To Stop Crime, Share Your Genes

This guy will go far, probably writing policy papers for the government.

From The New York Times:

PERHAPS the only thing more surprising than President Obama’s decision to give an interview for “America’s Most Wanted” last weekend was his apparent agreement with the program’s host, John Walsh, that there should be a national DNA database with profiles of every person arrested, whether convicted or not. Many Americans feel that this proposal flies in the face of our “innocent until proven guilty” ethos, and given that African-Americans are far more likely to be arrested than whites, critics refer to such genetic collection as creating “Jim Crow’s database.”

In truth, however, this is an issue where both sides are partly right. The president was correct in saying that we need a more robust DNA database, available to law enforcement in every state, to “continue to tighten the grip around folks who have perpetrated these crimes.” But critics have a point that genetic police work, like the sampling of arrestees, is fraught with bias. A better solution: to keep every American’s DNA profile on file.

Your sensitive genetic information would be safe. A DNA profile distills a person’s complex genomic information down to a set of 26 numerical values, each characterizing the length of a certain repeated sequence of “junk” DNA that differs from person to person. Although these genetic differences are biologically meaningless — they don’t correlate with any observable characteristics — tabulating the number of repeats creates a unique identifier, a DNA “fingerprint.”

The genetic privacy risk from such profiling is virtually nil, because these records include none of the health and biological data present in one’s genome as a whole. Aside from the ability in some cases to determine whether two individuals are closely related, DNA profiles have nothing sensitive to disclose.

But for law enforcement, the profiles are hugely important: DNA samples collected from crime scenes are compared against a standing database of profiles, and matches are investigated. Obviously, the more individuals profiled in the database, the more likely a crime-scene sample can be identified, hence the president’s enthusiasm to expand the nationwide repository.

The current federal law-enforcement database, the Combined DNA Index System, or Codis, was designed for profiles of convicted criminals. When it became operational in 1998, only certain classes of convicted criminals (for instance, sex offenders) were profiled. Over the past decade, the list of qualifying crimes has quietly grown (states make their own laws on collection). And last year, the F.B.I. joined more than a dozen states and moved to include DNA profiles from arrestees not yet convicted.

There are several key problems with this approach to expanding the database. First, the national DNA database is racially skewed, as blacks and Hispanics are far more likely than whites to be convicted of crimes. Creating profiles of arrestees only adds to that imbalance.

Second, several states, including California and Colorado, have embraced a controversial new technique called familial DNA search, which exploits the fact that close relatives share substantial fractions of their DNA. If efforts to find a DNA match come up empty — that is, if the perpetrator is not yet profiled in the database — the police in these states can search for partial matches between crime-scene samples and offenders in their record base. If they find a partial match, they can zero in on relatives of the profiled person as possible suspects.

This sounds elegant, and it occasionally works: in Britain, a handful of high-profile cases have been solved using familial search. But this approach is crippled by a very high false positive rate — many partial matches turn up people unrelated to the actual perpetrator. And it raises serious legal questions: how can we justify the de facto inclusion in DNA databases of criminals’ family members who have been neither arrested nor convicted?

Moreover, familial search threatens to skew racial bias further still: by effectively including all close relatives of profiled individuals, the database could approach universal population coverage for certain races or groups and not others. Even if this bias is found to be legally permissible, it may still prove politically unpalatable.

A much fairer system would be to store DNA profiles for each and every one of us. This would eliminate any racial bias, negate the need for the questionable technique of familial search, and of course be a far stronger tool for law enforcement than even an arrestee database.

This universal database is tenable from a privacy perspective because of the very limited information content of DNA profiles: whereas the genome itself poses a serious privacy risk, Codis-style profiles do not.

A universal record would be a strong deterrent to first-time offenders — after all, any DNA sample left behind would be a smoking gun for the police — and would enable the police to more quickly apprehend repeat criminals. It would also help prevent wrongful convictions.

As a practical matter, universal DNA collection is fairly easy: it could be done alongside blood tests on newborns, or through painless cheek swabs as a prerequisite to obtaining a driver’s license or Social Security card. Once a biological sample was obtained, its use must be limited to generating a DNA profile only, and afterward the sample would be destroyed. Access to the DNA database would remain limited to law enforcement officers investigating serious crimes.

Since every American would have a stake in keeping the data private and ensuring that only the limited content vital to law enforcement was recorded, there would be far less likelihood of government misuse than in the case of a more selective database.

Provided our privacy remains secure, there is no excuse not to use every bit of science we can in the fight against crime. The key is making sure that all Americans contribute their share.