When the principal, and perhaps sole, proof of guilt is a forensic test, such as DNA or fingerprint matching, there is a risk that credibility inflation or a "CSI" effect might unduly influence the jury.[FOOTNOTE 1] This is a cautionary tale about the dangers of allowing convictions to rest primarily on a single type of forensic identification evidence without meaningful corroboration.
New concepts like "bacterial profiles" (allowing identification from the bacteria left by a person's hands) and "forensic molecular photofitting" (generating an image from a dna sample
might one day join the growing catalog of forensic detecting and identification tools. And before more additions to this genre of evidence attain an air of infallibility, it is necessary to consider the limitations of current "gold standard" forensics as stand-alone evidence of guilt.
In studies of public reactions to DNA profiling compared with other forms of evidence, a group of scholars from different disciplines discovered inflation and overflow effects based on popular perceptions of genetic profile matching:[FOOTNOTE 3] "[A]fter damaging cross-examination testimony and jury instructions detailing how to prudently use scientific evidence testimony, jurors were still more likely to convict when DNA evidence existed compared to other types of evidence."
And this effect encompassed different kinds of forensic proof that fit the DNA mold: "[P]reconceived beliefs regarding DNA evidence may be so powerful that they affect the interpretation of other types of evidence."[FOOTNOTE 4]
In their analysis of these studies, the researchers concluded: "The strong and largely invariant impact of DNA evidence across experimental conditions suggests that this type of scientific evidence may be so persuasive that its mere introduction in a criminal case is sufficient to seriously impede defense challenges."[FOOTNOTE 5]
In other words, jurors' imaginations will do the heavy lifting to meet their expectations about presumptively unimpeachable evidence. At the same time, this phenomenon indirectly lowers or shifts the prosecution's burden of proof.
In a national survey of prosecutors about their use of DNA evidence, it was revealed that genetic matching was most often introduced in sexual assault cases, and a small percentage of murder prosecutions.[FOOTNOTE 6] Moreover, "[r]espondents reported few defense challenges to the use of DNA evidence."
Challenges occurred in only 8.5 percent of cases, and more than half of the offices indicated there had been none in the year preceding the survey. "On the whole, prosecutors did not experience frequent challenges to the use of DNA evidence, suggesting that the validity of the technology itself is not now contentious."
The absence of reported challenges to particular forensic tests creates a "CSI"-like aura that the evidence is unassailable.
Frye/Daubert is not a free pass for all purposes. Once a scientific method or forensic technique makes it through the door, due process requires that the application and interpretation of that analysis withstand examination in the crucible of the courts.
Super forensic tests, like DNA, need vetting that holds up to due process scrutiny.
For example, DNA profiles are dependent on careful collection and preservation of genetic material, scrupulous adherence to standards for examination and testing, and finally impartial analysis and comparison.[FOOTNOTE 7] And still there remains plenty of room for human errors born of subjectivity, questionable analysis,[FOOTNOTE 8] and potentially fraud that can contaminate the identification.[FOOTNOTE 9]
Until some kind of categorical or effective screening measures are enacted that will filter out unreliable scientifically interpreted evidence before presentation to a grand jury or trial court, special attention must be paid to the most seemingly persuasive kinds of evidence when it stands alone.[FOOTNOTE 10]
In the first published New York decision to address this issue, People v. Rush, 242 A.D.2d 108 (2nd Dep't 1998), a defendant charged with rape and robbery confronted two kinds of identification evidence: eyewitness and DNA.
The victim had picked the defendant from a photo array and a lineup soon after the event, but misidentified him at trial. Then there was the match between the defendant's DNA and genetic material from a semen sample obtained from the victim. The prosecution's expert concluded that the odds that another person would have had the same profile as defendant were one in 500 million.
And the chief issue on appeal: "DNA evidence cannot serve as the sole evidence supporting his conviction because it is circumstantial in nature and is not absolute or infallible."
