The need for forensic reform in our justice system
By Brandon Garrett, Roy L. and Rosamond Woodruff Morgan Professor of Law, University of Virginia.
Forensics on the Hill: Part I
Two stray hairs sealed Donald Eugene Gates' fate. He was a suspect in the rape and murder of a college student in Rock Creek Park, in Washington D.C. There were no eyewitnesses. The revolver used to shoot the victim was never found. At his trial in 1982, an FBI analyst testified -- an "eight year veteran" who said he had conducted over 10,000 examinations of crime scene hairs. He said that only twice in his career had he ever been unable to distinguish different people's hair. And he said he compared Gates' hairs to two hairs found on the victim. He found the hairs "microscopically indistinguishable." It was "highly unlikely" that those hairs could have come from anyone else. On the strength of that finding, Gates was arrested. At his trial, the analyst gave the powerful forensic testimony about the hairs, a paid informant claimed Gates had confessed to him, and Gates was convicted.
This seemingly airtight murder conviction would unravel decades later to uncover a national scandal. The forensics in his case were deeply flawed, and this past year they led lawyers and journalists to help to uncover additional wrongful convictions and still more possible errors in thousands of old FBI cases spread across the country.
Gates' exoneration was a long time coming. He had maintained his innocence for years. He sought DNA testing in 1988 -- but DNA technology was new and the results were inconclusive. Meanwhile, the FBI analyst who had testified with such confidence at Gates' trial himself came under investigation in the late 1990s, after a whistle-blower repeatedly spoke out about shoddy work at the FBI lab. A 1997 review by the Department of Justice found that the analyst in Gates' case and 13 other FBI analysts reached false results and used inaccurate methods in a host of cases. The report found that the FBI analyst in Gates' case had "resorted to fabrication rather than admitting he did not know the answer." Having found his work in Gates' case unsupported, prosecutors were told to reinvestigate the case. However, nothing happened and Gates' lawyers were never told of any of this. The report on the FBI lab was kept secret. Gates' lawyers were never told that the FBI analyst had also been found to have testified falsely in many other cases, including death penalty cases.
The lid did not begin to come off the scandal until a decade later, in 2007, when Gates sought DNA testing again. This time he got DNA results that cleared him and matched another man. He was exonerated in 2009, after spending 28 years in prison. The judge called it "outrageous" that the report of botched FBI cases had been kept a secret and ordered an investigation into all FBI hair cases in Washington D.C.
Why didn't anyone tell Gates over a decade before that the hair analysis in his case was all wrong? In fact, the hair testimony was scientifically unsound. There is no research on how common microscopic hair characteristics are. Hair is not like DNA or even a fingerprint. Hairs vary quite a bit on the same person's body, which is why analysts would try to compare large sets of hair under a microscope. The FBI knew this well, having convened an entire symposium on the subject in the 1980s, to make clear that conclusions about hair comparisons must be extremely cautious. Yet analysts still crossed the line, and judges still let analysts offer exaggerated and scientific-sounding conclusions to the jury.
I saw countless examples of cases just like Gates', when I researched the cases of the first 250 people exonerated by DNA tests for a book, Convicting the Innocent: Where Criminal Prosecutions Go Wrong. I found that similarly overstated, unreliable, and invalid forensics were present in a majority of those criminal trials of the innocent. There were dozens more cases like Gates' involving botched hair comparisons.
One of the best known of those cases is that of Ron Williamson, who was sent to death row in Oklahoma. John Grisham wrote a non-fiction book about the case. In 1995, not knowing Williamson was innocent, a federal judge threw out his conviction, ruling that microscopic hair analysis should not even be allowed in court, finding it too susceptible to error, subjective and "scientifically unreliable."
When I looked at the criminal trials of DNA exonerees, I found similar problems with serology analysis, bite mark analysis, fiber comparisons, and other traditional forensics. I even saw several more recent cases where analysts botched DNA tests. Analysts even concealed evidence of defendant's innocence. Had DNA tests not been done years later, these people might have spent the rest of their lives in prison. Still worse, some forensic analysts mistakenly ruled out people who were later shown by DNA to have been the actual culprits.
