UNITED STATES DISTRICT COURT FOR THE WESTERN DISTRICT OF NEW YORK
May 1, 1996
UNITED STATES OF AMERICA, Plaintiff,
STEVEN JORDAN, a/k/a "BOOBA", Defendant.
The opinion of the court was delivered by: LARIMER
DECISION AND ORDER
In this criminal case defendant Steven Jordan is charged with armed bank robbery, bank robbery and bank larceny (in violation of 18 U.S.C. § 2113(a),(b), and (d)), as well as using or carrying a firearm (in violation of 18 U.S.C. § 924(c)). The Government represented that, at trial, it would rely in significant part on the testimony of victim bank teller Gayle McKinley. McKinley identified the defendant first from a photo array 47 days after the robbery and again at a preliminary hearing almost 2 years later. It was anticipated that McKinley would make an in-court identification at trial.
The Government also indicated that two alleged accomplices would testify, no physical evidence existed (fingerprints, gun, etc.), and that the Government did not intend to introduce any surveillance photographs.
In his defense, defendant seeks to have admitted the testimony of Dr. Kathleen Chen, an "expert in the area of human memory and perception," who will opine on the reliability and accuracy of eyewitness identification. Presently before me is the Government's motion in limine to exclude this expert testimony. The parties have submitted memoranda on the issue and the Court has conducted a hearing to allow both sides to conduct voir dire of the witness.
In the past, federal trial courts typically have been unwilling to admit the testimony of "eyewitness identification" experts. In United States v. Serna, 799 F.2d 842, 850 (2d Cir. 1986), the Second Circuit upheld the district court's exclusion of eyewitness identification expert testimony in a drug case where an informant identified the defendant from 2 photo arrays. Without extensive discussion, the Second Circuit noted that the voir dire of the expert was not clearly erroneous; the judge had broad discretion under Federal Rules of Evidence ("FRE") 702 and 403; the expert was ignorant of the conditions under which the informant identified the defendant; the expert's testimony consisted primarily of generalized pronouncements about the lack of reliability of eyewitness identification, particularly cross-racial identification; and the expert admitted that many of his conclusions coincided with common sense. The Circuit concluded that this expert's testimony would have done nothing but "muddy the waters."
More recently, federal courts have become more receptive to eyewitness identification expert testimony and, while the Second Circuit has not spoken on this issue since Serna, a number of other Circuits have expressly stated that, under the right circumstances, such testimony should be admitted. See, e.g., United States v. Brien, 59 F.3d 274, 277 (1st Cir.) (noting increased acceptance of eyewitness identification experts because "there is more expert literature on the subject, more experts pressing to testify, and possibly more skepticism about the reliability of eyewitnesses"), cert. denied, 133 L. Ed. 2d 320, 116 S. Ct. 401 (1995); United States v. Rincon, 28 F.3d 921, 926 (9th Cir.) (noting that where the eyewitness identification expert's opinion is based upon scientific knowledge which is both reliable and helpful to the jury in a given case, then it should be admitted), cert. denied, 130 L. Ed. 2d 516, 115 S. Ct. 605 (1994).
THE LEGAL STANDARD
In order to be admitted in federal court, any proposed expert testimony, including eyewitness identification testimony, must meet the standards set forth in FRE 702. See Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d 469 (1993). As explained in Daubert, scientific expert testimony is admissible if (1) it is based upon "scientific knowledge" and (2) it will "assist the trier of fact to understand or determine a fact in issue." Daubert, supra, at 2796. "This entails a preliminary assessment [by the trial judge] of whether the reasoning or methodology underlying the testimony is scientifically valid and of whether that reasoning or methodology properly can be applied to the facts in issue." Id.
In Daubert, the Supreme Court provided district courts with a list of "general observations" to assist them in determining whether a theory or technique is based upon "scientific knowledge."
-- Has the theory or technique been tested? Or can it be?
-- Has the theory or technique been subjected to peer review and publication? (relevant but not dispositive)
-- In the case of a technique, what is the known or potential rate of error?
