Washington v. Derri (Majority)

Annotate this Case

In March 2017, three North Seattle banks were robbed. Police administered a variety of photomontages to witnesses. Some aspects of the photomontage process complied with best practices generally recognized by new scientific research; some aspects of that process did not; and some aspects of that process fell into a gray area on which the scientific literature was in dispute. Defendant John Stites, a/k/a Christopher Lee Derri, moved to suppress the identifications resulting from those photomontages on federal constitutional grounds; the trial court denied his motion, and he was convicted of all three robberies. The issue this case presented for the Washington Supreme Court’s review asked whether trial courts had to consider new scientific research, developed after the 1977 Manson v. Brathwaite, 432 U.S. 98, decision, when applying that federal due process clause test. To this, the Supreme Court answered yes: courts must consider new, relevant, widely accepted scientific research when determining the suggestiveness and reliability of eyewitness identifications under Brathwaite. Considering this research, the Supreme Court concluded all three of the challenged identification procedures were suggestive. Under the totality of circumstances, however, the identifications were nonetheless reliable. Defendant’s convictions were affirmed.

Download PDF

FILE THIS OPINION WAS FILED FOR RECORD AT 8 A.M. ON JUNE 23, 2022 IN CLERK’S OFFICE SUPREME COURT, STATE OF WASHINGTON JUNE 23, 2022 ERIN L. LENNON SUPREME COURT CLERK IN THE SUPREME COURT OF THE STATE OF WASHINGTON STATE OF WASHINGTON, Respondent, v. NO. 100038-3 EN BANC Filed: June 23, 2022 CHRISTOPHER LEE DERRI, a/k/a JOHN STITES, Petitioner. GORDON McCLOUD, J.—In March 2017, three bank robberies occurred in North Seattle. Police administered a variety of photomontages to witnesses. Some aspects of the photomontage process complied with best practices generally recognized by new scientific research; some aspects of that process did not; and some aspects of that process fell into a gray area on which the scientific literature is in dispute. Defendant John Stites 1 moved to suppress the identifications resulting from those photomontages on federal constitutional grounds; the trial court denied his motion, and he was convicted of all three robberies. The State charged the petitioner as “Christopher Lee Derri, aka John Stites.” 1 Clerk’s Papers (CP) at 310-11. At trial, the parties used the two names interchangeably. However, the briefs on appeal refer to him as “John Stites.” We refer to the petitioner by that name to avoid confusion and intend no disrespect. 1 No. 100038-3 “[M]istaken eyewitness identification is a leading cause of wrongful conviction.” State v. Riofta, 166 Wn.2d 358, 371, 209 P.3d 467 (2009) (citing Brandon L. Garrett, Judging Innocence, 108 COLUM. L. REV. 55, 60 (2008)). At least eight Washingtonians have been exonerated after being convicted, in part, based on mistaken eyewitness evidence, but the number of people wrongly convicted on this basis is likely much higher.2 The due process clause of the Fourteenth Amendment offers some protection against this problem: it bars the admission of eyewitness identification evidence obtained through suggestive police procedures, unless the evidence is nevertheless reliable under the totality of circumstances. U.S. CONST. amend. XIV; Manson v. Brathwaite, 432 U.S. 98, 114, 97 S. Ct. 2243, 53 L. Ed. 2d 140 (1977). This case asks us to decide whether trial courts must consider new scientific research, developed after the 1977 Brathwaite decision, when applying that federal due process clause test.3 The answer is yes. We hold that courts must consider See Amici Curiae Br. of the Innocence Project Inc. & Wash. Innocence Project at 23 (citing National Registry of Exonerations Database, https://www.law.umich.edu/special/exoneration/Pages/detaillist.aspx). 2 The State filed a motion to strike portions of Stites’ supplemental brief on the ground that it addressed issues outside the scope of review. Specifically, the State moved to strike Stites’ arguments (1) that the state due process clause and/or the evidence rules provide more protection than the federal due process clause and (2) that suggestive behavior by nonstate actors implicates the state due process clause. Prior to oral argument, we granted the motion to strike the second argument and passed the remainder of the motion to the merits. We now grant the remainder of the motion to strike because 3 2 No. 100038-3 new, relevant, widely accepted scientific research when determining the suggestiveness and reliability of eyewitness identifications under Brathwaite. Considering this research, we conclude that all three of the challenged identification procedures were suggestive. Under the totality of circumstances, however, the identifications were nonetheless reliable. Stites also challenges the sufficiency of the charging information. Where, as here, such a challenge is first raised on appeal, we read the information liberally in favor of validity. State v. Kjorsvik, 117 Wn.2d 93, 102, 812 P.2d 86 (1991). Under that test, the information contains the necessary facts and Stites fails to show prejudice from its wording. We therefore affirm the convictions. FACTS4 Three Bank Robberies Occur in North Seattle In March 2017, two banks were robbed in North Seattle. Chase Bank was robbed on March 1 and HomeStreet Bank was robbed twice, once on March 7 and once on March 11. 1 Clerk’s Papers (CP) at 5-7. these arguments either were not raised or were not fully briefed until after review had been granted. See Pet. for Review at 6-12. Because the factual issue in this case relates to the denial of Stites’ pretrial motion to suppress eyewitness identifications, the facts below are taken from the record before the trial court when it ruled on that motion. 4 3 No. 100038-3 A. March 1, 2017: Robbery at Chase Bank On March 1, 2017, branch manager David Fletcher and teller Jacob Price were working at Chase Bank in North Seattle. Id. at 33. Around 3 p.m., a man entered the bank, approached Fletcher, and demanded cash. Id. at 33, 236. The man asked for wrapped $20 bills from the bottom drawer, requested the merchant teller, and said he didn’t want any “dye packs.” Id. at 34. Fletcher and Price began putting money on the counter. Id. at 34, 236. The robber began stuffing the money into a duffel bag. Id. at 36. The robber asked for more money, and the employees put loose coins on the counter. Id. at 236. The robber then left. Id. Police responded shortly thereafter, and Fletcher described the robber as a male, about 5 feet 11 inches tall, “very thin with a sunken in face wearing a thick olive green winter coat with the hood of his jacket pulled up over his head.” Id. The police then took Price to view a possible suspect who had been detained nearby, but Price told them it was the wrong man. Id. at 39. Around 4 p.m. that same day, Detective Len Carver obtained photographs of the robber from the bank’s surveillance cameras. Id. at 218, 227. After police disseminated the photos to a large Seattle Police Department e-mail list, an officer e-mailed Carver, saying, “I think a good suspect is John T. Stites . . . AKA Christopher L. Derri.” Id. at 229. Detective Carver then located a 2015 jail booking photo of Stites. Id. at 218. 4 No. 100038-3 i. March 2, 2017: Fletcher and Price fail to make a pick from a photomontage The next day, March 2, 2017, Carver interviewed Price. Price described the robber as about 5 feet 10 inches to 6 feet tall, 30 to 40 years old, with brown eyes, “very thin, very pale,” not clean-shaven, with “very sunken in cheeks.” Id. at 38. Detective Carver read a standard admonition to Price, then showed Price a sixphoto, sequential montage that included the 2015 booking photo of Stites.5 Id. at 44-50. Price did not make a pick. Id. at 44. The same day, Fletcher was shown the same six-photo montage. Id. at 5258. The record does not contain any transcript of an interview of Fletcher by Detective Carver. Fletcher was shown the photos sequentially and in a different order from Price. Id. He did not make a pick. Id. at 52. The standard admonition reads as follows: “In a moment I am going to show you a group of photographs. This group of photographs may or may not contain a picture of the person who committed the crime now being investigated. Keep in mind that hair styles, beards, and mustaches may be easily changed. Also, photographs may not depict the true complexion of a person—it may be lighter or darker than shown in the photograph. Pay no attention to any markings or numbers that may appear on the photos or to any differences in the style or type of photograph. When you have looked at all of the photos, tell me whether or not you see the person who committed the crime. Do not tell other witnesses that you have or have not identified anyone or the composition of the montage.” 1 CP at 44. 