, for reasons that are understandable but lamentable. But the lack of sustained focus on reform doesn't obviate the need for it.

Data reporting compliance doc reveal dismissal totals?
Among the recent links the Legislative Reference Library put out related to criminal records was a document Grits hadn't seen before: The "Fourteenth Report Examining Reporting Compliance to the Texas Computerized Criminal History System," published January 2016 using FT2014 data. The document was created to monitor  and improve reporting of arrest and case disposition information by local jurisdictions, which long-time readers know was a significant problem for years before the governor cracked down. Case disposition totals are broken out by county, adult and juvenile; arrests by local agency (it's a lengthy document). Most of this was already being reported elsewhere. But one interesting thing I hadn't seen before was the total number of cases referred to prosecutors that they dispose of on their own, presumably by dismissing the cases, vs. those whose final disposition occurred in court. The Office of Court Administration reports dismissal numbers but breaks them out by court type without aggregating county-wide. If that really is a datapoint on the number of cases dismissed by agency - I need to research it more to understand what they're measuring - that's kind of an interesting tidbit Grits hadn't seen.

Risk assessment intolerance
A couple of interesting articles came out this week from ProPublica and the Marshall Project about alleged benefits, biases and failures of risk assessment tools. Read them. For me, there's room for differentiating how risk assessments are used. Grits would not support using them at sentencing, just as I'm not a fan of a lot of  the junky "future dangerousness" testimony in capital cases' sentencing phase. But to advise a judge regarding pretrial release, when otherwise the default would be to keep the defendant in jail? That  seems less problematic to me. That said, we need good instruments. Grits is optimistic about some of the work the Laura and John Arnold Foundation has been funding on that score, but they'll have to work through the concerns raised in the above-lined articles. Meanwhile, an even more complex set of issues arises around risk assessments attempting to predict who will be shot and who will do the shooting, as the NY Times reported this week is happening at the Chicago PD. Salon grouped the method together with other "dystopian" police methods, and the missus quipped that this tactic smacked of a plot line from the movie Minority Report. But I'm more sympathetic to the idea of maximizing scarce resources that way and haven't fully formed an opinion yet. Hell, just because you identify a risk doesn't mean those on it will be subject to a crackdown. Some cities have experimented with creating a list like the one in Chicago then just paying those on it not to commit crimes! Assessing risk does not necessarily pre-determine how one reacts to it.

Problems go away if you ignore them, right?
Speaking of predictions, here's one which shouldn't have been suppressed. "A government report, blocked from publication a decade ago, presciently warned of an advancing, double-barreled health crisis of mental illness and substance abuse that has currently swamped the nation’s vast prison systems." In particular, "The 2006 document,  prepared by then-Surgeon General Richard Carmona, urged government and community leaders to formulate a treatment strategy for thousands of sick and addicted inmates that also would assist them after release or risk worsening public health care burdens." But "The 49-page report, Carmona said, was quashed at the time by George W. Bush administration officials who feared that such an acknowledgement would require a financial commitment that the administration was not willing to make," reported USA Today this week.

BJS: Lots more older inmates entering prison
Prisoner populations are aging, reported the Bureau of Justice Statistics. Two things are driving this. First: Longer sentences. "40 percent of state prisoners age 55 or older on December 31, 2013, had been imprisoned for at least 10 years, compared to 9 percent in 1993." Second, increased admissions of older prisoners: "Admission to prison of people age 55 or older increased 82 percent between 2003 and 2013. People age 55 or older accounted for 1 percent of state prison admissions in 1993, 2 percent in 2003 and 4 percent in 2013." Nationally, "More than four times as many prisoners age 55 or older were admitted to state prisons in 2013 (25,700) than in 1993 (6,300)."

On the limits of drug war revisionism
Grits pretty much agreed with this response from Bernadette Rabuy of the Prison Policy Initiative to revisionist arguments claiming the drug war was irrelevant to or a minor cause of mass incarceration. Good job. See related Grits posts.

Research bolsters arguments for protecting cell-phone metadata
From TechCrunch: "More proof, if proof were needed, of the privacy-stripping power of metadata. A multi-year crowdsourced study, conducted by Stanford scientists and published this week, underlines how much information can be inferred from basic phone logs cross-referenced with other public datasets." The Texas Lege for two sessions has considered bills protecting privacy of cell-phone location data, which is an important subset of metadata, but the bill couldn't make it through the House Calendars Committee despite two-thirds of House members signing on as co-authors. The reason was the bill author's squabbles over unrelated issues with House leadership and particular members of the Calendars Committee. So with Bryan Hughes either out of the Lege or moving to the senate (we find out which today), maybe that will free up a chance to create a different dynamic and finally pass a bill protecting personal phone metadata.

6 comments:

  1. The CCH report is interesting in that there are multiple LE agencies that are not listed anywhere on the report (including in the county breakdown). But also, it would be nice to know if cases disposed by the prosecutors were done so due to lack of evidence, a no bill, or pre-trial diversion.

    ReplyDelete
  2. "Two have been canceled and a third one may be because they can't get even ten qualified applicants to apply. The story puts the blame on low wages at DPD - that other area departments pay more, Fort Worth is poaching officers, etc.. I wonder if it's that simple?"

    Dallas recently eliminated their DROP retirement program for new employees as well, a substantial reduction in total compensation that is not often mentioned in news reports. While their compensation package is still substantially better than places like the city of Houston, when you are surrounded by communities that pay better and also have trouble recruiting quality candidates, you have to make choices. When Houston was faced with the same dilemma, it eased up on their hiring requirements and became less strict in their racial hiring quotas/preferences but still regularly couldn't fill a class of candidates, far more of those the city accepts flunk out than ever before.

    Some claim the answer is to pay more, others claim changing recruiting practices to be more inclusive for those with past drug convictions or other types of crimes is the way to go, but the only thing most will agree upon is that the status quo is not going to cut it.

    ReplyDelete
  3. TDCJ treat visitation as an inconvenience at best, and a security situation at worst. While many individual staff are professional and occasionally welcoming, the majority in my experience just wish it didn't happen.

    ReplyDelete
  4. "BJS: Lots more older inmates entering prison" --- We have to keep the "For Profit" prisons full somehow !! (Sarcasm)

    ReplyDelete
  5. A better use of risk assessment with sentencing outcomes is to look at the differences among low, moderate, and high risk offenders on LSI-Rs or other tools when given the same sentences (dispositions or prison lengths). You can establish, as we did when I ran the research unit in OK DOC, that sentences to prison and/or to certain lengths actually have diminishing or negative returns on the prison investments which can then be used for arguments about where fiscal inputs can be maximized. We found that for the major nonviolent offenders with low and moderate risk scores, you started getting greater recidivism after 1-2 year times served in prison. We also found that for low and moderate risk offenders, you got no better survival rates with prison only or split sentences than you did with probation. High risk offenders, otoh, had the highest survival rates/lowest recidivism with prison only sentences. By focusing on the aggregate outcomes and applications, you avoid the individual application problems that are legit concerns but you also inform policymakers better as they try to figure out where triage needs to occur in their laws and policies, something that will be more necessary and prevalent in the future, not less. Needless to say, this evidence went nowhere in OK but that doesn't mean the utility and value were lacking. Just that OK was.

    ReplyDelete
  6. Putting on a badge increases ones risk of going to prison exponentially.

    ReplyDelete