ABA Connection

The Uncertain Science of Evidence

  • Print.

On the morning of Dec. 29, 1991, the owner of the CBS lounge in Phoenix went to his bar to meet a repairman. Normally, the bar would have been closed and locked at the time, but he found the front door unlocked and the lights still on.

Looking around, he discovered the naked body of a bartender, Kim Ancona, sprawled on the floor of the men’s restroom. She had been stabbed 11 times. The killer left behind few clues. There were no fingerprints or semen. Hair found on the body was consistent with the victim’s. There was blood–lots of blood–but it was identified as type O, the same as the victim’s–and some 43 percent of the world’s population.

Police investigators didn’t check for DNA evidence because the technology to do so wasn’t widely used in 1991.

But the body did yield one potentially crucial piece of evidence. The victim had been bitten on the neck and on the left breast, through a tank top she had been wearing, with sufficient force to leave teeth marks. Police hoped to match these marks with the killer’s teeth.

Suspicion soon focused on Ray Krone, a 35-year-old mailman with no criminal record who knew the victim and had been a regular customer at the bar. After learning that the victim had told a friend that Krone was going to help her close the bar on the night of the murder, police asked Krone for a Styrofoam impression of his teeth for comparison. Two days later, Krone was charged with murder, kidnapping and sexual assault.

Misleading Impression

At his 1992 trial, Krone maintained his innocence, arguing that he was home sleeping at the time of the murder. But an expert witness for the prosecution testified that the bite marks found on Ancona’s body matched the Styrofoam impression made by Krone’s teeth. Krone was convicted and sentenced to death.

Three years later, the Arizona Supreme Court overturned Krone’s conviction and granted him a new trial on grounds that a videotape used by the state’s bite mark expert during his testimony wasn’t disclosed to the defense until the day before trial. But Krone was subsequently convicted again, largely on the basis of the bite mark evidence, and sentenced to life in prison.

It wasn’t until 2002, after Krone had spent more than 10 years behind bars, that DNA testing, which by then had become routine in criminal cases, was conducted on saliva found on the victim’s tank top. Not only did the results exonerate Krone, but they also implicated another Arizona prison inmate serving time for attempted child molestation.

Experts say this is one more in a growing line of cases that affirm the power of forensic DNA evidence. But another lesson of this and similar cases, they say, is that–as the credibility of some types of science-based evidence becomes stronger–courts should treat forensic evidence based on more uncertain science with greater skepticism and scrutiny. But that doesn’t always happen, say experts in the forensic evidence field.

“The courts, as a rule, have not done a very good job of separating junk science from emerging science and accepted science,” says Andre Moenssens, a professor emeritus at the University of Missouri-Kansas City School of Law.

Granted, several major developments in the past decade or so have dramatically changed the way courts look at scientific evidence in criminal cases, Moenssens and other legal experts say.

At the top of the list is the advent of DNA profiling, the most significant advance in forensic science since the development of fingerprinting in the early 1900s. With its exacting precision, DNA profiling allows scientists to positively identify one person to the exclusion of everybody else in the world except an identical twin.

Flimsy Foundations, Wrongful Convictions

At the same time, DNA profiling has helped expose the shaky scientific foundations upon which other forensic science rests.

DNA profiling has demonstrated how to conduct a scientifically respectable forensic identification, says Michael J. Saks, a professor at the College of Law at Arizona State University. It also is a dramatic reminder, he says, that many of those wrongful convictions were based on other types of forensic evidence.

(The number of convicted prisoners who have been exonerated through post-conviction DNA testing stood at 159 at the beginning of June, according to the Innocence Project at Yeshiva University’s Benjamin N. Cardozo School of Law in New York City.)

“DNA analysis now sets the gold standard against which other forensic sciences are measured. And it usually makes other forensic sciences look deficient by comparison,” says Paul C. Giannelli, a law professor at Case Western Reserve University in Cleveland who chairs the Committee on Science and Technology in the ABA Section of Criminal Justice.

The U.S. Supreme Court’s landmark 1993 decision in Daubert v. Merrell Dow Pharmaceuticals Inc., 509 U.S. 579, also has fundamentally altered the way courts deal with scientific evidence in criminal cases.