5 5 No. 100038-3 ii. March 10: Fletcher and Price are shown a second photomontage, and Fletcher selects Stites On March 10, 2017, Detective Carver showed Fletcher and Price a sixphoto, sequential montage that included Stites once again—but this time the detective used a more recent photograph of Stites. See id. at 186-192, 77-83, 85-91. Detective Carver read the montage admonition to Price. Id. at 77-83. Price initially selected Stites’ photo. Id. at 94. However, when he noticed the tattoo on Stites’ neck, he changed his mind—because he didn’t remember the robber having any tattoos. Id. at 94-95. Ultimately, Price did not make a pick. The record contains no interview with Fletcher during this second photomontage, either, but Fletcher did sign a “Montage Identification Sheet” that contained the admonition. Id. at 77. This time, Fletcher selected Stites’ photo and indicated that he recognized Stites as the robber. Fletcher was 90 percent certain. Id. at 78. iii. March 7, 2017: Robbery at HomeStreet Bank On March 7, 2017, a similar robbery occurred at HomeStreet Bank in North Seattle. Id. at 241. Around 5:15 p.m., a man entered and approached tellers Hannah Amdahl and Andrew Hilen. Id. at 241, 293. The man “said . . . something along the lines of . . . this is a robbery, . . . give me your money.” Id. at 279, 295. He asked for the merchant teller and also told the tellers not to include “dye packs” or 6 No. 100038-3 “devices.” Id. at 279, 281. The tellers began putting money on the counter, and the man began putting it in his pockets. Id. at 279. Eventually, Amdahl and Hilen backed up and put their hands up to indicate that was all the money they had. Id. at 280. The robber then looked at both tellers and apologized before walking out the door. Id. at 296. Hilen said that the robber spent about 15 seconds in front of each teller. Id. After the robber left, Amdahl called 911. Id. at 297. Amdahl and Hilen described the robber to police as an adult white male, wearing a hat, a “jacket with the hood up and cinched down over the hat,” baggy jeans, and gardening gloves. Id. at 241. Amdahl described him as “tall, pale and thin.” Id. Both Amdahl and Hilen also told police that they recognized the robber as a man who had come into the bank on February 23 or 24 to ask about opening a checking account. Id. at 242, 284, 299-300. Amdahl had discussed accounts with him, and at the end of their conversation, the man introduced himself as “John Stites.” Id. at 283. Amdahl wrote down the man’s name so she would remember their conversation in case Stites did come back to open an account. Id. at 284. She didn’t write down the date at the time, but about a week later, she wrote “2/24?” on the note. Id. at 147, 282. Amdahl said she recognized Stites as the March 7 robber because he had the same “sunken in eyes and . . . he looked very emaciated in the face.” Id. at 285. 7 No. 100038-3 Bank manager Dustin Foss learned of the robbery and went to the bank around 6 p.m. on March 7. Id. at 251. He learned from an officer that the suspect’s name was John Stites. Id. at 257. Recognizing the name as someone with whom he had attended elementary school, Foss searched Facebook and found a profile photo for a John Stites. Id. at 257-58. Foss said he showed the photo to Amdahl and Hilen either the night of the robbery or the following day and asked them if it looked like the robber. Id. at 259. 6 Foss said that both tellers identified the photo as the robber. Id. No copy of the Facebook photo appears in the record. iv. March 8: Amdahl selects Stites from a photomontage On March 8, 2017, Detective Carver interviewed Amdahl. She described the robbery and said that the robber was wearing a black Windbreaker with the hood “pulled up and the strings pulled so it was tight against his face, so you could only see his face.” Id. at 279. Detective Carver had assembled a seven-photo montage using the more recent photograph of Stites. Id. at 60-67. He read Amdahl the standard admonition, then showed the photos sequentially. Id. at 286-87. At the pretrial hearing, the prosecutor stated that in her defense interview, Amdahl “adamantly denie[d] that she saw any Facebook photos” before selecting Stites’ photo. 1 Verbatim Report of Proceedings (VRP) (Mar. 19, 2019) at 41. That interview is not in the record. On the other hand, there was no objection to the trial court considering the prosecutor’s hearsay statements. The trial court did not resolve this factual discrepancy. 6 8 No. 100038-3 Amdahl thought the “cheekbones and the eyes” of one pictured individual looked “very similar” to the robber. Id. at 287. Then she got to the photo of Stites and said, “[T]hat’s him.” Id. She said she recognized the neck tattoo from the time when he came into the bank in February. Id. She had not mentioned any neck tattoo until seeing the photo. Amdahl said she recognized Stites as the robber with 100 percent confidence. Id. at 67, 288. v. March 9: Hilen selects Stites from a photomontage On March 9, Detective Carver interviewed Hilen about the robbery. Hilen described the robber as “very skinny” with “sandy blonde hair,” “a very weathered face,” and a “stutter[ing]” way of speaking. Id. at 299. After asking questions about the March 7 robbery, Detective Carver said, “I’m gonna show you some pictures from a March 1 bank robbery.” Id. at 302. He then showed Hilen some surveillance photos from the Chase Bank robbery and asked Hilen if the man in the photos “look[ed] like the same guy” who had robbed HomeStreet. Id. Hilen said it was hard to be sure because the pictures were of such poor quality, but that the person in the photos had a chin similar to the HomeStreet robber’s chin. Id. Detective Carver read Hilen the standard admonition and then showed Hilen a six-person photomontage that included the more recent photo of Stites. Id. at 6875, 303-04. When Hilen saw the fourth photo, which was of Stites, he said, “That’s 9 No. 100038-3 him.” Id. at 304. Carver asked Hilen to look at the rest of the photos “[t]o make sure.” Id. After doing so, Hilen selected Stites’ photo and stated that he recognized Stites as the robber with 98 to 99 percent confidence. Id. at 72, 304. Hilen said that “the tattoo . . . definitely gives it away,” along with “just the general boney structure of his chin.” Id. at 304. Carver said, “[W]e didn’t talk about a tattoo with you yet,” and Hilen elaborated that seeing the photo “reminded me of the first time he was here that I could see his tattoo.” Id. at 304-05. Hilen had not mentioned the tattoo in his description of the robber before seeing the photomontage, either. Id. at 292-303. B. March 11: Second robbery at HomeStreet Bank On March 11, 2017, Foss and Amdahl were working at HomeStreet Bank. Id. at 261. Close to 2 p.m., a person approached the door. Id. When he reached the door, Foss and Amdahl recognized him as the same robber from the incident a few days earlier. Id. at 262. Amdahl immediately pressed her silent alarm button. Id. The man then pulled a mesh mask over his face and opened the door. Id. He approached Amdahl’s teller station and demanded cash. Id. at 263. Amdahl gave the robber some cash, and he left the bank. Id. at 264. After the robber left, Amdahl called 911. Id. at 265. Foss said he recognized the robber from the Facebook photo of John Stites, and he provided the name “John Stites” to the responding officer. Id. at 273-74. The record is silent about 10 No. 100038-3 whether Amdahl or Foss viewed photomontages or gave statements to the police after this robbery. C. The State charges “Christopher Derri, aka John Stites” with the robberies On March 13, 2017, officers arrested Stites in a parking lot in North Seattle. 2 Verbatim Report of Proceedings (VRP) (Mar. 27, 2019) at 615. The State charged “Christopher Lee Derri, aka John Stites” with three counts of first degree robbery. 7 1 CP at 310. PROCEDURAL HISTORY A. Stites moves to suppress the eyewitness identifications Stites moved to suppress the identifications made by Amdahl, Fletcher, and Hilen and to prevent them from making in-court identifications. Id. at 9-27; 1 VRP (Mar. 19, 2019) at 12. He did not seek to call any witnesses at the pretrial hearing but relied on the interviews, police reports, and montage records attached to his motion. Stites argued that each montage procedure was impermissibly suggestive and the resulting identifications were unreliable under the totality of the circumstances, requiring suppression under the federal due process clause, citing Brathwaite, 432 U.S. at 114. 1 CP at 16-24; 1 VRP (Mar. 19, 2019) at 12-41. He The State initially charged “Christopher Lee Derri” with the crimes but amended the information to add “aka John Stites” to the accused’s name. 1 CP at 1, 310. 7 11 No. 100038-3 also argued that any in-court identification of Stites by these witnesses was tainted by the prior identification procedures and should also be suppressed. 1 CP at 24. With regard to Fletcher’s identification, Stites argued that the procedure employed by police was impermissibly suggestive because of double exposure: Fletcher was shown two different montages and Stites was the only person pictured in both of them. 1 VRP (Mar. 19, 2019) at 14, 40. With regard to Amdahl and Hilen’s identifications, Stites argued that the procedure employed by police was impermissibly suggestive because of double exposure and improper highlighting of Stites: (1) Stites was the only person in the montage with a tattoo and (2) Foss claimed he had shown both tellers a Facebook photo of Stites before they viewed the photomontages. 1 CP at 19; 1 VRP (Mar. 19, 2019) at 12-17. Additionally, Stites argued that Hilen’s procedure was impermissibly suggestive because of triple exposure: Detective Carver showed Hilen a surveillance photo of the March 1 robber prior to showing Hilen the montage. 