Thursday, June 19, 2008
Trawl Fishing?: Supreme Court of California Affirms Use of Product Rule To Match Defendant's DNA In Cold Case
Earlier, this week, the Supreme Court of California affirmed the first degree felony murder conviction of Dennis Louis Nelson in People v. Nelson, 2008 WL 2404949 (Cal. 2008), in the process affirming the use of the "product rule" in Nelson's "cold hit" case. But was the ruling correct?
In Nelson, on February 25, 1976, 19 year-old college student Ollie George was found dead, raped and drowned in mud in Sacramento County, California. George had disappeared two days earlier, shortly after witnesses saw her inside a car at a shopping center. A couple of weeks after George's death, one of these witnesses saw what he believed to be that same car and contacted police, who traced the car to Nelson and questioned him. The witness' tip, however, was merely one of hundreds of tips and did not lead to any charges being brought, making the George murder a cold case.
It remained that way until October 2000, when California allocated funds to enable local law enforcement agencies to utilize DNA to solve sexual assault cases that lacked suspects. In July 2001, a review of Ollie George's death determined that the case had biological evidence that warranted analysis. An analyst conducted a vaginal swab on George to develop a DNA profile and then compared it with the state's convicted offender databank, resulted in a match with Nelson (Cases like Nelson's are sometimes called "trawl cases" because the match was discovered by searching a database of previously obtained DNA samples).
A criminalist then applied the "product rule," which is a statistical method to calculate the rarity of a given DNA sample in the relevant population, expressed as the probability of a single random person possessing the same DNA profile as the person whose DNA was found at the crime scene. Comparing 15 loci, the criminalist determined, inter alia, that the DNA profile on the vaginal swab would occur at random among unrelated individuals in about one in 950 sextillion African-Americans (Nelson is African-American, and there are 21 zeros in a sextillion). The prosecution admitted this evidence over Nelson's objection, and he appealed this ruling and others on his appeal, which eventually reached the California Supremes.
Now, first, let's look at what the appeal did not concern. Nelson did not claim that there was any problem with the "product rule" as applied to a single suspect. As the court noted, Nelson agreed with the state "that using the product rule to calculate the random match probability makes sense when comparing one suspect's profile with the crime scene evidence because, as he explains, the random match probability 'estimates the chance that any single, random person drawn from the relevant population would have the same DNA profile as that of the unknown person whose DNA was found at the crime scene.'" Indeed, such a challenge would be difficult because, "[a]t present, courts across the nation have accepted the results of DNA typing into evidence and ruled the product rule, and the frequency tables underlying it, an acceptable way to represent the significance of a match." Erin Murphy, The New Forensics: Criminal Justice, False Certainty, and the Second Generation of Scientific Evidence, 95 Cal. L. Rev. 721, 781 (2007).
Instead, Nelson's argument was that while the product rule makes sense in a single suspect case, there is a problem with applying the product method to a trawl/cold case because, when a suspect is found by a search of a large DNA database, the chance of a coincidental match is increased because "a single genetic profile (from the crime scene evidence) is compared to the very large number of profiles in these databases." He therefore claimed that application of the product rule to a trawl/cold case was a new scientific technique, meaning that it had to pass California's Kelly test, which requires that the technique must have gained general acceptance in the field to which it belongs. The Supreme Court of California, however, rejected this argument, finding that "[t]he Court of Appeal in this case and other courts that have considered this question have concluded that use of the product rule in a cold hit case is not the application of a new scientific technique;" it also rejected Nelson's argument that evidence obtained from application of the product rule to a trawl/cold case was irrelevant.
Reviewing these cases and Erin Murphy's aformentioned article, it's tough to argue with the court's opinion, but the Murphy article reveals that it's possible that the times they are a-changin'. First, with regard to the product rule in general, Murphy notes that "recent evidence calls into question the accuracy of using the product rule to convey match probabilities." She specifically cites to, inter alia, some eye opening research conducted by an alert analyst and emerging independent research indicating that an assumption underpinning the product rule - that of independence at the various loci - may not hold true for all populations.
And what about Nelson's argument about the difference between single suspect and trawl/cold cases? Well, according to Murphy, there is considerable disagreement on the issue. Some suggest that the results of a “trawl” are more reliable than in a simple confirmation case, because the analyst has compared the genetic profile to a database and excluded a large number of persons whereas some contend that the likelihood of a "false positive" increases as the analyst looks in a database for a match, and thus the statistical probability should be accordingly discounted by this risk. All of which leads me to concur with Murphy's conclusion that "[r]ather than render admission of a methodology a one-time question that, once answered, is rarely asked again, the law should affirmatively require the government to provide evidence verifying the technique's continued viability."