|
At the conclusion of the Daubert-Kumho Tire hearing, the court decided that Dr. Saks' arguments did not properly apply to the evidence presented in this case and, in general, was contradicted in many respects by the Government's experts whom the court found more persuasive. The court concluded that the opinion evidence offered by the Government's experts met the requirements for admissibility postulated in the United States Supreme Court opinions. At the trial, the defendant did not call an expert to challenge the Government expert's conclusions. Because the opinion of the District Judge examines concisely the various approaches courts have taken when dealing with the issue of admitting expert opinion evidence in cases of disputed authorship, we print the edited version of the court's decision below, omitting the introductory legal discussions about the state of the law on admissibility of handwriting identification prior to and through the Daubert and Kumho Tire cases, since these cases are discussed elsewhere in the website. UNITED STATES v. PRIME On October 9, 2001, Michael S. Prime ("Prime") moved in limine to exclude expert testimony on handwriting identification at his trial or, in the alternative, for a hearing to determine the admissibility of such evidence pursuant to Daubert v. Merrell Dow Pharmaceuticals and Kumho Tire v. Carmichael. Prime's motion brought into issue the testimony of Kathleen Storer ("Storer"), a forensic document examiner ("FDE" or "examiner") working for the United States Secret Service in Washington, D.C. Storer was to testify for the government that, in her opinion, Prime's handwriting appeared on counterfeit money orders and other documents. In its response to Prime's motion, the United States insisted that expert testimony regarding handwriting analysis met the Daubert test and that no hearing was necessary. The Court held a Daubert hearing on March 18, 2002, and issued an order denying Prime's motion on April 3, 2002. This ruling provides the reasoning behind the Court's conclusion that Storer's testimony was properly included at trial. I. THE APPLICABLE STANDARD [The discussion of Frye v. United States (1923), Daubert v. Merrill Dow Pharmaceuticals (1993) and Kumho Tire v. Carmichael (1999) is omitted. Editor.] II. STORER'S ANALYSIS In the case before the Court, the questioned documents comprised 76 exhibits such as envelopes, postal forms, money orders, post-it notes, express mail labels and postal box applications. The "known" handwriting came from three suspects in the case: 114 pages from David Hiestand ("Hiestand"), 14 pages from Jeff Hardy ("Hardy") and 112 pages from Prime. In Storer's opinion, Hiestand wrote portions of eight documents, Hardy wrote portions of one of the questioned documents and Prime wrote portions of 45 documents. These opinions ranked the highest in the nine-point scale (i.e.identification). As to portions of 14 questioned documents, Storer opined that Prime was the "probable" writer. As to portions of two exhibits, Storer rendered an "indications" opinion, i.e., that there were "few features which are of significance for handwriting comparison purposes...." As to 38 signatures, Storer rendered a "could not be determined" opinion and the remainder of the material had "no evidence of significance" according to her. In December 2001, Storer issued a second report based on additional questioned material (two brown envelopes with hand printing). She was of the opinion that Prime was the writer of the material on the new exhibits. III. COURT TREATMENT OF HANDWRITING IDENTIFICATION. By 1960, Ninth Circuit noted that it was "well settled" that "an expert in handwriting may testify and state his opinion as to whether different documents or signatures were written by the same person or are similar or dissimilar in respect of handwriting or whether a particular handwriting is genuine or disguised." Robles v. United States, 279 F.2d 401, 404-05 (9th Cir.1960). The world appears to have changed with Daubert, after which district courts began to cast a suspicious eye at the discipline of forensic document analysis. After Daubert, but prior to Kumho Tire, district courts had the option of analyzing handwriting comparison testimony under two alternative strands: They could either look at the area of forensic document examination as being grounded in scientific knowledge and apply Daubert, or treat it as nonscientific expert testimony, i.e., falling under the "technical, or other specialized knowledge" prong of Fed.R.Evid. 702. Analysis under both approaches was conducted in United States v. Starzecpyzel, 880 F.Supp. 1027 (S.D.N.Y.1995). After a Daubert hearing in which Judge McKenna himself occasionally questioned the examiner, the court treated the testimony as science-based and then applied the Daubert factors. Noting that two ongoing studies were being conducted by the U.S. Postal Laboratory and the Immigration and the Naturalization Service ("INS") Laboratory, the court ruled, first, that the discipline of handwriting analysis was "amenable to testing." Second, the court found that the field had not, in fact, actually been subjected to rigorous testing and, therefore, the error rate as to any conclusions testified to by examiners was unknown. As to the third Daubert factor, the court noted that FDEs publish in several