Edward Imwinkelried is a law professor at the University of California-Davis and visiting professor at University College Dublin.

However, it would be a mistake to leap to the conclusion that these revelations should prompt the wholesale exclusion of forensic testimony. Singling out expert testimony for that treatment might well prove to be counterproductive. To the extent that we discriminate against expert testimony, we force the courts to rely on alternative types of evidence, notably lay testimony such as eyewitness identifications. (Foreword, Convicted by Juries, Exonerated by Science: Case Studies in the Use of DNA Evidence to Establish Innocence after Trial xiii-xiv (1996).) To date, the various innocence projects have documented more than 300 cases of wrongful conviction in the United States. Nearly two-thirds have involved mistaken identification testimony by lay witnesses. (The Innocence Project, 5 RES IPSA 2 (Spring 2002).) The large number of mistaken eyewitness reports in the recent Washington, D.C.-area sniper investigation is further proof that lay witness identification is frequently untrustworthy. (Sharon Begley, Eyewitnesses to Crime Are Often Blinded by Shock, Adrenaline, Wall St. J., Oct. 25, 2002, at B1).) Reflexively toughening the standards for introducing expert testimony could compel the courts to depend even more on lay testimony. In that light, the challenge in reforming the admissibility standards for expert testimony is to strike the right balance—framing standards that improve the caliber of the expert testimony admitted without raising the bar so high that by default, the courts will have to rely even more heavily on suspect eyewitness evidence.

The purpose of this short article is not to chronicle every instance in which erroneous expert testimony has contributed to a wrongful conviction. Rather, this article undertakes to give the reader a general sense of the magnitude of the problem of flawed expert testimony and then to compare that to the size of the problem caused by erroneous lay testimony. That comparison points to the real challenge that the revelations of flaws in expert testimony pose for the judicial system. Any fair comparison of lay and expert testimony will lead to the conclusion that both types of evidence are imperfect. Although our prior naive faith in science makes the recent revelations of flawed expert testimony particularly shocking, the shock effect should not blind us to the weaknesses of lay evidence. The thesis of this article is that it will do more harm than good to react to the revelations by excluding expert testimony in a sweeping fashion. Our response must be more measured, designed to identify reforms that will enhance the reliability of the expert evidence that is admitted to supplement lay testimony at trial.

Flawed expert testimony. What is the magnitude of the problem of flawed expert testimony? For several decades there has been mounting evidence of a substantial margin of error in expert analysis. Numerous proficiency studies of laboratories have documented that expert analysis is far from infallible. (Hansen, Caudill & Boone, Crisis in Drug Testing: Results of CDC Blind Study, 253 J. Am. Med. Ass’n 2382 (1985) (a Centers for Disease Control test of the proficiency of laboratories engaged in immunoassay drug testing); J. Peterson, E. Fabricant & K. Field, "Crime Laboratory Proficiency Testing Research Program" (1978) (a study involving 21 tests and 240 laboratories); Paul C. Giannelli, The Admissibility of Laboratory Reports in Criminal Trials: The Reliability of Scientific Proof, 49 Ohio St. L.J. 671, 675–92 (1988) (discussing the Laboratory Proficiency Testing Program conducted by the Law Enforcement Assistance Administration); D. Michael Risinger, Mark P. Denbeaux & Michael J. Saks, Exorcism of Ignorance as a Proxy for Rational Knowledge: The Lessons of Handwriting Identification "Expertise," 137 U. Pa. L. Rev. 731, 743–51 (1989) (proficiency tests of questioned document examiners, conducted by the Forensic Science Foundation); Grieve, Possession of Truth, 46 J. Forensic Ident. 521 (1996) (a proficiency test of fingerprint examiners).)