The Appellate Division cited with approval the lower court's evaluation of this "strong evidence": "[t]he DNA evidence at the trial was the product of careful evaluation by a recognized expert in a prominent laboratory using a scientific technique to determine a statistical probability" (People v. Rush, 165 Misc. 2d 821, 825).
The appeals court held the DNA proof to be "highly probative," allowing the jury to conclude that the defendant was guilty beyond a reasonable doubt, despite the conflicting eyewitness identification evidence. In other words, a DNA match trumped a bad eyewitness identification. But what trumps a bad DNA match?
Notably, the prosecution's expert evidence was "unrebutted" and the claim that the "scientific expert improperly combined several racial databases in reaching his statistical probability conclusions" had not been preserved for appeal -- the Appellate Division indicated that it would not have credited the argument in any case.
1ST DEPARTMENT CASE
In 2005, the 1st Department in People v. Harrison, 22 A.D.3d 236, reached a similar conclusion where the "proof connecting defendant with the crime consisted almost entirely of DNA evidence." Like Rush there was also a legal preservation problem.
The appeals court only remarked that the evidence was "particularly powerful," citing the Court of Appeals decision in People v. Wesley, 83 N.Y.2d 417, initially recognizing the scientific acceptance of DNA profiling, and People v. Rush, 242 A.D.2d 108.
The defendant in People v. Taylor, 6 A.D.3d 556 (2nd Dep't 2004), indicted for murder, also failed to preserve a sufficiency of the evidence argument at trial, but the judge neglected to give a circumstantial evidence instruction. While the appeals court found that the proof of guilt was legally sufficient, it could not ignore its circumstantial nature.
Since the DNA identification in Taylor, along with the rest of the evidence, was all circumstantial, the failure to charge the jury was not harmless error. The conviction was reversed and a new trial ordered. This decision highlights the realization that while impressive, DNA is only one piece of indirect evidence.
Lastly, a federal magistrate judge in Harrison v. Walsh, 2007 U.S. Dist. LEXIS 39616 (S.D.N.Y. June 1, 2007), summarized the current view of New York courts in this way: "A state could require that there be 'corroboration' of DNA evidence by other evidence. ... New York, however, has not established such a requirement. Thus, under New York law, 'uncorroborated' DNA evidence -- just like a fingerprint or any other forensic evidence linking a defendant to the crime -- is sufficient to prove a defendant's guilt beyond a reasonable doubt."
Still, in cases where exclusive reliance on DNA has been claimed, there may have been some other kind of evidence serving as de facto corroboration or a failure to preserve the issue. In other words, the conclusion that DNA alone was sufficient to sustain convictions may not have been fully and fairly considered in a case free of any other supporting evidence. So the question of whether DNA by itself was enough to meet proof beyond a reasonable doubt may not have been squarely confronted.
The other concern is that the popular views of some forensic tests can over-impress a jury and bolster weak evidence that might have been insufficient on its own. In other words, the "CSI" effect could transform "gold standard" forensics into a tide that lifts all the boats.
DNA-heavy prosecutions have been scrutinized in other states revealing the shortcomings of stand-alone forensics as the principal evidence of guilt.
In State v. Toomes, 191 S.W.3d 122, 131 Fn. 4 (Tenn. Crim. App. 2005), the defendant's twin had to be eliminated as a source of genetic material. The appeals court reasoned that "reliable" forensic techniques, such as fingerprints, did not equal "absolutely infallible" evidence and a defendant's accessibility to an object where the identifying material was found had to be vetted of any innocent reasons for his fingerprints or genetic material being there.
Ultimately, the judges upheld the conviction (as it turned out his brother had been in jail at the time of the offense excluding him as a contributor) but footnoted this admonition: "By our holding, we are not announcing an iron-clad principal that DNA evidence, without corroboration, is always sufficient to support a conviction. Practically infinite factual variations can arise, and we do not intend by our opinion to prejudge other factual scenarios."