-- "General acceptance" -- though not required, is certainly useful where it exists. Id. at 2796-97.
The inquiry is supposed to be a flexible one, focusing on the "scientific validity -- and thus the evidentiary relevance and reliability -- of the principles that underlie a proposed submission. The focus ... must be solely on principles and methodology, not on ... conclusions ..." Id. at 2797.
Thus, before admitting scientific expert testimony, federal trial courts must carefully consider the proffered testimony to ensure that it "rests on a reliable foundation and is relevant to the task at hand." Id. at 2799.
Decisions from other Circuits offer guidance on applying the FRE 702 standards in the context of expert testimony concerning eye witness identification. In United States v. Brien, supra, the First Circuit upheld the district court's exclusion of eyewitness identification expert testimony because the proffer of evidence did not meet the first Daubert prong -- i.e., establishing that the testimony was based upon scientific knowledge. Brien was an armed bank robbery case. Defendant was identified by 3 different tellers, 2 of whom were directly confronted by the robber. Defendant was also identified by 2 motel employees as having stayed at the motel for several days immediately after the robbery.
The defendant's first proffer of expert testimony (a three page statement) was rejected as being too general and lacking the foundational requirements of Daubert. In the second proffer (an 8 page affidavit containing about 11 paragraphs of substantive opinion), the expert stated that he would not opine as to whether a specific witness' identification was accurate, but he would discuss the elements of memory (observation, retention and retrieval) and the factors that undermine memory, such as stress, delay, and improper reinforcement through police questioning. This proffer too was rejected as being too general and lacking foundational requirements, and the testimony was excluded.
The First Circuit declined to adopt a blanket rule of exclusion or inclusion, noting that the determination was within the considerable discretion of the trial court and that such matters should be considered on a case by case basis.
Obvious concerns are the reliability and helpfulness of the proposed expert testimony, the importance and the quality of the eyewitness evidence it addresses, and any threat of confusion, misleading of the jury, or unnecessary delay. United States v. Brien, supra, at 277.
The First Circuit expressly noted that given an appropriate proffer, a trial court might well decide to admit some or all of the testimony of this nature. But in this case, the proffer was so lacking in detail about the bases for the testimony that the district court's decision was correct.
Similarly, in United States v. Rincon, supra, the expert's testimony was excluded because the proffer did not adequately identify the bases for the expert's conclusions. Rincon was an unarmed bank robbery case. The state's case consisted primarily of eyewitness identification (there were several eyewitnesses to two separate robberies) and one surveillance photo. The defendant's original proffer of expert testimony indicated that the expert would testify about the three phases of eyewitness identification including perceiving and encoding, storage and retention, and retrieval. The expert would also testify as to the effect of various factors on each phase, including stress, state of mind, suddenness, suggestibility, and cross-ethnic identifications. The expert would offer no opinion concerning the reliability or certainty of the witnesses in that particular case.
While the attorney's declaration submitted in support of the proffer stated repeatedly that there is a "wealth of research" on the issue, none of it was submitted or described so that the district court could determine if the studies were indeed scientific: i.e., "whether the reasoning or methodology underlying the testimony is scientifically valid..." See Daubert, supra. Thus, the district court excluded the expert testimony on eyewitness identification. The Ninth Circuit affirmed.
Following Daubert, the Supreme Court remanded the case for further inquiry. On remand, the defendant supplemented the proffer with one background article, containing the views of 63 experts on various topics related to eyewitness testimony. However, the article only identified the topics; it did not discuss the research in any detail. Thus, the district court again excluded the expert's testimony because the court could not determine if the research was scientifically valid.
