DNA Evidence Frees the Innocent

But when we're not careful, this powerful technology can help imprison the wrongly accused


Sir Alec Jeffreys didn't know what awaited him when he entered his lab in Leicester, England, on September 10, 1984. "My life changed on Monday morning at 9:05 a.m.," he later told an interviewer at the University of Leicester. "In science, it is unusual to have such a 'eureka' moment." But the discovery he made that day in the science of DNA would turn criminal justice on its head.

The first genetic fingerprint was discovered "purely by accident," says Jeffreys. He and his team were studying the evolution of mammalian globin genes when they discovered repetitive DNA sequences spread through the genomes of almost all the species they tested. Further examination showed that they were present in most mammals and that the patterns appeared to be different in each animal. Thinking that these might act as markers that would help pinpoint the genes responsible for particular traits, they set up an experiment to look for them in human beings.

In September 1984, the experiment revealed several markers in the human genome, which appeared to be passed down through families and unique to each individual. The team first published its findings in 1985; a second article that year concluded the probability of two individuals having the same DNA fingerprint was less than one in 33 billion. The researchers realized immediately that their discovery would have radical implication for at least two areas: criminal investigations and paternity cases.

DNA evidence quickly found its way into American courtrooms, but the technology wasn't always used responsibly. It was soon challenged in court, a charge led by defense attorneys Barry Scheck and Peter Neufeld. The pair went on to found the Innocence Project, which is most famous for using DNA evidence to exonerate 350 people who were convicted of crimes they didn't commit. Their challenges to the early use of DNA fingerprinting in criminal trials encouraged labs to be more modest and precise in their claims, but it also sowed seeds of doubt about the accuracy of the techniques, and about the criminal justice system more broadly.

The First Criminal Case In November 1983, the body of 15-year-old Lydia Mann was found in the village of Narborough, England. She had been raped and strangled. The crime sparked a major investigation but went unsolved. Three years later, in July 1986, a similar crime occurred in the neighboring village of Enderby, where another 15-year-old, Dawn Ashworth, was found beaten, raped, and strangled. This time investigators had a suspect: Richard Buckland, a 17-year-old who worked in the kitchen of a local mental hospital.

Buckland had a very low I.Q. Upon questioning, he confessed to Ashworth's murder but not to Mann's. So investigators asked Jeffreys—who was in the spotlight after helping settle a paternity dispute in a high-profile immigration case the previous year—to compare Buckland's DNA with a sample taken from the Mann crime scene, which he did in September 1986.

The results showed that the samples did not match. Surprised and unsure how to proceed, police asked Jeffreys to test Buckland's sample against material from the Ashworth investigation, the crime to which he had confessed. The tests showed that the two crime-scene samples matched each other, but neither one matched Buckland, shocking not only the police but also Jeffreys, who found the result "blood-chilling." After the Home Office's Forensic Science Services confirmed the results, all charges were dropped against Buckland. Thus, Jeffreys said, "the first time DNA profiling was used in criminology, it was to prove innocence." It would later help identify the real killer, Colin Pitchfork.

The DNA Industry Jeffreys and his team were soon inundated with requests for assistance. Jeffreys described this time as "simply insane." In response to the incredible demand, Imperial Chemical Industries (ICI), the largest chemical company in the United Kingdom, opened its first DNA testing laboratory in the summer of 1987.

ICI Americas started offering DNA tests in the United States in October 1987 as Cellmark Diagnostics USA. Although the company initially planned to focus on paternity testing, it soon began accepting forensic casework. Concerned about the stricter admissibility standards in U.S. courtrooms, Cellmark hired a former chief of the Bureau of Alcohol, Tobacco, and Firearms' forensic laboratory, Daniel Garner, who immediately began preparing the company to introduce its tests into court, hiring scientists and building a network of expert advisers to shore up the scientific and technical side of things.

