Last week, Los Angeles Police Department (LAPD) officers arrested a suspect they believe is the elusive "Grim Sleeper" serial killer responsible for 11 murders across South Los Angeles, dating as far back as 1985. The arrest made national news, in part because the LAPD broke the case by using a controversial new technique known as "familial DNA" searching. Prosecutors are now using this success to argue that familial DNA should be used more widely.

There's no question that the arrest in the "Grim Sleeper" case was an enormous relief for both the victims' families and for the South Los Angeles community that has lived under the shadow of these serial murders. But whether we should expand familial searching isn't just about the success in this case: it's about whether familial DNA searching is really the silver bullet prosecutors suggest, and whether privacy and civil rights concerns have been adequately addressed. The answer to both questions, for the moment, is no.

With traditional DNA forensic analysis, police take a DNA sample recovered from the crime scene, analyze various portions of it, and try to find an exact match in a database of DNA taken, for the most part, from people convicted of felonies. There is enough variation in human DNA that an exact match provides a very high likelihood that the crime-scene sample came from the same person as the match in the database, which usually provides potent evidence of guilt (or innocence, if the samples do not match). Familial DNA searching, on the other hand, goes a step further and works off the principle that if the DNA is a partial but not exact match, then the sample may be taken from somebody who is related to the database match. So all the police have to do is look at all that person's relatives and see if one of them might be the perpetrator, right? Not so simple.

First, there are serious privacy concerns. Currently, there are inadequate regulations about the storage of DNA in databases and their potential uses. DNA is much more than just a record of where we've been dropping stray hairs. It's our genetic information. It contains information about diseases and medical conditions we have, as well as those we might one day get. Genetic information has been linked not just to physical traits like hair and eye color, but aggression, susceptibility to addiction, and sexual orientation. And while the portions of DNA that are analyzed for inclusion in the database probably don't code for any traits (though that's still being debated by scientists), most jurisdictions retain the original biological samples and so have access to the full genetic information of people in the database.

Second, expanding the use of DNA investigations to familial searches widens the net of innocent individuals who could fall under suspicion and be investigated by police. The allure of DNA databases used in traditional forensic analyses has been that they can identify people who are guilty of a crime. But familial searching identifies only people who police know are not guilty of the crime, and potentially encourages police to go knocking on doors asking about family members, their histories, and their likelihood to commit a crime. While proponents of traditional DNA investigations justify their use by arguing that the people included in them have been convicted of crimes (usually felonies) so their civil liberties should weigh differently against public safety, such arguments do not hold water with familial searching, where people who have never even had a speeding ticket can be reached through the database if one of their relatives has committed a crime. This is all the more problematic as the familial searches aren't very accurate. And they usually don't result in arrests — California has run familial searches of their DNA databases in eight other cases since 2008, and once before in the Grim Sleeper case. None of those nine searches ever led to an arrest.

Compounding this problem, because our criminal justice system disproportionately arrests and convicts people of color, those populations (particularly African-Americans) are vastly overrepresented in the DNA databases. Because familial searches extend the reach of databases beyond the offender to the offenders' family, the result is a genetic map of communities of color, and a law enforcement tool that is much more likely to solve crimes committed by black offenders than white offenders. That raises serious concerns of racial equity.

In the "Grim Sleeper" investigation, the LAPD took great care to consider privacy and civil rights in its investigation, but there are no assurances or guidelines to make sure that such practices are widespread. If the police are going to widen the use familial DNA searching, then the "Grim Sleeper" case is a good example of the kind of case that it should be used in, and the way it should be used.

California Attorney General Jerry Brown has issued guidelines limiting familial searching to the most serious, violent cases, only when all other investigative leads have been exhausted and there's an ongoing threat to public safety. And the authorities in the "Grim Sleeper" case didn't go knocking on doors of every partial match — they winnowed the list based on additional DNA tests, gathered circumstantial evidence that showed the suspect lived in the area where the killings occurred and was the right age to have committed the crimes over the course of more than two decades, and they obtained an identical DNA match before going public. But in California, as in other states, the standards that led to this kind of caution are voluntary — there's no statutory scheme governing the use of familial searches, and no oversight mechanism to ensure the rules are followed. We need both.

Finally, the success in the "Grim Sleeper" case doesn't justify expanding the reach of DNA databases — for example, to include people who have been merely arrested for a felony, but not convicted or even charged with a crime. The Grim Sleeper arrest was made using DNA taken from a person convicted of a felony, not a mere arrestee. Even familial DNA searching shouldn't have been necessary, because the Grim Sleeper suspect has at least two prior felony convictions. California and 14 other states currently take DNA samples from people merely arrested — not convicted — for a felony, making residents' civil liberties and genetic privacy turn on the discretion of one police officer as he or she writes his or her report. Fully one-third of Californians arrested for a felony are never convicted of a crime. That's not just bad policy, it raises serious constitutional concerns. The ACLU of Northern California is currently challenging the DNA profiling of arrestees under the Fourth Amendment.

(Originally posted on Daily Kos.)

Permalink |	Defending Targets of Discrimination, Racial Justice, Technology and Liberty