MEMORANDUM OPINION RE EXCLUSION OF PROPOSED EYEWITNESS MEMORY EXPERT DR. SCOTT FRASER
WILLIAM ALSUP, District Judge.
INTRODUCTION
The classic way to question the accuracy of an eyewitness identification is to cross-examine the eyewitness and to present other fact witnesses to establish particulars at the scene that would have weakened the accuracy of the identification, such as obstacles, lighting, distance, police suggestion and so on. All of this is fact based. We traditionally rely on the common experience of jurors, once apprised of all the factual particulars, to assess the reliability of the identification. In recent times, however, criminal defense counsel have sought to lay before the jury opinions by academics and professional witnesses to elaborate on weaknesses in human perception and memory so as to draw into question the reliability of eyewitness testimony. To this effort, prosecutors have generally responded that the opinions are based on junk science and should be excluded under
Daubert.
Contrary to the government, the Court believes that there are aspects of the discipline that are based on genuine science. That does not mean, however, that all or any such testimony should be admitted.
In this RICO/VICAR prosecution, defendants Angel Noel Guevara and Guillermo Herrera seek to introduce expert opinion testimony regarding eyewitness memory and eyewitness identifications. After an evidentiary hearing and much argument and briefing, the undersigned excluded the proffered testimony of Dr. Scott Fraser and reserved judgment on whether certain aspects of Dr. Deborah Davis’ testimony would be allowed (Dkt. No. 3863). The determination regarding Dr. Davis (proposed by defendant Guevara) will be made after the Court hears the testimony of the eyewitnesses and determines what circumstances might warrant aspects of the testimony, a balancing decision that will be postponed until more fact evidence has been laid before the jury. The instant memorandum opinion explains why the testimony of Dr. Fraser (propounded by defendant Herrera) should not be admitted at all.
STATEMENT
Defendant Guillermo Herrera was identified by an eyewitness as the shooter in
the Armando Estrada homicide (Dkt. No. 3243). The July 2008 homicide occurred mid-day in clear conditions on Mission Street in San Francisco. The eyewitness was inside a restaurant across the street from the shooting. To challenge the accuracy of this eyewitness identification, defendant Herrera would introduce the testimony of Dr. Scott Fraser.
Defendant Herrera’s Rule 16(b)(1)(c) expert notice specified that Dr. Fraser would testify to his conclusion that, “based on a review of the discovery and based on well-established studies in the field,” the identification of defendant Herrera as the shooter in the Estrada homicide was made “under circumstances likely to render his identification unreliable”
(id.
at 3). The notice explained that Dr. Fraser came to this conclusion because most or all of supposed causes of an inaccurate eyewitness identification were present in the Estrada homicide. The notice stated that Dr. Fraser would testify that the following factors “influenced the accuracy of [the identification]”: “perceptual obstructions, divided attention, multiple targets, distance, weapons focus, kinetic distortions, and physiological arousal”
(ibid.).
The government moved to preclude the testimony, arguing that Dr. Fraser’s proposed testimony was unscientific, irrelevant, invasive of the province of the jury, and unreliable (Dkt. No. 3371). The government also protested that — in violation of Rule 16 — it had not been provided with a sufficient summary of Dr. Fraser’s opinions.
Oral argument on the government’s motion was heard during the second day of the final pretrial conference (Dkt. No. 3569). At that time, it was determined that a
Daubert
evidentiary hearing was appropriate (Dkt. No. 3522). At the pretrial conference, the Court specifically raised the issue of whether the testimony should be allowed under Rule 403 and noted the issue would be considered at the evidentiary hearing.
The evidentiary hearing was held during the four-week interim between the commencement of jury selection and opening statements. At the evidentiary hearing, counsel was given an opportunity to demonstrate the relevance and reliability of Dr. Fraser’s expert opinions. Dr. Deborah Davis — an eyewitness memory expert proposed by another defendant — was also examined at the evidentiary hearing. Counsel for defendant Herrera was even permitted to cross-examine Dr. Davis where her testimony tended to undermine the testimony of Dr. Fraser.