The Supreme Court expanded on its decision in Daubert–which Giannelli says may be “the most important evidence case ever decided”–with rulings in General Electric Co. v. Joiner in 1997 and Kumho Tire Co. v. Carmichael in 1999. “If DNA evidence revolutionized forensic science,” Giannelli says, “Daubert and its progeny have revolutionized the admissibility of evidence based on forensic science.”

Prior to Daubert, most courts followed a test for admissibility of scientific evidence set forth by the Supreme Court in its 1923 decision in Frye v. United States. Under the Frye rule, expert testimony may be admitted only if it is based on a well-recognized scientific principle or discovery that has “gained general acceptance in the particular field in which it belongs.”

But under Daubert, admissibility must be based on scientific methodology–meaning the evidence has a known or potential error rate and can be validated through testing, peer review and publication.

In General Electric Co. v. Joiner, 522 U.S. 136, the Supreme Court ruled that such admissibility decisions are subject to appellate review under an abuse of discretion standard, which gives wide latitude to the trial judge. And in Kumho Tire v. Carmichael, 526 U.S. 137, the court ruled that the Daubert standard applies not only to expert scientific evidence but also to expert evidence based on “technical or other specialized knowledge.”

Giannelli says it wasn’t at first clear whether Daubert would impose a more permissive or more restrictive standard of admissibility on scientific evidence than the Frye test. “There is language in the opinion that cuts both ways,” he says. But in practice, Giannelli and other experts say, Daubert has evolved into a rigorous standard for judging the admissibility of scientific evidence.

Some federal courts have even read the Daubert trilogy as inviting a re-examination of “generally accepted, venerable, technical fields.” Especially after the Kumho decision, experts say, many previously accepted forensic techniques–including handwriting analysis, hair comparisons, fingerprint examinations, tool mark identification, bite mark evidence and comparative bullet-lead analysis–have come under sharper scrutiny from courts.

That trend hasn’t been lost on criminal defense lawyers, who have become far more aggressive in challenging the admissibility of forensic evidence and far more willing to hire qualified experts to assist them in doing so.

In most cases, these challenges have failed to block admissibility of evidence, experts say, but they have exposed the lack of empirical support for many commonly employed forensic techniques. And even when defense lawyers don’t succeed in excluding evidence, they are increasingly challenging its weight and value before juries.

“All of these fields are vulnerable,” Giannelli says, “because they’ve never done the kind of empirical research that would be demanded today under Daubert.” But the Frye test is far from extinct, at least among the states.

At least 14 states have expressly rejected Frye’s “general acceptance” test in favor of the Daubert approach. But at least 16 states have retained the Frye standard. And many of those states–which include California, Florida, Illinois, New York and Pennsylvania are among the nation’s most active criminal law jurisdictions, according to Giannelli.

Some courts have held that Frye offers greater protection for defendants than Daubert. But the Daubert standard has also helped transform Frye into a more exacting test of admissibility, experts say.

An example of that trend is Ramirez v. Florida, in which the Florida Supreme Court rejected the testimony of five experts who claimed general acceptance for the technique of matching a knife with a cartilage wound in a murder victim. 810 So. 2d 836 (2001). The court noted that the procedure had never been tested, that meaningful peer review was lacking, that the error rate had not been quantified and that objective standards had not been developed.

“In short,” Giannelli says, “the Daubert factors were used. Ramirez represents an invigorated Frye test, and it is not alone.”

Whether under an increasingly stringent Daubert standard or a reinvigorated Frye test, the courts are scrutinizing scientific evidence more closely than ever, experts say. Federal courts in a number of jurisdictions, for instance, now limit the scope of testimony by handwriting analysis experts. While those courts have permitted experts to describe points of comparison between handwriting samples, they have refused to allow experts to testify as to the authorship of the handwriting sample in question.

However, as Moenssens points out, seven circuits of the U.S. Court of Appeals have held that handwriting analysis meets the Daubert criteria; no circuit has held that it doesn’t. Expert testimony concerning hair comparisons, tool mark identifications and bite mark evidence has also been subject to similar scrutiny by the courts.

Indeed, with the advent of mitochondrial DNA testing on hair, teeth and bones, evidence based on microscopic hair comparisons has all but disappeared from courtrooms. “Hair comparison evidence may be used today to exclude an individual, but I don’t think any modern day forensic scientist is going to say the opposite,” says Joseph Peterson, acting director of the Department of Criminal Justice at the University of Illinois-Chicago.

