DNA evidence and why we trust the government
The LA Times released a story two days ago which raises new questions about the reliability of DNA evidence in court. Well, what strikes me about this story is not the reliability of DNA evidence so much as the ethics of the FBI.
Government experts routinely explain to juries how certain a DNA match is, although I hear different numbers each time. Pick a random number that is too big for the brain to comprehend; one in that number is the odds of another DNA match occurring. I in 113 billion. 1 in 100 billion. 1 in 108 trillion. 1 in 1 quadrillion. FBI experts have testified that certain DNA profiles are unique, to a reasonable degree of certainty.
The problem is, FBI analysts have never tested 100 billion, trillion, or quadrillion DNA samples and looked for matches. They are giving their best guess, and undoubtedly the need to obtain convictions transforms a "best guess" into a "reasonable degree of certainty."
In the 1990s, FBI scientists estimated the rarity of each genetic marker by extrapolating from sample populations of a few hundred people from various ethnic or racial groups. The estimates for each marker are multiplied across all 13 loci to come up with a rarity estimate for the entire profile.
However, the growing number of samples contained in state DNA databases is making real research into DNA matches possible. An Arizona crime lab analyst began finding dozens of DNA matches in the state database. Court ordered searches in two states found 1000 pairs that matched (9 of 13 points matching). A court ordered search in an Arizona case found 122 matches in 65,000 samples tested.
The odds that the new searches indicate are still substantial, and likely are still not only admissible but persuasive in court. The ever growing DNA databases offer new unique opportunities for research that may benefit not only the court systems but the sciences as well. And yet, rather than open up the DNA databases for testing, or even conduct their own testing, the FBI went on a campaign to prevent any such tests.
They threatened state labs who complied with court orders to conduct the tests. They schemed to figure out ways to trick judges into not granting defense counsels' requests for testing. They told judges that they would cut off state labs from the national databases if the courts ordered the tests, and consulted experts who would explain to judges that the tests were not scientific and not necessary. They argued that the tests would violate the privacy rights of convicted felons and that they would result in the systems being overloaded and shut down. They argued that, under federal law, the databases were exclusively for the use of law enforcement agencies and not defense lawyers.
Why not run the tests, get accurate probabilities based on real numbers, and tell the truth to juries? The probabilities are surely going to be impressive even if they are not as ridiculously large as the FBI has been claiming. I doubt that it will result in more acquittals. To me this story was not about the revelation of finding DNA matches so much as the FBI's attempt to hide the truth and preserve their inflated estimates. I hope that this national discussion sparked by the Times' story leads to more research and testing in the area of DNA analysis, free from the prosecutorial bias and scientific repression of the FBI.
