Juvie sex-assault sentence overturned based on recantation, junk science

Slowly but surely, in leaps and starts (punctuated by inexplicable pauses and disappointing regressions), junk science long accepted in Texas courts is beginning to be more carefully vetted, at least in fringe fields where the number of cases affected is relatively small. In the latest example, Chuck Lindell at the Austin Statesman ("Central Texas man freed after conviction overturned on bad science," June 2) reported on Michael Arena's homecoming yesterday after he was released and his sentence overturned in response to a recantation by his accuser and the use of junk science at his trial. Since he was sixteen when convicted, Arena was sentenced as a juvenile and transferred to TDCJ where the 29-year old remained until yesterday:

Michael Arena, almost 13 years into a 20-year prison sentence for molesting a young cousin who later said the incident never happened, was released from prison Friday evening.

The Texas Supreme Court threw out Arena's sentence last month, ruling that a prosecution witness provided damaging false testimony during his 1999 trial — labeling Arena a pedophile based on a misused psychological test with a 35 percent error rate.

Friday morning, state District Judge Gordon Adams ordered prison officials to release Arena "forthwith pending a new (sentencing) hearing" in Bell County.

At 5:25 p.m., Arena — a mesh bag stuffed with belongings in each hand — walked out of his Dilley-area prison and into a prolonged embrace with his father, Robert Arena, freshly arrived after a giddy 220-mile drive from his Harker Heights home.

Arena has been freed, but not necessarily declared innocent:

Stephanie Arena, the cousin who accused Arena of molesting her as a 7-year-old, is prepared to testify in Arena's favor and has submitted sworn affidavits saying she lied about being sexually assaulted at the urging of her mother, who was embroiled in a bitter custody battle.

In addition, prosecutors cannot present testimony from psychologist Fred Willoughby, who in 1999 classified Arena as a pedophile based on a test that required the teen to click through images of swimsuit-clad people of various ages while the computer secretly measured how long he viewed each photo.

While testifying at Arena's trial, Willoughby overstated the test's 65 percent accuracy rate and improperly testified that a Brigham Young University study certified its accuracy. Instead, the study raised serious questions about the test, saying its ability to identify pedophiles was no better than chance.

The Supreme Court's May 18 ruling, however, rejected Arena's request to be declared innocent of aggravated sexual assault.

The Supreme Court said it could not credit Stephanie Arena's version of events because a Bell County judge determined that her recantation lacked credibility, finding that it was apparently the result of pressure by Michael Arena's family. The Supreme Court typically defers to lower courts on such judgments.

That outcome muted Friday's celebration of Arena's freedom.

"Unfortunately, despite mountains of evidence that the charges against him were not true, the court refused to find him actually innocent," defense lawyer Clint Broden said. "We are happy that Michael has won his freedom but sad that he has been deprived of 13 years of his life and that the truth has been suppressed."

What's important here from Grits' perspective is the court overturning the conviction in part because of the invalidity of the (supposedly) scientific assessment. How often has the same "Abel Assessment" been used in other cases and have they been vetted to ensure there was other, sufficient supporting evidence to justify the charges? Since these assessment functions aren't performed at accredited crime labs, they don't fall under the jurisdiction of the Forensic Science Commission, but there needs to be that sort of big-picture review that the FSC commissioned for arson cases regarding the science behind the Abel Assessment and whether it's improperly influenced the outcomes of other cases.

See prior, related Grits posts:

• New sentencing hearing ordered based on junk sex-offender science
• 'Inmate challenges pedophilia test as junk science.'

'Inmate challenges pedophilia test as junk science'

The Austin Statesman's Chuck Lindell reports ("Inmate challenges pedophilia test as junk science," March 2) on a legal challenge by TDCJ inmate Michael Arena - who was convicted of sexually assaulting two young cousins when he was a teenager - to use of a particularly insidious brand of junk science aimed at sex offenders called the Abel Assessment, which is based on similar principles to the penile plethysmograph. (Try saying that three times fast.) Neither forensic technique is allowed to be used at trial under modern evidence standards - and probably shouldn't have been in this case - but the plethysmograph is still used rather frequently in post-conviction settings in Texas, particularly among parolees who're assigned sex-offender conditions. The Abel Assessment suffers from an error rate of 35-48%, according to various estimates in Lindell's article. One study found "a 42 percent false-positive rate when non-molesters were tested."

In Arena's case, his claims about the test's inaccuracy are bolstered by the recantation of both alleged victims, who say they were encouraged to lie by their mother who was going through a bitter divorce. Writes Lindell:

Though interest in Arena's case tends to focus on his claim of innocence, his attack on the psychological test could influence future attempts to challenge allegedly bad science in the courtroom — a continuing problem that the nation's appellate courts have struggled with for decades.

The test, defense lawyers say, had an unacceptably high 35 percent error rate that was not disclosed to Arena's judge and jury. It was never intended to be used to identify pedophiles, they claim, and a university study found that its results were little better "than chance" when trying to distinguish pedophiles from non-pedophiles.

In addition, the psychologist who examined Arena inflated the test's effectiveness and scientific support when he testified at Arena's trial, leading to a reprimand from a state regulatory agency four years later, court records show.

Issues with the test seemed to resonate with many of the Supreme Court's nine justices during oral arguments in January.

Justice Nathan Hecht dismissed the test as "bordering on hokum" and less likely to yield valid results than lie-detector tests, which are not admissible in criminal court proceedings.

Justice David Medina noted that a 65 percent accuracy rate would have earned the test an F in a school setting. "To incarcerate somebody for one day, you use a standard that's not even A-plus?" Medina asked lawyers for Bell County. "That seems wrong on its face."

Lawyer John Gauntt Jr. with the Bell County attorney's office acknowledged that a 35 percent error rate was "not a suitable standard" for use in court. He also acknowledged that the prosecution's expert, Georgetown psychologist Fred Willoughby, provided false testimony about the test's effectiveness in the Arena case.

Even so, Gauntt told the court, the 20-year sentence should not be overturned because Arena cannot prove he was harmed by Willoughby's testimony — a necessary step toward earning a new sentencing trial.

Lindell includes an interview with Michael Arena, speaking from a Texas prison:


Notably, this case is being heard by the Texas Supreme Court instead of the Court of Criminal Appeals because Arena was convicted as a juvenile. That's important because civil courts generally have much more rigorous standards for scientific evidence, in part because both sides in civil litigation have money to routinely hire experts. By contrast, in most criminal cases the prosecution has most if not all of the access to lab resources and forensic expertise unless the defendant pays for it or a judge orders it. These are judges who're used to seeing science in the courtroom held to much higher standards than their counterparts on the Court of Criminal Appeals, and Lindell's account of oral arguments in Arena's case shows the Abel Assessment is getting a more skeptical reception from the Texas Supreme Court than junk science sometimes receives from Judge Keller and Co. on the other side of the building.