DNA matching continues to raise questions about the accuracy and reliability of identifications based on the choice of statistics used and the interpretation of the testing methods.[FOOTNOTE 11]
Treating this issue for the first time, the Colorado Court of Appeals in People v. Clark, 214 P.3d 531, 537-538 (Colo. 2009), dilated upon an important point raised in Toomes: "Human biology and common sense thus dictate that the evidentiary value of DNA obtained from a hair recovered in a public place is far different from that of DNA extracted from semen found in a private residence. Hence, where connection to a crime is established by DNA evidence alone, it is necessary to consider the type of biological material from which the crime scene DNA sample was obtained and its susceptibility to transfer by innocent means."
From this reasoning they concluded that corroboration was required: "Guided by the Ray standard [People v. Ray, 626 P.2d 167, 170-71 (Colo. 1981), describing limits of uncorroborated fingerprint evidence], we conclude that where, as here, the only direct evidence connecting an accused person to the crime is the presence of DNA at the scene of a crime, the evidence, to be legally sufficient to sustain a conviction, must be coupled with evidence of other circumstances tending to reasonably exclude the hypothesis that the DNA was deposited at a time other than that of the crime. Such other circumstances may include the source material of the DNA and its susceptibility to transfer, the location of the DNA, the character of the place or premises where it was found, the accessibility of that place or premises to the general public, and the object upon which the DNA was found."
Notably, the Supreme Court of Colorado granted certiorari (Clark v. People, 2009 Colo. LEXIS 785) on the question of whether stand-alone DNA identification without the presence of other "significant corroborative proof," was legally sufficient evidence of guilt.
The Tennessee and Colorado courts have recognized that there are many factual variations that can affect stand-alone DNA evidence, and by extension other forensic techniques and methods.[FOOTNOTE 12] However, DNA identification and other forensic tests are often treated as super-circumstantial evidence of guilt. And the mystique of "gold standard" proof continues to warp due process.
The reputation of a single impressive forensic test might cause decision makers to overlook mistakes and misconduct that could have tainted the results. DNA identification, like other forensic evidence, is only one element of a case and alone should not amount to proof beyond a reasonable doubt or be permitted to inflate weak evidence that might accompany it.
Ken Strutin is director of legal information services at the New York State Defenders Association.
FN1 See generally "Convictions Predicated on DNA Evidence Alone: How Reliable Evidence Became Infallible Citation," 38 Cumb. L. Rev. 313 (2007-2008); "DNA Identification Evidence in Criminal Prosecutions," Law Library Resource Exchange, March 7, 2010.
FN2 See, e.g., "An Individual's Unique Germs Could Be the Fingerprint of the Future," ABA Journal Law News Now, March 16, 2010; "To Sketch a Thief: Genes Draw Likeness of Suspects," Wall Street Journal, March 27, 2009.
FN3 See generally "Gold Versus Platinum: Do Jurors Recognize the Superiority and Limitations of DNA Evidence Compared to Other Types of Forensic Evidence?" 14 Psych. Pub. Pol. and L. 27 (2008).
FN4 Id. at 44.
FN5 Id. at 58.
FN6 See "Use of Forensic DNA Evidence in Prosecutors' Offices," 35 J. L. Med. & Ethics 310, 313-314 (2007).
FN7 See generally ABA Standards on DNA Evidence (3d ed. 2007).
FN8 See, e.g., FBI DNA Laboratory: A Review of Protocol and Practice Vulnerabilities (Office of the Inspector General 2004); "Art in the Science of DNA: A Layperson's Guide to the Subjectivity Inherent in Forensic DNA Typing," 58 Emory L.J. 489 (2008).
FN9 See "Authentication of Forensic DNA Samples," Forensic Sci. Int. Genet., Vol. 4, Issue 2, pp. 95-103 (Feb. 2010).
FN10 See, e.g., "U.S. Judge Urges Skepticism on Forensic Evidence," Boston Globe, March 29, 2010.
FN11 See, e.g., R. v. Watters  EWCA Crim 89; see generally DNA's Identity Crisis, Nature, March 18, 2010, at 347.