Between day one and day two of the evidentiary hearing, counsel for defendant Herrera submitted an “amended” notice for Dr. Fraser’s opinions, perhaps seeking to address a concern that Dr. Fraser’s noticed opinion would usurp the province of the jury and the fact eventually surfacing that he had considered only a few of the actual particulars of the Estrada identification (Dkt. No. 3729). The amended notice, however, was identical to the original notice except it: (1) specified that the eyewitness “may” have been' — -rather than “was” — confronted with factors rendering his identification unreliable; and (2) replaced its assertion that relevant studies strongly suggested the identification
was
unreliable with the assertion that scientific studies indicated that a confluence of error-inducing factors during the Estrada homicide “has consistently been associated with the finding of unreliable recognition.”
After the evidentiary-hearing, both sides were given an opportunity to submit supplemental, post-hearing proffers without page restrictions (Dkt. No. 3741). Defendant Herrera declined to provide any supplemental proffer, stating he “has not and will not supplement the showing made at
the hearing” (Dkt. No. 3808). Despite this assertion, defendant Herrera later requested to join in a supplemental proffer submitted by defendant Guevara (Dkt. No. 3839).
This memorandum opinion addresses Dr. Fraser and why he should be and has been excluded.
ANALYSIS
Even assuming
arguendo
that Dr. Fraser has some specialized knowledge, training, and experience and that the core discipline of eyewitness memory is scientific, Dr. Fraser’s testimony should be and has been excluded as unreliable, unhelpful to the jury, and substantially more prejudicial than probative.
District courts have a continuing duty to act as vigilant gatekeepers to ensure expert testimony is based upon scientific knowledge that is both reliable and helpful to the jury.
Daubert v. Merrell Dow Pharm., Inc.,
509 U.S. 579, 597, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993);
United States v. Rincon,
28 F.3d 921, 926 (9th Cir.1994). An opinion is unreliable under
Daubert
where the witness unjustifiably extrapolates from an accepted premise to an unfounded conclusion.
General Elec. Co. v. Joiner,
Free access — add to your briefcase to read the full text and ask questions with AI
MEMORANDUM OPINION RE EXCLUSION OF PROPOSED EYEWITNESS MEMORY EXPERT DR. SCOTT FRASER
WILLIAM ALSUP, District Judge.
INTRODUCTION
The classic way to question the accuracy of an eyewitness identification is to cross-examine the eyewitness and to present other fact witnesses to establish particulars at the scene that would have weakened the accuracy of the identification, such as obstacles, lighting, distance, police suggestion and so on. All of this is fact based. We traditionally rely on the common experience of jurors, once apprised of all the factual particulars, to assess the reliability of the identification. In recent times, however, criminal defense counsel have sought to lay before the jury opinions by academics and professional witnesses to elaborate on weaknesses in human perception and memory so as to draw into question the reliability of eyewitness testimony. To this effort, prosecutors have generally responded that the opinions are based on junk science and should be excluded under
Daubert.
Contrary to the government, the Court believes that there are aspects of the discipline that are based on genuine science. That does not mean, however, that all or any such testimony should be admitted.
In this RICO/VICAR prosecution, defendants Angel Noel Guevara and Guillermo Herrera seek to introduce expert opinion testimony regarding eyewitness memory and eyewitness identifications. After an evidentiary hearing and much argument and briefing, the undersigned excluded the proffered testimony of Dr. Scott Fraser and reserved judgment on whether certain aspects of Dr. Deborah Davis’ testimony would be allowed (Dkt. No. 3863). The determination regarding Dr. Davis (proposed by defendant Guevara) will be made after the Court hears the testimony of the eyewitnesses and determines what circumstances might warrant aspects of the testimony, a balancing decision that will be postponed until more fact evidence has been laid before the jury. The instant memorandum opinion explains why the testimony of Dr. Fraser (propounded by defendant Herrera) should not be admitted at all.
STATEMENT
Defendant Guillermo Herrera was identified by an eyewitness as the shooter in
the Armando Estrada homicide (Dkt. No. 3243). The July 2008 homicide occurred mid-day in clear conditions on Mission Street in San Francisco. The eyewitness was inside a restaurant across the street from the shooting. To challenge the accuracy of this eyewitness identification, defendant Herrera would introduce the testimony of Dr. Scott Fraser.
Defendant Herrera’s Rule 16(b)(1)(c) expert notice specified that Dr. Fraser would testify to his conclusion that, “based on a review of the discovery and based on well-established studies in the field,” the identification of defendant Herrera as the shooter in the Estrada homicide was made “under circumstances likely to render his identification unreliable”
(id.
at 3). The notice explained that Dr. Fraser came to this conclusion because most or all of supposed causes of an inaccurate eyewitness identification were present in the Estrada homicide. The notice stated that Dr. Fraser would testify that the following factors “influenced the accuracy of [the identification]”: “perceptual obstructions, divided attention, multiple targets, distance, weapons focus, kinetic distortions, and physiological arousal”
(ibid.).
