United States v. Jenkins

Citations: 887 A.2d 1013; 2005 D.C. App. LEXIS 647; 2005 WL 3434075Docket: 05-CO-333

Court: District of Columbia Court of Appeals; December 15, 2005; District Of Columbia; State Supreme Court

SummaryOriginalCytator Case TreatmentBackward CitatorAnalysis

EnglishEspañolSimplified EnglishEspañol Fácil

The United States government appeals a pre-trial ruling that barred the introduction of DNA evidence against Raymond A. Jenkins, asserting that the trial court erred in determining that the statistical methodology for DNA match significance, particularly from a 'cold hit,' lacked general acceptance in the scientific community under the Frye standard. The trial court's decision, which also excluded the DNA match evidence itself due to the absence of statistical significance, prevented the jury from knowing that Jenkins' DNA matched that found at the crime scene.

The court acknowledged that while statistical evidence enhances the probative value of DNA matches, it disagreed with the trial court's characterization of the scientific consensus regarding the statistical methodology, asserting that the methodology and its results are widely accepted in the relevant scientific community. Consequently, the appellate court reversed the trial court's decision.

The case background involves the investigation of Dennis Dolinger's murder, where his body was found with multiple stab wounds and signs indicating the assailant was also injured. Blood evidence was traced throughout the crime scene, and items belonging to Dolinger were later linked to Stephen Watson, who was arrested for felony murder. The FBI analyzed DNA samples from the scene, determining that while Dolinger's DNA matched many samples, none matched Watson or other known individuals.

The FBI DNA laboratory identified an unknown individual as the sole contributor of blood in multiple evidence samples. On November 16, 1999, the government requested the Virginia Department of Criminal Justice Services (DCJS) to match this profile against Virginia's DNA database of 101,905 offenders. DCJS found a match with Robert P. Garrett, an alias of Raymond Anthony Jenkins, leading to a focused investigation on him. A search warrant for Jenkins' blood was executed, and a thirteen loci DNA profile was generated, matching the evidence from the crime scene.

In March 2001, Jenkins filed a motion in limine to exclude the DNA evidence, arguing that the FBI's method of presenting rarity statistics for DNA matches in "cold hit" cases was not generally accepted in the scientific community, thus inadmissible under the Frye standard. Jenkins cited a significant debate regarding the appropriate calculation methods for cold hit significance. An evidentiary hearing followed, where both sides presented expert testimony and scholarly articles. The FBI's approach, based on the product rule, involves calculating the frequencies of genetic variations at thirteen loci, which are then multiplied to determine the profile's rarity in the population.

The government asserted that in non-cold hit cases, the derived number indicates both the rarity of the DNA profile and the probability of a random match. However, they conceded that in cold hit scenarios, this number does not accurately reflect the probability of a chance match.

Many profiles being searched enhances the likelihood of finding a match, with 'database match probability' providing a more precise estimate for cold hit matches. This method, established in a 1996 National Research Council report, aims to mitigate 'ascertainment bias'—the bias that arises when searching for rare items in a fixed database. The database match probability is determined by multiplying random match probability by the database size, and it can be calculated without a match occurring. The government argues that while a database search alters the match probability, it does not affect the rarity of a specific profile in the general population. It asserts that rarity statistics are more significant than database match probabilities when assessing thirteen individual loci. For Mr. Jenkins' profile, the rarity statistics are extremely low across various populations: 1 in 26 quintillion (African-American), 1 in 870 quintillion (Caucasian), 1 in 1 sextillion (Southeastern Hispanic), and 1 in 4 sextillion (Southwestern Hispanic). 

Mr. Jenkins claims there is ongoing debate about the most appropriate method for expressing cold hit probabilities, citing four perspectives: the 1992 National Research Council recommendation (confirmatory loci approach), the 1996 recommendation, and the Balding-Donnelly approach. The 1992 approach mitigates ascertainment bias by retesting with different loci, while the 1996 report suggests using database match probability to address bias but lacks clarity on whether it should be presented alone or alongside rarity statistics. The FBI’s DNA Advisory Board interprets the 1996 recommendation as advocating for both statistics. The Balding-Donnelly approach, however, emphasizes the significance of a match by considering the elimination of profiles during searches, positing that larger databases increase match significance.