journals. However, the court found the articles "to be significantly different from scholarly articles in such fields as medicine or physics, in their lack of critical scholarship." Finally, the court found that the field did indeed receive general acceptance within the community of examiners and the legal community, but not from "financially disinterested parties, such as academics." In sum, the court concluded that "forensic document examination, despite the existence of a certification program, professional journals and other trappings of science, cannot, after Daubert, be regarded as 'scientific ... knowledge.' " (citing Fed.R.Evid. 702). However, this did not result in an automatic ruling of inadmissibility: The court went on to state that Daubert did not apply to forensic document examiner testimony. The court ruled that such testimony was not based on science but on "technical, or other specialized knowledge." After outlining what FDEs actually do, the Court held that forensic document examiner testimony was admissible, largely on the grounds that (1) the jury could visually confirm the first part of an FDE's analysis in which the examiner identifies significant similarities and differences between genuine and challenged handwriting examples and (2) the other, unverifiable portion of analysis, in which the examiner draws inferences, was dependent on the first part, and the testimony was, in any event, subject to cross examination. After Kumho Tire all expert testimony, whether based on science or not, is subjected to the Daubert screen. Circuit courts, admonished by the Supreme Court to review a district court's decision deferentially, generally have upheld district courts' decisions. See, e.g., . . . United States v. Paul, 175 F.3d 906, 912 (11th Cir.1999) (upholding district court's decision to exclude testimony of law professor critical of forensic document examination). But see United States v. Velasquez, 64 F.3d 844, 852 (3rd Cir.1995) (overturning decision to exclude testimony of same law professor). Among district courts, handwriting comparison testimony has fared unevenly since Kumho Tire. Much of the evidence presented to the courts is the same (and indeed, mirrors that presented to this Court). Yet, after applying Daubert, courts have reached varying conclusions as to the reliability of such testimony. In United States v. Gricco, 2002 WL 746037 (E.D.Pa.2002) (issued after the Daubert hearing in the case before the Court) the court found that testimony of an expert's opinion that there was a handwriting match between the defendant's exemplars and two government exhibits, including a handwritten list of materials allegedly used in manufacturing methamphetamine and a handwritten list of alleged laboratory supplies, was "sufficiently reliable for purposes of Rule 702." The court applied factors elucidated by the Third Circuit, In re Paoli Railroad Yard PCB Litigation, 35 F.3d 717, 742 n. 8 (3d Cir.1994), which incorporate the Daubert analysis, and found that the case for admissibility was clear. In reaching its conclusion that handwriting analysis was "based on valid reasoning and reliable methodology," the court noted the pedigree of such evidence in courtrooms across the country that had been established under the approving eye of the circuit courts. In contrast, in U.S. v. Saelee, 162 F.Supp.2d 1097 (D.Alaska 2001), the trial court ruled such evidence inadmissible. The questioned writing in the case involved address labels on packages, which the court presumed "would be considered a very small quantity of printing" and the defendant was an Asian whose first language was not known. In that case, the government sought to have its expert testify only about the similarities and differences between the known writing and the questioned documents and not to have the expert testify about his ultimate conclusions as to whether defendant authored the questioned documents. Acknowledging that it was taking one step further than other courts, the court concluded that after applying Daubert that the expert testimony "is as likely to mislead a jury as to assist it in determining the facts of this case. It is therefore excluded entirely." In United States v. Rutherford, 104 F.Supp.2d 1190 (D.Neb.2000), the court found it helpful to break down the expert testimony into two components: (1) The comparison of the stylistics of the defendant's handwriting with the handwritings on the questioned documents and (2) conclusions that the defendant was the author of a signature or other writing on a questioned document. The former was not challenged by the defendant. The court found that the latter did not meet Daubert/Kumho requirements and was therefore excluded. Similarly, in United States v. Santillan, 1999 WL 1201765 (N.D.Cal.1999), the district court limited the testimony to the mechanics and characteristics of handwriting and testimony as to comparison of similarities between defendant's known handwriting and the handwriting on the questioned documents, and barred any testimony concerning the expert's belief that the handwriting on the questioned documents is in fact handwriting of the defendant. . . . Trial court rulings reach one of three conclusions: (1) exclusion of all forms of expert testimony on handwriting comparison; (2) inclusion of the testimony on similarities and