Time will tell whether a national solution comes out of Washington, with important federal forensics reform legislation pending when lawmakers return this month. In the two pieces to follow I will talk about more fall-out from Donald Gates' case, forensics scandals across the country at other crime labs -- and why this federal legislation to improve forensics is so urgently needed.
http://www.huffingtonpost.com/brandon-l-garrett/forensics-on-the-hill-par_b_1857330.html
Where is the path forward for forensics? Part II
Donald Eugene Gates, convicted of murder in Washington D.C., would spend 28 years in prison before he was exonerated in 2009 through the tireless work of his lawyers at the Public Defender Service for the District of Columbia - but it was not until this Spring that the flawed FBI forensics supposedly "matching" his hair to the crime scene would finally come under broader scrutiny.
The case brought still worse problems to light. This Spring, the Washington Post released the results of a remarkable investigation, uncovering a buried 1990's Department of Justice inquiry into FBI hair cases, and finding hundreds more cases with flawed forensics. Some of the cases were death row cases -- for example, a Texas man was executed in the late 1990s, but if it had not been for the FBI's flawed hair analysis, he would not have been even eligible for the death penalty.
This Spring, the Public Defender Service also cleared two more people who had flawed FBI hair testimony, using DNA tests. One, Kirk Odom had served 22 years in prison for a crime he did not commit and has been exonerated. The second man, Santae Tribble, spent 28 years in prison, and is still waiting for a ruling on an uncontested motion for exoneration.
After the Post reporting and two more exonerations, the FBI agreed to an unprecedented national case review of thousands of old FBI hair comparison cases, spearheaded by the Department of Justice. The audit has begun. Why did it take so long? Why didn't police or prosecutors or lab analysts ensure that this seriously flawed forensics be properly investigated? These were elite FBI lab analysts and elite federal prosecutors. Yet none of this should be a surprise.
There are multiple crime lab scandals all across the country at any given time, with entire labs closed down and ongoing audits of old cases. The latest news comes out of Massachusetts, where the Governor just shut down a crime lab after revelations after a chemist was violating protocols and possible engaging in "malfeasance," potentially affecting 34,000 criminal cases. Nassau County, New York, continues to grapple with the shutdown of its crime lab last year, and many thousands of old cases are being examined. Hundreds or even thousands of cases in San Francisco have been called into question due to misconduct in its lab. Washington. D.C. has also had problems with accurate use of its breathalyzer tests.
Those are cases that involve fairly routine drug testing. How about cases where forensic evidence was hidden from the defense? How about forensic techniques that have not been validated and like hair comparisons, that depend on the judgment-calls of the analyst? The FBI in 2005 stopped using bullet-lead analysis, which purported to trace bullets back to the particular manufacturing batch that they came from, finding it scientifically unsupported -- and sparking an audit to look into old cases. The FBI has revised its procedures for examining fingerprints, in response to the high-profile mistake in the Madrid bombings investigation, where a Portland Oregon lawyer had his prints falsely linked to that terrorist incident. Most forensic analysts, however, have not changed their practices -- and judges are reluctant to limit adventurous uses of forensics in their courtrooms.
Scientists have called for a path forward -- what is the path? Congress asked the National Academy of Sciences to lead a comprehensive scientific review of the state of forensics. Their 2009 report, ''Strengthening Forensic Science in the United States: A Path Forward," called for the creation of a National Institute of Forensic Science to set standards and fund research. That report concluded that the entire system of forensics needs to be overhauled. The underlying science used for many types of forensics needs more support. On the subject of hair comparisons like that used in Gates' case, the report said that there are no "scientifically accepted statistics" on microscopic hair characteristics and no scientific support for using hair to individualize. But the report was not limited to hair or fiber comparisons, which are not used so often today. The report said that for a host of forensics ranging from bite mark to tool mark to fingerprint comparisons -- in fact all forensic used to identify individuals aside from nuclear DNA testing -- basic scientific research is needed to find out how reliable the techniques really are.