1 VRP (Mar. 19, 2019) at 18. Stites further argued that under the totality of the circumstances, the suggestive procedures created a substantial likelihood of irreparable misidentification as to all three witnesses. Id. at 33-39; 1 CP at 19. 12 No. 100038-3 The trial court orally ruled on the suppression motion. 8 With regard to the Fletcher identification, it ruled that showing Fletcher two photomontages with Stites’ as the only repeat photo did not make the procedures impermissibly suggestive because “the photos are so different from one another that it did not taint the second montage or draw attention to that picture.” 1 VRP (Mar. 19, 2019) at 108-09. The trial court ruled that the Amdahl and Hilen identification procedures were not impermissibly suggestive, either, despite the fact that Stites was the only person pictured with a tattoo. Id. at 109. It opined that because no witness described the robber as having a tattoo, “there was no obligation on the part of the police to provide another photograph of individuals with tattoos.” Id. The court said that Amdahl “is expected to testify she was never shown a Facebook photograph” of Stites but continued that even if Amdahl had viewed the Facebook photo, “the mere fact that a witness may have looked at a photograph provided by a non-law enforcement officer outside the investigation does not make the police procedure that followed impermissibly suggestive.” Id. at 110. No written findings of fact and conclusions of law on the CrR 3.6 motion appear in the record, though CrR 3.6 requires the entry of such an order. In its oral ruling, the trial court did not explain which of the controverted facts it credited and did not distinguish between findings of fact and conclusions of law. 1 VRP (Mar. 19, 2019) at 106-10. 8 13 No. 100038-3 The court did not address Hilen’s alleged exposure to the Facebook photo or the fact that Detective Carver showed Hilen the March 1 suspect immediately before administering the photomontage. See id. at 108-10. As to all three identifications, the court concluded that even if the police procedures were impermissibly suggestive, the identifications were still sufficiently reliable under the Biggers9 factors and “the totality of the circumstances does not give rise to a substantial likelihood of irreparable misidentification.” Id. at 110. The court denied the motion to suppress but noted that it was “a fairly close issue.” Id. at 106; 1 CP at 315. B. The jury convicts Stites on all three counts At trial, Amdahl, Hilen, and Fletcher testified about the out-of-court identifications, and each made an in-court identification of Stites. 1 VRP (Mar. 26, 2019) at 340, 362-66; 448-49, 458-60; 2 VRP (Mar. 27, 2019) at 528, 539-45. Stites did not call any expert witnesses to testify about eyewitness identifications. In closing, Stites argued that the eyewitness identifications were unreliable and that he was not the robber. 2 VRP (Apr. 1, 2019) at 911-39. Stites did not propose any jury instructions relating to eyewitness evidence, and none were given. 1 CP at 337-60. The jury found Stites guilty as charged. Id. at 361-63. 9 Neil v. Biggers, 409 U.S. 188, 198, 93 S. Ct. 375, 34 L. Ed. 2d 401 (1972). 14 No. 100038-3 C. The Court of Appeals affirms Stites appealed. He assigned error to the court’s decisions to admit both “the impermissibly suggestive, unreliable out-of-court identification of Mr. Stites” and “the unreliable in-court identifications of Mr. Stites where the State failed to establish they were untainted by the unduly suggestive and unreliable out-of-court procedures.” Br. of Appellant (Wash. Ct. App. No. 80396-4-I (2020)) at 2. He also argued that the information was constitutionally deficient for failing to include all the essential elements of first degree robbery. Id. at 1. The Court of Appeals affirmed. State v. Derri, 17 Wn. App. 2d 376, 486 P.3d 901 (2021). Relevant here, that court held (1) that even if the identification procedures were suggestive, they were sufficiently reliable, so the trial court did not abuse its discretion in denying Stites’ motion to suppress and (2) that the information was sufficient. Id. In a concurrence, Judge Coburn opined that the identifications were all impermissibly suggestive because Stites was the only person pictured with a neck tattoo but concluded that the identifications were nonetheless reliable under the totality of circumstances. Id. at 412 (Coburn, J., concurring). We granted review of (1) whether the eyewitness identification evidence should have been suppressed on the basis of suggestive photomontage procedures 15 No. 100038-3 and (2) whether the charging document was deficient. State v. Derri, 198 Wn.2d 1017 (2021). ANALYSIS I. Courts must consider the current scientific understanding of the fallibility of eyewitness identification when deciding suggestiveness and reliability issues under Brathwaite As mentioned above, “mistaken eyewitness identification is a leading cause of wrongful conviction.” Riofta, 166 Wn.2d at 371 (citing Garrett, supra, at 60). That is true even here, in Washington. 10 Courts have long recognized the potential unreliability of eyewitness testimony and the unique risks to reliability posed by suggestive police procedures. See, e.g., United States v. Wade, 388 U.S. 218, 228, 87 S. Ct. 1926, 18 L. Ed. 2d 1149 (1967); Stovall v. Denno, 388 U.S. 293, 297-98, 87 S. Ct. 1967, 18 L. Ed. 2d 1199 (1967). In 1977, the United States Supreme Court held that the due process clause of the Fourteenth Amendment compels exclusion of eyewitness identification evidence that (1) was obtained by an unnecessarily suggestive police procedure and (2) lacks reliability under the totality of circumstances. Brathwaite, 432 U.S. at 114; see also U.S. CONST. amend. XIV; Biggers, 409 U.S. at 198; Simmons v. United States, 390 U.S. 377, 384, 88 S. Ct. 967, 19 L. Ed. 2d 1247 (1968); State v. Hilliard, 89 Wn.2d 430, 438-39, 573 P.2d 22 (1977) (adopting 10 See supra note 2. 16 No. 100038-3 Brathwaite test); State v. Vickers, 148 Wn.2d 91, 118, 59 P.3d 58 (2002) (applying Brathwaite test). Under this Brathwaite test, the defendant has the burden to establish, by a preponderance of evidence, that a police-administered identification procedure was unnecessarily suggestive. Vickers, 148 Wn.2d at 118. As the movant, it is in the defendant’s interest to fully develop the record on the issue of suggestiveness. 11 If the defendant shows that the police procedure was unnecessarily suggestive, then the court must consider whether, under the totality of the circumstances, the unnecessarily suggestive procedure created “‘a very substantial likelihood of irreparable misidentification.’” Brathwaite, 432 U.S. at 116 (quoting Simmons, 390 U.S. at 384). The United States Supreme Court explained that “[t]he admission of testimony concerning a suggestive and unnecessary identification procedure does not violate due process so long as the identification possesses sufficient aspects of reliability.” Id. at 106. The United States Supreme Court held that those “aspects of reliability” include the five factors set out in Biggers: (1) the opportunity of the witness to In this case, the defense attached written evidence about the identification procedures, including police interviews with witnesses and one scientific study. However, the defense did not call any witnesses at the suppression hearing and failed to develop the record on certain aspects of the suggestiveness issue. The State challenged some of the assertions in those documents, and the trial court did not resolve factual discrepancies. 11 17 No. 100038-3 view the criminal at the time of the crime, (2) the witness’ degree of attention, (3) the accuracy of the witness’ prior description of the criminal, (4) the level of certainty demonstrated at the procedure, and (5) the time between the crime and the identification procedure. Id. at 114 (citing Biggers, 409 U.S. at 199-200). Where these “aspects of reliability” are “outweighed by the corrupting effect” of law enforcement suggestion, the identification should be suppressed. Id. at 116. Many of Brathwaite’s conclusions about eyewitness identifications relied on the limited available empirical research on the subject. E.g., id. at 116 (citing Simmons, 432 U.S. at 383 n.2 (citing PATRICK M. WALL, EYE-WITNESS IDENTIFICATION IN CRIMINAL CASES 74-77 (1965))). In the 45 years since Brathwaite, researchers have conducted hundreds more empirical studies relating to the reliability of eyewitness evidence. That research has greatly enriched what we know about the accuracy and reliability of witness memory and recall under various conditions. For example, we now know that cross-racial identifications can be particularly unreliable—studies show that rates of error in making identifications are much higher when a person is asked to identify someone of another race. 12 See Christian A. Meissner & John C. Brigham, Thirty Years of Investigating the Own-Race Bias in Memory for Faces: A Meta-Analytic Review, 7 PSYCHOL., PUB. POL’Y & L. 3, 15 (2001); State v. Allen, 176 Wn.2d 611, 625, 294 P.3d 679 (2013) (plurality opinion). 