The government moved to preclude the testimony, arguing that Dr. Fraser’s proposed testimony was unscientific, irrelevant, invasive of the province of the jury, and unreliable (Dkt. No. 3371). The government also protested that — in violation of Rule 16 — it had not been provided with a sufficient summary of Dr. Fraser’s opinions.
Oral argument on the government’s motion was heard during the second day of the final pretrial conference (Dkt. No. 3569). At that time, it was determined that a
Daubert
evidentiary hearing was appropriate (Dkt. No. 3522). At the pretrial conference, the Court specifically raised the issue of whether the testimony should be allowed under Rule 403 and noted the issue would be considered at the evidentiary hearing.
The evidentiary hearing was held during the four-week interim between the commencement of jury selection and opening statements. At the evidentiary hearing, counsel was given an opportunity to demonstrate the relevance and reliability of Dr. Fraser’s expert opinions. Dr. Deborah Davis — an eyewitness memory expert proposed by another defendant — was also examined at the evidentiary hearing. Counsel for defendant Herrera was even permitted to cross-examine Dr. Davis where her testimony tended to undermine the testimony of Dr. Fraser.
Between day one and day two of the evidentiary hearing, counsel for defendant Herrera submitted an “amended” notice for Dr. Fraser’s opinions, perhaps seeking to address a concern that Dr. Fraser’s noticed opinion would usurp the province of the jury and the fact eventually surfacing that he had considered only a few of the actual particulars of the Estrada identification (Dkt. No. 3729). The amended notice, however, was identical to the original notice except it: (1) specified that the eyewitness “may” have been' — -rather than “was” — confronted with factors rendering his identification unreliable; and (2) replaced its assertion that relevant studies strongly suggested the identification
was
unreliable with the assertion that scientific studies indicated that a confluence of error-inducing factors during the Estrada homicide “has consistently been associated with the finding of unreliable recognition.”
After the evidentiary-hearing, both sides were given an opportunity to submit supplemental, post-hearing proffers without page restrictions (Dkt. No. 3741). Defendant Herrera declined to provide any supplemental proffer, stating he “has not and will not supplement the showing made at
the hearing” (Dkt. No. 3808). Despite this assertion, defendant Herrera later requested to join in a supplemental proffer submitted by defendant Guevara (Dkt. No. 3839).
This memorandum opinion addresses Dr. Fraser and why he should be and has been excluded.
ANALYSIS
Even assuming
arguendo
that Dr. Fraser has some specialized knowledge, training, and experience and that the core discipline of eyewitness memory is scientific, Dr. Fraser’s testimony should be and has been excluded as unreliable, unhelpful to the jury, and substantially more prejudicial than probative.
District courts have a continuing duty to act as vigilant gatekeepers to ensure expert testimony is based upon scientific knowledge that is both reliable and helpful to the jury.
Daubert v. Merrell Dow Pharm., Inc.,
509 U.S. 579, 597, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993);
United States v. Rincon,
28 F.3d 921, 926 (9th Cir.1994). An opinion is unreliable under
Daubert
where the witness unjustifiably extrapolates from an accepted premise to an unfounded conclusion.
General Elec. Co. v. Joiner,
522 U.S. 136, 146, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997) (finding a trial court “may conclude that there is simply too great an analytical gap between the data and the opinion proffered.”)
The disconnects between Dr. Fraser’s proposed opinions and the bases from which they were purportedly derived are so severe that Dr. Fraser’s testimony would only confuse and mislead the jury.
A prime example is Dr. Fraser’s proposed testimony regarding the “Rule of 15.” On day one of the evidentiary hearing, on questioning by the proponent of the witness, Dr. Fraser repeatedly testified that scientific studies in his field have demonstrated that if a witness views a stranger from 15 meters away (49 feet) in good conditions, the chance of a correct identification is only five percent or less (March 15 Tr. 170-72, 175). As the statistic seemed a remarkable one, the Court repeatedly asked Dr. Fraser about this assertion. Dr. Fraser interchangeably referred to the five percent rate as a “reliability rate,” “accuracy rate,” or “likelihood of a correct identification.” He left the clear and repeated impression that studies in his field directly supported his claim that less than five percent of eyewitness identifications of strangers made from 15 meters away or more are correct. Variations of the following were repeated throughout the evidentiary hearing:
THE COURT: I want to make sure I understand this 15 meter thing. Are you saying that if somebody sees somebody else head on, 49 feet away [15 meters], that the chances that they can make a correct identification on average is 5 percent or less? Did I hear you?