differences but exclusion of the expert's conclusions; and (3) and inclusion of comparison and expert testimony. IV. APPLICATION OF DAUBERT/KUMHO TIRE IN THE CASE BEFORE THE COURT The heart of Prime's challenge, however, goes to the government's claim that handwriting and handprinting identification testimony meets the requirements of Daubert and Kumho Tire. Before the Court applies the Daubert factors to assess the admissibility of Storer's testimony, a few general observations are in order. First: Daubert and Kumho Tire were opinions issued in response to the increasing efforts to introduce novel theories in civil trials. By issuing these landmark opinions, the Supreme Court attempted to strike a balance between wholesale exclusion of most forms of non scientific expert testimony, even that based on sound principles, versus liberal inclusion of such testimony, including that based on untested theories of highly dubious merit. Therefore, in Daubert, the Supreme Court overturned the trial court's exclusion of expert evidence based on the general acceptance test, but in Kumho Tire, upheld the exclusion of testimony based on a tire failure expert's obviously flawed methodology. These rulings have been used by some trial courts to exclude not just novel theories, but also time-tested techniques used almost universally by law enforcement, such as fingerprint and handwriting analysis. The Court believes that outright exclusion of such evidence is a mistake. While the Court agrees that the Daubert analysis needs to be applied to all expert testimony, the test, depends upon the particular circumstances of the particular case at issue. Where a novel theory is presented to a court, it makes sense to demand proof of statistically significant results and strict compliance with scientific methods. However, where a technique has been repeatedly applied and tested by law enforcement and by courts for over a century, the Court does not believe that the absence of scientific data, without more, should be the death knell for such testimony. In United States v. Llera Plaza, 188 F.Supp.2d 549 (E.D.Pa.2002) Judge Louis H. Pollak reached a similar conclusion in an analogous situation where the testimony at stake was fingerprint identification based on a regimen known as "ACE-V." Judge Pollak acknowledged that such testimony did not satisfy all the Daubert factors. Nevertheless, he concluded that such testimony was admissible based largely on its historical acceptance by law enforcement and by English and American courts. Second: The Court also believes that the Daubert inquiry is not intended to ask the "larger question" regarding the reliability of a particular technique in general. Rather, the inquiry is case-specific. In other words, all applications of handwriting identification testimony are not at issue in the motion before the Court. Rather, the Court will evaluate the reliability of handwriting expert testimony within the confines of the facts of this case. When the focus of the analysis shifts from the general to the specific, the number and nature of samples of questioned and known documents become important, as does the particular experience and training of the expert and the soundness of the method actually employed. This narrowed look does not, of course, detract from the standards of intellectual rigor to be demanded from the testimony; however, the Court believes that the Supreme Court adopted a sound approach when it stated in Kumho Tire that the "specific issue before the court was not the reasonableness in general of a tire expert's use of a visual and tactile inspection to [reach his conclusion]. Rather, it was the reasonableness of using such an approach, along with Carlson's particular method of analyzing the data thereby obtained, to draw a conclusion regarding the particular matter to which the expert testimony was directly relevant." The Supreme Court made clear, therefore, that even in the context of evaluating expert testimony, the focus of a trial court should be on the facts of the dispute before it rather than on theoretical issues that may be relevant to discussions in other contexts. With these principles in mind, the Court will proceed to apply the Daubert factors to the expert testimony at issue. The most important line of inquiry under Daubert for present purposes--and one that appears to have foiled the testimony of many a handwriting expert in other district courts--is whether the theory and technique of handwriting identification has been, or is capable of being, tested and whether handwriting identification has an acceptable error rate. The Court understands this inquiry to include the following prongs: (a) whether the premises of handwriting identification are sound given the facts of this case; i.e., given the number of known samples of handwriting in this case, whether characteristics individual to Prime can be identified and used by a trained forensic examiner such as Storer; and (b) whether the work of professional forensic document examiners, such as Storer, can be, and has been, tested for accuracy and proficiency and what the error rates are in such tests. The government and Storer have cited to several different sources as support for the proposition that handwriting is unique. Among the evidence were results from research conducted by the Center