Sunday, August 20, 2017

Judges fumbling forensics decisions, meet-and-confer challenged, prisons and rural economies, and other stories

Here are a few odds and ends that may interest Grits readers while I'm focused elsewhere:

3,500 and counting support Just Liberty petition to limit trooper arrests
With the Public Safety Commission scheduled to consider the petition for rulemaking from Just Liberty and 15 allied organizations on Thursday, more than 3,500 Texans so far have emailed DPS Director Steve McCraw in support of more restrictive arrest policies for Class C misdemeanors. Go here to join them.

Prisons and rural economies: A case study
Prisons prop up the economies of small towns which otherwise would not have enough employment for those who live there. Palestine, with five nearby units employing 2,645 people, is a prime example. In the early '90s, then-Gov. Ann Richards touted tripling the size of the prison system as a means of rural economic development which she hoped would help stave off a GOP takeover of the state. Rural voters welcomed her investment but rejected her politics, voting for George W. Bush in the 1994 election. Meanwhile, the prison bonds she championed financed prison construction throughout the rest of the decade. In recent years, Texas has closed eight prison units, but most of them have been in areas where the local economy had shifted to other industries. Where locals rely on prisons as primary employers, closing them remains difficult.

The case for a warrant requirement on cell-phone location data
Dallas attorney Mike Lowe made the case for a warrant requirement for cell-phone location data in a recent blog post anticipating a case at SCOTUS, US v. Carpenter. See also an amicus brief from Fourth Amendment scholars on the subject.

Contemplating end of 'meet and confer' at Austin PD
Accountability advocates in Austin are suggesting the city pull out of its "meet and confer" agreement and decide funding for police as part of its annual budget:
Critics of Austin police dominated a city budgeting session Thursday evening as the City Council heard from scores of residents on what 2018 financing priorities should be. 
Those should be less money for policing and more money for social services, a large group of activists argued. They urged the council to abandon contract negotiations with the police union and instead return to funding police annually without a multiyear union contract ... 
The Austin Justice Coalition, a nonprofit organization focused on helping minorities, outlined eight changes it wants to see made to how police misconduct cases are handled. They include considering past misconduct in future discipline, eliminating the statute of limitations for discipline, considering misconduct in weighing promotions and stopping the practice of sealing officer misconduct files. 
The group said the current police contract protects misconduct and rewards a policing system in need of reform.
'How the poor get locked up and the rich go free'
An LA Times editorial included a discussion of Harris County's bail litigation.

Judges fumbling forensics decisions
Judges keep allowing disavowed science into evidence. Texas' junk science writ might help wrongfully convicted defendants seek relief on the back end, but only judges can prevent non-probative, non-scientific "forensic" analysis into evidence in the first place. The framework from the Daubert case, which laid out modern standards for admitting expert testimony, has simply broken down in the wake of evidence that many forensic methods are less reliable than long assumed. It simply doesn't provide enough leeway for judges to exclude non-probative testimony even after its falsity has been established. E.g., in Just Liberty's Reasonably Suspicious podcast the other day, we discussed a case where forensic DNA analysts excluded a suspect using one analysis while those using another type of mathematical calculation, based on the exact same data, matched the suspect. The 12th Court of Appeals said the solution was to let in both results into evidence and let the jury decide! That's hardly playing much of a "gate keeper" function.

31 comments:

KBCraig said...

Federal prison placement-as-social-experiment hasn’t worked out so well.

When congressional delegations win placement of a federal prison in an economically depressed area, they tout it as an economic lifeboat.

It doesn’t turn out that way. Employment requirements exclude most local residents. Scant housing and shopping and local schools that are often poor don’t help attract current BOP employees — at least not those who intend to stay longer than 12 months to punch a notch and get another promotion out.

Anonymous said...

Hopefully DNA analysts are making progress in recognizing when a sample is just too crappy to report an inclusion/exclusion result. There are much worse things to say in a report than 'the sample was not suitable for comparison due to insufficient quantity/quality DNA'. Trying to make silk purses out of sow's ears has plagued forensic 'science' disciplines for decades, well before DNA testing.

