A Massachusetts man imprisoned 30 years for rape has been granted a retrial, marking the first conviction in the country overturned solely because of recent FBI admissions that its evidence experts for decades overstated forensic hair links to crimes.
Previous exonerations based on flawed hair evidence — including several in the District — included new DNA results or came in appeals launched before the FBI and Justice Department confirmed in April that FBI lab reports and testimony incriminating hundreds “exceeded the limits of science.”
The decision by the Massachusetts state judge had been awaited as a bellwether for how state and legal authorities respond to the errors, made by almost every examiner in an elite microscopic hair analysis unit in nearly all trial testimony before 2000, when DNA testing of hair had become routine.
The judge’s ruling also comes as a White House-appointed commission, Justice Department and FBI are debating how to strengthen forensic science standards, including how results are reported by law enforcement.
The federal review of convictions including FBI hair matches began in 2012 and has uncovered errors in more than 1,300 cases, including many in which defendants pleaded guilty.
Flaws have been identified in 33 of 35 death row cases, including 14 cases in which the defendant was executed or died in prison. Defenders and prosecutors are being notified to determine whether there is other evidence of guilt or grounds for appeal.
One notice went in 2014 to George Perrot, 48, a Springfield, Mass., man who was 17 when he accused of rape during a burglary and later convicted by a jury that heard now-recanted FBI testimony matching Perrot to hair found on a bedsheet.
In a 79-page opinion filed Jan. 26, Hampden County Superior Court Judge Robert J. Kane found that the FBI acknowledgment of errors on its own constituted newly discovered evidence regarding evolving science and granted Perrot a retrial.
“Justice may not have been done’” in Perrot’s conviction for the Nov. 30, 1985, rape of a 78-year-old woman in Springfield, Kane wrote, because of errors “not authoritatively recognized and addressed” earlier.
Had “any” of that hair evidence been offered today, the judge wrote, “it would have been excluded.”
Some legal observers called the ruling a milestone in how courts tackle questions about final convictions involving later-disputed or debunked forensic practices.
“This decision, although it does not represent precedent elsewhere, can be a very persuasive analysis to other trial judges facing similar questions,” said Kirsten Mayer, a partner with the Ropes & Gray law firm in Boston that represented Perrot pro bono.
In 2009, a panel of the National Academy of Sciences commissioned by Congress reported that although forensic examiners have long claimed to be able to match pattern evidence — such as hair samples, shoe and tire treads, bitemarks or marks on fired bullets — to a source with “absolute” or “scientific certainty,” only DNA analysis had been validated through statistical research.
Courts have continued to admit the other types of forensic evidence citing precedent, however, and rarely reopen related convictions in the absence of new DNA results.
That has posed a challenge to appeals of older cases in which biological evidence often is lost, destroyed or degraded.
The National Commission on Forensic Science this month is weighing whether to call on U.S. Attorney General Loretta E. Lynch to direct federal prosecutors, law enforcement agents and forensic scientists to abandon such assertions of “scientific certainty” in forensic testimony in the future, saying “these terms have no scientific meaning and may mislead” judges and juries — as happened previously with hair testimony.
The Innocence Project and National Association of Criminal Defense Lawyers — two outside legal groups assisting the FBI review — said Perrot’s was the first retrial granted and that it would help speed greater scrutiny of flawed techniques. They praised the FBI for acknowledging a “duty to correct” scientific errors and said they hoped state labs whose examiners were trained by the FBI would join that effort.
Texas, North Carolina and Massachusetts are reviewing their hair-examiner cases.
NACDL President E.G. “Gerry” Morris said the group “looks forward” to other courts recognizing the FBI concessions as new evidence.
M. Chris Fabricant, director of strategic litigation for the Innocence Project, praised Kane for holding an evidentiary hearing and basing the decision on current science “rather than a reflexive reliance on precedent.”
In a statement, the FBI said the purpose of its ongoing review “is to ensure that FBI Laboratory examiner testimony regarding microscopic hair comparison analysis . . . met accepted scientific standards,” adding: “In cases in which those standards were not met, remedial action may be taken.”
The issues raised by Perrot’s case are controversial within the FBI lab and across law enforcement because of concerns by some practitioners that strict testimony limits may do more harm than good by unduly weakening valuable techniques.
In September, during arguments in Perrot’s case, local prosecutors called an FBI Laboratory chief scientist, Marc A. LeBeau, who testified the FBI review is identifying only “potential errors” that might have been offset elsewhere in agents’ testimony.
That month, FBI Laboratory Director Christopher Todd Doss wrote the Senate Judiciary Committee that “statements have been designated as ‘errors’ regardless of whether or not they were later clarified or corrected with context.”
The following day, Sept. 15, a senior Justice Department official assured senators that federal authorities were not attempting to walk back or undermine concessions of flawed testimony. The government “will not dispute that the erroneous statements should be treated as false evidence,” Assistant Attorney General Peter Kadzik wrote the Senate committee.
Federal authorities began their investigation after The Washington Post reported in 2012 that flawed forensic hair matches might have led to convictions of hundreds of potentially innocent people since at least the 1970s, typically for murder, rape and other violent crimes.
Perrot was among the first to ask for new trial without DNA evidence in a contested proceeding. Prosecutors raised procedural objections and argued there was other evidence of guilt, although in his ruling, the Massachusetts judge noted the defense had also countered there may have been police misconduct.
Perrot, 17 at the time with multiple burglary arrests, was identified in a purse-snatching in December 1985. After lengthy interrogation, Perrot confessed to burglarizing the homes of two elderly women but denied attacking one who said she also had been raped.
The victim, since deceased, testified that although she was not wearing her eyeglasses the bearded Perrot could not have been her clean-shaven attacker. But at trial, an FBI examiner testified that a bedsheet hair matched Perrot’s, and said that it would be “extremely rare” for the examiner to be unable to distinguish hair samples from different people.
The Hampden District Attorney’s Office said it was weighing whether to challenge the judge’s decision or retry Perrot and oppose his attorneys’ request to have him released. A bail hearing is set for Feb. 8. James F. Leydon, a spokesman for District Attorney Anthony D. Gulluni, noted that state appeals courts have previously upheld the conviction, and added that the judge did not overturn convictions for a burglary to which Perrot confessed or to a previous robbery of an elderly couple.
In a statement, Senate Judiciary Committee Chairman Charles E. Grassley (R-Iowa) called the Perrot case "a prime example" of the review’s importance and of why the ongoing effort "falls short." The FBI was scheduled to brief the committee Wednesday with updates on the review.
"This case is more proof that the FBI's review of cases where work performed by its crime lab may have led to a miscarriage of justice is important and critical, and the agency needs to approach this task with the utmost care," Grassley said in a statement before the briefing. "People's lives might depend on it."
The FBI said as of September it had completed reviews of 1,537 cases in which examiners had declared hair matches before 2000 and in which results were used in convictions. As of Jan. 29, the FBI was unable to review another 389 cases because prosecutors did not respond to requests for information.
Grassley and ranking Judiciary Committee member Patrick J. Leahy (D-Vt.) have faulted the bureau’s efforts as incomplete because it has not yet identified hundreds of FBI-taught state and local crime lab analysts; checked cases before 1985 when FBI files were computerized; or conducted in-person visits or released information to help track the 389 cases with no-responses.
In response to the senators, the FBI said Monday it has put out a formal contract request for information to launch "a root cause analysis" to uncover why problems continued for so long, and has completed new written standards that define scientifically appropriate ways to explain results in testimony and lab reports for 14 of 18 forensic disciplines across the lab. Standards for the four others are in the process of being developed.
This story has been updated.