David Smith v. Cynthia Davis

• Court: Court of Appeals for the Sixth Circuit
• Decided: July 31, 2024
• Docket: 23-3604
• Status: Unpublished

Opinion

NOT RECOMMENDED FOR PUBLICATION File Name: 24a0333n.06

No. 23-3604

UNITED STATES COURT OF APPEALS FOR THE SIXTH CIRCUIT FILED Jul 31, 2024 KELLY L. STEPHENS, Clerk ) DAVID M. SMITH, ) Petitioner-Appellant, ) ON APPEAL FROM THE ) UNITED STATES DISTRICT v. ) COURT FOR THE NORTHERN ) DISTRICT OF OHIO CYNTHIA DAVIS, Warden, ) Respondent-Appellee. ) OPINION ) )

Before: COLE, CLAY, and THAPAR, Circuit Judges.

CLAY, J., delivered the opinion of the court in which COLE, J., joined. THAPAR, J. (pp. 27–35), delivered a separate dissenting opinion.

CLAY, Circuit Judge. While investigating the beating and robbery of Quortney Tolliver,

a law enforcement officer presented her with a single photograph of Petitioner David Smith and

told her that he committed the crime, that he had been previously convicted for attempted murder,

and that he wanted her dead. Although Tolliver did not identify Smith as her assailant at the time,

she did positively identify him several months later. Based on the corruptive influence of law

enforcement’s unduly suggestive procedures on Tolliver, Smith filed a motion to suppress

Tolliver’s identification of him as her assailant. The trial court denied Smith’s motion. After

being presented with Tolliver’s eyewitness identification during trial, a jury convicted Smith of

attempted murder, felonious assault, aggravated robbery, and aggravated burglary.

Following a direct appeal of the trial court’s suppression ruling and exhaustion of his state

court options, Smith filed a petition for a writ of habeas corpus in federal court, pursuant to No. 23-3604, Smith v. Davis

28 U.S.C. § 2254. Smith now appeals from the district court’s denial of his habeas petition. For

the reasons set forth below, Smith is entitled to relief on his claim that his due process rights were

violated by the admission of Tolliver’s identification. The identification was obtained through

unduly suggestive means conducive to irreparable mistaken identification, and lacked any strong

indicia of reliability. Because the Ohio Court of Appeals unreasonably applied Supreme Court

precedent related to eyewitness identifications, we REVERSE the judgment of the district court

and GRANT Smith’s federal habeas petition, meaning that Smith is entitled to a new trial.

I. BACKGROUND

A. Factual Background

On October 16, 2015, Quortney Tolliver was attacked with a hammer in her mobile home

in Portage County, Ohio. For approximately two weeks after the incident, Tolliver was

hospitalized and placed in a medically-induced coma due to the severity of her head injuries. As

soon as Tolliver regained consciousness, law enforcement began to interview her to glean what

she remembered—if anything—from the day of her attack. These interviews, occurring in

November 2015, December 2015, and February 2016, are the subject of the suppression motion

on appeal.

The police first attempted to speak with Tolliver on November 2, 2015, approximately two

weeks after the attack. At the time, Tolliver was limited in her communication due to her injuries

and could not verbalize audibly. Communicating through hand signals and writing, Tolliver

reviewed a photo array of 24 black men. These sets of photos did not include one of Smith, who

was not yet a suspect, and Tolliver did not signal that she recognized any of the individuals from

the array. When the police asked if Tolliver had any memory about the incident that occurred on

October 16, 2015, she replied that she had none. At this time, Tolliver also wrote a note to her

2 No. 23-3604, Smith v. Davis

mother that stated, “who did this to me?” There was no indication at this initial interview that

Tolliver knew who assaulted her.

Once Tolliver’s recovery progressed, the police met with her for a second time on

December 9, 2015 “to confirm the identity of a person that [they] had identified as the suspect in

the incident.” Tr. Suppression Hr’g, R. 10-1, Page ID #761. Smith had become the lead suspect,

as his DNA was found to be preliminarily associated with the DNA found mixed with Tolliver’s

blood in the trailer.1 Lieutenant Greg Johnson, Chief of Detectives of the Portage County Sheriff’s

Office, conducted the interview with Tolliver and surreptitiously recorded the conversation.

Tolliver’s mother was also present for this conversation.

Upon entering Tolliver’s room, Johnson declared that he “found out who did this to [her].”

Id. at Page ID #768. Using a large photograph of Smith, Johnson then told Tolliver that the person

was David Smith and asked if she recognized him. Tolliver did not recognize the man in the photo

and asked, “who is that?” Id. at Page ID #825. She eventually admitted that she had met Smith at

least once before and vaguely knew him through a mutual friend. However, Tolliver did not

suggest that the man in the photo could be her assailant. Undeterred, Johnson then began to paint

a disparaging picture of Smith, stating that he had already interviewed Smith, and Smith did not

“have anything good to say about [Tolliver].” Id. at Page ID #771. At that point, Tolliver stated

1 Preliminary association is not conclusive evidence that Smith’s DNA was in the trailer; instead, information derived from mere genetic associations is used by law enforcement only as an investigative lead. In fact, the officer that received this information, Lieutenant Greg Johnson, was warned that preliminary notifications about DNA are “not intended to replace the forensic laboratory’s reported document” because “an additional DNA sample . . . must be obtained for verification by the forensic laboratory.” Tr. Suppression Hr’g, R. 10-1, Page ID #780. Johnson agreed that the preliminary association was far from conclusive, yet he presented the information to Tolliver as though Smith’s DNA was definitively linked to her trailer.