Worse still, there is now solid evidence that erroneous expert testimony is causing a significant number of erroneous convictions. At one time, we might have been able to kid ourselves into thinking that although forensic experts sometimes erred, any errors would be unmasked at trial, and the errors therefore could not cause miscarriages of justice. The forensic analyst might err in the laboratory, but we believed that in the courtroom an astute cross-examiner would expose the error. However, we can no longer entertain that belief in good conscience. That belief was shattered by the U.S. Department of Justice’s celebrated 1996 report, Convicted by Juries, Exonerated by Science. The report reviewed 28 cases of wrongful conviction. In several of those cases, flawed expert testimony, in particular hair evidence, had contributed to the conviction. Although later DNA testing established their innocence, at trial defendants such as Edward Honaker, Roger Coleman, Ron Williamson, Steven Linscott, David Vasquez, and Dennis Williams all were convicted, and all were found guilty by juries that had heard supposedly inculpatory expert hair testimony. (Giannelli & West, supra at 514–18.) If expert errors were confined to the laboratory, they would be of concern solely to the scientific community. However, today it is undeniable that erroneous expert testimony is sometimes offered at trial—and that on at least some occasions, the error is not exposed.

Mistaken lay eyewitness testimony. The question naturally arises: If errors in expert testimony are causing miscarriages of justice, should the courts respond by barring the testimony or at least making it markedly more difficult to introduce expert testimony? No. That simplistic "cure" would be worse than the disease.

Although some expert testimony is flawed and, on occasion, such testimony contributes to wrongful convictions, the problem of erroneous expert testimony is smaller than and more tractable than the problem of mistaken eyewitness testimony by laypersons.

To get a sense of relative magnitudes of the problems, consider the set of cases reviewed in the DOJ’s Convicted by Juries, Exonerated by Science. The study found seven of the 28 cases of wrongful conviction featured flawed expert hair testimony. (Id. at 16–17.) In contrast, most of the remaining cases involved eyewitness identifications by victims, percipient witnesses, or both. (Id.) More broadly, as previously stated, one estimate is that mistaken eyewitness testimony factored in approximately two-thirds of the wrongful conviction cases identified to date by the various innocence projects in the United States. (The Innocence Project, supra.) In the experience of the Center for Wrongful Convictions at Northwestern University, faulty eyewitness identifications are "by far the most ubiquitous factor." (Steve Chapman, Your Lyin’ Eyes: What to Do About Eyewitnesses Who Get It Wrong, May 14, 2002, available at http://slate.msn.com/id/2065761/.) Similarly, most of the wrongful convictions identified by the Innocence Project at the Benjamin Cardozo School of Law have "hinged on faulty eyewitness testimony." (Barry Scheck, Peter Neufeld & Jim Dwyer, Actual Innocence (2000).) If this set of cases is at all representative of the universe, mistaken lay testimony is a much more common cause of wrongful conviction.

Although some types of expert testimony have significant error margins, for the most part the reported error margins are relatively small. "The point is not that most laboratory test results are erroneous. . . . Indeed, the opposite is true." (Giannelli, The Admissibility of Laboratory Reports in Criminal Trials, supra, at 692.) For instance, on most of the 21 sets of samples involved in the Laboratory Proficiency Testing Program, the vast majority of laboratories reached correct, complete findings. (Id. at 689 n.155.) For many sets, the accuracy rate was well over 90 percent. (Id. (test samples 1, 3, 4, 6, 7, 12, 13(A), 13(B), and 20(A)).) Contrast the error rates reported by many of the eyewitness testimony researchers. There are tens of studies finding an alarmingly high level of error in eyewitness identifications. In one classic experiment, fewer than 15 percent of the lay witnesses to a simulated crime correctly identified the perpetrator. (Robert Buckout & Mark Greenwald, Witness Psychology, in Sci. and Expert Evidence 1291 (2d ed. 1981).)

Consider, for example, the current controversy over the admissibility of questioned document examination testimony. That controversy is a microcosm of the broader dispute over the standards for admitting nonscientific forensic testimony. The critics of questioned document testimony have presented a persuasive, even compelling, case that it does not qualify as science. (D. Michael Risinger, Mark Denbeaux & Michael Saks, Brave New "Post-Daubert World"—A Reply to Professor Moenssens, 29 Seton Hall L. Rev. 405 (1998); D. Michael Risinger & Michael Saks, Science and Nonscience in the Courts: Daubert Meets Handwriting Identification Expertise, 82 Iowa L. Rev. 21 (1996).) As they have quite properly pointed out, many of the discipline’s basic premises have never been validated by controlled scientific experimentation and induction. (D. Michael Risinger, Mark Denbeaux & Michael Saks, Exorcism of Ignorance as a Proxy for Rational Knowledge: The Lessons of Handwriting Identification "Expertise," 137 U. Pa. L. Rev. 731 (1989).) Should questioned document examiners be permitted to testify as "scientists"? On the current state of the record, the answer must be "no." (United States v. Starzezpyzel, 880 F. Supp. 1027 (S.D.N.Y. 1995).)

