WASHINGTON — The military's premier crime lab has botched more of its evidence testing than has been previously known, raising broader questions about the quality of the forensic work relied on to convict soldiers, sailors, airmen and Marines.
Now, the Supreme Court could weigh in, while two senators want the Pentagon to open a full-blown investigation. If they start looking, Pentagon officials will find that the crime lab's problems extend beyond one discredited analyst.
The scrutiny comes after McClatchy published a series of stories detailing how a former long-time forensics analyst at the Army Criminal Investigation Laboratory made false statements and mishandled dozens of tests.
A follow-up McClatchy investigation reveals that a second lab analyst, responsible for firearms, was quietly fired for making a false statement and destroying evidence. The lab subsequently had to review 541 firearms cases to make sure they were thorough, properly conducted and met legal requirements. Ultimately, officials determined that none of them needed full retesting.
More recently, a third lab analyst, who handles fingerprints, was found to have erred in at least three cases, one involving murder.
But the previously undisclosed problems go beyond discredited or flawed individual analysts. Some lab employees "do not like...(the) leadership style" of a top lab manager, an Army official conceded in a court deposition in March. Six formal discrimination or retaliation complaints have been filed against lab management in the past three years. One was filed by the lab's former chief attorney, who had helped oversee prior internal investigations into the lab's mistakes.
"The problem is not with just one person, but systemic," said David Sheldon, an attorney for a former Navy man who's challenging the lab's work in an appeal to the Supreme Court. "It's as if (the lab) has had no oversight, and one has to seriously question whether or not it can effectively police itself."
The Atlanta-based lab, commonly known as USACIL, serves all the military branches, handling evidence in more than 3,000 cases annually. The director, Larry Chelko, has been in charge since 1993.
On May 12, the Senate Judiciary Committee chairman, a Democrat, joined a Republican member of the committee in asking the Defense Department's inspector general to investigate the alleged misconduct of one analyst, Phillip Mills.
"Falsified lab tests could have contributed to criminals remaining free and innocent people being wrongfully convicted," wrote Democratic Sen. Patrick Leahy of Vermont and Republican Sen. Charles Grassley of Iowa. "The failure to address these issues in a timely manner could damage the nation's trust in the military justice system."
Lab officials have vigorously defended their oversight and the handling of mistakes and misconduct.
"As with all crime labs across the county, human error does occur from time to time," the Army's Criminal Investigation Command, which oversees the lab, said in a statement. "The control mechanisms designed to identify and correct those issues have proven very effective."
The Army added that the lab's leadership, including Chelko, handled the problems appropriately.
Still, lab officials have been close-mouthed with the news media and slow to inform defense attorneys about mistakes.
The lab never publicly acknowledged the extent of the problem with Mills' work. Many defense attorneys never learned of the massive retesting effort. When McClatchy asked whether other analysts had made mistakes, the military initially refused to say, acknowledging problems only when confronted with details obtained from outside sources.
The military also turned down or responded narrowly to numerous open records requests McClatchy filed that sought more information about Mills' cases, citing privacy rights.
Military officials say they can't comment in detail about the most recent mistakes for fear of tainting jurors. They also can't discuss the discrimination allegations because of privacy laws.
McClatchy pieced together a fuller picture after independently obtaining military documents and conducting dozens of interviews.
In March, McClatchy revealed the military's three-year, $1.4 million internal review of Mills' work. The reviewers disagreed with Mills' DNA results 55 percent of the time. More broadly, they found that Mills neglected tests and overlooked evidence. Making matters worse, officials found that evidence had been routinely destroyed by investigators in hundreds of other cases and therefore couldn't be retested. The investigators had been following policy at the time.
"The result," forensics analyst Robert Shaler noted in an independent review of the lab, "might have led to a miscarriage of justice."
While investigating Mills, lab officials learned of firearms analyst Michael Brooks' mistake.
Brooks had said he'd examined a hat for gunshot residue and concluded that the weapon had been fired at close range. His supervisor, however, later discovered the hat hadn't been tested. The victim allegedly had shot himself in the right temple, but the hole in the hat was on the left side.
Brooks, who could not be reached for comment, later destroyed evidence from the case file and lied about his actions, investigators concluded.
The lab fired him in 2006.
