Indiscriminate DNA testing could put innocent Oklahomans in prison

For several years, Rep. Lee Denney (R-Cushing) has proposed legislation to require taking DNA samples from everyone who has been arrested in Oklahoma (more recent attempts scaled it back to those arrested and held over for trial). These samples would be checked against a large database to see if the DNA shows up in crime scenes nationwide. 

The bill has never garnered enough votes to pass, though Rep. Denney is continuing to push — this year she’s holding an interim study to build support for the idea. Meanwhile, a recent Journal Record column by law professor Andrew Spiropoulos attempts to back her up.

Rep. Denney and Professor Spiropoulos argue that widespread DNA testing will help to catch criminals. As Rep. Denney told the Oklahoman, “I think there’s a lot of people out there that are repeat offenders. Right now, DNA is collected from people once they are convicted and incarcerated. If we could capture DNA earlier in the process, we could stop future crimes committed by people who are habitual offenders, and we could exonerate the innocent.”

Meanwhile, opponents of the idea cite concerns about the loss of civil liberties by allowing what amounts to a bodily search without a warrant.

Both sides miss an even greater problem. Contrary to popular beliefs about DNA evidence, running a DNA sample against a large enough database actually has a fairly high chance of incriminating an innocent person.

The misconception that DNA is a foolproof marker of identity comes from a basic misunderstanding about probability. It’s true that the likelihood of two samples of DNA from different people matching is very small. For a full DNA sample, the odds of matching can be one in many trillions. However, DNA recovered from crime scenes often don’t preserve all of the markers used for comparison. In some cases, the odds of a coincidental DNA match can grow to one in a million.

Those odds still sound unlikely, right? Yet we must also take into account another factor — if a one in a million match is run against a database with 500,000 DNA profiles, the chances of a coincidental match increase to one in two. And the FBI DNA database contains more than 500,000 forensic profiles.

The chance of a coincidental match rises even higher if the database is disproportionately filled with individuals who have a similar genetic heritage. That’s a major concern because of racial disparities in who gets arrested and charged. As reported in Jurist, “African-Americans are 13 percent of the population but make up about 40 percent of those whose DNA is included in the nationwide set of DNA databanks.”

If a person has already come under suspicion through traditional investigative techniques, DNA can be a useful supplement to build a case against them. But so-called “cold hits,” where someone is put under suspicion by a DNA match alone, can destroy an innocent person’s life. The problem is not just theoretical — an increasing number of convictions are relying on DNA evidence alone, and juries often have the same misguided faith in DNA evidence as the police. A Washington Monthly investigation shared a troubling example of a man convicted of murder based solely on a flimsy DNA match.

Another reason to be skeptical of DNA evidence is ever-present human error. California approved an initiative in 2004 requiring DNA samples be taken from anyone arrested on suspicion of a felony or serious misdemeanor. Forensics labs subsequently were swamped with a huge spike in the number of samples to be processed, and with the higher workload came more contamination of samples. The Los Angeles Times uncovered numerous instances in California and across the country of innocent people being charged and sometimes convicted before it was discovered their DNA sample was cross-contaminated in the lab.

Meanwhile, Oklahoma’s state medical examiner has been without accreditation for years due to chronic understaffing and a crumbling facility, and Oklahoma City police forensics have run into trouble in the not too distant past for falsifying evidence reports. Other states with more resources for forensics have been unable to cope with the workload of mass DNA testing. Can we expect Oklahoma forensics labs to do any better?