The Ninth Circuit again affirmed, noting that the expert's proffer failed to meet both prongs of the Daubert standard: (1) it was not shown to be based upon "scientific knowledge" and (2) it would not assist the jury. Specifically, the Circuit noted that, per Daubert, a district court must assess whether the reasoning and/or methodology underlying expert opinions is scientifically valid. In that case, the proffer failed to provide the necessary information. Moreover, the Circuit found that the testimony would not assist the jury. Indeed, the Circuit found that it would confuse the jury and thus the testimony was also inadmissible under FRE 403. See also United States v. Harris, 995 F.2d 532 (4th Cir. 1993) (affirming exclusion of eyewitness identification expert testimony where it was deemed irrelevant to the facts of the case: i.e., the identification was not cross-racial, there was no delay in the identification, there was no stress associated with the original identification); United States v. Stevens, 935 F.2d 1380 (3d Cir. 1991) (reversing the district court's decision, in part, and admitting eyewitness identification expert testimony on the subject of the correlation between confidence and accuracy, because this area is one not easily understood by jurors and, thus, appropriate for expert amplification).
As these cases demonstrate, it is not enough that a proffer of expert testimony include the substantive areas of the proposed testimony. A proffer must also include sufficient information regarding the bases for the expert's opinion. Also, the testimony must be relevant to the facts of the particular case. Finally, the testimony must assist the jury.
Other Circuits offer additional guidance. The 5th, 6th and 8th Circuits have found that there is no abuse of discretion where trial courts exclude eyewitness identification expert testimony in cases where the eyewitness identification is only a small part of the Government's case. These cases suggest that where a prosecutor relies wholly on eyewitness identifications, such expert testimony should probably be admitted. See United States v. Blade, 811 F.2d 461 (8th Cir.) (finding no abuse of discretion in decision to exclude expert testimony where Government's case relied on more than only the eyewitness identification), cert. denied, 484 U.S. 839, 98 L. Ed. 2d 82, 108 S. Ct. 124 (1987); United States v. Moore, 786 F.2d 1308 (finding no abuse of discretion in decision to exclude expert testimony where eyewitness identification only a small part of Government's case but noting that admission would be proper where such identification is the only evidence), reh'g denied, 791 F.2d 928 (5th Cir. 1986); United States v. Smith, 736 F.2d 1103 (6th Cir.) (finding harmless error in decision to exclude expert testimony where there were 3 separate eyewitnesses and a definitive palm print), cert. denied, 469 U.S. 868 (1984); see also United States v. Alexander, 816 F.2d 164 (5th Cir. 1987) (reversing as clearly erroneous decision to exclude expert testimony on photographic comparison where entire case turned on photographic identification).
THE PROFFER PRESENTLY BEFORE ME
In the case before me, defendant seeks to have admitted the expert testimony of Dr. Kathleen Chen. Defendant submitted a proffer consisting of his attorney's affidavit (setting forth the areas of proposed testimony), Dr. Chen's curriculum vitae, and portions of trial transcripts from two state court proceedings at which Dr. Chen recently testified.
In addition to this written proffer, a voir dire hearing was conducted of Dr. Chen for the purpose of further developing the foundation for her testimony.
Based upon the written proffer and the voir dire hearing, I find that Dr. Chen's testimony is (1) based upon scientific knowledge and (2) would be helpful to the jury. See Daubert, supra; and FRE 702.
Dr. Chen has an extensive curriculum vitae. She has a Masters Degree and Ph.D. in Experimental Psychology and has been a professor at the Rochester Institute for Technology ("RIT") for 24 years. She is currently the Chairman of the RIT Psychology Department and has taught at both the undergraduate and graduate level. She is stated to be familiar with all of the recognized publications, reports and studies in the field of human memory, perception and recollection. She has conducted replications of others' studies on the correlation between confidence and accuracy of eyewitness identification.
The specific areas of Dr. Chen's proffered testimony are described as follows:
- Mental Processes Involving Recollection -- there is a three stage process by which identification occurs: (1) sensory perception; (2) transfer of sensory perception into short-term memory; (3) transfer from short-term memory into long-term memory for later recall.
- Selective Perception -- the human mind will select certain items upon which to focus.
- Stress -- stress impacts a witness' sensory perception and/or subsequent recollection.
- Retention Interval -- the time from the original observation to the subsequent identification is a factor in the accuracy of eyewitness identification.