12 18 No. 100038-3 In this case, we must determine whether the trial court erred in declining to consider widely accepted scientific data on the fallibility of eyewitness identification procedures, based on the written factual record presented by the parties and the lack of trial-level adversarial testing of some of the new scientific assertions. We hold that when a trial court uses the Brathwaite test, it must apply relevant, widely accepted modern science on eyewitness identification at each step of the test. See State v. O’Dell, 183 Wn.2d 680, 695, 358 P.3d 359 (2015) (court may adapt legal frameworks by considering “advances in the scientific literature”); State v. Bowman, 198 Wn.2d 609, 633, 498 P.3d 478 (2021) (Yu, J., concurring) (court should look to “empirical data . . . to support and expand on our jurisprudence where appropriate”); Wyman v. Wallace, 94 Wn.2d 99, 102, 615 P.2d 452 (1980) (“[A] court can take notice of scholarly works, scientific studies, and social facts.”).13 A. The three challenged identification procedures were impermissibly suggestive When reviewing the denial of a CrR 3.6 suppression motion, we review the trial court’s findings of fact for substantial evidence and its conclusions of law de As for factors on which there is debate within the scientific community, that debate may continue in the trial court on a case by case basis. 13 19 No. 100038-3 novo. State v. Garvin, 166 Wn.2d 242, 249, 207 P.3d 1266 (2009). We first consider Stites’ argument that the identification procedures used with Amdahl, Fletcher, and Hilen were impermissibly suggestive.14 Researchers have extensively studied the variables that affect the reliability of eyewitness identifications and generally place them in two groups: “system variables” and “estimator variables.” Gary L. Wells et al., Policy and Procedure Recommendations for the Collection and Preservation of Eyewitness Identification Evidence, 44 L. & HUMAN BEHAV. 3, 6-7 (2020) [https://perma.cc/LVQ3-EEV8]. “System variables” are variables under police control when administering identification procedures, meaning they are relevant to the question whether the government used a suggestive identification procedure. State v. Henderson, 208 N.J. 208, 218, 27 A.3d 872 (2011). “Estimator variables” are environmental or individual variables not under the control of the police but “equally capable of affecting an eyewitness’ ability to perceive and remember an event.” Id. at 261. Brathwaite uses the phrase “unnecessarily suggestive.” We have appeared to use the term “impermissibly suggestive” interchangeably with “unnecessarily suggestive.” E.g., Vickers, 148 Wn.2d at 118 (“An out-of-court photographic identification violates due process if it is ‘so impermissibly suggestive as to give rise to a substantial likelihood of irreparable misidentification.’” (quoting State v. Linares, 98 Wn. App. 397, 401, 989 P.2d 591 (1999))). To the extent that these standards are not synonymous, the photomontage procedures here were also unnecessarily suggestive. Certain inherently suggestive procedures may be justified where they are required by exigent circumstances. Brathwaite, 432 U.S. at 109. However, in this case, there were no exigent circumstances justifying the use of suggestive procedures. 14 20 No. 100038-3 Research on system variables has resulted in many clear and widely accepted conclusions. Relevant here, we now know that identification procedures should be administered in double-blind fashion, meaning the administrator does not know who the suspect is. Police should present photomontages sequentially, rather than simultaneously. They should give preidentification admonitions informing the witness that the perpetrator may or may not be in the montage and the witness should not feel compelled to make a selection. They should never show the same suspect to the same witness over the course of multiple identification procedures. They should construct a photomontage in such a way that the suspect is not the only individual pictured who closely matches the description of the perpetrator. And they should avoid giving feedback to witnesses that might inflate confidence levels. 15 In this case, the detective did several things right: he read the witnesses an admonition and administered the montages sequentially. But, as discussed below, the administration of the montages fell short in other significant ways. We hold See, e.g., Wells et al., supra, at 7-28; John T. Wixted & Gary L. Wells, The Relationship Between Eyewitness Confidence and Identification Accuracy: A New Synthesis, 18 PSYCH. SCI. IN PUB. INTEREST 10, 14-17 (2017); Margaret Bull Kovera & Andrew J. Evelo, The Case for Double-Blind Lineup Administration, 23 PSYCH., PUB. POL’Y & L. 421 (2017); Kenneth A. Deffenbacher et al., Mugshot Exposure Effects: Retroactive Interference, Mugshot Commitment, Source Confusion, and Unconscious Transference, 30 LAW & HUM. BEHAV. 287, 299 (2006); Henderson, 208 N.J. at 248-61 (discussing each system variable extensively). 15 21 No. 100038-3 that the montages were impermissibly suggestive for those reasons, which we will discuss below. However, we also hold that on this record, the trial court did not err in determining that the montages were not impermissibly suggestive because Stites was the only person shown with a tattoo. i. Neck tattoo Stites’ main argument is that the identification procedures were suggestive because Stites was the only person pictured with a neck tattoo. We agree with the Court of Appeals that this is a close issue; but we also agree with the Court of Appeals that on this record, the trial court did not err in determining that this factor did not render the identifications suggestive. The Court of Appeals has broadly stated that a photomontage is impermissibly suggestive if it “directs undue attention to a particular photo.” State v. Eacret, 94 Wn. App. 282, 283, 971 P.2d 109 (1999) (per curiam). But that rule is generally applied when the “undue attention” stems from a distinctive feature of the defendant that the witness previously described; in other words, that rule has generally been applied “when the defendant is the only possible choice given the witness’s earlier description.” State v. Ramires, 109 Wn. App. 749, 761, 37 P.3d 343 (2002)) (emphasis added); see, e.g., State v. Kinard, 109 Wn. App. 428, 431, 433, 36 P.3d 573 (2001) (photomontage was impermissibly suggestive where the witness had described the perpetrator as having tooth gap and the witness was 22 No. 100038-3 presented with a photomontage in which only one individual had that feature); State v. Traweek, 43 Wn. App. 99, 103, 715 P.2d 1148 (1986) (lineup was suggestive where witness described perpetrator as blond and defendant was sole blond person in lineup)); United States v. Diaz, 986 F.3d 202, 207 (2d Cir. 2021) (photo array was unduly suggestive where witness described perpetrator as having a large neck tattoo and defendant was only person shown with a neck tattoo); United States v. Kelsey, 440 U.S. App. D.C. 47, 917 F.3d 740, 750 (2019); United States v. Morgan, 690 F. Supp. 2d 274, 289-92 & n.75 (S.D.N.Y. 2010). But see State v. Burrell, 28 Wn. App. 606, 611, 625 P.2d 726 (1981) (photomontage was impermissibly suggestive as to both witnesses where defendant was only person pictured with unique hairstyle described by one witness because that distinctive characteristic made the defendant stand out); Derri, 17 Wn. App. 2d at 414 (Coburn, J., concurring) (discussing Burrell). In this case, Stites was the only person in the photomontages who had a neck tattoo. That is certainly a distinctive characteristic. But it was not previously described by any witness. The witnesses’ failure to describe a neck tattoo makes sense: the robber’s neck would not have been visible during the robberies given the witnesses’ description that his hood was pulled up and cinched around his face. Stites argues that the tattoo was impermissibly suggestive anyway because it made him stand out from the others. And current eyewitness identification 23 No. 100038-3 research agrees that lineup administrators should avoid constructing a lineup so that any photo stands out from the others. E.g., Wells et al., supra, at 8, 19. Most of the research supporting this conclusion, however, considers photomontage pictures showing a suspect’s distinctive physical characteristic that was previously described by, or at least visible to, a witness. See id. at 8, 18-20; Gary L. Wells & Deah S. Quinlivan, Suggestive Eyewitness Identification Procedures and the Supreme Court’s Reliability Test in Light of Eyewitness Science: 30 Years Later, 33 L. & HUMAN BEHAV. 