However, the harder question is whether they ought to be permitted to testify at all. Should they be allowed to appear as nonscientific experts, perhaps with a cautionary instruction telling the jury that in evaluating the weight of their testimony, the jurors should consider that the witness is not a full-fledged scientist? The critics are correct in not only pointing to the dearth of validating research, but also in faulting the discipline for failing to conduct the research. The research is feasible, and there is no good justification for the discipline’s neglect to critically examine its own basic premises. However, there is hard evidence that although their techniques are not as reliable as the courts previously assumed them to be, experienced questioned document examiners can perform identification tasks far more reliably than laypersons—such as the persons who sit as jurors. Professor Inbau’s classic study, "Lay Witness Identification of Handwriting," 34 Illinois Law Review 433 (1939), demonstrated the unreliability of lay testimony on this subject. More recently, Dr. Moshe Kam and his colleagues have conducted a series of studies examining the question of whether experienced examiners can determine authorship more reliably than laypersons. (Moshe Kam, K. Gummadidala, G. Fielding & R. Conn, Signature Authentication by Forensic Document Examiners, 46 J. Forensic Sci. 884 (2001); Moshe Kam, G. Fielding & R. Conn, Effects of Monetary Incentives on Performance by Non-professionals in Document Examination Proficiency Tests, 43 J. Forensic Sci. 1000 (1998); Moshe Kam et al., Writer Identification by Professional Document Examiners, 42 J. Forensic Sci. 778 (1997); Moshe Kam, J. Welstein & R. Conn, Proficiency of Questioned Document Examiners in Writers Identification, 39 J. Forensic Sci. 5 (1994).) In their 1997 study, the researchers found that the percentage of matching errors by laypersons—38.6 percent—dwarfed the percentage of errors by questioned document (QD) examiners—6.5 percent. (Kam et al., supra, at 779.) In the most recent study released in 2001, the researchers reported that "the error rates exhibited by the FDEs (forensic document examiners) were much smaller than those of the laypersons." (Kam, Gummadidala, Fielding & Conn, supra, at 884.) "Laypersons wrongly classified nongenuine signatures as ‘genuine’ 13 times more often than the FDEs." (Id. at 884–85.) Australian research, published in 2002, largely confirms Dr. Kam’s findings. (J. Sita, B. Found & D.K. Rogers, Forensic Handwriting Examiners’ Expertise for Signature Comparison, 47 J. Forensic Sci., 1117, 1121 (2002) (the laypersons committed six times as many errors as the questioned document examiners.).) Admittedly, Dr. Kam’s research falls far short of establishing the scientific status of questioned document examination. However, the results of these experiments furnish convincing evidence that professional examiners can make the authorship determination far more accurately than laypersons. From a systemic perspective, it would be a step backward to bar forensic expertise and force the courts to rely on lay determinations of authorship.

It is not just that lay testimony seems at least as prone to error as expert evidence. To make matters worse, the causes of mistaken eyewitness identification are much less tractable than the causes of erroneous expert analysis. Inherent deficiencies in the human processes of perception and memory are the leading causes of flawed eyewitness testimony. (Felice J. Levine & June Louin Tapp, The Psychology of Criminal Identification: The Gap from Wade to Kirby, 121 U. Pa. L. Rev. 1079, 1095–1103 (1973).) Sadly, "very little can be done to improve" those processes. (Id. at 1130.) To be sure, we can "tinker" a bit with eyewitness testimony. For example, a court can mandate that lineups be conducted sequentially. (In re Thomas, 733 N.Y.S.2d 591 (S. Ct. 2001) or in a double-blind manner. (In re Wilson, 741 N.Y.S.2d 831 (S. Ct. 2002).) In addition, a court may admit expert psychological testimony to help the jurors appreciate the extent and causes of mistaken eyewitness identifications. (United States v. Mathis, 264 F.3d 321 (3d Cir. 2001), cert. denied, 152 L. Ed. 2d 148, 122 S. Ct. 1211 (2002).) To a degree, those procedures can reduce the impact of deficient perception and memory, but they cannot eliminate the root causes of mistaken eyewitness testimony. Little can be done to control the fortuitous events such as crimes that laypersons unexpectedly witness. There is no time machine to enable the witness to travel back in history and observe the event more carefully a second time.