Who is flying the plane? Who steps in if there is a scandal at a crime lab and bad work may have tainted old cases? On the research front, we have a National Institute of Health to oversee research for medicine -- why not an institute for forensics? We are willing to spend vast sums on forensics. Every year we continue to expand DNA databanks, for example. Yet very few criminal cases have testable DNA. Most criminal cases must rely on traditional forensics using the same sort of subjective comparison under a microscope that led to Gates' wrongful conviction. Scientific standards and oversight are urgently needed. That is what new federal legislation aims to do -- and that will be the subject of my third piece about forensics on the Hill. We could spend a fraction of what we spend on our massive system of DNA databanks to put the bulk of our forensics on firmer scientific ground.
So far, the scientific recommendations for a path forward for forensics have gone ignored in the halls of Congress. This Summer, Odom and Tribble, the two men who had just been cleared by the Public Defender Service, both stood before the U.S. Senate as Peter Neufeld of the Innocence Project testified in hearings to consider forensics reform legislation. They knew all too well that forensics errors are not isolated. When you start to look backwards at just one case, like Donald Eugene Gates' case, you find out that unreliable methods and poor standards were in place, you find even more wrongful convictions, and then you realize there is a real system problem with forensics. This fall, just maybe, Congress will take the path forward.
http://www.huffingtonpost.com/brandon-l-garrett/where-is-the-path-forward_b_1864341.html
Science, standards and forensics: Part III
Our entire system of forensics needs to be shored up -- and Congress may finally be responding to the crisis. This summer, Senator Jay Rockefeller introduced important legislation in the Senate, introduced in the House by Reps. Eddie Bernice Johnson, Donna Edwards and Daniel Lipinski, called the Forensic Science and Standards Act of 2012 (S. 3378 and H.R. 6106). This bill asks that the National Science Foundation (NSF) fund research and the National Institute for Standards and Technology (NIST) set out scientific standards for forensics based on the research.
The bill would allocate tens of millions of dollars in grant money for forensics research, setting out a "unified" federal research strategy and new Forensic Science Research Centers. We need research to find out what is the frequency of microscopic characteristics that forensic analysts rely on. We may find out that some techniques like fingerprint comparisons can be improved, and we may find out that some techniques, like hair comparisons, simply do not hold water. The Attorney General would have the responsibility to put these new scientific standards into practice, including by requiring federal labs to adopt them and encouraging their adoption by state and local crime labs. (Another bill, by Senator Patrick Leahy, the Criminal Justice and Forensic Science Reform Act of 2011, S. 132, has been kicking around even longer. It also calls for new grants for forensics, but it would center oversight in the Department of Justice.)
All of this help is desperately needed -- though it is not as effective or direct as what the National Academy of Sciences Committee proposed: creating a new, independent and scientific agency for forensics. Still, the Rockefeller bill will go a long way. It is remarkable that there still is no nationally standard terminology for forensic analysis, no scientific standards and a lack of basic research asking whether a host of forensic techniques are valid and accurate. There is also no standard process to conduct an inquiry when something terribly wrong comes to light, like in the FBI hair cases.
More work is being done just down the street from the hallowed halls of Congress. Washington D.C. is building a new city crime lab, and passed legislation to implement the National Academy of Science Report's suggestions at the local level, to make sure that its lab has scientific oversight and clear scientific standards for its work. Maybe that local lab can provide a model for the Feds.
Gates can never really be made whole for losing so much of his life based on false charges and faulty FBI forensics. His life has certainly improved now that he is free for the first time since the 1980s. The same is not true for forensics. Things are still not very different from the 1980s in the world of forensics. Despite the remarkable feats that DNA tests can sometimes make possible, and contrary to what you see on C.S.I.-type shows, the vast bulk of forensics work still usually hinges on the conclusions of largely unregulated lab analysts, often peering into old-fashioned microscopes, and often not required to carefully document their conclusions. Maybe that is good enough for government work. But if we are going to use forensics to put people in prison for years, Congress should pass legislation to make forensics far more of a science. Forensics may not be anything close to a big election year issue, but maybe that is a good thing. It is not a partisan issue, but an issue of science and of justice: forensics change is long overdue.
http://www.huffingtonpost.com/brandon-l-garrett/science-standards-and-forensics_b_1870255.html?utm_hp_ref=crime