Cellmark also began to connect with criminal justice professionals through direct mail campaigns, advertisements in legal publications, and presentations at professional meetings. It presented its DNA fingerprinting technology as a definitive way to link offenders to a crime scene. In one 1988 advertisement, for instance, the company claimed that "criminals leave more conclusive evidence than ever before" and that DNA fingerprinting "may be the difference between conviction and acquittal." The advertisement went on to say that the company could obtain "conclusive results in only one test" using "samples that are months or even years old."

During this same period, another company, called Lifecodes, developed a technology for commercializing the DNA fingerprint and applying it to forensic analysis. The company's scientists published two research articles in 1986 concluding that its technology could successfully identify individuals from dried bloodstains and biological fluids recovered after the fact. Like Cellmark, Lifecodes began to pitch its new test to criminal justice professionals, advertising it as "exquisitely accurate." It also presented it as relatively simple and ready to be used in the courtroom.

By 1989, DNA test results had been admitted in more than 100 cases in the United States and had likely been used to acquire confessions in many more, all without substantial challenge from defense attorneys. As Neufeld put it at the time, "Mostly the evidence has come in without any objections, because the lawyers haven't known how to respond to it." The prevailing wisdom that the tests would produce either the correct answer or no answer at all—with no real possibility of a false positive—echoed throughout the early cases.

Setting Precedent In 1984, Neufeld and Scheck got their first introduction to DNA evidence. The defense attorneys, who had previously worked together at the Bronx Legal Aid Society in New York, received a case involving a man named Marion Coakley, who had been convicted of rape and robbery. The serologist who was involved in the case was going to work for Lifecodes and told the lawyers about the company's DNA testing, suggesting that they try it in the Coakley case.

There wasn't enough evidence to get conclusive results. Fortunately, other evidence pointed to Coakley's innocence, including a palm print that did not match, and he was exonerated.

But the idea of using DNA evidence was a powerful one, and Neufeld and Scheck kept thinking about the technology. "We realized then that DNA was going to be extremely important, and it had not really been involved in the courts yet," Scheck said in a 2003 interview at the University of California, Berkeley. "So we held a forum at Cardozo Law School involving DNA testing—the first one, I think, in any American academic institution for its forensic application." Scheck knew, he says, "from how that was set up, that there were some classic problems in the way they were trying to transfer this technology from medical and research purposes" to the forensic arena. They soon became involved in a number of legal battles over DNA in court; the first was the case of Joseph Castro.

On February 5, 1987, Vilma Ponce, 20 years old and seven months pregnant, was found stabbed to death along with her 2-year-old daughter. The husband told police that he had seen a man leaving the building with blood on his hands. The man turned out to be 38-year-old Joseph Castro, a janitor's helper and neighborhood handyman who lived in the adjacent building. Ponce's husband could not identify Castro in a photo array, but the investigation continued.

The "grim but unexceptional homicide" took a turn when detectives discovered a small spot of blood on Castro's watch. Risa Sugarman, an assistant district attorney in the Bronx, ordered a genetic test; the watch, along with samples taken from the two victims, was sent to Lifecodes for analysis. On July 22, the company issued a definitive, unambiguous report to the Bronx District Attorney's Office: "The DNA-PRINT pattern from the blood of Ponce matches that of the watch with three DNA probes. The frequency of these patterns in the general public is 1:189,200,000."

Castro was charged with two counts of second-degree murder, and prosecutors sought to have the test results admitted as evidence. Sugarman knew that this meant a Frye hearing, the standard procedure whereby courts determine whether to admit evidence obtained through novel scientific means or controversial methodology. Until this point, DNA identification evidence had never been declared inadmissible.

Unlike many of their colleagues in the legal and criminal justice worlds, Neufeld and Scheck were skeptical of the oft-praised DNA evidence and concerned about its use in the court. "We therefore decided to make the Castro case the first in which there would be a comprehensive inquiry into the various issues that comprise DNA typing," Neufeld told Science in 1989. Although case law seemed to be against them, they forged on.