of Excellence for Document Analysis and Recognition ("CEDAR") at the State University of New York at Buffalo. The CEDAR research results were accepted for publication in the Journal of Forensic Sciences prior to the Daubert hearing. (Sargur N. Srihari ("Srihari"), Sung- Hyuk Cha, Hina Arora and Sangjik Lee, Individuality of Handwriting, J. Forensic Sci (approved for publication May, 2002).) Professor Srihari testified about his research at the Daubert hearing. (3/18/2002 Transcript of Daubert Hearing.) His project was undertaken with the purpose of testing the hypotheses underlying forensic document examination. A portion of Professor Srihari's study looked at the process of examining two writing samples and determining whether they were written by the same or different writer. Professor Srihari and his colleagues extracted features of handwriting--both "macro" features such as slant, and "micro" features such as the presence of openings in characters--and utilized these to compare documents using computer software. According to Professor Srihari, he obtained a 96 percent accuracy rate within his sample, which was "statistically inferable over the entire population." His conclusion was that "handwriting is individualistic." The government also refers to databases maintained by the Secret Service and the German law enforcement agency, Bundeskriminalamt ("BKA"). These so- called "Forensic Information System for Handwriting" ("FISH") databases of letters convert handwriting features into mathematical algorithms. The government claims that of the 90,000 writers in the German database, "the system has determined that no two writers write alike, nor do they share the same combination of handwriting characteristics." The same conclusion can be reached from an analysis of the Secret Service's slightly smaller database of 9,000 writers, according to the government. In addition, Storer refers in her affidavit to studies which show that the handwriting of twins can be distinguished. She cites numerous articles published in forensic science journals that conclude that handwriting is a distinguishable, individual trait. She also testified that her own personal experience showed that "every writer does have their [sic] own combination of individual characteristics." As the Court already has noted, it need not address the reliability of handwriting evidence generally. However, within the confines of this case, the Court has no trouble concluding that the premises of handwriting identification are sound. Storer states she received as many as 112 pages containing specimen writing from Prime, samples that the defense itself characterizes as "extensive." Storer's training credentials are, furthermore, impeccable: She received a Master of Forensic Science degree from George Washington University in 1988. From 1989 to the present she has been employed as an FDE with the Secret Service in Washington, D.C. At the Secret Service, she underwent a three-year apprenticeship or training program in document examination leading up to a certification on July 1, 1992. The training program in the Secret Service involves writing 18 research papers in the area of document examination and presenting them to peers for discussion. It also entails working alongside senior examiners who impart knowledge of their craft to the apprentices. Storer continues to take internal proficiency tests twice a year. Storer testified that she had never failed any of the in-house tests she was required to take by the Secret Service. In 1997, she was certified by the certifying body for forensic document examiners, the American Board of Forensic Document Examiners. The certification process included a three-part test: a practical, written and an oral test. With this extensive level of training by the examiner and the array of available writing samples, the Court has no trouble concluding that unique characteristics of Prime's handwriting may be established. [At this point in the opinion, the judge included a lengthy footnote (footnote 5) that it may be relevant to print: "The Court's conclusion is supported in general by the results of Professor Srihari's research. However, the Court acknowledges the limitations on the inferences that can be drawn from the study. Even if handwriting is individualistic when the examination is conducted by a computer, this does not necessarily establish that it will be so when subjected to human examination. Moreover, the evidence provided of the BKA and Secret Service databases, while marginally probative, appears to beg the question it is presented to answer, as pointed out by Professor Michael J. Saks ("Saks"), professor of law and of psychology at Arizona State University, who testified at the Daubert hearing for the defendant. The uniqueness or individuality of handwriting cannot be established simply by stating that different writers generate different algorithms. It is clear that individuality is an attribute that depends on the criteria used to judge the writing's characteristics: the more thorough the examination, the more likely that writings will appear unique (even if written by the same person). Dealing with the uniqueness question in a particular situation requires determining the criteria used to determine uniqueness, whether such criteria are reliable and whether these criteria were