1, 7 (2009). The literature is unanimous that such an identification procedure may be suggestive. That is not the situation presented in this case. And the record lacks sufficient evidence for us to draw the same conclusion about the precise situation that occurred here, where no witness described the characteristic. ii. Double-blind administration Other aspects of the montage procedures were potentially suggestive, though. The first potentially suggestive aspect of the procedures that we consider is the fact that they were administered in a single-blind, rather than in a doubleblind, fashion. In a “double-blind procedure,” neither the lineup administrator nor the witness knows which photo shows the suspect and which photos show only fillers. In a “single-blind procedure,” the administrator, but not the witness, knows which 24 No. 100038-3 photo is the suspect. Research strongly suggests that police should employ doubleblind procedures in administering photomontages. Wells et al., supra, at 14-17. Only double-blind procedures “prevent the tester from unintentionally influencing the outcome of the results.” John T. Wixted & Gary L. Wells, The Relationship Between Eyewitness Confidence and Identification Accuracy: A New Synthesis, 18 PSYCH. SCI. IN PUB. INTEREST 10, 17 (2017). A lineup is a type of experiment in which the police have a theory (that the suspect is the culprit) and develop a hypothesis to test that theory (that the witness will recognize the suspect from the lineup). Margaret Bull Kovera & Andrew J. Evelo, The Case for Double-Blind Lineup Administration, 23 PSYCH., PUB. POL’Y & L. 421, at 423 (2017). As in any experimental scenario, the expectations of the tester (the police) can cue the behavior of the person tested (the witness). Id. Decades of research have found that cues from the tester “can be subtle, transferred unconsciously, and result simply from the expectations of the experimenters.” Id. (citing Robert Rosenthal, Covert communication in classrooms, clinics, courtrooms, and cubicles, 57 AM. PSYCH. 839-849 (2002)).16 Numerous studies show that “single-blind administration of See also Wixted & Wells, supra, at 17; Ryann M. Haw & Ronald P. Fisher, Effects of Administrator–Witness Contact on Eyewitness Identification Accuracy, 89 J. APPLIED PSYCHOL. 1106, 1107 (2004); Steven E. Clark et al., Lineup Administrator Influences on Eyewitness Identification Decisions, 15 J. EXPERIMENTAL PSYCHOL.: APPLIED 63, 66-73 (2009); Sarah M. Greathouse & Margaret Bull Kovera, Instruction Bias and Lineup Presentation Moderate the Effects of Administrator Knowledge on Eyewitness Identification, 33 LAW & HUM. BEHAV. 70, 71 (2009). 16 25 No. 100038-3 lineups increases the likelihood that witnesses will identify the suspect . . . irrespective of whether the suspect is the culprit or an innocent suspect.” Wells et al., supra, at 14 (citations omitted) (citing studies). This effect is found even where the administrator does not consciously believe they are cuing the witness and where the witness does not consciously register any cues. See, e.g., Steven E. Clark et al., Lineup Administrator Influences on Eyewitness Identification Decisions, 15 J. EXPERIMENTAL PSYCHOL.: APPLIED 63, 66-73 (2009). There is also a connection between nonblind procedures and reported rates of witness confidence: “lineup administrators’ own expectations are likely to influence the confidence of the witness even when the lineup administrators are given an objective script to follow and are instructed to not deviate from that script.” Wixted & Wells, supra, at 18; Kovera & Evelo, supra, at 423 (“Two metaanalyses have supported that confirming feedback inflates witnesses’ confidence in their identifications and their reports of the quality of the witnessing conditions. . . . Confidence inflation can occur even if administrators do not directly tell witnesses that they identified the suspect.”). These findings underscore the importance of double-blind procedures. 26 No. 100038-3 Overall, widely accepted research clearly shows that double-blind procedures protect against suggestiveness.17 Other jurisdictions have also accepted this science and held that double-blind administration is a system variable that should be considered when determining whether a lineup was suggestive or reliable. E.g., State v. Lawson, 352 Or. 724, 741, 291 P.3d 673 (2012); Henderson, 208 N.J. at 250. In this case, the lineups were not administered in a double-blind fashion. Detective Carver, the lineup administrator, knew that Stites was the suspect. He administered the lineups in the same room with the witnesses. The record is unclear as to whether Detective Carver knew the order in which the photos were presented to the witnesses. At least one of the detective’s statements, though—the statement to witness Hilen after he made his pick—suggests unconscious confidence-bolstering. 18 If a double-blind procedure is impossible, a “blinding” procedure can be used, in which “an officer who knows the suspect’s identity places single lineup photographs into different envelopes, shuffles them, and presents them to the witness. The officer/administrator then refrains from looking at the envelopes or pictures while the witness makes an identification. This ‘blinding’ technique is cost-effective and can be used when resource constraints make it impractical to perform double-blind administration.” Henderson, 208 N.J. at 249-50. 17 As discussed above, when Hilen made his pick, he said, “the tattoo . . . definitely gives it away,” along with, “just the general boney structure of his chin.” 1 CP at 304. Detective Carver responded, in part, “[W]e didn’t talk about a tattoo with you yet.” Id. at 304-05 (emphasis added). 18 27 No. 100038-3 iii. Repetition of photo of same suspect, or “double exposure” Another potentially suggestive aspect of the eyewitness identification procedures in this case was double exposure. Fletcher viewed two separate montages; Stites’ photo was the only one repeated in both of them. To be sure, the photo contained in the first montage was older, depicting Stites with longer hair and a fuller face compared to the newer photo used in the second montage. The trial court concluded that there was no legal problem with this procedure, opining that “the photos are so different from one another” that the fact that Fletcher was shown two montages featuring the same suspect “did not taint the second montage or draw attention to that picture.” 1 VRP (Mar. 19, 2019) at 108-09. But “social science research indicates that false identification rates increase, and accuracy on the whole decreases, when there are multiple identification procedures.” Young v. Conway, 698 F.3d 69, 78 (2d Cir. 2012) (citing Ryan D. Godfrey & Steven E. Clark, Repeated Eyewitness Identification Procedures: Memory, Decision Making, and Probative Value, 34 LAW & HUM. BEHAV. 241, 241, 256 (2010)). 19 “[S]uccessive views of the same person can make it difficult to See also Henderson, 208 N.J. at 255-56 (citing Deffenbacher et al., supra, at 299 (“[A] meta-analysis of multiple studies revealed that although 15% of witnesses mistakenly identified an innocent person viewed in a lineup for the first time, that percentage increased to 37% if the witness had seen the innocent person in a prior mugshot.”)); Wells et al., supra, at 25 (discussing studies showing that “[s]imply being 19 28 No. 100038-3 know whether the later identification stems from a memory of the original event or a memory of the earlier identification procedure.” Henderson, 208 N.J. at 255 (citing Kenneth A. Deffenbacher et al., Mugshot Exposure Effects: Retroactive Interference, Mugshot Commitment, Source Confusion, and Unconscious Transference, 30 LAW & HUM. BEHAV. 287, 299 (2006)). “This phenomenon is especially pronounced where, as here, the witness initially makes no identification from a photo array, but then selects someone whose picture was included in the photo array during a later identification procedure.” Young, 698 F.3d at 82 (citing Godfrey & Clark, supra, at 247). In fact, the general recommendation against multiple exposures to the same suspect “holds no matter how compelling the argument in favor of a second identification might seem (e.g., the original photo of the suspect was not as good as it could have been).” Wells et al., supra, at 25. Indeed, the Seattle Police Department forbids this practice absent prosecutorial approval. 1 CP at 107.20 exposed to an innocent suspect in a mug book, showup, or a lineup increases the chances of that person being identified in a later lineup even if the witness did not choose the person in the first identification procedure”). The record does not indicate that any such approval was obtained in this case before Detective Carver showed Fletcher and Price an additional photomontage featuring Stites. 20 29 No. 100038-3 Numerous courts have recognized the suggestive effects of multiple viewings of the same suspect. 21 Simmons, 390 U.S. at 383 (danger of misidentification is increased where police show the witness pictures of different persons “among which the photograph of a single . . . individual recurs”); State v. Haugen, 361 Or. 284, 308, 392 P.3d 306 (2017) (noting that “the victim’s identification actually became less reliable through multiple viewings” of the suspect (citing Lawson, 352 Or. at 745)); Jones v. United States, 262 A.3d 1114, 1127 & n.15 (D.C. 2021); State v. Green, 239 N.J. 88, 106, 216 A.3d 104 (2019); Young v. State, 374 P.3d 395, 421 (Alaska 2016); Dennis v. Sec’y, Pa. Dep’t of Corr., 834 F.3d 263, 328 (3d Cir. 2016); Henderson, 208 N.J. at 255; Young, 698 F.3d at 78; Commonwealth v. Gomes, 470 Mass. 352, 376, 22 N.E.3d 897 (2015); There appears to be little recent Washington case law on this subject. In 2012, the Court of Appeals rejected the contention that including a photo of a defendant in two separate photo identification procedures was impermissibly suggestive and faulted the appellant for identifying “no authority . . . that presenting a picture to a witness twice is improper.” State v. Sanchez, 171 Wn. App. 518, 582, 288 P.3d 351 (2012). The Sanchez court cited with approval a 1973 Court of Appeals case that held that an 11-photo array that included two photographs of the defendant, both of which were larger in size than the other photos shown, was not impermissibly suggestive. Id. (citing State v. Smith, 9 Wn. App. 279, 511 P.2d 1032 (1973)). We overrule this holding of Smith because it conflicts both with later-decided, well-reasoned cases discussing why multiple photo identification procedures involving the same suspect are suggestive and with the scientific evidence discussed herein. 