More recently, a latent fingerprint examiner has been found to have missed several fingerprints. The examiner isn't accused of misconduct, but her mistakes could cast doubt on the overall quality of the lab's work.
The examiner's mistakes came up earlier this year in a pre-indictment hearing of Army Spc. Neftaly Platero. He's accused of killing two of his 3rd Infantry Division roommates and wounding a third in Iraq last year.
The forensic evidence is especially important in Platero's case because the victim who survived doesn't remember the events, said Platero's attorney Guy Womack.
Womack received two separate fingerprint reports with troubling differences prepared by lab examiner Shauna Steffan. In the second report, she identified Platero's prints where she hadn't noticed them before. The lab wasn't required to directly inform Womack of the reasons for the disparity, but he uncovered it during cross-examination.
"I asked why it was retested," Womack said, "and she started squirming."
A review of Steffan's other work found that she'd made similar mistakes in two other cases, Womack said.
When contacted by McClatchy, Steffan said she was instructed not to talk to the media. The Army later said that Steffan didn't wish to be interviewed.
Known as "missed identifications," Steffan's mistakes are an "inevitable part of forensic latent-print casework in all laboratories," the Army said in a statement. Military officials stressed that Steffan didn't make the more serious error in the murder case of misidentifying prints and a corrected report was issued before trial.
As it happens, Womack also represented one of two convicted murderers whose evidence Mills examined.
Mills was found to have missed bloodstains in the 2002 murder trial of Army Pfcs. Jonathan Schroeder and Andrew Humiston, who pleaded guilty to beating a fellow soldier to death. The lapse called into question the quality of Mills' work and the lab's thoroughness in informing defendants of what had gone wrong.
Womack, Humiston's attorney, said he didn't know about the long-ago retesting of the evidence until McClatchy informed him recently.
"It's shady dealing on their part," Womack said. "If the lab was a standup organization, you would have thought they would have reported that."
Womack isn't the only one who's been left in the dark.
As McClatchy revealed in March, naval officers Samuel Harris and Roger House were convicted of charges related to a sexual misconduct case and later exonerated by the Mills retesting effort, but they learned about the retesting only by chance.
Military officials said the lab followed proper evidence-sharing requirements in Brooks' and Mills' cases by sending advisory memos to each military service's legal division.
Nonetheless, the nonpartisan National Institute of Military Justice worries that the military hasn't made enough of an effort to inform defendants.
The institute recently asked the Pentagon's top lawyers to track down all former service members who were convicted with evidence Mills handled. From 1995 to 2005, Mills handled evidence in some 465 cases.
On Thursday, the Pentagon declined the request.
"A system that purports to be just must bring sunlight to this issue in order to put it to rest," said Michelle Lindo McCluer, the institute's executive director.
Despite the Pentagon's reluctance to take a closer look at the lab, others might.
Lab worker Albert Bell has filed a discrimination complaint with the Equal Employment Opportunity Commission contending that lab supervisors engaged in a campaign of "harassment and intimidation" because he's African-American.
"My experience with the lab overwhelmingly tells me that the primary mode is to cover up any misfeasance," his attorney Pete Lown said.
And on May 27, Supreme Court justices will decide whether to hear an appeal that attorney Sheldon filed on behalf of Ivor Luke, a former Navy hospital corpsman second class.
Luke was convicted of indecent assault in 1999, in part with the help of Mills' testimony. Luke served his prison term. Because the evidence in his case had been destroyed, it couldn't be retested after Mills' errors emerged.
Although Luke's appeal could be a long shot, it raises a fundamental question about the military crime lab: How many mistakes can any one analyst, or any one lab, make before all the results must be second-guessed?
The U.S. Court of Appeals for the Armed Forces offered one answer when it rejected Luke's appeal last January. The court reasoned that there was no indication Mills botched Luke's specific case.
Luke's lawyer, though, argues that some government witnesses can be so thoroughly discredited that they can't be relied on at all.
In a 1956 Supreme Court ruling, for instance, the justices threw out convictions based on the testimony of an anti-communist informant who officials later came to believe had made false and "bizarre" claims in other cases.
The witness "by his testimony, has poisoned the water in this reservoir, and the reservoir cannot be cleansed without first draining it of all impurity," Chief Justice Earl Warren wrote.
(Tish Wells contributed to this report.)
DOCUMENTS RELATED TO THIS STORY
MORE FROM MCCLATCHY