in fact applied in the case. As far as the proficiency and accuracy of FDEs are concerned, directly relevant to these issues are studies conducted in the 1990s by Professor Moshe Kam of the Electrical and Computer Engineering Department at Drexel University. In his studies, Professor Kam compared the performance of professional forensic document examiners with non-professionals in matching handwriting. Professor Kam testified in court that the first of his studies showed that lay persons made far more types of errors than professional examiners. The second study showed that as a group, examiners' performance was different from that of lay persons: Lay persons rivaled professional examiners in being able to select different documents written by one person. However, lay persons also claimed erroneously that documents written by different people had the handwriting of the same person 38 percent of the time, whereas experts made the same mistake 6.5 percent of the time. As Professor Kam stated: "It struck me very quickly that lay persons tend to see similarities and jump to a conclusion ... whereas document examiners always started the analysis--when I asked why did you make the decision--by trying to show me [sic] what's different." The third study showed that a different incentive scheme did not make a difference in the results; it apparently also showed an unexplained improvement in the ability of laypersons to avoid false positives. The fourth study showed that professionals and laypersons did not differ significantly in detecting forgeries, but professionals were better at finding genuine signatures. Professionals erroneously concluded that forgeries were genuine 0.5 percent of the time whereas lay persons did so 6.5 percent of the time; professionals mistakenly concluded that genuine signatures were forgeries 7.1 percent of the times, lay persons did so 26.1 percent of the time. The Kam studies indicate that handwriting identification is not error- free; however, the differences in error rates and results between professionals versus laypersons show that the field is one that is amenable to developing an expertise and that, with proper training, professionals can improve their accuracy. For the purposes of this case, the Court considers the expertise and testimony of Storer to be adequately tested. Further scientific testing on handwriting comparison and identification would undoubtedly aid in gauging the field's legitimacy; however, as a legal matter, the field has been sufficiently tested by its long-established use and the research that already has been conducted. Daubert does not require more. The test of admissibility is not whether a particular scientific opinion has the best foundation or whether it is demonstrably correct. Rather, the test is whether the particular opinion is based on valid reasoning and reliable methodology. The Saelee court's problem with the Kam studies was that they did not conclusively establish that forensic document examiners can reliably do what they say they can do." However, the context of the Saelee court's ruling was entirely different: As already noted, the court was dealing with a writer whose native language was not known and with a small quantity of questioned documents. The Saelee court specifically noted that: In any event, the Court disagrees with the Saelee court's apparent assumption that Daubert requires that the reliability of a process or technique be established conclusively. As noted by the Supreme Court in Daubert itself: "It would be unreasonable to conclude that the subject of scientific testimony must be 'known' to a certainty; arguably, there are no certainties in science." To the extent that there are gaps in the research--and there are- they need to be filled. The Court encourages the profession to respond forthrightly to Professor Saks' criticism and urges Professor Kam to reveal his data for the purpose of re analysis. However, the fact that additional research can be done does not mean that FDE testimony should now be inadmissible on the ground that it has not been adequately tested. Such a ruling "would be to make the best the enemy of the good." Llera Plaza, 188 F.Supp.2d at 572. The other Daubert factors also are satisfied in the case before the Court. It is clear to the Court that the forensic sciences, including document examination, are subject to extensive peer review. Storer testified at the Daubert hearing that journals publishing articles in this area include: The Journal of Forensic Sciences, the Journal of the American Society of Questioned Document Examiners, the International Journal of Forensic Document Examiners, the Canadian Society of Forensic Science Journal, the Journal of Forensic Identification and Forensic Science International. Articles sent for publication in the Journal of Forensic Science are reviewed not just by handwriting experts but by others in the forensic sciences community. Even if this form of peer review is not conducted by academics, it does not mean that it is devoid of utility. As Professor Kam's testimony shows, forensic document examiners have a legitimate expertise based on years of experience and training. Their review of articles submitted for publication provides oversight on research in the field. The Court agrees with Judge Pollak in Llera Plaza that just because