Questions Remain About DNA Evidence After a Decade of Use

It was one of those chance-in-a-million things.

Not long after she was raped, an Indiana woman thought she spotted her assailant on the street. Police arrested the man, took his blood and sent it to a crime lab for DNA analysis.

Robert Flowers said he hadn’t committed the crime, and lab results showed he was telling the truth: His DNA didn’t match the rapist’s. But it came so close, the laboratory just had to ask--could Robert Flowers have a brother? He could, and he did. And, eventually, Robert’s look-alike brother was convicted of the rape.

In cases like that, DNA evidence seems to have almost supernatural powers to clear the innocent and convict the guilty. But, 10 years after it first was used in a U.S. courtroom, questions persist about how such evidence is employed.

“People come up to me and they ask, ‘Is this stuff any good?’ ” says Harlan Levy, a former assistant district attorney in New York City and a staunch supporter of DNA fingerprinting. “I think it’s amazing that 10 years into this technology, people are still asking that.”

After years of debate over population variability, the genetic patterns of isolated ethnic groups and other arcane issues, lawyers and scientists finally agree that DNA testing can match the blood or semen recovered at a crime scene to a given individual.

What they don’t agree on is what that match means.

“I’m not against DNA technology; I’m just against stupid uses of it,” said William Thompson, a law professor at UC Irvine and a member of O.J. Simpson’s defense team.

Simpson’s defense illustrates how the debate has changed in the last decade. In the early days, the defense might have argued that the blood matched to Simpson through DNA analysis could have belonged to somebody else.

But that’s not common anymore; DNA testing has proven itself as a means of identification comparable in reliability to fingerprinting. So defense lawyers didn’t waste time arguing that the incriminating samples attributed to Simpson really belonged to somebody else. Instead, they contended that every bit of DNA evidence could be explained by mishandling, laboratory contamination or the planting of evidence.

That argument raised questions the prosecution couldn’t answer, Thompson said, which is why the jury acquitted Simpson.

But things might have turned out differently a decade ago, when lawyers were less sophisticated about DNA evidence and the debate centered on the probability of a coincidental match between blood or semen found at a crime scene and that of an innocent defendant.

That issue, which has become much less important in recent years, frequently leads to misconceptions about DNA evidence by attorneys, judges and juries.

First of all, a DNA match doesn’t prove a person committed a crime; additional evidence almost always is needed for a conviction. But defense lawyers worry that juries give a match more weight than it deserves.

When a DNA match is introduced in court, it almost always is presented with a number known as a random-match probability. But nobody, DNA experts complain, seems to understand what that number means.

The random-match probability is an estimate of the chance that a randomly selected person will share the DNA pattern present in both the defendant and the sample from the crime scene. In most cases, the random-match probability is one in a million, one in a billion or higher.

The random-match probability also says nothing about the other issues in the case. Although laboratory error or planting of evidence also could explain a match, those possibilities aren’t accounted for in that one-in-a-billion number.

Jurors rarely understand that, said Jonathan Koehler, a professor of behavioral science at the University of Texas at Austin.

His research shows that jurors, perhaps dazzled by figures in the millions or billions, focus on the random-match probability and often ignore laboratory error as a possible explanation for the correlation between the defendant’s DNA and that from the crime scene.

When the Orange County Sheriff’s Department sent 50 blood samples to three laboratories in the early days of DNA testing, the accuracy rate was about 98%. Presumably, labs have improved since then, but nobody knows how much. Subsequent studies have found that laboratories falsely match samples once in a few hundred to a thousand times.

“We are concerned that this effect, in combination with misleading assurances from forensic science experts that laboratory errors are impossible or nearly impossible, could lead to convictions where acquittals might otherwise result,” Koehler and two colleagues wrote in the Winter 1995 issue of Jurimetrics Journal.

But despite the complaints of critics, in 10 years, no one has been shown to have been falsely convicted by DNA evidence. That may be because laboratory errors are relatively rare and tend to be so easy to spot when they do happen that they need not be considered in most cases.

“DNA is the best tool for identifying the innocent and the guilty that we have,” said former New York prosecutor Levy, whose book, “And the Blood Cried Out,” examines the method in a series of case studies.

In one of those cases, a Baltimore woman was raped by a man who broke into her house at night and put a pillowcase over her head so she couldn’t see him. Carol Sanders identified the rapist as her ex-boyfriend, a man who had been stalking her for months, and insisted she knew him well enough to recognize him even without getting a good look at him.

When police caught up with John Davis, however, he swore he hadn’t committed the crime. He even went so far as to suggest police investigate an acquaintance, Gregory Ritter. Ritter looked just like him, Davis said, and knew the victim because he had once dated her roommate.

Accustomed to such claims, investigators ignored Davis’ attempts to clear himself. They arrested him and charged him with the rape, taking a blood sample for DNA testing.

Four months later, Davis was free, because his DNA didn’t match that of the rapist. And Ritter was convicted and sentenced to 40 years in prison.

It’s not uncommon for eyewitnesses to accuse the wrong person in a rape case. In fact, DNA fingerprinting has shown that eyewitness identifications are notoriously unreliable.

Eyewitness testimony helped convict 24 of the 28 people described in a 1996 National Institute of Justice report whose cases were later overturned by DNA evidence. And other forensic methods, such as hair analysis, were involved in about two-thirds of those wrongful convictions.

DNA fingerprinting is about to get even more powerful, its proponents say. Most states now have databanks containing DNA profiles of people already convicted of sexual or related offenses. When police begin investigating a new crime, they can test DNA collected at the scene to see if it matches anybody in the databank with a history of a similar offense.

“With what we know about the rates of recidivism,” said James Wooley, an assistant U.S. attorney in Ohio, “it can only be a tremendous help.”

And, as DNA evidence has in so many other cases, databanking also can help clear the innocent.

Most states haven’t yet entered many offenders into their databases. But as the databases grow, law enforcement officials expect to catch a lot more criminals with the DNA they leave behind.

That hasn’t escaped the attention of criminals. Some rapists now use condoms the same way a burglar uses gloves--to avoid leaving identifying marks at the crime scene.

That may be the most convincing testimonial to DNA fingerprinting and how far it has come in 10 years.

“This testing,” Levy said, “is going to become more and more like the fingerprint it was originally promised to be.”