Obtaining a match in a DNA database search eliminates other profiles as potential sources, thereby increasing the likelihood that the identified individual is the actual source of the DNA. In the model proposed by Balding and Donnelly, the probability that a person identified is the source of the DNA is slightly higher than the random match probability. Mr. Jenkins did not argue that the government's statistical methods were flawed; instead, he relied on the existence of a scientific debate to justify the exclusion of evidence under the Frye standard. The trial court concurred, noting that the debate regarding the calculation of statistical significance for cold hits necessitated the exclusion of both the probabilities and rarities associated with DNA evidence. The court offered the government a chance to present a more conservative probability formulation from the National Research Council's 1992 report, which the government rejected, prompting an appeal. The government contended that there was no controversy warranting exclusion under Frye and sought to introduce the rarity statistic, without objection to the database match probability. Frye v. United States serves as the foundational case for assessing the admissibility of scientific evidence, and the trial court typically exercises discretion in admitting expert testimony, especially concerning new scientific techniques, which undergo de novo review. Expert testimony is generally admissible when the subject matter is beyond common knowledge, and while courts favor admitting testimony based on established scientific principles, they also require that such principles have gained general acceptance in the relevant scientific community. The proponent of a new scientific method must demonstrate its acceptance by a preponderance of the evidence.

The excerpt emphasizes the distinction between consensus and controversy in the context of scientific methodologies used in legal proceedings, particularly under the Frye standard. It asserts that the Frye test requires acceptance of a scientific methodology by the relevant scientific community, not merely a consensus among courts. If a significant number of scientists oppose a technique as unreliable, it fails to meet Frye's standards. The excerpt critiques a trial court's failure to focus on methodology, highlighting that disputes within the scientific community preclude the admission of evidence. Specifically, it discusses the accuracy of various statistical approaches related to cold hit DNA matches, noting that these methodologies do not challenge the underlying mathematics but rather differ in their relevance to specific questions. The excerpt concludes that, since there is no controversy regarding the validity of the calculations, the trial court improperly excluded the DNA evidence, citing Porter as precedent which reinforced the need for a general consensus on methodology rather than results.

The controversy centers on the validity of assumptions used in calculating DNA match statistics, particularly regarding mating patterns within racial groups and the independence of DNA fragments. Critics of the FBI's rarity calculations argue that broad database usage can yield significantly inaccurate rarity figures for specific subgroups. While historical debates focused on these foundational assumptions, the current issue pertains to the relevance of presented statistics rather than their underlying calculations. Proponents of the government's view advocate for presenting rarity statistics independently, while others argue that such statistics alone do not adequately represent the significance of matches from database searches. A third perspective emphasizes assessing the probability of a match being correct, given known exclusions. Despite differing views on relevance, all parties acknowledge the accuracy of each other's calculations. There is ongoing debate about the appropriateness of using rarity statistics, database match probability, or Balding-Donnelly calculations in the context of cold hit DNA matches. Testimonies confirm no new mathematical methods are being proposed, with experts recognizing both the rarity of the DNA profile and the probability of a match as relevant factors. Dr. Krane, an expert for Mr. Jenkins, conceded that juries might want to know both rarity and match probability, and affirmed that the science behind calculating rarity remains uncontentious, unaffected by initial database searches.

Dr. Krane is capable of calculating rarity and database match probability, and can explain their distinctions to a jury. He asserts that in cold-hit cases, the database match probability is the primary concern, considering it more relevant than rarity statistics. Each of the rarity statistic, database match probability, and Balding-Donnelly formulation addresses different questions related to cold hits. The rarity statistic evaluates the uniqueness of a genetic combination, while the database match probability assesses the likelihood of a match within a specific database, and the Balding-Donnelly calculation estimates the probability that a person is the source of a DNA sample after eliminating others.