The laboratory is a far more controllable environment. Laboratories have tools such as scanning electron microscopes, which can magnify in excess of 200,000 times, for enhancing the powers of perception. Similarly, laboratories can employ photographic techniques to record events and remove concerns about the quality of memory. Experiments and tests can be planned in advance to allow time for meticulous observation and recordation.

The bottom line is that it is misleading to focus solely on the weaknesses of expert testimony. The judgment must be comparative. To the extent that we discriminate against expert testimony, subjecting it to uniquely restrictive rules, we pressure the courts to depend more heavily on other types of evidence. In all likelihood, a wholesale tightening of the standards for admitting expert testimony will result in greater judicial reliance on lay testimony. The massive evidence of the myriad weaknesses in lay testimony should give us pause.

If we consider the reform of admissibility standards for expert testimony, we should reject appeals for new blunderbuss restrictions on expert testimony.

Outright bans on certain types of expert testimony. In some cases, legislatures have intervened to ban particular types of expert testimony such as polygraphy. (E.g., Cal. Evid. Code
§ 351.1.) Convicted by Juries, Exonerated by Science identified seven cases in which expert hair testimony contributed to a wrongful conviction. Should we ban all testimony about microscopic hair analyses? In another seven cases, non-DNA blood testimony was a contributing factor. Should we exclude all genetic marker analyses other than DNA evidence? These bans would certainly be ill conceived. Even in the current state of these forensic disciplines, their findings appear more trustworthy than many eyewitness identifications. Further, research is ongoing in these disciplines. Even if a ban were justified based on the current state of these disciplines, the disciplines are constantly evolving; and research could easily overtake the ban in a matter of years or months. As the National Research Council’s Committee on DNA Technology in Forensic Science cautioned, it is a chancy proposition to "lock in" or freeze admissibility standards for specific types of expert testimony. (Comm. DNA Technology in Forensic Science, Nat’l Research Council, DNA Technology in Forensic Science § 2–18 (1992).) As then-Judge Ming Chin observed, the "scientific landscape" is constantly changing. (People v. Barney, 10 Cal. Rptr. 2d 731, 744 (Ct. App. 1992).) An outright ban is warranted only in the most extreme cases.

A requirement for validation by controlled scientific experimentation. Short of outright bans on particular types of expert testimony, should the courts take the position that the only way to validate expert testimony is by extensive, controlled scientific experimentation? If the courts were to rigorously apply that view, many types of expert testimony, such as questioned document examination and hair analysis, would not only not qualify as scientific testimony, they would be altogether excluded from testimony. (Randolph Jonakait, The Meaning of Daubert and What That Means for Forensic Science, 15 Cardozo L. Rev. 2103 (1994).) There are passages in Daubert that lend themselves to that reading. For example, at the very beginning of the list of factors he thought trial judges should consider in evaluating the reliability of proffered scientific testimony, Justice Blackmun stated that a "key question" is "whether [the theory or technique] can be (and has been) tested." (Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 593 (1993).) Justice Blackmun also approvingly quoted Popper’s declaration that the "criterion of the scientific status of a theory is its falsifiability, or refutability, or testability." (Id.)

