When a defendant's DNA matches a sample found at a crime scene, how compelling is the match? To answer this question, DNA analysts typically use relative frequencies, random-match probabilities or likelihood ratios. They compute these quantities for the major racial or ethnic groups in the United States, supplying prosecutors with such mind-boggling figures as ``one in nine hundred and fifty sextillion African Americans, one in one hundred and thirty septillion Caucasians, and one in nine hundred and thirty sextillion Hispanics." In People v. Prince, a California Court of Appeals rejected this practice on the theory that only the perpetrator's race is relevant to the crime; hence, it is impermissible to introduce statistics about other races. This paper critiques this reasoning. Relying on the concept of likelihood, it presents a logical justification for referring to a range of races and identifies some problems with the one-race-only rule. The paper also notes some ways to express the probative value of a DNA match quantitatively without referring to variations in DNA profile frequencies among races or ethnic groups.
For two decades, the legal system has grappled with DNA evidence. The most difficult issues have not been technological, but statistical and logical. When a defendant's DNA matches that found at a crime scene, DNA analysts typically apply a population-genetics model to derive genotype frequencies from estimates of allele frequencies in the major racial or ethnic groups in the United States. But these estimates have been controversial. There have been objections to the quality and quantity of the data on DNA allele frequencies, the realism of the populationgenetics models, and the manner in which the genotype frequencies are presented to judges and jurors. 2 The case of People v. Prince 3 involves the problem of presentation rather than computation. In Prince, the California Court of Appeals stated that a genotype frequency estimate for a racial or ethnic group, no matter how well founded, is not relevant to determining whether the defendant is the source of the crime-scene DNA 290 D. H. Kaye unless there is "independent evidence" that the perpetrator is of the same race or ethnicity as the defendant. At the same time, the court held that sufficient evidence (other than the DNA match) that the defendant is the culprit can establish this preliminary fact and thus make the frequency in the defendant's racial or ethnic group relevant.
This chapter maintains that both these propositions are false. Relying on a likelihood-based theory of relevance, I first show that the population statistics for various “races” or ethnic groups ordinarily are relevant without regard to the defendant’s race or ethnicity. They are relevant because they indicate that it is improbable that an innocent defendant would have a DNA genotype that just happens to match that of the actual perpetrator.
Second, I explain that, if it were true that frequencies in major subpopulations were only conditionally relevant, this condition normally could not be satisfied by other evidence that the defendant is guilty. Instead, evidence that makes it more probable that some other member of his racial or ethnic group (as opposed to any other such group) is the culprit would be necessary. Evidence that only singles out the defendant but otherwise provides no information about the race or ethnicity of the perpetrator does not help satisfy the court’s condition.
After criticizing the reasoning in People v. Prince as to conditional relevance, I consider another objection to presenting the genotype frequency estimates for specific racial or ethnic groups. This is the possibility, emphasized in Prince, that a juror will jump to the conclusion that because the frequency in the defendant’s racial and ethnic group is mentioned, the perpetrator must belong to this subpopulation. I argue that this concern, while appropriate in some cases, does not justify excluding reasonable estimates of genotype frequencies by racial or ethnic groupings. 4
Patrick Paul Prince was charged with twelve counts of burglary, assault, and sexual crimes against five victims. In two of the attacks, “DNA matching his was found on a mask that each girl identified as having been worn by her attacker.” 5 Comparable evidence was not available in the other crimes, but the prosecution maintained that the modus operandi was so distinctive that all had to have been committed by the same individual. A criminalist found DNA on the mask and discovered that it matched Prince’s at nine STR loci. She testified about “a likelihood ratio that compared two different alternative possibilities, i.e., either the individual contributing the known reference sample contributed the evidence DNA and that is why the profiles matched; or the evidence DNA was contributed by some unknown, unrelated individual who happened to have the same DNA profile.” 6 Using data from the FBI on three samples of about 200 Caucasians, 200 Hispanics, and 200 African-Americans, she concluded that “for the Caucasian population, the evidence DNA 4 The Supreme Court of California reached the same conclusion and also rejected the view that subgroup frequencies are only conditionally relevant in People v. Wilson, 136 P.3d 864 (Cal. 2006). The state supreme court also granted review in Prince and then dismissed the grant of review “[i]n light of our decision in People v. Wilson . . . .” 142 P.3d 1183 (Cal. 2006). This resolves a split between the different divisions of the California Court of Appeals. In light of the opinion in Wilson and the granting of review in Prince itself, the Court of Appeals’ opinion in Prince no longer has precedential value in California. The case remains noteworthy, however, both because courts in other jurisdictions might find it attractive and because it is a stark example of how seemingly simple statistics can generate considerable confusion in the legal system.
5 36 Cal.Rptr.3d at 303.
6 Id. at 310.
profile was approximately 1.9 trillion times more likely to match appellant’s DNA profile if he was the
This content is AI-processed based on open access ArXiv data.