In 1997, the Mississippi Supreme Court reversed a lower court decision admitting expert testimony concerning bite mark comparisons, noting that numerous scholarly authorities had criticized the technique. The court also said that there was little consensus in the scientific community on the number of points that must match before any positive identification could be claimed. Howard v. State, 701 So. 2d 274.

Evidence under the Gun

In March, a New Jersey appeals court reversed the murder and armed robbery conviction of a man on grounds that an FBI crime lab technique that matches bullets to crimes by analyzing their lead content is scientifically flawed. New Jersey v. Behn, 868 A.2d 329. The decision is believed to be among the first to overturn a conviction based on a challenge to the FBI’s so-called comparative bullet-lead-analysis evidence since the release of a study by the National Research Council of the National Academies last year questioning the scientific validity of the technique. (See “Bullet Proof,” September 2004 ABA Journal, page 58.)

Even fingerprint evidence, which courts routinely have admitted for nearly 100 years, is not immune from challenges in the courts. In what appears to be the first reported case in federal court, a district judge in 2002 initially excluded expert testimony on fingerprint analysis, only to admit it later on reconsideration.

Ruling initially in United States v. Llera Plaza, 179 F. Supp. 2d 492, U.S. District Judge Louis H. Pollak of Philadelphia said the method of comparison used by fingerprint examiners did not meet the Daubert reliability standards.

The judge reversed himself a few months later after the government came forward with extensive expert testimony regarding the history and technique of fingerprint analysis; the training, certification and annual testing of FBI-certified experts; and the common standard used around the world for analyzing fingerprints. 188 F. Supp. 2d 549.

In 2004, a man was freed from prison in Massachusetts after having served more than six years when prosecutors acknowledged that his fingerprint was wrongly identified on a glass mug used by a man who shot and wounded a police sergeant in 1997. A few months later, the FBI said it had misidentified a fingerprint found on a plastic bag of detonator caps near the scene of a terrorist bombing at a Madrid train station last year as that of a lawyer in Oregon. Spanish authorities identified the fingerprint as that of an Algerian national.

That mistake apparently led the National Institute of Justice, the research arm of the Justice Department, to solicit proposals earlier this year for a research project designed to improve the fingerprint-examination process.

The credibility of scientific evidence in criminal cases has been further undermined by a rash of high-profile fraud and abuse cases involving perjury by expert witnesses, faked laboratory reports and testimony based on unproven techniques.

Those cases have forced the courts, the legal community and the public to recognize that scientific evidence is not infallible, says Betty Layne DesPortes, a defense lawyer in Richmond, Va., who chairs the jurisprudence section of the American Academy of Forensic Sciences.

More recent disclosures of laboratory scandals, such as those uncovered at the police crime lab in Houston, are reminders that errors occur through inadequate training and insufficient quality controls, DesPortes says.

“The search for truth in criminal cases has increasingly relied on the forensic science community,” she says. “If forensic evidence is not objectively tested, analyzed and interpreted by adequately trained scientists, the search for truth will potentially be compromised, if not defeated.”

Probably the most egregious of those cases involved Fred Zain, a police serologist at crime labs in West Virginia and Texas in the 1980s and early 1990s. Zain, who died in 2002, was found to have fabricated and misrepresented evidence in as many as 136 rape and murder cases in the two states over a 14-year period. Together, West Virginia and Texas eventually paid more than $7 million to settle false imprisonment suits filed by wrongfully convicted inmates.

Such cases, experts say, are starting to have an effect on admissibility decisions. In Ramirez, for example, the Florida Supreme Court cited the need to “preserve the integrity of the criminal justice system” in the face of rising criticism of forensic evidence by applying the Frye test in a prudent manner “to cull scientific fiction and junk science from fact.”

“Any doubt as to admissibility under Frye should be resolved in a manner that minimizes the chance of a wrongful conviction, especially in a capital case,” the court said.

Some prosecutors argue that fingerprint analysis is extremely accurate–assuming it is done properly. The misidentification in the Madrid bombing case was the result of failure by the FBI and a court-appointed independent expert to examine the original print rather than relying on a faxed copy, Clatsop County, Ore., District Attorney Joshua Marquis says.