Experts were initially hesitant to work with Neufeld and Scheck, because they were concerned that an attack on the new technology would harm its image and compromise its potential value. But that perspective began to change. In a 1989 story for American Lawyer, Neufeld said they eventually concluded that "it was the sloppy work of some of these laboratories that was going to give DNA technology a bad name." They had "a professional obligation to do some house-cleaning."

In November 1988, Scheck and Neufeld attended a symposium on forensic DNA typing at Cold Spring Harbor Laboratory on Long Island in New York. Neufeld described it as "the first conference where there was any discussion of potential problems with the technique. Previously, all you'd hear was how great it was." The meeting included not only defense attorneys but also prosecutors, FBI representatives, molecular biologists, and other scientists.

At the conference, Lifecodes scientist Michael Baird, who had by then testified in more than 20 hearings in the United States, gave a presentation in which he showed several autoradiographs, or autorads—the x-ray images examined to determine if the patterns on two samples match—and claimed that he would sometimes call a match even when the patterns did not quite line up but were similar. Furthermore, he said, he did not always conduct the proper control tests but knew from experience that the samples matched.

Buckland had a very low I.Q. Upon questioning, he confessed to one of the murders but not the other. DNA tests showed that the two crime-scene samples matched each other, but neither one matched Buckland—shocking not only the police but also Jeffreys, who found the result "blood-chilling."

Eric Lander, a human geneticist and mathematician at the Whitehead Institute, "was a little disturbed" by Baird's claim. Neufeld met Lander at the meeting and showed him one of the autorads from the Castro case. When Lander called some colleagues over and asked them if they thought it was a match, two of the three replied, "Garbage."

Soon after the conference, Lander was asked to serve as an expert in the Castro case. "Distressed and appalled" by Baird's responses to the defense attorney's questions, Lander reluctantly agreed to testify, though he declined to accept the expert witness fees. He ended up dedicating more than 350 hours of work to the case, preparing a 50-page report, and spending six days on the witness stand.

In Lander, Scheck and Neufeld knew they had found somebody "right up [their] alley." Scheck described him in the American Lawyer story as "this fast-talking New York Jew from Brooklyn who loved to argue and knew a lot of law." Throughout the hearings, Lander and the defense team found several problems with the methods used by Lifecodes in the Castro case.

The following April there was another meeting at Cold Spring Harbor, this one on genome mapping and sequencing. There, Lander met the biochemist and molecular biologist Richard Roberts. The two discussed the Castro case; Lander shared his reservations and left his report with Roberts, who "quickly became rather concerned" and "soon realized that something had to be done."

A few days after reading the report, Roberts suggested that the expert witnesses in the case, both from the prosecution and from the defense, meet and discuss the issue as scientists, with "none of this lawyerly talk." There were 10 experts in total, five for each side; only four were able to attend the meeting, but the remaining six gave their approval.

The experts met and ended up drafting a two-page consensus statement concluding that the Lifecodes tests in the case were "not scientifically reliable enough to support the assertion that the samples…do or do not match." Nine of the 10 experts agreed with this conclusion. Baird was the only holdout.

Neufeld describes the meeting as "unprecedented in the annals of law." The Castro decision held that while DNA fingerprinting was generally acceptable and admissible, the specific analysis performed by Lifecodes in the case was not. That is, the theory underlying forensic DNA typing was accepted, but Lifecodes failed to perform all the necessary tests. The prosecution conceded that the results were unreliable, with Sugarman noting that this outcome "was a better service to the criminal justice system."

To have a legal actor, particularly a prosecutor, share such a sentiment was no small feat. After all, it had been only a few years since forensic DNA fingerprinting was introduced to the U.S. with the promise of revolutionizing the search for truth—a new type of scientific evidence that was "incapable of giving a wrong answer" and would "reduce to insignificance the standard alibi defense." The profession's perception of DNA evidence was shifting, from an infallible crime-fighting tool to one that should be questioned deeply and systematically.