21 30 No. 100038-3 Commonwealth v. Collins, 470 Mass. 255, 262 n.9, 21 N.E.3d 528 (2014); Lawson, 352 Or. at 743. Based on this scientific research and legal authority, we conclude that the trial court erred in ruling that this was not really a multiple exposure case. Certainly, the 2015 picture differs from the newer photo: it shows Stites with longer hair and a fuller face. The older photo also reveals just a small section of his neck tattoo. But in the more recent picture, Stites’ distinctive facial characteristics are still present. His deep-set eyes, facial structure, and chin are recognizable. These similarities are sufficient to direct undue attention to Stites’ photo during Fletcher’s second identification. iv. Single-suspect identification In Hilen’s case, the double exposure problem took the form of a preliminary single-suspect “showup.” The detective showed Hilen a photo of the suspect from the March 1 Chase Bank robbery and asked him if the March 1 photo “look[ed] like the same guy” who had robbed HomeStreet Bank. 1 CP at 302. Immediately afterward, the detective showed Hilen the photomontage featuring Stites. Id. Although Stites argued this issue at the suppression hearing, the trial court did not address it. 1 VRP (Mar. 19, 2019) at 37-38. However, this procedure was suggestive in several ways. 31 No. 100038-3 First, showing Hilen the photo of the March 1 robber individually constituted a single suspect identification or “showup,” a practice that has long been “widely condemned.” State v. Rogers, 44 Wn. App. 510, 515, 722 P.2d 1349 (1986). “The showing of a single photograph is, like all identification procedures involving a single suspect, highly suggestive.” Mason v. United States, 134 U.S. App. D.C. 280, 414 F.2d 1176, 1182 (1969). “It is hard to imagine a situation more clearly conveying the suggestion to the witness that the one presented is believed guilty by the police.” Wade, 388 U.S. at 234; State v. Hanson, 46 Wn. App. 656, 666, 731 P.2d 1140 (1987). The Seattle Police Department specifically bars its officers from using single suspect photo identifications. 1 CP at 107. Following the single suspect identification with a montage that included a photo of the person police believed to be the same suspect only increased the suggestiveness of this procedure. It suggested to Hilen that despite any contrary admonition, the police believed that a photo of the suspect—the person whose photo they had just shown him—appeared in the montage. Simmons, 390 U.S. at 383 (danger of misidentification is increased “if the police indicate to the witness that they have other evidence that one of the persons pictured committed the crime” (citing WALL, supra, at 82-83)). Although the surveillance photos were not close-up face shots of the suspect, they were clear enough for Hilen to see that the robber in the photos had a chin similar to that of the March 7 robber. Hilen then 32 No. 100038-3 identified “the general boney structure of his chin” as a reason for selecting the montage photo of Stites. 1 CP at 304. We hold that each identification procedure was suggestive for one or more of the reasons discussed above: the failure to employ a double-blind procedure, multiple exposures to the same suspect, and use of a single suspect showup. But the inquiry does not end there. Because the identifications were impermissibly suggestive, we move to the next step of the Brathwaite test: whether, under the totality of the circumstances, the identifications were so unnecessarily suggestive as to create “‘a very substantial likelihood of irreparable misidentification.’” Brathwaite, 432 U.S. at 116 (quoting Simmons, 390 U.S. at 384). B. Under the totality of the circumstances, all three identifications were nonetheless sufficiently reliable Here, examining the totality of the circumstances, we conclude that there were sufficient indicia of reliability to outweigh the suggestiveness of the procedures. We emphasize that the Biggers factors are not exclusive and that updated scientific evidence relating to the reliability of eyewitness identification (including estimator variables) must be considered when analyzing this step of the Brathwaite test.22 As noted, “estimator variables” are environmental or individual variables not under the control of the police that “are equally capable of affecting an eyewitness’ ability to perceive and remember an event.” Henderson, 208 N.J. at 261. These include, but are not limited to: distance and lighting, witness characteristics (age, visual acuity, 22 33 No. 100038-3 i. Opportunity to view This Biggers factor overlaps with estimator variables relating to duration of observation and to distance and lighting. Each robbery lasted at least a few minutes and occurred in a well-lit bank. 1 CP at 236, 279, 296. During each of the robberies, the robber was directly in front of the tellers for a period of time and his face was visible. Id. at 34, 294, 301. Further, Amdahl and Hilen told police that they recognized the robber as a man who had come into the bank a few weeks before to discuss opening an account. Id. at 300. Amdahl reported having an indepth conversation with the man and writing down his name, John Stites. Id. at 284. Hilen also observed Amdahl and Stites’ conversation at that time and heard Stites’ voice and “stutter[ing]” manner of speaking. Id. at 299. He told police that Stites’ “voice was the same” as the robber’s voice. Id. at 301. With regard to Amdahl and Hilen, this prior interaction with Stites weighs even further toward reliability. intoxication, etc.), stress, presence of a weapon, duration of observation, characteristics of the perpetrator (such as wearing a disguise or mask), delay between observation and identification, cross-racial identification, and suggestions from “co-witnesses and others not connected to the State.” Id. at 261-72 (discussing each variable extensively). 34 No. 100038-3 ii. Degree of attention This Biggers factor overlaps with estimator variables relating to witness characteristics, stress, and presence of a weapon.23 Rather than improve one’s ability to accurately recall, “studies have shown consistently that high degrees of stress actually impair the ability to remember.” Henderson, 208 N.J. at 244 (citing Kenneth A. Deffenbacher et al., A Meta-Analytic Review of the Effects of High Stress on Eyewitness Memory, 28 LAW & HUM. BEHAV. 687, 699 (2004)). Undoubtedly, witnessing a robbery is a stressful event, and Amdahl and Hilen reported experiencing fear during the robbery. 1 CP at 280, 298. However, the witnesses were all able to provide a detailed description of the robber’s appearance, including facial features, height, clothing, and voice, and no witness reported a visible weapon. This factor weighs toward reliability for all three witnesses. iii. Accuracy of prior description of criminal The trial court found that “the descriptions of the bank robber are all sufficiently consistent with the attributes of Mr. [Stites].” 1 VRP (Mar. 19, 2019) at 111. This is accurate, in that Stites appears from his most recent photomontage Research shows that the presence of a visible weapon may also weigh against the reliability of a witness’ memory of a perpetrator because the fear instilled by the weapon directs attention to the weapon rather than the perpetrator. Id. at 263. Here, no teller reported seeing a weapon. 23 35 No. 100038-3 photo to be a thin, adult white man with a “gaunt” or “sunken in” face. All three witnesses provided a reasonably detailed description of the robber before viewing the suggestive photomontages. 1 CP at 258, 236. The fact that no witness described a neck tattoo on the robber is consistent with the witnesses’ reports that the robber had a hood pulled up or cinched around his face. Some researchers have criticized reliance on this factor. E.g., Wells & Quinlivan, supra, at 13. In a case where the witness’ initial description differs from that witness’ later selection from a montage or lineup, the difference would weigh against reliability. But the fact that the witness’ initial description coincides with that witness’ later selection does not necessarily indicate reliability: witnesses “tend to select the person who looks most like their memory of the culprit and will readily select an innocent person if that person fits the eyewitness’ pre-lineup description better than do the lineup fillers.” Id. (citation omitted) (emphasis added). This consistency factor should therefore receive limited weight. iv. Level of certainty at time of procedure All three witnesses reported high levels of certainty—ranging from 90 to 100 percent—at the time they selected Stites’ photo. But while the Biggers and Brathwaite Courts assumed that high levels of certainty correlated with high levels of accuracy, we now know that is not always true. Research indicates suggestive police procedures “severely compromise” the correlation between witness certainty 36 No. 100038-3 and accuracy. Wixted & Wells, supra, at 50. Specifically, suggestive procedures— including the failure to administer a lineup in double-blind fashion—can artificially inflate a witness’ certainty in their identification. Id. at 48. For that reason, high levels of witness certainty should be given little, if any, weight at this step of the Brathwaite analysis, where it has already been determined that the procedure employed was suggestive. 