Friday, January 11, 2019

In praise of pretrial-detention algorithms, the 'failure to appear fallacy,' Louisiana ↓ technical revocations, and other stories

Several national items merit Grits readers' attention here during the calm before the 86th Texas Legislature storm.

Debtors-prison reform: Economic populism and the justice system
The New York Times this week published a lengthy, excellent feature on how criminal fines and fees oppress the poor. Good analysis and background.

Best evidence supports use of risk assessments in pretrial release decisions
Some prominent heavyweight corrections researchers responded to criticisms in the press, not to mention by left-wing advocacy groups (I'm talking to you, ACLU of Texas), of risk-assessment algorithms used to aid pretrial detention decisions. One of the authors, Jennifer Skeem, has presented to Texas probation audiences, and is someone whose research Grits has relied on for years. Everyone concerned with the use of risk assessments in pretrial release determination should read this 20-page analysis. They show that the best evidence available - taking into account all the available studies about what works - supports use of risk assessments, which generate better safety outcomes and less unfairness than judges flying by the seat of their pants. The authors document consistent research-based findings that a structured decision making process, in which judges formally take pretrial risk assessments into account, produces the best results. If you care about bail reform, read this. Their analysis more or less coincides with Grits' views, perhaps because mine are to a significant degree influenced by Skeem's scholarship: Racial disparities created by validated risk assessments in pretrial detention decisions (the analysis differs for sentencing, predictive policing, and other risk-assessment uses) are measurably less problematic than disparities and injustices such assessments prevent.

The Failure to Appear Fallacy
A lot goes in to Failure To Appear (FTA) rates, and this extensive article from The Appeal offers one of the the more nuanced looks you'll see on the topic, including a detailed description of how judges in Harris County - most of whom were ousted in the last election cycle - sabotaged bail reform in an effort to artificially drive up FTA rates and use them as a "political football." Excellent background for bail reformers.

Louisiana successfully decreased technical-probation revocations; why can't Texas?
After the Texas Legislature created Intermediate Sanctions Facilities as part of the state's widely lauded 2007 probation reforms, the parole system was able to radically reduce the number of people sent to prison for technical violations. However, the probation system could never accomplish it: half of revoked probationers in Texas were sent away for technical violations, not because they committed a new crime. So I was interested to see the Pew Charitable Trusts researching causes for a big reduction in technical probation violations in Louisiana. The Bayou State was able to reduce the amount of time revoked probationers were incarcerated, reduce the number of revocations for new crimes (read: reduce crime), and save millions of dollars in incarceration costs, Pew found. If they can do it, why can't Texas?

Prisons during shut down
The Marshall Project has the best coverage I've seen of the brutal effects on federal prisons from the government shut down. MORE: From the Washington Post.

Bail injustice worst case
For all of the flaws which may exist in Texas' pretrial-detention system, they're minimal compared to the Philippines, which is the only other nation in the world besides (some states in) the U.S. that uses wealth-based bail to determine pretrial detention. Money quote: “When you are detained in Philippine jails, you are being tortured.”


Anonymous said...

Intermediate Sanction Facilities are a terrible solution though.

From the beginning a probationer is taken into custody and held for an indefinite period in a county jail waiting for a bed to become available. While waiting they are not continuing with their treatment plans, merely living off the county until ISF has the ability to receive them. Returning from ISF Is also subject to the counties ability to collect them. The effect is a 90 day sanction could stretch as long as 150 days with almost half that time spent waiting for transport and doing nothing. Probationers can't preserve housing or employment on unreliable schedules like that, and I would argue that the "treatment" provided at an ISF facility is worthless because it doesn't continue any previous treatment and lasts too short to establish any useful habbits. The practical effect of ISF is to encourage homelessness and unemployment, a condition I would think is worse than any technical violation.

If the state was reliably entering and exiting probationers that would be one thing, if the treatment was somehow measureable and useful that would be another. But the real result is that probationers who likely never spent apn extended period locked up before are returned in a more desperate situation than they started their sentence with in the beginning.

I doubt anyone who works with drug offenders is going to say "getting them fired and out if their homes is exactly what they need to become clean and compliant!"

What amazes me is the lack of conversation about the usefulness of any of the original probation sanctions to begin with. I'd like to see a Pareto Chart of violations by county and hear what the root cause analysys of the significant issues are.

Gritsforbreakfast said...

One can grant all that, 1:08, and still also believe that a stint in an ISF is better (both for the individual and the taxpayer) than being revoked to prison for one's full sentence. They will have been fired from their job and lose their homes if incarcerated for a decade over a dirty UA OR if they are incarcerated for 3 months. But revoking them for the decade is still the worse option.

Unlike many commenters, I try to base my analysis on what is, not some pie-in-the-sky vision of what I wish could be. And under the system we have (as opposed to what I wish it were), it would be better if probation used ISFs more instead of revoking so many people to prison for tech violations.

Gadfly said...

I can think of a way to slightly reduce the stress on federal prisons.