Long before the Austin Police Department’s DNA crime lab was shuttered, prosecutors and defense lawyers were second-guessing its work.
In 2009, prosecutors raised red flags about DNA results in a rape case, later dismissed the charges against the suspect and eventually hired two private labs to double check the department’s results before refiling them.
In 2015, a rape suspect’s trial expert said the lab’s results were so riddled with corrections and bad math that the defense should order new tests.
And in early 2016, prosecutors hired an expert to review 17 criminal cases for potential problems. That expert found mistakes in all of them.
“The sloppiness of the lab has been pretty much known for a while,” said Austin defense attorney Nate Stark.
Prosecutors, defense attorneys and others in the criminal justice system say it is too soon to determine how missteps in DNA analysis could impact thousands of criminal cases that have already flowed through the courts.
But multiple cases identified by the American-Statesman through court records provide insight into how problems are beginning to play out — including suspicion of the wrong person in a sexual assault case — and provide a hint at what the future may hold in the Travis County criminal justice system.
Despite long-term skepticism about the lab’s credibility, the facility continued to operate largely unchecked until police officials abruptly closed it in June amid an audit by the Texas Forensic Science Commission that faulted the lab for shoddy work that could impact thousands of cases.
It is unclear how — and if — questions about the lab’s performance flowed from the courtroom to the top brass at Austin police headquarters. Former prosecutors say they were in regular communication with lab managers to sound the alarm with them.
Interim Police Chief Brian Manley said he’s been part of discussions about lab staffing during budget talks in recent years. However, he said concerns about their calculations didn’t land on his desk until last year, when he learned the Forensic Science Commission was troubled by what auditors were uncovering. Since then, he said, he has treated it as a top priority.
The Forensic Science Commission’s report slammed the Austin Police Department lab for using outdated calculations that overstated the certainty of the results and were “neither scientifically valid nor supported by the DNA forensic community.” The use of old procedures to interpret test results means an expert witness theoretically could tell jurors that the chances are 1 in more than a billion that the genetic material in question belonged to someone other than the defendant, when those odds really are more like 1 in 100.
The commission said employees weren’t properly trained and that the lab’s leaders refused to acknowledge their mistakes. They also identified one case in which mistakes put the wrong person under suspicion in a rape case.
That case, which involves a sexual assault that occurred in 2008, still hasn’t gone to trial. And it may never.
Implicating the innocent
In the case of Tyrone Robinson, Austin DNA lab staff contaminated a sample that could have led police to charge the wrong man with the crime.
On Oct. 12, 2008, a woman who had spent the evening partying downtown was walking to Fourth Street to hail a cab. A man in a Dodge Charger pulled up beside her at about 2:30 a.m. and made sexually derogatory remarks, according to police documents. The man then allegedly pulled his car in front of the victim and jumped out.
“The suspect grabbed the victim by both arms and forced her into the rear seat area of the vehicle,” the document states. “The victim tried to get out of the car but the doors would not open for her and appeared to be locked. The suspect turned up the radio in order for people not to hear the victim’s cry for help.”
The attacker allegedly took the woman to the Drury Inn & Suites on North Interstate 35 and raped her. The woman later escaped and went to the hospital.
Police quickly located a vehicle at the hotel that matched the victim’s description and interrogated the owner, who denied committing the assault. The man agreed to a DNA test to see if the victim’s DNA was on him.
Austin Police Department lab tests told a confusing story. The suspect’s swab included the DNA of three people: the suspect, the victim and a third unknown person. That third person was identified to be Robinson.
But the victim insisted that the first suspect, who was Hispanic, couldn’t be the rapist. Her attacker, she said, was black.
The DNA results were wrong, an investigation later showed, because the lab had contaminated the sample. During testing, the DNA of the suspect was placed adjacent to the DNA of the victim, and traces of DNA likely moved from one sample to the next, the report states — investigators referred to it as “carryover contamination.”
Prosecutors raised red flags at that time, but the lab did not appear to have addressed those concerns, the report said.
“The case file provides no documentation that questions were raised by the DA’s office or vetted by the APD DNA Section,” the report states.
Police charged Robinson with aggravated kidnapping. But in late 2013, after years of delays, prosecutors dropped the charges against Robinson “pending further investigation.” While prosecutors won’t detail why they did that, their actions over the next six months offer some clues.
In October 2013, police took another DNA sample from Robinson and the results connected him to the rape, an arrest affidavit for Robinson states. But this time, they didn’t just run it through the Austin Police Department lab. Prosecutors hired two independent labs to confirm the Police Department’s conclusion, which they did.
In March 2014, Robinson was charged with sexual assault in Harris County and that case is expected to go to trial next month.
In May 2014, prosecutors refiled charges against him, this time for aggravated kidnapping and aggravated sexual assault. He was released on bail and was charged with another sexual assault in Harris County this summer.