- Level of Confidence in Accuracy -- there is no correlation between a witness' confidence and the actual accuracy of the identification.
- Head-Shoulders Photo vs. Corporal Viewing -- viewing a photographic array of only the head and shoulders is not as effective as a full body viewing.
- Post-Event Suggestion -- events following the actual incident tend to be incorporated into the witness' memory/recollection.
- Perception Activity -- the activity in which a witness engages during the event will have an effect on the accuracy of the witness' recollection.
- Weapons Focus -- a witness will tend to focus not upon the perpetrator but upon the weapon.
- Exposure Time -- the time in which the witness is exposed to the perpetrator will positively correlate to the accuracy of the identification. Most witnesses overestimate how much time they had to identify an individual.
- Motivation of Victim to Make a Correct Identification -- witnesses are motivated to make some identification even if it may be incorrect.
Dr. Chen has testified on the topic of eyewitness identification in four state court proceedings. As noted, defendant has provided the transcripts of Dr. Chen's testimony in two of those cases and indicates that her testimony here will be comparable. Based upon this prior state court testimony alone, I could not find that Dr. Chen's opinions are based upon scientific knowledge. Her testimony in these cases was mostly conclusory, and based upon a handful of studies which she mentioned but did not describe in any detail.
However, Dr. Chen's live testimony on voir dire remedied these shortcomings. At the hearing, Dr. Chen spoke at greater length about the various studies on which her opinions are based. Specifically, Dr. Chen discussed and described in some detail a book by Dr. Elizabeth Loftus, of the University of Washington, which compiled numerous studies on eyewitness testimony; a report published in the "American Psychologist" indicating significant consensus among psychologists concerning many factors related to eyewitness identification; the so-called "Yerkes-Dodson" Law -- measuring the impact of stress on perception; a study by authors Loftus, Loftus & Messo on the subject of weapons focus; a study by author Daffenbacher on the correlation (or lack thereof) between confidence and accuracy in identification; and other studies by psychologists on issues such as exposure time and corporal vs. head/shoulder identification.
Dr. Chen also described a handful of experimental studies she has done herself, replicating the studies of these leading psychologists. The results of her studies were consistent with those of the other psychologists.
Dr. Chen's testimony during the voir dire testimony clarified and expanded upon the numerous bases for her opinions. In short, I am satisfied that Dr. Chen's opinions are based upon "scientific knowledge" as demonstrated by these numerous studies and psychological principles.
I also find that her testimony would satisfy the second Daubert prong -- it would be helpful to the jury. The Government's case depends largely on the testimony of a single eyewitness who was the victim teller. There were other tellers and victims present but none was able to make an identification. The Government's other witnesses are alleged accomplices turned informants. One has pled guilty to nine bank robberies; the other to four. Their testimony, thus, will be colored by questions of credibility and motive.
Furthermore, there is no physical proof in this case. There is no gun, no fingerprints. The Government has chosen not to introduce surveillance photographs. Thus, this is a case where the testimony of the victim teller is of singular importance.
This victim teller's experience was marked by significant stress: she was directly confronted by a gun-wielding robber. The time for observation was short in duration -- the whole robbery took place in minutes and there were two robbers causing significant commotion. Moreover, the teller's first identification of the defendant was made 47 days after the robbery -- a significant passage of time. Thus, there are many issues surrounding her identification: what is the impact of her being under significant stress? what is the impact of the 47 day delay in the accuracy of her identification? what is the significance of her confidence in her identification? what is the significance of the presence of a gun? These are matters on which Dr. Chen's testimony might assist the jury.
Accordingly, for all the above reasons, I find that Dr. Chen's testimony is scientifically based and would be helpful to the jury. The Government's motion to exclude her testimony is hereby DENIED.
IT IS SO ORDERED.
DAVID G. LARIMER
UNITED STATES DISTRICT COURT
Dated: Rochester, New York
May 1, 1996
© 1992-2004 VersusLaw Inc.