However, in other passages, the Court pointed to a broader test. Justice Blackmun wrote that "[t]he inquiry envisioned by [Federal] Rule [of Evidence] 702 is, we emphasize, a flexible one." (Id. at 590.) Ultimately, the Justice explicitly demanded "appropriate validation" (id.), not controlled scientific experimentation and induction. Kumho, the Court’s 1999 opinion, undercuts any contention that empirical testing and induction are the mandatory means of validating all expert theories and techniques. On its face, Rule 702 refers in the alternative to "scientific, technical or other specialized knowledge." In his lead opinion, Justice Breyer appeared to concur with the solicitor general’s contention that: "[T]here are many different kinds of experts, and many different kinds of expertise. See Brief for United States as Amicus Curiae 18–19 and n.5 (citing cases involving experts in drug terms, handwriting analysis, criminal modus operandi, . . . and others)." (Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 141 (1999).) Rule 702 was amended effective December 1, 2000. The new advisory committee note expressly states that "[s]ome types of expert testimony will not rely on anything like a scientific method . . . ."

This expansive view of "appropriate validation" leaves open the possibility of admitting fingerprint testimony as nonscientific expertise even if, in the current state of the research record, fingerprint analysis cannot qualify as full-fledged science. The critics of fingerprint evidence have argued that as in the case of questioned document examination, fingerprinting’s essential premises have not been subjected to rigorous, scientific scrutiny. (Robert Epstein, Fingerprints Meet Daubert: The Myth of Fingerprint "Science" Is Revealed, 75 So. Cal. L. Rev. 605 (2002); Jennifer Mnookin, Fingerprint Evidence in an Age of DNA Profiling, 67 Brooklyn L. Rev. 13 (2001).) However, the question is whether fingerprint analyses by FBI examiners are more likely to be reliable than the lay evidence that the trier of fact would otherwise have to rely on. (United States v. Llera Plaza, 188 F. Supp. 2d 549, 572 (E.D. Pa. 2002).) If we are interested in the reliability of judicial fact finding, the prospect of denying the jury fingerprint testimony and forcing it to depend on eyewitness testimony is not a comforting one. Like outright legislative bans on various types of expert testimony, an invariable requirement for validation by controlled scientific experimentation would probably do more harm than good.

Rather than resorting to a blunderbuss approach to the problem of flawed expert testimony, we should seek reforms that improve the caliber of expert testimony without making the courts unduly dependent on eyewitness testimony. Two reforms are
illustrative.

The abolition of the exemptions from judicial scrutiny. One is the abolition of the complete exemption from judicial scrutiny that certain types of expert testimony currently enjoy in various jurisdictions. For the most part, these exemptions exist in states still adhering to the traditional general acceptance test for the admissibility of scientific testimony. In Frye v. United States, 293 F. 1013 (D.C. Cir. 1923), the court enunciated the general acceptance standard. Frye involved a novel "instrumental" scientific technique, namely, the measurement of systolic blood pressure in order to detect conscious attempts at deception. Note the implication: If the scope of the test is strictly confined to hard science techniques, then neither soft science nor nonscientific expertise is subject to scrutiny.

In a large number of Frye jurisdictions, the courts have embraced that implication and formally exempted either soft science such as psychology or nonscientific expertise. For instance, many Frye courts have ruled that the test is inapplicable to noninstrumental or "software" scientific techniques such as psychological analysis. In People v. McDonald, 690 P.2d 709 (Cal. 1984), the California Supreme Court held that the test does not apply to psychological testimony about the unreliability of eyewitness identification. The court suggested that the test controls only "when the evidence is produced by a machine." (Id. at 724.) More recently, in People v. Wilson, 86 Cal. Rptr. 2d 204 (App. 1999), an intermediate California court refused to extend the test to psychiatric testimony that a defendant displays signs of deviance. In State v. Trager, 974 P.2d 750 (Or. App. 1999), the Oregon Court of Appeals concluded that the test was inapplicable to a physician’s opinion diagnosing two young girls as having been sexually abused. Trager, Wilson, and Kelly are not isolated cases. One commentator asserts that this limitation is the majority view among Frye jurisdictions. (Roger S. Hanson, James Alphonzo Frye Is Sixty-Five Years Old; Should He Retire? 16 W. St. U. L. Rev. 357, 414 n.192 (1989).)