DR. FRASER: That’s what the research shows, your Honor. Research studies.
THE COURT: - So a baseball pitcher on the mound looking at the batter 60 feet away, it’s going to be even less than 5 percent that they could identify that person later?
DR. FRASER: If that was a stranger?
THE COURT: A stranger, yeah.
DR. FRASER: A stranger that they were looking at and then subsequently were tested to recognize them by their face. You can see the face, and what are called boundary conditions. From first base to home plate, 90 feet, you can see the player. And you know it’s the first baseman. They have block letters, 18 inches high on their uniform, or very large, and from this information, you can tell that it is—
THE COURT: No, no. Answer my question. Just the face alone. Forget the uniform; forget everything. Just looking at the face, you’re telling me that the pitcher would have one in 20 times or less would be able to remember and recognize who they were throwing the ball to?
DR. FRASER: On that singular encounter to a stranger. And that’s what the studies showed. And that’s consistent—
THE COURT: I want to see that study. What’s the name of that study?
DR. FRASER: That’s a study by Wagenaar, W-a-g-e-n-a-a-r, and Van der Schrier, published 1996, Psychology Crime & Law. Subsequently replicated in 2003, I believe, maybe 2005 by Jong, J-o-n-g, and Wagenaar, with familiar faces. Interesting enough.
THE COURT: Do you have that — when you submit that, is that going to be one of the ones you submit to the prosecutor?
DR. FRASER: I’ll be happy to give both of them to him.
THE COURT: I’d like to see that, and please bring that tomorrow to court.
(Mar. 15 Tr. 170-72).
THE COURT: Let’s say it’s a famous baseball player or a famous movie star, you’re at the mall and you see them 60 feet away.
DR. FRASER: And then you tested them immediately as to which of these individuals they actually saw.
THE COURT: What would be the accuracy rate?
DR. FRASER: And again, the accuracy rate appears to be less than 5 percent for those.
(Mar. 15 Tr. 175).
DEFENSE COUNSEL: You were telling us about the Rule of 15. If you could continue, Dr. Fraser, please?
DR. FRASER: The ability of the human eye to accommodate the detection of facial features necessary for reliable recognition shows that beyond 15 meters, even though the rest of the viewing conditions are optimal — good lighting; over 1,000 lux — high attentional focus, high in motivation to remember the face, and immediate testing afterwards, no time delay, shows that beyond 15 meters, the rates of reliably recognizing among an array of similar alternatives the individuals seen is less than 5 percent.
THE WITNESS: 15 meters. 49 feet.
THE COURT: All right.
THE WITNESS: Beyond that, you can see the person, you can see the body, but you cannot, the research shows, de
tect — the human eye cannot, regular person with 20/20 vision — cannot detect the features that are used for reliable recognition even under otherwise optimal conditions, all right. That may be relevant to this case.
(Mar. 15 Tr. 168).
It turned out that the so-called “Rule of 15” was merely an idea advanced in a single article, one by Willem Wagenaar and Juliette Van der Schrier. Only after requesting the article and reviewing it did the Court discover that it did not support Dr. Fraser’s characterization of the Rule of 15. The .article described a university experiment wherein participants were asked to first view targets at different distances and illumination levels and then to identify the targets in photographic lineups.
The targets and individuals depicted in the line-ups did not have conspicuous facial hair, marks, varying dress, or other particularly notable physical characteristics. The article summarized the results of the experiment in terms of “hits” (correct identifications of the target) as compared to “false alarms” (false identifications). A table from the article with these results immediately follows.
[[Image here]]
As indicated in the table, where the target was 12 meters away (the cell of data closest to 15 meters) and viewed in well-lit conditions (3000 lux), the participants
correctly
identified the target 61 percent of the time and only incorrectly identified the target one percent of the time. The remainder of the participants (38 percent) declined to make any identification — meaning that those participants who
attempted to make an identification
(that is, excluding those who did not attempt to make an identification), made a correct identification at a rate of 61 out of 62 times (98.3 percent). In the Court’s mind, this is a solid and reliable accuracy rate.