24 However, witness certainty is not wholly irrelevant. Research also shows that under any conditions, a low level of certainty always weighs against reliability. Id. at 14, 20. For that reason, witness certainty should not be entirely eliminated from consideration under the totality of circumstances. See id. at 49. Here, the witnesses’ expressions of certainty were made after being exposed to suggestive identification procedures. The trial court erred in concluding that this factor weighed toward reliability. v. Time between crime and identification This factor overlaps with the estimator variable known as “memory decay.” Memory deteriorates after viewing an event and never improves. Henderson, 208 N.J. at 267 (citing Kenneth A. Deffenbacher et al., Forgetting the Once-Seen Face: Some studies suggest that when nonsuggestive or “pristine” procedures are used, “eyewitness confidence is a highly informative indicator of accuracy.” Wixted & Wells, supra, at 11. However, we reach the second step of the Brathwaite test only once suggestiveness is established, so this finding is not relevant to our analysis. 24 37 No. 100038-3 Estimating the Strength of an Eyewitness’s Memory Representation, 14 J. EXPERIMENTAL PSYCHOL.: APPLIED 139, 142 (2008)). One study suggests that misidentifications substantially increase from 2 to 24 hours after an event. Id. (citing Carol Krafka & Steven Penrod, Reinstatement of Context in a Field Experiment on Eyewitness Identification, 49 J. PERSONALITY & SOC. PSYCHOL. 58 (1985)). Thus, identifications made immediately after a crime may be more accurate, but as time passes, this factor becomes less useful in determining reliability. However, researchers have not pinpointed a precise time at which memory becomes unreliable. Id. Here, Hilen and Amdahl made their identifications one day after the robbery. Fletcher made his identification nine days after the robbery. In this case, this factor does not weigh against reliability. vi. Additional estimator variables Researchers have identified many other estimator variables that can affect reliability. Where relevant, these variables should be considered as part of the totality of the circumstances.25 One such estimator variable is cowitness suggestion. Although it does not affect the suggestiveness of the police procedure under the Brathwaite framework, suggestive behavior by cowitnesses or other nonstate actors can affect the reliability of a witness’ later identification because it is a form of multiple exposure to the suspect that “may cause a person to form a false memory of details that he or she never actually observed.” Henderson, 208 N.J. at 268 (citing Elin M. Skagerberg, Co-Witness Feedback in Lineups, 21 APPLIED COGNITIVE PSYCHOL. 489, 494 (2007)); see also State v. Chen, 208 N.J. 307, 27 A.3d 930 (2011). Stites argues this variable is relevant here because Foss, the HomeStreet Bank manager, claimed that he showed Amdahl and Hilen a Facebook photo 25 38 No. 100038-3 In sum, with regard to Amdahl’s and Hilen’s identifications, the fact that both witnesses claimed to recognize the robber as John Stites, the man who came in to the bank about two weeks before, tips the scale toward reliability. Amdahl interacted with the man for several minutes and wrote down his name. Hilen observed the interaction and heard the man’s distinctive voice. Although Fletcher did not previously interact with the robber, we find that his ability to observe the robber, the degree of attention expressed in his detailed description of the robber, the match between his description and Stites, and the relatively short time between the robbery and the photomontage in which he selected Stites render his identification sufficiently reliable. After carefully examining the totality of the circumstances, we conclude that “the corrupting effect of the suggestive procedure” does not outweigh the additional indicia of reliability present with regard to each witness. Brathwaite, 432 U.S. at 98. of Stites before those witnesses viewed the photomontage. At the suppression hearing, the State proffered that Amdahl was expected to testify that she did not see the photo until after she viewed the photomontage. As the movant, it was Stites’ burden to establish the merit of the motion. He failed to call Foss or Amdahl as witnesses to resolve the inconsistencies in their proffered statements. The trial court did not make an explicit finding regarding either Amdahl’s or Foss’ contradictory statements on this topic but appeared to credit the State’s proffer. On this record, we cannot say the trial court erred in declining to consider the alleged cowitness suggestion. However, if it had been established that Foss showed a photo of Stites to Amdahl and Hilen before they were shown a photomontage, this would weigh against the reliability of Amdahl’s and Hilen’s identifications. 39 No. 100038-3 II. Under Kjorsvik, the information can be fairly construed to contain the essential elements of robbery and Stites does not show actual prejudice from the omission of certain statutory language The second issue is whether the charging document was constitutionally adequate. Stites argues that the information omitted an essential element of the crime of robbery because it did not specify that “‘force or fear must be used to obtain or retain possession of the property, or to prevent or overcome resistance to the taking.’” Pet. for Review at 2 (quoting RCW 9A.56.190). Stites raised this argument for the first time on appeal. An information is constitutionally adequate under the federal and state constitutions “only if it sets forth all essential elements of the crime, statutory or otherwise, and the particular facts supporting them.” State v. Hugdahl, 195 Wn.2d 319, 324, 458 P.3d 760 (2020); U.S. CONST. amend. VI; WASH. CONST. art. I, § 22; CrR 2.1(b). “Essential elements” are “the facts that the State must prove beyond a reasonable doubt to establish that the defendant committed the charged crime.” State v. Recuenco, 163 Wn.2d 428, 434, 180 P.3d 1276 (2008) (citing State v. Johnstone, 96 Wn. App. 839, 844, 982 P.2d 119 (1999)). However, “[c]harging documents which are not challenged until after the verdict will be more liberally construed in favor of validity than those challenged before or during trial.” Kjorsvik, 117 Wn.2d at 102. Under Kjorsvik, we first ask whether the necessary facts appear in any form in the charging document or 40 No. 100038-3 whether they can be found by fair construction therein. Id. at 105. If so, the conviction will not be reversed unless the defendant can show “that he or she was nonetheless actually prejudiced by the inartful language which caused a lack of notice.” Id. at 106. 26 A. The second sentence of the robbery statute constitutes an essential element In relevant part, the amended information alleged: That the defendant . . . did unlawfully and with intent to commit theft take personal property of another, to-wit: U.S. currency, from the person and in the presence of David Fletcher and Chase Bank, who had an ownership, representative, or possessory interest in that property, against his will, by the use or threatened use of immediate force, violence and fear of injury to such person or his property and to the person or property of another, and that he did commit the robbery within and against a financial institution defined in RCW 7.88.010 or RCW 35.38.060, to-wit: Chase Bank; Contrary to RCW 9A.56.200(1)(b) and 9A.56.190. 1 CP at 310.27 If the necessary facts cannot be found by a liberal construction of the charging document, “prejudice is presumed and reversal is necessary.” State v. Zillyette, 173 Wn.2d 784, 786, 270 P.3d 589 (2012) (citing State v. McCarty, 140 Wn.2d 420, 425, 998 P.2d 296 (2000); Kjorsvik, 117 Wn.2d at 105-06). 26 The language in the second and third charges was identical, except that the second count was alleged to have occurred on March 7, 2017, and “from the person and in the presence of Hannah Amdahl, Andrew Hilen, and HomeStreet Bank” and the third count was alleged to have occurred on March 11, 2017, and “from the person and in the presence of Hannah Amdahl.” 1 CP at 310-11. 