Internet Explorer 11 is not supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

Evidence of Failure

Overwhelmed by a flood of DNA evidence, public crime labs are performing poorly.

Warning to viewers: The television series "CSI" does not mirror life at public crime laboratories. On TV, crime scene investigators bring evidence into forensic crime labs outfitted with the latest technology. Lab tools splice and dice evidence to come up quickly with findings that finger whoever dunnit. "Every time I see "CSI," I cringe," says Paul Ferrara, director of Virginia's Division of Forensic Science. "Everyone's waiting to work on this one case. In 44 minutes, all that work is done."

Not so at Ferrara's lab, where 100,000 pieces of evidence now await scrutiny from only about 45 examiners. The average testing turnaround, from the time a piece of DNA evidence comes into the lab to the release of results, is more than six months. DNA examiners average six to eight cases a month.

Virginia is not alone. The backlog in most public crime labs is four to eight months or longer. There are staff shortages and only enough resources to work on the most egregious crimes, if those.

The working conditions and other problems in labs can lead to serious errors. At the city crime laboratory in Philadelphia, for example, there was a mix-up of DNA samples from the accused and the victim in a rape case. At a New York State police crime lab, the cocaine rocks tested as evidence were not the actual drugs that prosecutors had alleged were sold to a police informant. Then there was the man who was convicted in Tulsa, Oklahoma, based in part on a DNA test, the results of which were misinterpreted by a lab worker.

The emergence of DNA as a key crime-solving tool has added to the problem. As investigators rely more and more on DNA for solving crimes, the crime labs are overloaded. "Laboratories like my own, whether federal, state or local, are all reeling under the onslaught of physical evidence hitting the labs," Ferrara says.

The overloading, in turn, leads to difficulties further down the line. If lab work doesn't get done on time, it slows down or even defeats efforts at moving cases forward. "Crime labs are equivalent to the pinch in the middle," says David Epstein, chief scientist at the National Forensic Science Technology Center.

Despite the errors and delays, Ferrara insists that the work done in labs today is better than it's ever been. There seem to be more problems, he suggests, because labs results are in the limelight more than ever, with many more experts scrutinizing the work that's done. "Whenever humans are involved, mistakes can and will occur," he says. "You try to minimize it."

THE HOUSTON DISASTER

When you hear about crime lab problems these days, the recurring refrain is, "We don't want to become another Houston." Problems got so bad at that crime lab, housed within the police department, that at least one innocent person was wrongly convicted of a crime based on incorrect lab results. The lab voluntarily disconnected itself from the FBI's DNA database--before the FBI could terminate the connection.

According to a needs-analysis report released in July by the National Forensic Science Technology Center, top-level staff was not providing adequate supervision or support and scientists promoted to supervisory positions were not given formal training in management. Deficiencies showed up in personnel qualifications, proficiency testing and other key areas. As further proof of inadequacies, while the needs analysis was being conducted, water was dripping from a leaky roof onto scientists working on evidence.

After the report came out, the head of the crime lab and the police chief resigned, and the DNA crime lab was shut down. The Houston Police Department has had to outsource 1,500 of its cases in the past 12 months. But help may be in sight: Last month, the lab posted a want ad for an experienced supervisor who could get the lab back on its feet.

The Houston lab still faces potential liability for court verdicts that used its results. "Hundreds of past cases could well be subject to court challenges, and rightly so," says William C. Thompson, a lawyer and professor at the University of California, Irvine.

Thompson believes Houston didn't have an adequate oversight system to expose problems in the crime lab. If, for instance, lab workers embellish findings to achieve what they believe are appropriate results in a case--"Some people become white-coat vigilantes," Thompson says--there is no impartial official or agency whose task it is to look into allegations of scientific misconduct or crime lab problems. If any state or city finds out that someone in their labs mishandled findings in one case, it's important, he points out, to have a neutral person or agency in place to look into other cases that person has handled.

TRIAL AND ERROR

There are proposed solutions to crime-lab problems. Critics say crime labs should be independent entities within the criminal justice system. They should not, as they do now, operate within police and sheriffs' departments or the offices of attorneys general, prosecutors or medical examiners. There can be a perception that they are not able to analyze evidence objectively.

Virginia is often held up as a model because its crime lab is not part of a law enforcement agency. Instead, it reports to the state Department of Criminal Justice Services, whose director reports to the governor's secretary of public safety. The crime scene investigators are trained by the lab, not by a law enforcement agency.

But that has not saved it from many of the problems labs face. As with labs all over the country, understaffing is a huge issue. "It's pretty bad," Ferrara says. "And we're probably one of the best-funded laboratories around."

Chronic staff shortages and lack of training contribute to mistakes everywhere. As more cases are dumped on lab workers, the pressure to finish cases too quickly increases markedly, as does pressure to "extend opinions beyond the scientific method" and to get a particular result, according to a national survey of forensic science lab directors, performed by Wendy S. Becker, an assistant professor at the State University of New York at Albany. "Staffing problems are systemic and pervasive, and impact the quality of labs and outcomes and effectiveness," she says.

It doesn't help that the starting pay for public lab analysts is $37,000 a year, compared with $50,000 for similar private-sector positions. When divisions attached to law enforcement fight for scarce dollars within their departments, "we get lost in the shuffle," says Barry Fisher, the crime lab director for Los Angeles County sheriff's department. "Most crime labs are pretty small compared to a much larger police function. They're easily overwhelmed by needs such as patrol and investigations."

Last year, the problems at the Houston crime lab led the Texas legislature to mandate accreditation by 2005 of all crime labs handling evidence in felony cases. When the law passed, about half of the 42 crime labs in the state were not accredited. Crime lab problems in Oklahoma led to a similar law there. New York has required for many years that all forensic labs be accredited, but few other states are moving in that direction.

The American Society of Crime Lab Directors strongly promotes accreditation for crime labs but not necessarily via state mandate. In general, however, accreditation makes managing easier, says Roger Kahn, president of the lab directors' association and deputy superintendent of the Ohio Bureau of Criminal Identification and Investigation. "No one wants non-compliance issues to explain," he says. "It's a powerful risk management technique for the entire criminal justice system."

The Ohio bureau was able to standardize operations and put in policies and procedures that gave lab workers a better handle on their jobs. Training is now a requirement despite budget swings. Equipment must be operating properly.

Accreditation may force improvements, but it is no guarantee that crime labs will be performing their work smoothly. The Houston Chronicle reviewed audits of crime labs that were conducted between 1999 and 2002 and found problems at Department of Public Safety labs in Austin, Corpus Christi, El Paso, Garland, Houston, Lubbock and McAllen. All of them were accredited.

Accreditation may be only the beginning of necessary reforms. The Innocence Project, an independent non-profit legal clinic and resource center, would like crime labs to have standards as good as, or better than, any professional organization. It would also like to see each jurisdiction put in place an oversight agency with the authority to regulate lab practices and set standards for the use of private labs when public labs need to outsource. Moreover, it calls for accreditation standards that include spot-checking of labs, rigorous quality control and periodic inspection by a regulatory body. Labs should go through proficiency testing and be rated on their performance. At trial, the Project suggests, labs should be required to present information on their controls and error rates for testing procedures.

Steps like these would help make crime lab results more reliable. They might also help correct an injustice in the American legal system.

Special Projects