How could this reliable accuracy rate of 61 out of 62 somehow be transformed into an unreliable accuracy rate of less than one out of 20?
This amazing transformation was the result of some liberties taken by the authors of the article, together with sleight of hand
by Dr. Fraser. In order to dilute the force of the 61/62 ratio, the authors invented a concept called “diagnostic value of the total evidence” (Dkt. No. 3729-1 at 329); stating a “diagnostic value” could be derived from the requirement that “20 guilty suspects should be set free against one innocent suspect being convicted.” To this was added the “simple legal rule” that, “one witness is not enough but two witnesses are.” And from these two ideas the authors purported to apply statistical theorems to arrive at a “diagnostic value.”
Before proceeding further, it is worth mentioning that it is difficult to extract from the article how the foregoing ideas were factored together to arrive at the so-called “diagnostic value.” Even so, at no place did the article undermine the solid, irrefutable fact that 98.3 percent of identifications made were correct at a distance of 12 meters, under lighting conditions comparable to the facts of our case.
For his part, Dr. Fraser tampered with science. He took the sentence quoted above — the “requirement” that “20 guilty suspects should be set free against one innocent suspect being convicted” and transmogrified this concept into a reliability and/or accuracy rate of one out of 20. There was no other reference in the article to a one-to-20 ratio. It bears emphasizing that Dr. Fraser affirmatively advanced the Rule of 15 and explained it as a five percent accuracy rate for targets viewed from a vantage point of 15 meters. He did not front or explain the article. On direct examination Dr. Fraser never explained how he transformed the 61/62 ratio to a one out of 20 ratio. But for the fortuity of a request to produce the article, the issue may not have surfaced in time for anyone to appreciate the shortcomings in Dr. Fraser’s proposed testimony.
Through cross-examination by government counsel on the second day, it became clear that Dr. Fraser’s five percent figure was not supported by the article. For the first time, he asserted that the article derived the five percent figure from the Wig-more standard. While the article referred to the Wigmore standard, however, it used Wigmore in a wholly different way. Even if we prefer that 20 guilty suspects go free before one innocent suspect is convicted, in no way could it follow that the accuracy rate at 15 meters is only one out of 20 (Mar. 16 Tr. 76-77). The one-and-20 ratio, as used by Dr. Fraser, was a severely unfounded assertion.
Indeed, the article specified that the Rule of 15 was simply a framework to assist non-experts in understanding the probability of achieving an identification which satisfied the authors’ view of the “diagnostic value” that
should
be required to convict (Dkt. No. 3729-1 at 329). The authors themselves noted that the decision regarding whether observation conditions were good enough to accept an eyewitness identification, however, should be reserved for the jury. This is a far cry from Dr. Fraser’s blanket claim that the article found that only five percent of eyewitnesses can make correct identifications when 15 meters away from a target.
In the Court’s view, no honest scientist could have made Dr. Fraser’s statements regarding the Rule of 15 under oath in good faith. No honest scientist could have transformed an accuracy rate of 61 out of 62 to less than one out of twenty.
To be sure, Dr. Fraser attempted to backpedal on day two. This change was
only made after he was exposed. Even in the midst of his attempts to rehabilitate his earlier, misleading testimony, however, Dr. Fraser was still unable to resist reverting to his assertion that the Rule of 15 contemplated a five percent accuracy rate:
DEFENSE COUNSEL: Could you tell the Court again what your conclusion was about the 49 feet, the 15 meters, the percentages — let’s get straight to that— what you told the Court yesterday about the percentages? Sorry to the court reporter.
DR. FRASER: What is said, the research that’s been done by Wagénaar and Van der Schrier and their associates indicates that beyond 15 meters — all right? — which is the cut point when the diagnostic value of the person’s accuracy under very optimal conditions of viewing — all right? — was greater than 15, which converts essentially into an error rate of less than five — of five percent. All right? So they used that as the standard, the Wigmore standard, that it’s better to let 19 guilty people go free than convict one innocent person. So that’s five percent. One innocent person is what you are trying to — erroneously convicted being the standard they employ. That’s five percent. So at what point of illumination and distance, vary both of them, would you consistently have — all right? — more — less than five percent accuracy? All right. Reliability in terms of selecting. Using that diagnostic value of 15, which is what it converts to.