27 41 No. 100038-3 In order to decide whether the information was sufficient, we must first determine the essential elements of robbery. RCW 9A.56.190 provides: A person commits robbery when he or she unlawfully takes personal property from the person of another or in his or her presence against his or her will by the use or threatened use of immediate force, violence, or fear of injury to that person or his or her property or the person or property of anyone. Such force or fear must be used to obtain or retain possession of the property, or to prevent or overcome resistance to the taking; in either of which cases the degree of force is immaterial. Such taking constitutes robbery whenever it appears that, although the taking was fully completed without the knowledge of the person from whom taken, such knowledge was prevented by the use of force or fear. (Emphasis added.) Stites argues that the first clause of the second sentence of the statute, emphasized above, is an essential element of robbery that needed to be included in the information. Pet. for Review at 13. The State argues that the italicized sentence is merely a definition, not an element. Answer to Pet. for Review at 11-12. Stites’ argument finds support in our case law. In State v. Johnson, Johnson took an item from a store and brought it outside in a shopping cart without paying for it. 155 Wn.2d 609, 610, 121 P.3d 91 (2005) (per curiam). After security guards confronted Johnson, he “abandoned the shopping cart and started to run away.” Id. A guard grabbed Johnson’s arm, and Johnson punched the guard and escaped. Id. Johnson was convicted of first-degree robbery. Id. 42 No. 100038-3 We reversed Johnson’s conviction due to insufficiency of the evidence. Id. at 611. While Johnson used force against the guard, there was no proof that Johnson used that force “‘to obtain or retain possession of the property, or to prevent or overcome resistance to the taking.’” Id. at 610-11 (emphasis omitted) (quoting RCW 9A.56.190). Rather, Johnson used force only after abandoning the property, in order to effect an escape without the property. Id. at 611. We explained that to constitute robbery, “the force must relate to the taking or retention of the property, either as force used directly in the taking or retention or as force used to prevent or overcome resistance ‘to the taking.’” Id. Johnson makes clear that the second sentence of the robbery statute states an essential element of the crime because we reversed the conviction due to failure of proof of that point. See also State v. Allen, 159 Wn.2d 1, 9, 147 P.3d 581 (2006) (holding that to prove aggravating factor of robbery in aggravated first degree murder prosecution, State must prove defendant “used force or fear to take” a cashbox from his mother “or to prevent his mother from resisting the taking”); State v. Todd, 200 Wn. App. 879, 885, 403 P.3d 867 (2017) (holding that “‘[s]uch force or fear must be used to obtain or retain possession of the property, or to prevent or overcome resistance to the taking’” is a statutory element of robbery in the context of a challenge to a to-convict instruction (quoting RCW 9A.56.190, and citing Allen, 159 Wn.2d at 9)). 43 No. 100038-3 The State argues that the clause is merely a definition of an element and thus need not be included in the information. Answer to Pet. for Review at 12 (information is not required to define essential elements). The State relies primarily on the Division One Court of Appeals decision State v. Phillips, 9 Wn. App. 2d 368, 444 P.3d 51 (2019), to support this contention. In Phillips, the court held that an information charging second degree robbery was not constitutionally deficient for failing to “specifically state that [the defendant] had used force or fear to obtain or retain possession of the property at issue.” Id. at 374. The court further held that “the statutory elements of robbery are set forth in the first sentence while sentence[] two . . . [is a] mere definitional statement[].” Id. at 377. According to the Phillips court, the second sentence “defines to ‘obtain’ or ‘retain’ as a form of ‘take,’ as used in sentence one.” The court said our decision in State v. Handburgh, 119 Wn.2d 284, 830 P.2d 641 (1992), supported this view. Phillips, 9 Wn. App. 2d at 377. In Handburgh, we “discussed the interplay between sentences one and two, concluding that ‘a forceful retention of stolen property in the owner’s presence is the type of “taking” contemplated by the robbery statute.’” Id. (quoting Handburgh, 119 Wn.2d at 290). The Phillips court concluded that this language indicates that a “‘retention’ is included within a ‘taking,’” so the purpose of the second sentence is to define “taking” as including 44 No. 100038-3 both obtaining and retaining. Id. The Phillips court also held that “[t]he second sentence defines ‘force,’ and ‘fear,’ as used in sentence one.” Id. Here, the Court of Appeals adhered to its decision in Phillips. Derri, 17 Wn. App. 2d at 387. The State also urges us to embrace Phillips’ view that the second sentence is definitional, but it disagrees slightly as to what the second sentence defines, explaining that “the second sentence defines ‘by the use . . . of’—it explains what it means to take personal property ‘by the use of’ force.” Suppl. Br. of Resp’t at 38-39 (alteration in original). On this view, one “takes” something “by force or fear” when one obtains or retains it by force or fear, or uses force or fear to overcome resistance to obtaining or retaining it. We conclude that Phillips’ reasoning is not persuasive. The second sentence of the robbery statute does more than provide a definition—it expands the range of activity criminalized as robbery. The history of the robbery statute supports this view. Under the common law view of robbery, “the force used during a robbery must be contemporaneous with the taking.” Johnson, 155 Wn.2d at 611 (discussing Handburgh). The first sentence of the robbery statute expresses the common law view: “A person commits robbery when he or she unlawfully takes personal property from the person of another or in his or her presence against his or her will by the use or threatened use of immediate force, violence, or fear.” RCW 9A.56.190. 45 No. 100038-3 But the second sentence of the statute makes clear that Washington has adopted the modern, transactional view of robbery, under which “a taking can be ongoing or continuing so that the later use of force to retain the property taken renders the actions a robbery.” Handburgh, 119 Wn.2d at 290 (discussing Washington’s 1975 amendments to the robbery statute). In other words, the second sentence of the robbery statute expands the range of behavior criminalized as robbery from the common law definition, making clear that “robbery” includes common law robbery (taking by force or fear), plus more (retaining by force or fear). Thus, the second sentence essentially indicates that robbery is an alternative means crime. There are at least two ways to rob someone—taking by force or fear, or retaining by force or fear—but the State must prove only one of those ways to obtain a conviction. Johnson, 155 Wn.2d 609; Allen, 159 Wn.2d 1. Notably, the 11 Washington Practice: Washington Pattern Jury Instructions: Criminal (5th ed. 2021) (WPIC) lists this second-sentence language as an alternative element. The WPIC to-convict instruction for robbery states the relevant elements as “[t]hat force or fear was used by the defendant [to obtain or retain possession of the property] [or] [to prevent or overcome resistance to the taking] [or] [to prevent knowledge of the taking].” 11 WPIC 37.02(4), at 773. The use of the disjunctive “or” supports our reading, too. While the WPICs do not 46 No. 100038-3 control our interpretation of statutes, they are informative. Kjorsvik, 117 Wn.2d at 102 & n.13. We therefore hold that “[s]uch force or fear must be used to obtain or retain possession of the property, or to prevent or overcome resistance to the taking,” RCW 9A.56.190, is a statutory element of robbery and that the State must charge at least one of these means of using force or fear in the charging instrument. B. The necessary facts appear in the information, and Stites shows no prejudice resulting from the language of the information However, even though the second sentence of the robbery statute states an essential element, and even though it was not repeated verbatim in the charging instrument, Stites still cannot meet the requirements of the Kjorsvik test. Under Kjorsvik, “even if there is an apparently missing element, it may be able to be fairly implied from language within the charging document.” 117 Wn.2d at 104 (citing United States v. Ellsworth, 647 F.2d 957, 962 (9th Cir. 1981)). That element can be fairly implied here. As noted, each count of the information alleged that Stites “did unlawfully and with intent to commit theft take personal property of another . . . by the use or threatened use of immediate force, violence and fear of injury to such person or [their] property and to the person or property of another.” 1 CP at 310-11 (emphasis added). Even though the information did not use the second sentence’s language “such force or fear must be 47 No. 100038-3 used to obtain or retain possession of the property, or to prevent or overcome resistance to the taking,” the information included the relevant part of that element—“take . . . by . . . immediate force . . . and fear.” Id. This language “fully informed” Stites of “the nature of the accusations” against him. Kjorsvik, 117 Wn.2d at 101. Stites cannot show prejudice from the State’s failure to recite the statute’s second sentence verbatim. CONCLUSION We hold that courts must consider relevant, widely accepted scientific evidence relating to each step of the Brathwaite test. Applying that test, each of the challenged identification procedures was unnecessarily suggestive. But there are sufficient indicia of reliability present to outweigh that suggestiveness as to all three challenged identification procedures. We further hold that the second sentence of RCW 9A.56.190 states essential, but alternative, elements of the crime of robbery. The necessary facts supporting one of those means—using force or fear to accomplish a taking—appears in the information. Stites’ challenge to the information therefore fails. We affirm the convictions. 48 No. 100038-3 WE CONCUR: Leach, J.P.T. 49