Other types of forensic testing that are now considered controversial, such as hair comparisons and bite mark evidence, have never been held out by prosecutors to be definitive, says Marquis, who is a member of the National District Attorneys Association board of directors.

Marquis also claims that it was prosecutors who pushed in the late 1980s and early 1990s to have DNA testing admitted in the courts as reliable scientific evidence. And, he says, there have been fewer cases of fraud and abuse in the forensic sciences in the past few years.

James E. Starrs, a professor of law and forensic science at George Washington University in Washington, D.C., says it’s a good thing if many self-proclaimed experts are running for cover from tougher scrutiny in the courts.

“If they’re going to stand by their guns,” he says, “they better make sure they’re loaded.”

Tune In To Evidence Issues

As important as Frye and Daubert may be to lawyers and judges, their impact may pale in the mind of the public–and that means potential jurors–when compared to the “CSI effect.” The term refers to the impact popular television shows are having on the public’s expectations of what forensic science can and cannot do.

“Such shows fuel America’s fascination with forensics,” says Carol Henderson, director of the National Clearinghouse for Science, Technology and the Law at Stetson University College of Law in St. Petersburg, Fla., and a law professor at Nova Southeastern University in Fort Lauderdale, Fla. “Jurors are disappointed if the technologies they see on TV are not used in court, even when those technologies are purely fictional.”

Some experts express skepticism as to whether the CSI effect actually exists.

“There’s ample evidence that people are mesmerized by what they see on TV,” says Starrs, who wrote an article on the subject in the March/April issue of the American Academy of Forensic Sciences’ Academy News. “But I’d like to see some empirical proof that it’s being translated into actual cases on a regular basis, which is so far lacking.”

Some criminal defense lawyers also dismiss the CSI effect, but some prosecutors swear it’s real. They claim to have experienced it firsthand.

“We’ve had numerous cases where jurors have told us they acquitted even though they thought the defendant was guilty because there was no scientific evidence to back us up,” says Patricia Jessamy, the state’s attorney in Baltimore, who serves with Marquis on the board of directors of the National District Attorneys Association.

Jessamy cites the example of a man who was acquitted on first-degree murder and other charges last year in the shooting death of another man that was witnessed by the victim’s wife and 11-year-old daughter, both of whom identified the defendant as the shooter. But one alternate juror told a local reporter after the verdict that prosecutors had failed to produce any physical evidence tying the defendant to the murder.

“We were kind of shocked because the eyewitness testimony was very strong,” Jessamy says. “When you can’t get a conviction of a person who’s obviously committed a crime, that’s a problem, for us and for society.”

Some prosecutors now ask prospective jurors during voir dire about their TV-watching habits. Others are asking judges to include instructions to the jury to the effect that they shouldn’t be held to the unrealistic expectations of a television program.

Still others have taken to putting on what they describe as “negative evidence witnesses” to explain to the jury why certain evidence was not available or was not analyzed. Marquis says prosecutors now routinely practice what he calls defensive law. “If we don’t prove we did every test, it will become the touchstone for the defense,” he says.

Starrs says defense attorneys engage in their own strategies to make the most of whatever CSI effect there may be. He says some defense lawyers, upon learning that a prosecutor’s case lacks forensic evidence, will attempt to stack the jury with fans of crime-scene shows. Others will present their own version of the negative witness: an expert who suggests to the jury that the prosecution’s case is fatally flawed in the absence of forensic evidence.

But Starrs uses the trial of Scott Peterson last fall in California to dismiss the CSI effect. Peterson was convicted and sentenced to death for the murder of his pregnant wife, Laci, and their unborn son.

Forensic evidence was almost nonexistent in the case, Starrs notes. Instead, he says, “the jury’s verdict was entirely based on circumstantial evidence of a nonscientific stripe on the order of what over the long run has been the warp and the woof of the criminal justice system, rather than a reliance on any dubious CSI effect.”


Because of an editor's error, in "The Uncertain Science of Evidence," July 2005, page 48, Frye v. United States, was not correctly identified as a ruling of the U.S. Court of Appeals for the District of Columbia Circuit. The Journal regrets the error.

Mark Hansen is a senior writer for the ABA Journal.

Give us feedback, share a story tip or update, or report an error.