The institute’s researchers found that in 5 percent of homicide and sexual assault cases, DNA testing ruled out the convicted person. If the scope is narrowed to just the sexual assault convictions, DNA testing eliminated between 8 percent and 15 percent of convicted offenders. The wrongful conviction rate previously had been estimated at 3 percent or less.
Although all of these tests were done on Virginia cases, a lead researcher said the results likely could be applied elsewhere.
“I believe that there’s nothing about the Virginia situation that is much different from what was going on across much of the United States at that time,” said John Roman of the Urban Institute. “I think that states have a responsibility to take these findings seriously in other places and investigate other cases that they have where they have retained evidence, because chances are they’re going to find far more wrongfully convicted people than they would have anticipated before this study.”
Researchers analyzed the results of new testing of DNA samples archived from 635 murder, sexual assault and non-negligent manslaughter cases that led to convictions. The cases stemmed from 715 offenses in Virginia between 1973 and 1987.
Virginia was able to do the testing because a state serologist and those she had trained had retained cotton swabs and clothing swatches. The samples mainly contained semen and blood samples from the cases during an era when DNA analysis wasn’t widely used as an investigative tool. After two men were exonerated following the discovery of the old evidence, the state in 2005 ordered each of the samples tested.
The report acknowledged certain limitations. For instance, it said that in two-thirds of the cases the samples didn’t have enough DNA for testing. Roman said that may mean the number of false convictions is much higher.
Roman said there likely are “dozens, if not hundreds, of people who were convicted erroneously; dozens, if not hundreds, of people who were not convicted of a crime they committed who may have gone on to commit new crimes; and there were dozens, if not hundreds, of people who thought they had justice as a victim of a horrible crime who didn’t.”
State files that were shared with the Urban Institute didn’t always contain context about the physical evidence or non-DNA facts about the case that would help determine the value of the DNA testing results.
That’s why the Department of Forensic Science always sends its results to the law enforcement agency that originally investigated the case, to the commonwealth’s attorney and to the suspect when he or she can be located, said Gail Jaspen, chief deputy director of the department.
“The laboratory’s job is to completely and expertly perform the DNA analysis and provide those results to the law enforcement agencies and the commonwealth’s attorneys who know the facts in the case,” Jaspen said. “DFS recognizes its own case files at best only contain some of the facts in the case, so we would not to presume to do the kind of analysis that the Urban Institute performed for the purposes of their study.”
Until now, the test results have been shielded from the public, but the General Assembly mandated they be made available on July 1. The department will comply, Jaspen said.
The department has revealed that it has 78 cases in which the tests have excluded the person who was convicted. However, the department has stressed all along that being excluded by DNA evidence does not prove innocence. For instance, someone could commit a rape or another crime without leaving behind biological evidence.
The Urban Institute report details 18 cases in which exoneration is a possibility if additional evidence was gathered. In some cases, there was no victim DNA with which to compare the sample, and in others the sample matched only the victim.
Among the 38 cases that the Urban Institute deemed likely for exoneration, 33 involved a DNA profile that did not match the suspect but did match a known person who was neither the person convicted nor the victim. In the other five cases, the convicted person was eliminated because the profile was shown to be from a male, but not from the suspect. Scientists can distinguish between male and female contributors in a DNA sample.
The U.S. Department of Justice’s National Institute of Justice gave the Virginia Department of Forensic Science $4.5 million to finance testing in the project, which aimed to compare evidence profiles to those of DNA profiles of known criminals. As part of the grant obligation, the department had to surrender its files to the Urban Institute for the study, Jaspen said.