Goldstein says forensic science faces two core problems: "questionable validity and questionable reliability." The "questionable validity" speaks to fundamental concerns whether the scientific basis underlying many forensic practices is truly sound: E.g., whether tool marks, ballistic matches or tire track comparisons should be considered dispositive identification in criminal courtrooms? Goldstein's suggestions for a state-centered regulatory structure do not speak to "validity" questions, and IMO it probably is a federal role to fund the sort of research needed to move that ball forward in the way that science, for example, has improved eyewitness ID techniques and arson investigations over the last two decades.
On questions of "reliability," though, where the underlying science is valid, Goldstein rightly calls state regulation of crime labs "an ignored resource." States inherently have a role regulating forensic science because forensic evidence is most often used in state courts. So in each state, either the legislature does the regulating, it hands off regulation authority to some executive-branch entity, or else the courts must do it themselves on an ad hoc basis. But somebody at the state level must do it, and Goldstein points out that states are "well positioned to implement reforms ... that target reliability." He argues that:
The NAS Report charged the federal government with implementing its recommendations. But, as mentioned above, this ignores the direct control that states have over their laboratories and criminal justice systems. Furthermore, federal reform requires uniformity and ignores the benefits of state experimentation. It ignores geographic differences in values and the differences in the ways that states administer their systems of criminal justice and criminal investigation. Finally, with the current political climate in Washington, federal reform may be difficult to pass. And even if it were to pass, it may reflect a compromise between differing interests rather than the most robust oversight possible.Goldstein thinks mandatory accreditation is insufficient (as do I) in part because "accreditation only addresses issues of compliance with existing scientific practices. It does not address the validity of the underlying science, identify cases of technician negligence or fraud, remedy past injustices, or necessarily advocate for the best possible laboratory practices." For those issues, he said, states have turned either to independent investigations in high-profile cases, a tactic made more attractive by the availability of federal funds for the purpose in recent years, or else some type of oversight board or independent investigative body like Texas' Forensic Science Commission or the North Carolina's Innocence Commission. Each has pros and cons, but these are all examples of state-level experimentation through which we're learning important overall lessons about what crime lab regulation should look like, argues Goldstein.
Prof. Paul Gianneli has a response (pdf) to this piece in the same issue of the Texas Law Review pointing out the need for federal support on improving forensic "validity" while agreeing that "The states are indeed responsible for their crimes laboratories, and it is a responsibility that they have often neglected." Meanwhile. CrimProf blog points to another response (pdf) by Jessica Gabel and Ashley Champion arguing that centralized federal oversight would be more "practical." Gianelli, though, notes the federal and state approaches are not mutually exclusive: "The proposed federal legislation ... would not preempt state innovation; it would fund accreditation, not require it. This funding, however, may never materialize, and the states (as Mr. Goldstein notes) are independently obligated to regulate their crime labs."
On one, key point, I wholeheartedly, vociferously disagree with Mr. Goldstein when he writes that "the NAS Report‘s call for the removal of laboratories from law enforcement is infeasible and unlikely to receive political support." That's a political assessment, not a data-driven one, and in my experience it's deeply flawed. The political class generally gets why it's a conflict for police to control crime labs, they just don't have a great set of options available to them or, at the moment, extra resources. But politicians in Houston, e.g., which is the central example in Goldstein's essay, mostly pay lip service to the idea of removing the crime lab out from under law enforcement, though the devil is of course in the details. Still, I've neither seen nor heard of opposition in the political sphere so intense that it makes me think the goal of separating crime lab management from law enforcement is categorically impossible. Certainly, can't never could.
That said, in all, this student note was a worthy contribution, locating the cutting edge of a debate that's presently very much a moving target. This post portrayed just a fraction of what's in these three pieces, so those with more than a passing interest in such matters should read them for yourselves.