The determination of which statistic is most relevant does not fall under the strict standards established by Frye, which governs the admissibility of scientific evidence. Instead, relevancy is at the discretion of the trial court, which must decide whether evidence makes a fact more or less probable without being substantially outweighed by unfair prejudice. In the context of DNA evidence, the most pertinent aspect for factfinders is whether the DNA from a crime scene matches that of a suspect; thus, rarity statistics remain highly probative.

The appellate record indicates no significant debate in the scientific community regarding the methodologies of the three calculations, allowing for their introduction as evidence. The appellate court reverses and remands the trial court’s order, emphasizing that the trial judge has the responsibility to determine relevance, which should not be subjected to the Frye standard. This ruling is limited to the admissibility of evidence related to the statistics in question, without addressing the broader investigation details, as these were not presented in the lower court.

Factual recitations from both the government’s and Mr. Jenkins' briefs are uncontested for this appeal. Loci refer to specific locations on DNA that hold genetic material, with the FBI consistently testing thirteen loci for the Combined DNA Index System (CODIS). The felony murder charge against Watson was dismissed, and at the time of the database search, only eight of the thirteen CODIS loci were utilized in the Virginia DNA database. The DNA Advisory Board, established under the DNA Identification Act of 1994, is responsible for creating quality assurance standards for DNA testing, and no claims of bias against this board have been made. 

The product rule is explained as a method for calculating the probability of two independent events occurring together, though it no longer accurately reflects random match probability due to changes in database searches. Random match probability and rarity, while numerically identical, represent different concepts; an increase in database size influences the likelihood of finding a match. 

Statistics illustrating the rarity of DNA profiles across different racial groups are provided, with probabilities expressed in quintillions and sextillions to highlight their rarity within the respective populations. The government’s witnesses confirmed that U.S. DNA laboratories typically report both rarity statistics and database match probabilities, although Dr. Chakraborty emphasized that presenting rarity alone does not convey the full context, particularly in cold hit cases. The government expressed willingness to provide both statistics if requested by the trial court but hesitated to disclose the database match probability due to potential prejudice regarding Mr. Jenkins' DNA profile being in the Virginia offender database.

U.S. DNA laboratories generally adhere to the DNA Advisory Board's methodology. In the case of Porter, it was established that in the absence of the government's preferred statistical evidence regarding DNA match significance, the government may present the most conservative probability estimate, as it remains relevant and scientifically supported. The court emphasized that even if the reliability of the probability figures is disputed, juries should be informed of the existence of a DNA match and the likelihood that the true perpetrator is not the defendant, even if the exact probability is uncertain. For instance, in scenarios where a consensus might agree on a 1 in 1,000,000 probability, but skepticism leads to a conservative estimate of 1 in 100,000, the jury should at least be aware of the latter figure.

The 1992 recommendation from the National Research Council to use confirmatory loci yields the most conservative estimate but is deemed outdated due to advancements in DNA profiling technology and changes in scientific consensus. Currently, DNA labs typically utilize a database match probability calculation to address ascertainment bias, a methodology that has replaced the confirmatory loci approach. The evidentiary strength of a cold hit from a database of 100,000 individuals further supports the reliability of DNA match probabilities, as it allows for the exclusion of 99,999 non-sources. The case law indicates that both rarity and database match probability are credible and significant statistics for judicial consideration.

Case law regarding the Balding-Donnelly calculation is underdeveloped, with the 1996 National Research Council report indicating that its use for explaining DNA match significance in criminal cases is infrequent. There is no opinion expressed on the probative value of the Balding-Donnelly statistic. During oral arguments, concerns were raised about potential jury prejudice from presenting the database match probability, suggesting it implies Mr. Jenkins may be a prior offender. This potential prejudice is comparable to what would arise from statistics based on the 1992 National Research Council recommendations for testing confirmatory loci, which would necessitate disclosing that Mr. Jenkins was identified via an offender database search. Despite this, numerous trial courts, including the one in this case, have accepted statistics from untainted loci testing, indicating that revealing the identification method has not been viewed as significantly prejudicial enough to outweigh the evidence's probative value.