But Neufeld and Scheck were not satisfied. Lifecodes had performed its flawed tests in almost 1,500 criminal cases to that point, and the ruling didn't mention that the flaws uncovered in Castro potentially tainted those earlier results.

In what came to be an important part of the innocence movement's mantra, Scheck described this issue as "a fundamental civil liberties problem." Even if all of the scientific problems are fixed, "it is still wrong that these past cases are based upon unreliable science, and it is not inconceivable that some of these people have been unjustly convicted." Some people, including Neufeld, thought Castro could lead to challenges to previous Lifecodes cases. After the ruling, he and Scheck received numerous requests for help from attorneys around the country. Although they were unable to assist directly with most of them, they did make briefs and transcripts from Castro available through the National Association of Criminal Defense Lawyers.

The Castro case—along with the Schwartz case in Minnesota, in which Cellmark's techniques were challenged—also led the media to adopt some skepticism toward DNA evidence. While the cases did not disprove the potential power and reliability of DNA testing, they showed that the evidence must be carefully examined before being deemed admissible.

The DNA Wars In December 1991, Science published two articles—one by Richard Lewontin and Daniel Hartl criticizing the common probability statements made regarding forensic DNA matching, and one by Ranajit Chakraborty and Kenneth Kidd suggesting that the methodology was sound and generally accepted. The debate between the two sets of scholars was fierce, but it was the circumstances surrounding the publications that really showed just how contentious things were.

Lewontin and Hartl's article had been accepted by the respected popular journal first. When word of the upcoming publication got out, rumors flew that members of the FBI planned to halt publication of the paper. Prosecutor James Wooley called Hartl and expressed his concerns about the article being published. Lewontin wrote back to Wooley, suggesting that it was inappropriate for a state official to try to intimidate a citizen. The editor of Science, Daniel Koshland, was aware of the controversy and took another look at the paper. He asked Lewontin for revisions, which upset the original authors. Koshland also solicited the rebuttal from Chakraborty and Kidd.

Accusations were thrown from every direction. The scientists criticized one another, suggesting that philosophical and ideological beliefs were getting in the way of the science. But the lawyers were also involved. Neufeld and Scheck accused numerous scientists and federal law enforcement officials of "meddl[ing] in the peer-review process at leading scientific journals" and suggested that the journal editors were in cahoots with the FBI. They also accused Wooley of unethical behavior. Wooley's allies fired back, suggesting that the scientists supporting the defense were simply hired guns.

The debate over DNA evidence remained heated, with Scheck testifying before several congressional committees in the early '90s. His remarks focused on the need for blind proficiency testing and suggested that the FBI could not be trusted to adequately regulate itself.

Soon the technology would be used by defense attorneys for the first time to prove that someone who had been convicted and incarcerated was actually innocent.

Testimony Recanted On July 9, 1977, Cathleen Crowell was standing on the side of a road in the Chicago suburb of Homewood when a police officer passed by. He noticed her dirty clothes, and the 16-year-old recounted the story of being thrown into a car by three young men and raped as she was leaving work at a local mall. The rapist, she said, had scratched letters into her stomach using a broken beer bottle.

Crowell was taken to a nearby hospital, and a rape examination was performed. Clothing containing what was believed to be a seminal stain was collected, along with several pubic hairs and a vaginal swab. She worked with police to develop a sketch of the rapist. When shown a book of mug shots, she identified one Gary Dotson, who was arrested the next day and identified in a lineup.

Dotson went to trial with two key witnesses against him. The first was Crowell, and the second was a forensic specialist who claimed that the stain from the victim's underpants came from a type B secretor, which the defendant was. Since only about 10 percent of the population fits this profile, there was a good chance that Dotson was the source. The specialist also testified that the hairs recovered from the victim were "microscopically similar" to Dotson's; the prosecutor would later claim that they actually matched.