Wolf said...

"How the poor get locked up and the rich go free".......
While the article focuses on people stuck in jail because of lack of money, the same issue applies to our prisons. They are filled with poor, undereducated individuals who could not afford top quality legal defense. Sure there are prisoners who need to be locked up, At the same time there are many serving long sentences who could be safely released.... if only the powers that be implemented the talking point that "prisons are for those we're afraid of, not the ones we're mad at,"

Karen Page said...

"Small town economics".... My friend's husband lost his job at one of the prisons that closed. I'm sorry about that, but I informed her that prisons are not to increase the local tax revenue or to keep the unemployment rate down. Politicians get pressure from local government to keep them open and full. After so many years of mass incarceration, society doesn't give the 'human element' a second thought. There are way too many people locked up that should, and could be released without being a danger to society. Let's build manufacturing plants, not prisons.

Anonymous said...

@7:42 says:

"There are much worse things to say in a report than 'the sample was not suitable for comparison due to insufficient quantity/quality DNA'."

Sometimes that statement is absolutely the worst thing that can be said.

The fact that a DNA profile is partial or complicated is not a satisfactory justification for ignoring it, or hiding the fact that it exists by not reporting it. The most incomplete DNA profile - a single allele at a single locus - can be used to reliably exclude a person, which is a fairly significant finding if you are that person.

Gritsforbreakfast said...

@7:42, 4:24 is right. Data that's not useful for inclusion may still be used for exclusion. That's fairly common, I'm told, and not just in DNA analysis.

Anonymous said...

@7:42-

Nope. They are getting dumber. This is because the leaders of the forensic community are covering up for their own past ignorance, and only retain those analysts in their labs that are "pliable". Like home-growing mushrooms, the lab analysts are fed sh*t and kept in the dark.

"...Smears [from rape kits] are not processed in a way that is intended to preserve them for later DNA testing. As an illustration, smears are fixed and stained using various solutions. However, none of these solutions are subjected to QC testing to determine if they are contaminated with human DNA...
...following analysis, smears are not stored in the freezer in a way that would maintain their testability using DNA analysis methods. They are stored at ambient temperature, where it is known that the DNA will ultimately degrade and become untestable...
...In the case of the smears, they are retained and not discarded because they are a stable record of the test result - they can be reexamined microscopically at a later time. They are not retained because they are suitable for DNA testing..."

-Courtesy of Dr. Jeff Barnard, Chair of the Texas Forensic Science Commission, August 2009. These statement were made because his forensic analysts were handling this biologically hazardous evidence with their bare hands (no gloves) - a common practice at SWIFS. And they continued to do so with total disregard for the evidence they were self-contaminating at SWIFS. Lots of confusing and unusable DNA profiles coming out of Dallas.

Anonymous said...

10:27,
Aren't the lab reports themselves the "stable record"? There's no need to rely on the actual slides from the kit if they've already been analyzed, so there's no need to retain them at the lab. Am I missing something?

Anonymous said...


The slides from the sexual assault kits CAN be DNA tested, even after they've sat in the lab for several years. There are a few exonerations from Dallas that have used these slides. Dr. Barnard should know this. But it sounds like, if the statements written by Dr. Barnard above are true, that those analysts are purposely mishandling the evidence or at the very least acting negligently for handling and storing evidence. Barnard is the one writing the protocols and training the lab analysts, but he's about a decade behind, scientifically.

Anonymous said...

Additionally, when low quantity or very complex DNA data is used for inclusions, the match statistics when correctly done will reflect that, resulting in a very low level of statistical significance. When a DNA profile matches someone with a 1-in-2 or 1-in-5 probability, it is clear to everyone that it is a weak association that shouldn't be relied on alone.

Anonymous said...

1:21-

"when correctly done" seems to be the forensics problem. No one knows what "correct" is, even the Chair of the TFSC who has documented "incorrect" several times over.

