The collection of DNA has mushroomed in the past five years. California has the third-largest forensic DNA database of any government entity in the world (behind Britain and the US government). All three collect DNA from arrestees regardless of guilt. All three have databases highly skewed by race and class. In Britain 42 percent of black men have had their DNA sampled and stored. Until recently, however, forensic DNA samples were retained only from convicted felons who had committed violent or sexual offenses. With very little oversight or consistency, local rules for collection in criminal cases have expanded haphazardly, often including anyone who is arrested even mistakenly. People may challenge the retention of their DNA and sue to have it expunged, but that process can take years.
On March 6 of this year President Obama appeared on America’s Most Wanted to announce support for taking DNA from all arrestees, for banking that data and for keeping it without restriction even if there is no subsequent conviction. Undoubtedly it would have been much more controversial if he’d advocated DNA collection for all citizens, but his proposal seems to have slipped under the radar when framed as "only" for arrestees. By this metric "innocence" becomes a matter of degree, with certain people less innocent than others, marked by suspicion, kept on file, looked at first, disproportionately subject to the vicissitudes of laboratory and storage error. And even if mistakes are rare, it is exceedingly hard to persuade anyone of that if you are tossed in with a category the public deems inherently suspicious.
Obama and John Walsh, the host of America’s Most Wanted, endorsed the patently false idea that DNA is no different from fingerprinting. DNA samples can potentially reveal data about all kinds of other aspects of the self, including probabilities for certain health conditions, and can potentially implicate close family members. Employers and insurance companies are already hungering for DNA analysis that may allow them to refuse jobs or coverage to those with the mere hint of genetic predisposition to certain medical conditions (never mind pre-existing conditions).
In addition to the invasion of privacy, we must also weigh the possibility of technical mistakes. As a report from the National Research Council observes, "There is no uniformity in the certification of forensic practitioners, or in the accreditation of crime laboratories." That measure of human error is never conveyed in narratives touting the perfect infallibility of DNA evidence.
Furthermore, while DNA has been vaunted as a revolutionary tool of exoneration–in ruling out suspects, in other words–it is much less precise as a mechanism of positive identification when dealing with "cold hits," that is, attempting to match DNA fragments from a crime scene against samples stored in massive databases. Bioethicist Erin Murphy clarifies the point by comparing DNA typing to blood typing: "Imagine a murder scene at which police find a blood sample certain to belong to the killer. Crime scene technicians…show that it is type O. Later, the police find and draw blood from two suspects. One suspect is type AB; the other is type O. We can, with unreserved confidence, say that the first suspect is not the killer; but regarding the second suspect, we can only say that she is included within the class of people that includes the killer. The probability that she is the actual killer turns on how many other people have that blood type, along with any other evidence that we might be able to adduce."