The evidence was enough to persuade the jury. Dotson was convicted of rape and aggravated kidnapping and given concurrent sentences of 25 to 50 years.

With the case seemingly closed, the victim moved on. She married (becoming Cathleen Webb) and moved to New Hampshire. It was not until March 1985—nearly eight years after the initial incident—that Webb expressed her guilt to her pastor. She informed him that she had fabricated the rape allegation as a cover story because she and her boyfriend had feared she was pregnant.

She contacted a lawyer, but the Cook County prosecutors wanted little to do with the old case. She eventually went to the press, and the Chicago Sun-Times covered the story. A clemency hearing was forthcoming, but the decision was far from a sure bet, as witness recantations rarely resulted in overturned convictions. Some people in the legal community, while respecting the precedent for not believing recantations, questioned the court's logic in this case, which was different from similar scenarios, since there was neither a threat nor money involved. One prominent Chicago lawyer bluntly stated, "The law has made an ass out of itself and the judge has helped it out, if you want to know the truth."

The national media picked up the story. A New York Times article suggested that the case could have wide ramifications for the criminal justice system: "Confidence in American justice cannot rest easily when [Dotson] is sent back to jail on the word of a woman who is, one way or the other, an acknowledged liar." Webb even said she would stand trial for perjury if it meant Dotson would be freed.

Judge Richard Samuels, however, believed the recantation to be "implausible" and denied clemency after a three-day hearing. The governor commuted Dotson's sentence to time served as a matter of "basic justice" and placed him on parole.

Journalist Civia Tamarkin recruited former assistant state's attorney Thomas Breen to take up the case. Edward Blake, then described as "the number-one forensic geneticist in the country," was brought on, and a new test was conducted on the seminal stain and blood samples from Dotson.

In August 1988, the test results excluded Dotson as a potential source of the semen. His lawyers requested a new trial, and prosecutors ultimately dropped the charges against him in 1989. Dotson was finally exonerated 10 years after his original conviction. By most accounts, he is considered the first American to be cleared using DNA evidence, though this is debatable; earlier the same year, a man named David Vasquez was exonerated of a rape and murder due, in part, to DNA testing.

A Movement Is Born Although Neufeld and Scheck were on the side challenging the scientific evidence in the Castro case, they did so not out of an aversion to the technology but because they recognized its potential: "In terms of how the innocence movement, through at least our work, began to form," Scheck says, "there was a focus at the very beginning on…the power of DNA technology…not just to exonerate the innocent but to identify those who had committed crimes, and that would really begin to expose all these other causes of wrongful convictions."

They already believed that some types of evidence, such as eyewitness identifications, were not always reliable, and that this technology would allow them to go back and examine questionable cases. While they had issues with the ways DNA was being handled, they never doubted its utility for exonerating innocents who had been wrongly convicted.

Just as DNA was a powerful investigatory tool that could be used to secure convictions, Scheck testified to Congress, it was also "an amazing tool to revisit old cases where people had been dragged out of the courtroom screaming, 'I'm innocent! I'm innocent! I'm innocent!'"

Scheck and Neufeld were taking potential innocence cases on an ad hoc basis and had some early successes. As word of their exploits spread, the letters piled up. So in 1992, they officially founded the Innocence Project as a clinical program at Cardozo Law School, creating an organized, systematic way to investigate old cases and use DNA to exonerate prisoners when possible. Initially the staff was just four people "processing thousands of letters," Neufeld says. He adds with a laugh that they were just "trying to hold the whole thing together with bubble gum and aluminum foil."

What existed at that point wasn't really a "movement"—just a handful of successful cases and a small circle of supporters and crusaders. But the next decade would bring greater attention to wrongful convictions, an expansion of the advocacy network, the formulation of an agenda, and, ultimately, the emergence of the modern innocence movement.

This essay is adapted from Exonerated: A History of the Innocence Movement by permission of New York University Press.