Further, a number of Frye states have limited the reach of the general acceptance test to purportedly scientific theories and techniques; if the proponent simply labels the testimony nonscientific, that expedient obviates the necessity for the proponent to establish that the expert’s theory or technique is generally accepted in the pertinent field. In the past, many jurisdictions have taken an essentially laissez-faire attitude toward nonscientific expertise. (John Strong, Language and Logic in Expert Testimony: Limiting Expert Testimony by Restrictions of Function, Reliability, and Form, 71 Or. L. Rev. 349, 361 (1992).) Florida is a case in point. On the one hand, unlike most Frye states, Florida does not broadly exempt soft science from Frye scrutiny. (Hadden v. State, 690 So. 2d 573 (Fla. 1997) (child sexual abuse accommodation syndrome); Flanagan v. State, 625 So. 2d 827 (Fla. 1993) (sex offender profile).) On the other hand, Florida has balked at broadening the scope of the test to extend to nonscientific evidence. In Hadden, the Florida Supreme Court wrote that the general acceptance test "is not applicable to an expert’s pure opinion testimony which is based solely on the expert’s training and experience." (Id. at 579–80.) In other jurisdictions, while the courts have not carved out a formal exemption from scrutiny for nonscientific expertise, they have accorded "technical" evidence such as fiber and hair analysis a presumption of reliability. (State v. Fukusaku, 946 P.2d 32, 43 (Haw. 1997).)

As previously stated, in the main it would be unwise to adopt legislative bans on such species of expert testimony as hair analysis and non-DNA blood evidence. However, it is equally unwise to sweepingly exempt nonscientific evidence from all judicial scrutiny. That loophole should be closed. The reliability of lay eyewitness testimony might be questionable, but it is hardly an antidote to admit nonscientific expert evidence as a matter of course without any inquiry into its reliability.

These exemptions from scrutiny affect a large number of trials. It is true that in Kumho the Supreme Court extended the requirement for a showing of reliability to all types of expertise in federal practice. It is also true that the Daubert-Kumho approach has spread to many states with the result that Frye is already a minority view in the United States. (1 Paul Giannelli et al., Sci. Evidence § 1–15 (3d ed. 1999).) However, the remaining Frye jurisdictions include some of the largest and most litigious states such as California, Florida, Illinois, New York, Pennsylvania, and Washington. The majority of state criminal trials in the United States are conducted in Frye jurisdictions (Evidentiary Balance, Nat’l L.J., May 13, 2002, at B11); and, as we have seen, this loophole looms large in many such states. In those states, in passing on the admissibility of the expert testimony, the judge may determine whether the witness qualifies as an expert and whether the subject matter is beyond the ken of the typical lay juror. However, there will be no meaningful inquiry into either the general acceptance or the reliability of the expert’s underlying theory.

A foundational requirement for a showing of sound test procedure. A second useful reform would be mandating that the proponent of scientific testimony demonstrate that on the occasion in question, the expert followed sound procedure in conducting the test. Sloppy test procedure is the Achilles’ heel of many types of expert evidence. Modernly, we have powerful technologies such as DNA typing, but the analysts who use these technologies are fallible human beings who are sometimes overworked and under-trained. Many of the published proficiency tests identify sloppy test procedure as a leading cause of erroneous findings. (J. Peterson, E. Fabricant & K. Field, supra, at 203–06, 223, 230, 239, 258; The Debate in the DNA Cases Over the Foundation for the Admission of Scientific Evidence: The Importance of Human Error as a Cause of Forensic Misanalysis, 69 Wash. U. L.Q. 19, 32 (1991).) In one sense, proof of proper test procedure is one of the most fundamental guarantees of the reliability of expert testimony. When the original validating studies are conducted, researchers control for certain variables. We can have confidence in the researchers’ findings only under those conditions. If at the time of the later forensic test the analyst does not control for the same variables, the supposedly validating research furnishes no assurance of the reliability of the test. It makes no sense to insist upon a validation of the expert’s technique while ignoring the question of whether the analyst followed correct test procedure in the sense of duplicating the conditions that obtained during the earlier validating research.

