No Surrogate Forensics

By Brandon L. Garrett, originally posted on American Constitution Society's blog, June 24, 2011.

Yesterday, the U.S. Supreme Court in Bullcoming v. New Mexico ruled that it violates the Confrontation Clause of the Sixth Amendment to permit a forensic report to be introduced at trial by putting on the stand a crime lab analyst who did not do the actual work in the case. The Court, in a majority opinion authored by Justice Ginsburg, reinforced the Court’s 2009 decision in Melendez-Diaz v. Massachusetts, by ruling that putting on such “surrogate testimony” denies the defense an opportunity for cross-examination. Some labs have long put on the stand forensic analysts who were not involved in the particular work done in a person’s case. Perhaps the analyst who actually did the work had left the office. Perhaps that analyst was overworked.  Or perhaps one analyst in the lab was especially effective in front of a jury.

Donald Bullcoming was arrested for driving while intoxicated, after rear-ending another pick-up truck. The central evidence against him at trial was a lab report, a “Certificate of Analyst,” stating that his blood alcohol level was well above the legal limit. The prosecution did not call the analyst who actually tested Bullcoming’s blood sample – he had been placed on unpaid leave for some undisclosed reason. (This could raise a red flag – though perhaps the reason for the unpaid leave was entirely unremarkable.) The State instead called someone else from the lab who was familiar with their testing procedures.


When ruling that forensic reports are testimonial in nature, and that the defense has a right to cross-examine the right analyst, the Court noted that one reason why the person who did the work should be placed on the stand is that the work of a forensic analyst should be scrutinized, even if the task to a non-scientist seems fairly routine. A juror might read the certificate and assume that the blood testing equipment is automatic and foolproof. However, the Court highlighted the role that human error can play in forensics, even forensics heavily relying on machines. Serious errors can result from carelessly maintaining or using the gas chromatograph machines used to conduct blood alcohol analysis. The Court noted that in neighboring Colorado, a single analyst produced over two hundred erroneous results in a three-year period, by failing to properly use the internal standard in the machine. Other crime labs have had similar problems, the full scope of which will only be known if and when proper audits are conducted to find out just how many hundreds or thousands of cases may have been botched.

What the Court did not say, since the issue was not raised, was that those underlying problems of quality control require much more than Confrontation Clause protections at a trial. Indeed, it was Justice Kennedy, in his dissent in Bullcoming, who highlighted how reliability of forensics is better protected by structural reforms such as: “free retesting for defendants; result-blind issuance of reports; testing by an independent agency; routine processes performed en masse, which reduce opportunities for targeted bias; and labs operating pursuant to scientific and professional norms and over-sight.”  Even in the cases that do go to a trial, the defense cannot meaningfully question the analyst about the work they did in the lab, unless judges do far more to ensure that the defense has access to underlying bench notes describing the lab work (and not just the final lab report announcing the result).  Defense lawyers are routinely denied forensic experts of their own, so they have no way to meaningfully examine the work that the prosecution analyst did or offer a second opinion.

Viewing this ruling in the perspective of the array of problems facing forensics today, its importance quickly fades. Scandals in our forensic crime labs have been persistent. Indeed, several scandals, including some still unfolding, such as those in the District of Columbia and in Nassau County, NY, involve laboratories that failed to properly perform blood alcohol tests and may have botched countless tests resulting in convictions. A scandal in Washington State is also on point, including misconduct where a supervisor had signed off on work she had not herself done.

I have studied the trials of innocent people later cleared by DNA tests, and encountered a parade of invalid, unreliable, and erroneous forensic testimony in their cases. Forensic analysts used forensic techniques ranging from serology, hair comparison, bite mark comparison, fiber comparison, to modern DNA testing. In those trials, the defense rarely had an expert. In a few cases, cross-examination really was effective at uncovering flawed forensic analysis. But in half of the cases, the defense did not even ask questions challenging errors by forensic analysts.  When they did, the analysts sometimes exaggerated their conclusions even more egregiously.  And in those trials of the innocent, judges rarely questioned the unscientific claims by the analysts. Still worse, in a disturbing number of cases, forensic evidence of innocence was hidden from the defense. Even when the unscientific forensics were challenged on appeal, judges frequently found any error to be harmless. Indeed in Bullcoming, the Court noted that on remand, the court might find the error to be harmless.

Outside intervention is needed, since it appears as if our criminal justice system has an ingrained junk science habit that it cannot break. Melendez-Diaz was not the only crucial development in forensics in 2009. Earlier that year, the landmark National Academy of Sciences report, "Strengthening Forensic Science in the United States: A Path Forward," explained how the bulk of forensic techniques in wide use today, with the exception of nuclear DNA testing, cannot consistently and validly be used to link evidence with a particular defendant.  The report called for a nationwide overhaul of forensics due to endemic scientific, methodological, and quality control deficiencies. So far, Congress has not had the appetite to tackle structural problems with much needed federal legislation, although judges have begun to more carefully scrutinize forensics in their courtrooms and scientists have begun the work of reassessing methods and standards. Progress is being made, perhaps too slowly, but one thing is for sure: it will take much more than one criminal procedure ruling to solve those underlying structural problems.

Brandon L. Garrett is a professor at the University of Virginia School of Law.