Travis County First Assistant District Attorney Mindy Montford declined to comment on the Robinson case because it is ongoing.
“The defendant is currently in custody in Harris County, and we are deferring prosecution to them,” she said.
Montford said Travis County prosecutors are assisting the Harris County district attorney’s office and will evaluate the local case again at the conclusion of Robinson’s upcoming trial in February.
Bad math called out
Suspicions of poor work had swirled around the lab for years. In 2010, former employee Cecily Hamilton accused a fellow lab technician — Diana Morales — of cheating on a proficiency exam and doing substandard work. Both internal and external reviews dismissed her claims, saying Morales appeared competent and qualified.
But complaints about the math at the DNA lab resurfaced again in 2015, when a defense witness raised alarms in court about the work of one of the department’s analysts. Those concerns didn’t stop the jury from finding Donald Lewis guilty of child sex assault. Instead, it was yet another airing of lab problems that were plain to experts who scrutinized the lab’s DNA analysis yet did not prompt a thorough review at the Austin Police Department.
The claims stemmed from the criminal case of Lewis, an Austin man accused of sexually assaulting a 13-year-old girl who ran away from a group home for troubled youth. As Lewis’ trial approached, his lawyer hired a private lab, Cellmark Forensics, to review more than 400 pages of documents from the case, including Austin Police Department lab reports and the curricula vitae of laboratory analysts at the facility.
Hiring an expert to malign the state’s DNA results is routine for defense lawyers. The expert did not dispute that Lewis’ DNA was on the victim. But what she did say foreshadowed flaws that would ultimately bring down the lab: sloppy work and bad math. She also specifically pointed out problems with Morales’ work, pointing out the high number of corrections in her documentation.
“In summary, the number of corrections to the paperwork from the screening and sampling of the evidence raises concerns regarding the quality of the work performed at the initial stages of analysis in this case. …” the expert wrote. “I also have unresolved concerns regarding the analysis and the statistical calculations performed in this case. Given the concerns regarding the work that was performed in this case, I would suggest independent testing by another laboratory to determine if the results obtained by the Austin Police Department Forensic Science Services Division Laboratory can be confirmed.”
The expert testified at Lewis’ trial. In April 2015, Lewis was convicted of one count of aggravated sexual assault and one count of indecency with a child by contact. He was sentenced to 40 years.
The conviction illustrates that while DNA evidence can be powerful, it is very rarely the only piece of evidence responsible for guilty verdicts.
“DNA is a tool,” said Randall Sims, president of the Texas District and County Attorneys Association. “We have a lot of tools in our toolkit, and DNA is just one of them. It sure is helpful, but by the same token, you don’t have to have it to win a case if you have enough evidence.”
But weaknesses in DNA evidence also open the prosecution’s case to attack by the defense, he said.
“That is obviously a prime target for them if they have it,” Sims said. “All they have to do is convince one juror that they don’t have the right guy.”
‘Bad for both sides’
In 2015, the Texas Forensic Science Commission was surveying the scientific methods of labs across the state when it quickly zeroed in on major problems that Austin officials and auditors had missed or ignored.
Prosecutors handpicked 17 ongoing cases and hired University of North Texas Health Science Center professor Bruce Budowle to review them. The DNA expert flagged problems in all of them.
Documents obtained by the Statesman don’t provide details on all of the cases Budowle reviewed. Some don’t include the full name of the defendants, the suspected problems with their evidence or the disposition of their cases.
But of those the newspaper reviewed, the expert’s main problem was the analysts’ statistical calculations determining the probability of a random person having the same DNA.
However, the numbers in the cases the Statesman saw were so high — in the octillions, in some cases — that the different calculations didn’t significantly alter the conclusions.
But there are thousands of other DNA analyses that might have been marred by bad protocol, said defense lawyer Darla Davis, who also is a former prosecutor. Travis County and the city of Austin will be forced to spend up to $14 million, according to some estimates, to review and reopen more than 3,600 DNA samples taken from the lab.
And in some of them, the results may significantly affect the conclusions, she said.
“I think what we’re going to find out is that it’s not just a giant number versus another giant number,” Davis said.
The statistical problems aren’t the only ones that have come to light in recent months that could affect cases. Other issues that have been identified include the contamination issue, the use of old protocols and the fact that lab workers weren’t following manufacturer’s instructions for a chemical used in the analysis. Meanwhile, a lab official failed to tell lawyers and police officials in March that the lab’s freezer broke for eight days, potentially affecting the hundreds of DNA samples it held.
Every case the Austin Police Department lab touched needs to be reviewed, Davis said.
“This has been bad,” she said. “It’s bad for both sides. It’s a catastrophe for the whole criminal justice system.”