11:15-
Here's the documented proof (and all kinds of impeachment material)...
https://goo.gl/cjag4e

Anonymous said...

@07:21:00 -

One of the things that it is sometimes difficult for non-scientists to fully understand is that in science there is rarely just one way to do things correctly. This is particularly true in statistics. Texas's state-wide review of DNA mixture interpretation procedures a couple years ago did not identify just one correct way to do the statistical interpretations of DNA profiles. Several correct, acceptable approaches were identified. Some were more robust than others, but they were all correct and acceptable. There were, however, several clearly incorrect approaches that some laboratories were using. Once identified, these incorrect approaches could be easily identified and corrected by Texas laboratories. (It is worth noting that these corrections have not yet been systematically addressed outside of Texas.)

Gritsforbreakfast said...

"Several correct, acceptable approaches were identified. Some were more robust than others, but they were all correct and acceptable"

Here, two of those methods generated contradictory results, 12:20. What then?

Anonymous said...

GFB -

@12:20 here.

This is a guess, because I haven't seen the data. But it sounds like the lab originally did an interpretation without a stochastic threshold, and therefore calculated an incorrect statistic. Then after the state-wide mixture review they over-corrected their interpretation. They didn't have a statistical procedure yet that would allow them to interpret low level profiles, and because they couldn't do the statistics they did the most conservative thing they could do, which was to say that the profile could not be interpreted. Then later they validated a procedure and software that could be used to do the calculation, and it gave the final results that were used.

That's my guess. Use it for what its worth.

Gritsforbreakfast said...

First, the "procedure and software" are secret so they cannot be independently validated. They're not revealing the secret sauce of their proprietary product.

Second, your phrase "interpret low level profiles" is suspect because what they're doing is taking incomplete data and running simulations to project what it'd say if there were "enough" data to intepret. Obama's Presidential Council of Advisers on Science on Technology said that tech isn't ready for prime time yet, but Texas appellate courts have yet to rule directly on its reliability as far as I can tell. Certainly this case did not do so.

Finally, the court did not reach any "final result" regarding the DNA forensics. They admitted in contradictory results and told the jury to pick the one they liked. Not quite a straightforward error correction in the way that you portray.

Anonymous said...

@1:37 here -

I would like to read the trial transcript. Apparently, the court made a determination that the probabilistic method passed the requirements for admissibility, and the appellate court agreed. In my understanding, conflicting testimony of any sort becomes a question of fact that the jury gets to decide. The judge gets to rule on questions of law, but not on matters of fact unless it is a trial before the court.

It may have been that the defense offered up the first (or second?) report that was most favorable to the defendant, and the prosecution offered up the last report. In which case, it seems appropriate that the jury would have to decide between them. But I am speculating, because I haven't seen the reports or the transcript.

And I don't think that the black box argument would or should go very far. Proprietary software lies behind lots of documentary evidence that is routinely admitted at trial. No one ever asks to evaluate the computer code for spreadsheet programs, or word processing programs, or cameras, or digital recording devices, or phones. I'm not sure why a program that measures the statistical weight of a DNA match would be treated any differently. But that is a legal question that I am not qualified to answer.

Anonymous said...

@12:20/1:37/3:42-
You stated, "One of the things that it is sometimes difficult for non-scientists to fully understand is that in science there is rarely just one way to do things correctly."

My Magic 8-Ball is "correct" some of the time. It may not be as robust as other methods, but it can provide a "correct" answer, and it falls within your definition of "correct."

Complicated DNA statistics aside, lab analysts disregarding written protocols is "correct" (see https://goo.gl/cjag4e)? How about lab analysts disregarding written protocols because the written protocols were not "correct" (see Austin Police Dept DNA Lab). How about lab analysts disregarding written protocols (even though they are "correct") because a Lab Supervisor declared it fiat (again, see Austin Police Dept DNA Lab)? Some protocols are more "correct" than others?