DEFENSE COUNSEL: All right. But you are not saying if a person is more than 49 feet away, that at that point, then under, as you put it, optimal conditions, that the chances of successfully identifying that individual is five percent?
DR. FRASER: That’s not what my testimony was yesterday. It’s not my testimony today. What the research shows is that, all right? — beyond that distance — all right? — the reliability of a correct selection is less than five percent. So, in other words, the false positives and the number of hits when you put them together — all right? — ends up that less than five percent of the time where they’re going to be accurate. As I said yesterday, that doesn’t mean it can’t happen and doesn’t happen.
DEFENSE COUNSEL: It’s still possible—
THE COURT: Can I — I do want to be clear on this. Under optimal conditions, if somebody is—
DR. FRASER: Forty-nine feet, your honor.
THE COURT: All right. So let’s say 50 feet away. ' DR; FRASER: Yeah.
THE COURT: If somebody, some unknown person is 50 feet away and they’re optimal viewing conditions, and you have 20 trials, you’re saying that in one out of 20 times would the eyewitness get it right?
DR. FRASER: No.
THE COURT: Isn’t one out of twenty five percent?
DR. FRASER; No. They’re using as their standard for what would be the cut point — all right? — five percent. You could pick some other point at which you said it would be sufficiently reliable for you to consider it to be useful or valid information. You could say they could be right only 40 percent of the time and that would be enough, or 60 percent of the time. What they chose as their cut points is at what distance and what illumination are you going to get — all right? — less than five percent accuracy under otherwise optimal conditions, less than five percent. That doesn’t mean
nobody gets it right. It just means less than five percent of the time they get it right.
THE COURT: You are saying 50 feet, under optimal conditions, when you would go from 40 feet, 45 to 50 feet, is that the accuracy rate of identifications is less than five percent?
THE WITNESS: Reliability. That’s what the research — and we provided a copy to Attorney Frentzen.
THE COURT: So that means the unreliability rate is 95 percent or more?
THE WITNESS: That’s right. When you put them together. When you put false positives and false alarms and misses, yep.
(Mar. 16 Tr. 62-65).
The confusing and misleading aspects of Dr. Fraser’s testimony were not limited to his discussion of the Rule of 15. Dr. Fraser’s unreliable methodology and tendency to mislead permeated his testimony at the evidentiary hearing. A few additional representative examples are now recounted.
Dr. Fraser testified that the Wagenaar and Van der Schrier article found that a witness 20 meters away from a target in brightly illuminated lighting conditions (3000 lux) would make an “incorrect” identification 43 percent of the time (Mar. 16 Tr. 72-75).
The article itself, however, contained no such figure — it only indicated that at 20 feet and 3000 lux, correct identifications were made 57 percent of the time and incorrect identifications were made five percent of the time (Dkt. No. 3729-1 at 325, Table 2). In other words, 92 percent of actual attempted identifications were correct (a 57/62 ratio). It appears Dr. Fraser arrived at the 43 percent figure by expanding the universe of “incorrect” identifications to include both: (1) instances where witnesses actually made incorrect identifications (proper); and (2) instances where witnesses
chose not to make any identification
(improper) (Mar. 16 Tr. 106-111). This unjustified revision of the article’s findings was misleading and would confuse a jury by suggesting that identifications were
incorrect
43 percent of the time when in reality actual attempted identifications were only incorrect five percent of the time. When a participant or a bystander declines to even try to make an identification, there is no identification at all, much less an incorrect one. Such a case is a recognition of his/her own limitations. In assessing the risk of wrong identifications, we must focus on identifications actually attempted and ask — of that universe — how many were wrong? Dr. Fraser’s approach was a gimmick to inflate the error rate.
Put differently, Dr. Fraser’s testimony ignores the eyewitness self-selection process. If those who cannot make an identification say so, they never appear in court. Similarly, some eyewitnesses may be fearful to testify against those accused of being violent gang members. This fear drives some to invent or exaggerate inability to recall, especially in prosecutions of violent street gangs where retribution is a genuine concern. By contrast, the subset of the general eyewitness population who are able and willing to identify a perpetrator in front of a jury have survived these impulses to self-select out of the process. This was ignored by Dr. Fraser.
Dr. Fraser also testified that studies have shown that 75 to 77 percent of eye
witness identifications made only four to six hours after viewing the target are “incorrect” (Mar. 16 Tr. at 13-14). Only after questioning by the government did Dr. Fraser admit that his statistics for “incorrect” identifications again incorporated instances where no identification was even attempted
(id.
at 16-17).
