New York Federal Criminal Practice Blog
October 30, 2007

Second Circuit Affirms Admission of Ballistics Expert Without a Daubert Hearing

The crime scene described at the beginning of United States v. Williams, 05-6036-cr (2d Cir. October 23, 2007), is just the kind in this age of CSI to have juries agog for forensic evidence to pinpoint the perpetrators: "a gelid night," "the ringing sound of gunshots," two fleeing mystery shooters, a Ford Bronco filled with "bullet-riddled bodies."  But as a recent article points out, Devil in A White Coat: The Temptation of Forensic Science in the Age of CSI, 41 New Eng. L. Rev. 503 (2007), forensic evidence is not scientific truth.  Forensic "experts" are usually technicians with little more than a college degree and often less.  They are trained in police laboratories that frequently lack any accreditation or peer review process, do not engage in blind testing, have shockingly high error rates (if such rates are tested at all), and fail to follow the basic principle of scientific inquiry: subjecting one's findings to repeated attempts at refutation.  In fact, as the article notes, quoting The Coming Paradigm Shift in Forensic Identification Science, 309 Science 892 (2005), "[f]alse or misleading testimony by forensic investigators is 'the second most common contributing factor to wrongful convictions, found in 63% of the cases.'"

But forensic practices based on subjective visual and impressionistic comparisons - such as ballistics, handwriting analysis, voice identification and hair comparison - are routinely admitted in criminal cases, and their routine admission becomes itself a self-justifying rationale.

In Williams, the district court permitted the government to present the testimony of a ballistics expert.  It also denied the defense request for a hearing under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), to challenge the reliability of that testimony.  Relying on a 2002 SDNY decision, among others, the district court held that ballistics is a proper subject of expert testimony, and the defense had failed to offer a reason "for us to depart from the reasoning of these cases." 

On appeal, the Second Circuit affirmed the ruling.  It found that the district court had adequately discharged its "gatekeeping" function under Daubert by relying both on the 2002 SDNY decision, which admitted a government ballistics expert based on the government's written submission, and the government's "exhaustive foundation for [the expert's] expertise" before presenting the expert's testimony to the jury.  The Court held that Daubert did not require the district court to hold a separate hearing to ascertain the reliability of the witness's methodology.

But lest the Court be deemed to have ratified echo-chamber jurisprudence, there is an important caveat in the decision that practitioners should be sure to quote in their Daubert motions: "We do not wish this opinion to be taken as saying that any proffered ballistic expert should be routinely admitted . . . [Daubert's] shift to a more permissive approach to expert testimony did not abrogate the district court's gatekeeping function.  Nor did it 'grandfather' or protect from Daubert scrutiny evidence that had previously been admitted [prior to the adoption of Rule 702].  Thus, expert testimony long assumed reliable before Rule 702 must nonetheless be subject to the careful examination that Daubert . . . require[s]."  And by implication, of course, expert testimony long assumed reliable post Rule 702 and Daubert, should similarly be subject to careful re-examination in light of emerging technology and expertise.

At the very least, given how significant forensic evidence in criminal cases has become, it is really time for forensic laboratories to institute a routine policy of blind testing (i.e. presenting the technician with a mixture of crime-related evidence and dummy evidence, without designating the items the investigating officer wants to match), similar to a line-up for identification purposes.  This will address one of the most frequently cited problems with forensic labs -- their lack of independence from their clients and associated "tunnel vision," where the analysis is skewed in favor of the match the client is seeking.

See Archives for all posts since September 2007.