Most of the non-scientists you're mentioning end up serving as jurors. How are they to discriminate, or even begin to understand, the nuances of the numerous statistical contortions that DNA data can dance through? Jurors rely on the absolute. Were protocols followed, or not? Are the protocols scientifically validated, or not? Is the DNA expert stating a fact, or guessing at an approximation (using his/her Magic 8-Ball). Any so-called DNA expert that reiterated your Scientific-Orwellian statement in a courtroom would be (and should be) laughed out of the witness chair. It serves no purpose other than to muddy the waters drowning a jury.

Gritsforbreakfast said...

No, 3:42, the appeal was on sufficiency of the evidence, not admissibility. Still waiting on that case.

Anonymous said...

@05:26 said -

"Most of the non-scientists you're mentioning end up serving as jurors. How are they to discriminate, or even begin to understand, the nuances of the numerous statistical contortions that DNA data can dance through? Jurors rely on the absolute. Were protocols followed, or not? Are the protocols scientifically validated, or not?"

That is not what a jury is asked to do. No jury is asked to become experts in science, or even to understand the science side of things. The jury is asked to consider the facts and opinions offered by the expert witness, and either believe them or not. If the jury believes them, then it is asked to evaluate the relevance of the facts and opinions in the determination of guilt and innocence. The defense can simply challenge the expert's testimony on cross. Or, if the defense wishes it can get an expert witness to reanalyze the evidence, or to reinterpret the data collected by the state. Then the defense's expert can also testify to facts and opinions, which may differ from those of the state's expert witness. Then the jury gets to decide what is true, and whose opinion to believe, and how that information bears upon the determination of guilt and innocence.

But the bottom line is the same: the jury isn't asked to perform as scientists, or to make scientific evaluations of the evidence. That is just not how it works.

Gritsforbreakfast said...

In this case they were, 12:18. Conflicting forensics were allowed into testimony and the court said "pick the one you like." The 12th Court of Appeals said they were treating it the same as if two eyewitnesses conflicted.

Anonymous said...

12:18-

"But the bottom line is the same: the jury isn't asked to perform as scientists, or to make scientific evaluations of the evidence. That is just not how it works."

With that logic, then there is no reason an expert needs to testify at all. A Prosecutor would only need a "scientific" lab report that states "He did it!". How the results were obtained and who generated the report would be irrelevant to a jury because they couldn't possibly understand it anyway.

When Prosecutors and Defense Attorneys fail to do the background research on the experts, the labs, or the scientific principles ... and the Judges forgo their responsibilities as "gatekeeper", the unscientific juries are left to do everyone else's job without ample time or computer to research the scientific principles, without time to research the background of the experts or the labs the work at, and without the leisure of directly asking questions of witnesses during testimony to fill in the significant facts overlooked by the Prosecutor and Defense Attorneys all while confined to a jury box. This is a disservice and a disgrace to the juries who diligently spend their time and efforts serving their community. This certainly is not how it is supposed to work.

Anonymous said...

If I'm a jury member, and I have to choose between two opposing eye witnesses (i.e. scientific experts), I'd be pissed if neither the Prosecutor nor Defense Attny asked the witnesses if they wore glasses, or were drunk, or high on drugs, or wearing a low-brimmed cap, or had long hair, or if it was daytime, or nighttime, raining, snowing, lightning, or in a crowd of people, from a distance of 5 feet or 500 feet, inside a moving car, inside a parked car, inside a parked car with tinted windows, inside a parked car with tinted windows in a rain storm, or if they were being paid or blackmailed for their testimony, etc. These unremarkable, mundane, and seemingly unimportant details (to the daft Prosecution or Defense Attny who didn't ask these questions during trial, but assumed a fact instead) could be quintessential to the reliability of the eye witnesses' account. If I find out post-verdict that the eyewitness omitted information, provided misleading information, or in fact lied (i.e. scientific expert disregarding written protocols and basic scientific axioms, or failed to document an error with evidence analysis) and this was known (or discoverable) by the Prosecutor or Defense Attny before or during the trial, I'd be irate as hell. A jury can't do its job properly if the parties upstream don't do theirs. This is nothing more than a waste of time, unfair to the defendant, and fiscally irresponsible to the taxpayers.