When asked if he had ever been barred from testifying as an expert in federal court, Dr. Fraser repeatedly answered “not to my knowledge” or “not to my recollection” (Mar. 16 Tr. 34-35). After being confronted with a 2009 published decision from the United States District Court in Kansas finding his proffered expert testimony unreliable and scientifically invalid,
P.S. ex rel. Nelson v. The Farm, Inc.,
658 F.Supp.2d 1281 (D.Kan.2009), Dr. Fraser claimed that he had no knowledge of the opinion and was only told by the defense attorneys in that case that they had decided not to call Dr. Fraser as an expert
(id.
at 36-37). Especially given that at least a part of Dr. Fraser’s living is made through his expert testimony, it stretches credulity to believe that he was unaware of the district court’s order excluding his expert testimony.
In the same vein, when Dr. Fraser was asked what ultimately happened to a California judge’s request that perjury charges be filed against him for his sworn statements in
People v. Adam Anthony Norie
ga, Case No. BA201786 (Cal.Sup.Ct.2001), Dr. Fraser stated that the Attorney General had “reviewed all the information and they refused to file for inadequacy” because his conduct was not perjury (Mar. 16 Tr. 100). Upon additional cross-examination, however, Dr. Fraser admitted that he was never directly told that the Attorney General made any such finding that the case against him was inadequate.
Dr. Fraser also tried to leave the impression that he had not been able to provide a full and complete expert evaluation of the actual viewing conditions in this case because the government had somehow failed to produce the records he needed to render such an evaluation
(id.
at 30-31, 39-41). This also proved to be untrue. Questioning by the Court and the government revealed that Dr. Fraser was in possession of needed material, but had simply failed to review the material prior to the evidentiary hearing
(id.
at 104-06). Dr. Fraser then asserted this was because he was denied adequate funding to conduct the review. This claim, however, also proved incorrect, as further CJA funding had been authorized and was available to Dr. Fraser over a month prior to the evidentiary hearing. Notice of this further funding authorization was sent to defense counsel and one of his staff members via email — receipt of which was acknowledged by the staff member shortly thereafter
(id.
at 105-06,113).
Dr. Fraser’s testimony offers little probative value. For example, Dr. Fraser’s expert disclosure specified that certain factors such as “perceptual obstructions” and “physiological characteristics” of the eyewitness to the Estrada homicide supported Dr. Fraser’s conclusion that the identification was made under circumstances likely to render it unreliable. During the evidentiary hearing, however, Dr. Fraser admitted that he had not interviewed the eyewitness and knew nothing about the witness’ physiological characteristics, where specifically the witness was located when the shooting occurred, and whether there were any perceptual obstructions at the scene
(id.
at 50-52, 54-58). An opinion based on unwarranted and unfounded premises and assumptions with no anchor in the actual facts would be pure argument and would call upon the jury to speculate.
Under
Daubert
and Rule 403, should a witness who persistently exaggerates, if
not prevaricates, be tolerated merely because in those instances he is caught he will come clean? The Court thinks not. In this instance, the witness, when caught, would not even come clean.
In stark contrast to the testimony’s limited probative value, the risk for juror confusion, unfair prejudice, and waste of time is very high. Because of his demonstrated tendency to leave false impressions, it would be confusing and wasteful of the jury’s time for cross examination at trial to try to winnow down Dr. Fraser’s testimony to those few nuggets that are more probative than prejudicial. There are some aspects of the field of eyewitness identification and memory that lend themselves to scientific treatment and this order does not condemn the entire discipline. Dr. Fraser’s proposed testimony, however, went well beyond any such science. At best, Dr. Fraser’s testimony was an amalgam of a whiff of science mixed with unjustified extrapolation.
See Joiner,
522 U.S. at 146, 118 S.Ct. 512 (holding a trial court “may conclude that there is simply too great an analytical gap between the data and the opinion proffered.”) It is like a small dose of medicine in a large bottle of snake oil. Yes, there is some medicine in there, but it is hard to separate from the snake oil. Under Rule 403, the jury should not be burdened with trying to drain the snake oil to find the medicine.
CONCLUSION
For the reasons stated herein, Dr. Fraser’s proposed testimony has been excluded in its entirety.
IT IS SO ORDERED.