Historic Audit of Virginia Crime Lab Errors in Earl Washington Jr.'s Capital Case
An unprecedented external audit conducted by the American Society of Crime Laboratory Directors / Laboratory Accreditation Board (ASCLD/LAB) concluded that the Virginia State Crime Lab made serious errors in DNA testing in Earl Washington, Jr.'s capital case.
According to the audit, not only did the laboratory's leading DNA analyst incorrectly identify a phantom DNA profile (which prevented then-Governor Gilmore from pardoning Mr. Washington on grounds of actual innocence), but the lab's technical reviewer also missed the errors, as did the crime lab's own internal audit just last year. This final failure demonstrates that the crime labs cannot police themselves.
"This laboratory that touts itself as the best DNA laboratory in the country generated erroneous test results in a capital case, twice, using two different DNA methods," said Peter Neufeld, co-director of the Innocence Project. "The audit reveals not only that the laboratory's most senior DNA analyst, responsible for DNA testing in many of the state's capital cases, made serious errors, but that the laboratory's system to catch these errors completely failed. This audit provides compelling evidence that crime labs cannot police themselves, and that only with the statutory requirement that they be subject to independent, expert oversight can we have faith that appropriate controls are in place."
The external audit was ordered by Governor Warner in the fall of 2004 after the state lab's director, Paul Ferrara, repeatedly refused the Innocence Project's demand that he order one. The public demand for an audit followed the filing of a civil lawsuit on Earl Washington's behalf by a team of lawyers including the law firm of Cochran Neufeld & Scheck. That lawsuit led to independent testing of the evidence in 2004, which revealed that a convicted serial rapist and not a “phantom” had raped the victim before killing her. Governor Warner should be congratulated for insisting on this audit and, more importantly, for adopting the remedies recommended by ASCLD/LAB.
The ASCLD audit found:
- The conclusions of Jeffrey Ban (the senior Virginia Crime Lab DNA analyst responsible for DNA testing in many of Virginia's capital prosecutions), in which he identified a nonexistent male as the source of the semen in the rape/murder victim's body, were "incorrect" (p. 16, finding 2). In contrast, Dr. Edward T. Blake of Forensic Science Associates, retained by Washington's civil attorneys, achieved "clear and definitive results," (P. 16 top of page). That is, unlike Ban, all of Blake's conclusions are supported by the data.
- The decision in 1993 to not grant Mr. Washington an absolute pardon was based on the erroneous reporting of a DNA allele which the auditors say should not have been included in Ban's 1993 report. Had it not been included then it would have been absolutely clear that there was only one semen donor, and that semen donor was not Earl Washington, (P. 19 Sec. 1B). Earl Washington spent an additional 7 years behind bars as a result of this mistake.
- One possible cause for the false results reported by Ban and incorrectly affirmed by the technical reviewer was "pressures from outside the laboratory and excessive managerial influence from within the laboratory," (P. 17, finding 6). The auditors quote Ferrara and Ban as referring to this as "not a normal case," (p. 16). This finding demonstrates that in "high profile" cases, in the Virginia crime labs, the risk of obtaining erroneous results is higher than in "normal cases." Since Ban was handling the most serious cases in the lab, his work in those cases is also thrown into doubt. Hence the recommendation for a more extensive audit of Ban's casework. As Ban remarked about this case, "inconclusive results were not an option," (P. 14). A rush to judgment by crime lab analysts has been responsible for many miscarriages of justice. If a lab scientist is told his experiment "must" produce dispositive results and he succumbs to that pressure, then he is no longer a scientist, and his work product can not fairly be called, "science."
- Virginia's peer-review system failed because the "technical reviewer did not observe the errors in the processes and the reported results," (P. 17, finding 7). The entire purpose of having a second examiner review data and reports at the time of testing is so that when errors occur, they can be caught and corrected in a timely fashion. Hence, Virginia's peer review system was not functioning properly and should be re-examined.
- The ASCLD/LAB external audit provides additional proof that crime labs cannot be allowed to police themselves when there are allegations of misconduct or serious negligence affecting the integrity of the results. In all other institutions when health, safety, or security are at stake, when something goes wrong we require an external audit. And for good reason. For instance, when the FBI fingerprint unit misidentified an Oregon lawyer as the source of the fingerprints in the Madrid train bombing, the internal findings of the FBI were deemed inadequate. Instead, the Department of Justice's Office of Inspector General has launched its own independent audit. Here, the ASCLD auditors also concluded that the Virginia Crime Lab's own internal audit of 2004 reached an erroneous conclusion when it defended the "correctness" of Ban's September 2000 analysis, (P. 13, pt. 2). When Cochran Neufeld & Scheck and the Innocence Project first presented Blake's 2004 report to lab director Paul Ferrara, suggesting that Ban may have been mistaken and that under all the circumstances, Ferrara should bring in "outside experts" to conduct an external audit, instead of welcoming the external review, he responded defensively.
1. Chicago Tribune, Oct. 21, 2004, "I'm not going to admit error when there is none." "As far as we're concerned, there is no error at all except in the minds of [critics]" and "When you are on the top of the heap, you are going to have someone trying to knock you down."
2. Virginian Pilot, June 20, 2004, "They can say what they want, we are standing by those results." "We don't need anybody to pass judgment on our results. You're talking about one of the best DNA labs in the country."
3. Washington Post, April 6, 2004. Ferrara, "said he is fully confident in the accuracy of the work performed."
The Auditors call for extensive remedial action including expansive reviews of Ban's casework and that of other DNA analysts in the Virginia Crime Lab.
1. Jeff Ban, the lab's leading DNA analyst, is to be immediately suspended from all cases involving low level DNA samples and/or mounted slides. The lab’s suggestion that Ban be pulled only from “low level” DNA samples is shortsighted since an examiner does not know for sure whether there is a low level until he extracts the DNA, in which case it may be too late to transfer it to another analyst.
2. Ban is to be suspended as a Technical Leader for the labs until corrective actions are completed. This is particularly significant because Ban sits on SWGDAM, the governing body set up by the FBI to establish standards and controls for all forensic DNA testing in America. We urge SWGDAM to remove Ban as well.
3. The ASCLD report calls for expanding the current audit, which was limited to one case (Earl Washington's), to all Ban's cases "from in and around 2000 and forward, particularly in cases in which there were low level DNA and/or mounted slides to determine if the conclusions are scientifically supported by the data." Since Ban as the most senior DNA analyst had responsibility for many, if not most of the capital cases processed by the DNA unit over the last five years, serious questions emerge concerning the integrity of those capital prosecutions.
4. To determine whether the deficiencies were limited to Ban, or were instead "endemic" to the laboratory, the report called for a "thorough examination of a minimum of 50 cases in the Virginia system," (p. 17, rec. 6). Since the auditors already determined that the "technical reviewer" back in September 2000 failed to catch Ban's errors, and since the 2004 internal auditors even with additional new information failed to admit the errors, the likelihood of laboratory-wide errors is likely.