The traditional, common-law view was that a showing of proper test procedure was a required element of the foundation for introducing scientific testimony. (State v. Schwartz, 447 N.W.2d 422 (Minn. 1989); People v. Castro, 545 N.Y.S.2d 985 (S. Ct. 1989).) However, after the enactment of the Federal Rules of Evidence, a trend emerged toward the view that errors in test procedure affected only the weight, not the admissibility, of scientific evidence. (Rockne Harmon, How Has DNA Evidence Fared? Beauty Is in the Eye of the Beholder, 1 Expert Evid. Rep. 149 (Feb. 1990).) The Daubert decision added fuel to that trend. There the Court relied on Federal Rule of Evidence 402 as the basis for holding that Frye did not survive the adoption of the federal rules. The Court construed Rule 402 as abolishing uncodified exclusionary rules of evidence. (Daubert, supra, at 586–87.) In pertinent part, Rule 402 announces: "All relevant evidence is admissible, except as otherwise provided by the Constitution of the United States, by Act of Congress, by these rules, or by other rules prescribed by the Supreme Court pursuant to statutory authority."

The Court approvingly quoted the statement of Professor Edward Cleary, the reporter for the Federal Rules of Evidence Advisory Committee, that "[i]n principle, under the Federal Rules no common law of evidence remains." (Id. at 588.) Frye was no longer good law because it was a creature of case law, and the statutory text of the federal rules did not contain any language that could reasonably bear the interpretation that it incorporated a general acceptance standard. By the same token, prior to the December 1, 2000, amendment to Rule 702, article VII did not include any language that seemed to prescribe a showing of proper test procedure. The argument runs that, like Frye’s general acceptance test, the foundational requirement for a showing of proper test procedure was impliedly abolished by Rule 402. The same argument can be raised in many of the more than 40 states with evidence codes patterned after the federal rules, including statutes tracking Rule 402. (Federal Rule of Evidence 402: The Second Revolution, 6 Rev. Litig. 129, 135-36 (1987).)

Fortunately, effective December 1, 2000, Rule 702 was amended to include a requirement for a foundational showing that "the witness has applied the principles and methods reliably to the facts of the case." The amendment closes the loopholes in federal practice, but the problem still persists in many states, almost none of which have amended their version of Rule 702 to insert the same language.

In the past, all too many jurisdictions took a lax approach to the admission of expert testimony. Several jurisdictions exempted the underlying premises of nonscientific expertise from judicial scrutiny. In addition, there was some authority that the foundation for the introduction of expert testimony did not require even a minimal showing that the expert had followed a proper test procedure. That laxity has not served the justice system well. Rather, that approach paved the way for the admission of "junk science." More to the point, there is now solid evidence that that lax approach has contributed to a number of wrongful convictions.

Yet, it would be a mistake to react in knee-jerk fashion to the revelations of flawed expert testimony. The shock of those revelations makes it tempting to leap to the polar extreme and adopt blunderbuss measures such as legislative bans on various types of expert testimony or announcing a categorical requirement that all expert testimony be validated by controlled scientific experimentation and induction. However, it is not at all clear that those measures would help. Quite to the contrary, it is likely that they would increase the courts’ dependence on lay witness testimony. Although expert testimony undeniably has its weaknesses, lay eyewitness evidence appears to be at least equally susceptible to error. In that light, the imposition of sweeping severe restrictions on expert testimony might well prove to be counterproductive. The result could easily be more rather than fewer inaccurate verdicts. The revelations in the past decade have hopefully disabused the courts from the misconception that forensic experts such as fingerprint analysts and questioned document examiners are infallible. However, it would be foolish to bar all expert testimony by fingerprint analysts and questioned document examiners.

The soundest—and the most challenging—approach is to devise reforms that improve the caliber of the expert testimony admitted without setting the admissibility hurdle too high. There are no magic bullets. However, ending the exemptions and mandating a showing of correct test procedure are sensible steps in the right direction. They hold the genuine promise of reducing the incidence of miscarriage of justice. It is not asking too much to expect both that every proponent of an expert theory or technique must demonstrate its reliability and that every expert will show that he or she followed sound scientific protocol. Indeed, it would be irresponsible to ask less. In this setting, the consequences of irresponsibility are both grave and predictable: wrongful convictions.

Return to Table of Contents - Spring 2003

Return to Criminal Justice magazine home page