Anonymous said...

Regarding Austin dropping their police contract, it would certainly benefit the rest of the state's taxpayers to do this, especially if it meant lowering compensation given Austin is widely considered the highest paid police agency in the state. If Austin can do this and San Antonio follows, it will definitely make an impact in the way other large departments address compensation, lowering it will free up all sorts of monies that can be used on better things.

Anonymous said...

@9:26 -

It is not the role of a jury to do what you are saying. You don't seem to understand how the jury trial process works.

Anonymous said...

@5:43-

Enlighten us with your "wisdom".

Anonymous said...

@2:37 -

https://en.m.wikipedia.org/wiki/United_States_criminal_procedure

Anonymous said...

@2:37 -

If the previous link was too detailed, this one might work better:

http://www.ducksters.com/history/us_government/serving_on_a_jury.php

Anonymous said...

@2:37 -

Or, if you want the very nutshell explanation from the second link, here it is:

5) Where are jury members supposed to get their information regarding a case?

A) Evidence at the trial
B) Internet
C) Social media
D) Newspapers
E) Television

The correct answer is A.

Anonymous said...

@10:54-

Obviously you missed the concept of the comment of @9:26.

When Prosecutors, Defense Attnys, and Judges don't do their jobs correctly or thoroughly, the jury is supposed to assemble an opinion without all the facts that COULD and SHOULD be presented. But the jury can't possibly get all the facts when confined to a jury box. They can't do someone else's job. It's an impossibility. The people with the least amount of information are supposed to determine the fate of a defendant. Thus a defendant can't get a fair trial when others don't do their jobs.

If you read the document at https://goo.gl/cjag4e, you would learn that the lab analysts were deviating from lab protocols on a regular basis, and without documentation for the better part of a decade. How many trials should this have affected?
The Prosecutors had this document.
The Defense Attnys may have found (or been presented with) this document, but unlikely. Also unlikely that they understood it.
The Judges may have stumbled across it, but highly unlikely.
The testifying lab analysts won't acknowledge it, or will provide fraudulent information knowing that the Defense Attny can't counter. (Deviations from written protocols is most problematic when a surrogate analyst has to testify for an absent analyst. How does the surrogate analyst know if the absent analyst followed written protocols? They don't, so they guess in favor of the Prosecution's case.)
But the Jury Members certainly do not get this document, nor hear anything about it. So they don't get the ultra-critical facts that lab analysts who handled and tested the evidence (used against the defendants) were frequently and casually deviating from written protocols, cutting corners (fudging) for scientific interpretations, and irreparably damaging the evidence for re-analysis. Example- APD analysts were miscalculating DNA statistics for 6 additional years before is was discovered. It is plausible that nobody knew?

Sadly, Prosecutors and Defense Attnys and Judges work in small circles and thus may have shared knowledge of these lab violations (information accrued from successive trials). But the Jury Members are new for each trial, and don't benefit from the information gained from past trials. For each new trial, the jury comes in blind while the Prosecutors, Defense Attnys and Judges know of these issues, but don't present them to the nascent Jury Members.

Anonymous said...

"The Jury Speaks" is a good series where ex-Jury Members from high-profile cases reassess the additional evidence (evidence not presented at trial, but should have been) as it is presented years later -- a "What If" scenario. Not widely known, many jury members are harassed, taunted, and threatened post-trial because of the unpopular verdicts they arrived at. Many jury members change their opinions post-trial after they are told of additional information withheld from them, but known by the public. Too little, too late. It's scary what the jury does not know...

Anonymous said...

@5:04 -

You say, "For each new trial, the jury comes in blind ..."

That is the way jury trials are supposed to work. The jury bases their determination of the facts based upon only the evidence presented, not upon their personal knowledge. They take an oath to do this. To be frank, you seem to be a bit clueless on this point.