Houston fingerprint lab plagued with errors, two-year backlog

Stunning news this week out of Houston PD about errors and possible misconduct in their fingerprint analysis unit: First an audit showed errors and problems in more than half of 548 randomly selected cases, now the Houston Chronicle is reporting that the fingerprint unit is the focus of a criminal probe; two of the three analysts on the hot seat have been with the department since the '70s. Reports the Chron:

Because the audit found workers often overlooked fingerprints or wrongly determined fingerprints weren't able to be analyzed, authorities are re-reviewing more than 4,000 violent crime cases involving fingerprints from the last six years. They're also trying to work through a 6,000-case backlog that dates back two years and includes violent and property crimes.

After disclosing the results of the audit Tuesday, authorities emphasized that no cases have been found in which suspects have been wrongly identified because of faulty fingerprint analysis.

They have declined to release the audit publicly, citing the ongoing investigation.

Oettmeier said nobody knows what the six-year review will find, but officials could go back further if evidence of vast irregularities emerge, such as numerous cases of criminals going free or any misidentifications.

“We're not saying that we don't believe there will be false identifications. We are saying we don't know. We have an obligation to look,” Oettmeier said.

We're apparently not talking about just a recent concern; the employees involved are long-timers:

One of the employees on administrative leave worked for the fingerprint unit for 14 years, another for 38 years and a third for 32 years.

The employee who resigned worked full time in the unit from 1972 to 2006 when he retired and was hired part time, police said.

The National Academies of Science included fingerprint analysis among the list of forensic disciplines not currently supported by scientific evidence. It could be studied scientifically and potentially verified, they said, but that work has never been done. There are no standardized best practices for fingerprint examiners and some studies have discovered significant error rates, particularly when analysts are given leading contextual information about what they're looking at.

Whether that's what happened at HPD is impossible to say because the audit has not been publicly released, but among those who've seen it, reports the Chron, it has:

spawned renewed discussion about the need for an independent center to analyze forensic work, possibly including fingerprints.

Much forensic evidence across the country, such as fingerprints or DNA, is analyzed by labs closely tied to law enforcement despite the inherent conflicts of interest, said UCLA law professor Jennifer Mnookin.

Another problem: There are varying standards for how fingerprint analysts are trained and prints are analyzed, meaning some labs are more accurate than others.

“This provides yet another example for why developing validated, research-based models for the field would be a very good idea,” Mnookin said.

Oettmeier said the department was open to independent analysis of evidence. And officials are closely scrutinizing all forensic units at the department, he said. The biggest challenge is rooting out a culture that has allowed such significant problems to fester for years, he said.

“What makes it difficult is some of these issues have probably been around for a long time,” he said.

Houston's is by far the largest police department in the state and it provides forensic support for jurisdictions all over and around Harris County. Depending on what's in that audit - and I can't imagine some defense attorney won't soon obtain the document on behalf of an accused client - this news could have quite far reaching implications.

Makes you glad Texas has a neutral, trusted Forensic Science Commission to investigate such problems, doesn't it?

See related Grits posts:

• Coming to grips with unscientific forensic practices
  
• NAS report: Many forensic disciplines prone to error
  
• Fallible Fingerprints: The Dustup Over Cognitive Bias
• USDOJ should embrace, not fight forensics based science
• US justice system commonly relies on shoddy forensic 'science'
• Politics push for expanded fingerprinting, biometric profiles