peer review is not conducted by scientists, this need not "militate against the utility of the identification procedures...." Furthermore, at least in the case of Secret Service examiners, the process of document identification goes through an "internal" peer review as well, since every document reviewed by such examiners is subject to a second, independent examination. Finally, it cannot be ignored that handwriting evidence has been tested and reviewed in the courtroom for decades. This usage itself provides some assurance of reliability. Cf. United States v. Havvard, 117 F.Supp.2d 848, 854 (S.D.Ind.2000) (noting that "latent fingerprint identification has been subject to adversarial testing for roughly 100 years," a "track record [which] provides far greater assurance of reliability than, for example, publication of one peer-reviewed article describing a novel theory about the cause of a particular disease at issue in a civil lawsuit.") Storer's testimony also showed that the field of document examination is moving toward establishing "standards controlling the technique's operation." For one, the Secret Service laboratory where she works has maintained its accreditation with the American Society of Crime Laboratory Directors since 1998. This accreditation process requires an annual external proficiency test. Further, the standard nine-point scale for expressing opinions by the FDEs was established under the auspices of the American Standards and Testing Organization ("ASTM"). Perhaps in response to the enhanced scrutiny the field was receiving, a working group was formed in 1997 by the industry in order to standardize many of the processes utilized. Standards already established by ASTM include the terminology used in the profession, and the practice for receiving, documenting, storing and retrieving evidence in a laboratory. According to Storer, eight proposed guidelines are undergoing peer review. One of the standards that is being formalized is the comparison process itself. Under these circumstances, the Court finds that forensic document examination is making strides toward standardization. The fact that the document examination process has not been completely standardized is not necessarily a bar to admissibility in court. Not all expert testimony must be backed up by a standard procedure. Moreover, if a fact-finder is fully apprized of the process that is actually followed, and the expert is subject to cross-examination and to being countered by other experts, the lack of standardization can hardly be said to require exclusion. Finally, it is clear to the Court that handwriting analysis has received broad acceptance. Law enforcement agencies such as Interpol, Scotland Yard, the Central Intelligence Agency, the Federal Bureau of Investigation and the United States Postal Inspection service use handwriting analysis. In addition, Storer listed 15 universities in the United States that offer Masters degrees in forensic science with courses that include document examination. As has already been noted, handwriting analysis has long been used in American courts. Even after Daubert and Kumho Tire, most district courts have admitted such evidence, albeit with limitations. Therefore, the general acceptance prong of Daubert is satisfied. In sum, the Court is persuaded of the reliability of Storer's testimony; it was properly admitted and presented to the jury at trial. The Court acknowledges that had it required results of extensive scientific testing as exists in other fields, forensic document examination would come up short at the present time. However, the Daubert hearing made it very clear that the profession is in the process of making giant strides toward objective testing and standardization. The question before the Court, then, is whether in the interim period in which complete data are not available, the Court should exclude all FDE testimony as inadmissible. The Court is persuaded that, under Daubert, such testimony, including conclusions based on examinations, is reliable and admissible. Prime can present his own expert to dispute Storer's findings and/or to attack the entire field of forensic document examination as illegitimate. However, the apparent recent trend to exclude FDE testimony is a result, the Court believes, of an excessively-rigid application of Daubert. [Emphasis by Editor.] Since Daubert applies in both criminal and civil cases, such an approach may, one day, result in unfortunate consequences for a criminal defendant who is denied the ability to present the best evidence that he did not author an extortion demand or pen a forged signature. The Court declines to follow this trend on the record before it. V. CONCLUSION
|
|
Additional Articles in Identification Evidence.......
Friction Ridge Evidence: Creating A Record on Critical Fingerprint “Scholarship”? New 06/16/07 Handwriting and Forensic Document Examination: Palmprint and Handwriting I.D. Satisfy Daubert Rule Bite Mark Identification: Man Convicted on Erroneous Bite Mark Identification Evidence Finally Free Firearm and Toolmark Evidence: Toolmark Identification Received A (Frye-Daubert) Body Blow In Florida Lip Prints, Ear Prints, and Other Less Well-known Marks: Alphonse Bertillon and Ear Prints Miscellaneous Identification and Biometric Evidence: Dog Scent Evidence...Is it Scientific? |