3 No. 23-3604, Smith v. Davis

that she recalled that Smith was supposed to give her a ride to Cleveland the morning of the attack

but still could not remember the day itself, waking up that day, or any details of the incident.

Still determined to obtain some form of a positive identification from Tolliver, Johnson

proceeded to tell Tolliver that Smith wanted her dead and had previously done time in prison for

attempted murder. He further described Smith as “very violent” and “cold-hearted,” explaining

that Smith believed Tolliver deserved to be attacked and that he had left Tolliver to die. Id. at Page

ID #772, 774. When disparagement was not enough to convince Tolliver that Smith committed

the crime, Johnson also told Tolliver that they found Smith’s DNA inside of her mobile home and

that Smith, in a prior interview, denied ever being in the home. Johnson then said, “how the heck

did [Smith’s] DNA get in there unless the DNA fairy [placed it in there], and there’s no such thing

as a DNA fairy.” Id. at Page ID #778. As Johnson continued to elaborate on the evidence that he

had uncovered purportedly linking Smith to the crime, he even stated that “there’s some things

I’m going to tell you, [and there’s] some things I can’t because I don’t want this to have a bad

effect on the trial.” Id. at Page ID #772. Johnson concluded the interview by assuring Tolliver

that he was going to get Smith arrested and would let her know right away when he did so. Overall,

Johnson did not merely suggest that Smith was the perpetrator, but rather explicitly informed

Tolliver several times that Smith committed the crime and tried to kill her.2

Despite Johnson’s relentless attempts to secure a positive identification from Tolliver, she

maintained throughout the interview that she barely knew Smith, had no problems with Smith, and

could not remember the day of the attack at all. Although, towards the end of the conversation,

2 Even Johnson himself admitted during the suppression hearing that, instead of allowing Tolliver to identify the witness herself, he explicitly told her that Smith was the person who attempted to murder her with a hammer.

4 No. 23-3604, Smith v. Davis

Tolliver eventually admitted that she had a dream that a bald black man maced her3 and then

assaulted her with a hammer, she still did not identify Smith as her assailant and continued to

represent that she did not remember October 16, 2015. And importantly, Tolliver did not mention

any dream until Johnson had already shown her a photograph of Smith, identified Smith as a

convicted felon with a history of attempted murder, and emphatically asserted to her that Smith

committed the crime.

Nearly three months later, the next conversation took place on February 29, 2016, soon

after Tolliver was sentenced to probation in her own drug trafficking case. During this

conversation, Tolliver claimed for the first time that she was now “one hundred percent sure” that,

after replaying the dream in her head, her dream accurately reflected the attack. Id. at Page ID

#798, 812. Tolliver explained that Smith was supposed to come over that morning to take her to

Cleveland for a drug deal, so Smith had to be her assailant. For the first time, Tolliver stated that

she actually knew all along that Smith was the perpetrator.

B. Procedural History

Prior to trial, Smith moved to suppress “any eyewitness identification and testimony by

Qourtney Tolliver identifying Defendant out of court and in-court as her assailant on October 16,

2015.” Mot. to Suppress, R. 7-1, Page ID #109. Smith argued that Johnson’s actions were so

impermissibly suggestive that there was a substantial risk of his misidentification. Even in the

state court proceedings, the government admitted that showing a single picture to a person for the

3 There was no evidence either during the investigation or at trial that mace was ever deployed during the attack on Tolliver. And, based on the suppression hearing transcript, it is unclear whether Tolliver’s dream was specifically about Smith or a man who merely looked like Smith. See Tr. Suppression Hr’g, R. 10-1, Page ID #805–06, 832. Further, Tolliver had “so many dreams,” and this dream involving mace was “one of them.” Id. at Page ID #805.

5 No. 23-3604, Smith v. Davis

purpose of identification “has been widely condemned.” Pl.’s Br. In Opposition to Suppression,

R. 7-1, Page ID #131. The state court agreed to hold a hearing on the motion on July 15, 2016.

At the suppression hearing, the various law enforcement officers involved in the interviews

with Tolliver testified. Additionally, Tolliver herself took the stand. She explained that she texted

and called Smith on October 16, 2015, looking for a ride to Cleveland to get some drugs. This

testimony directly contradicted an earlier statement of Tolliver, recalling that Smith initiated

contact on October 16, 2015 by calling her first. Regarding the December 9, 2015 interview,

Tolliver stated that she misled Johnson and actually knew exactly who was depicted in the photo.

She testified that she lied to prevent her mother, who was in the room during the conversation,

from knowing that she was dealing drugs.4 Tolliver maintained throughout the suppression

hearing that, despite the Johnson’s suggestive commentary, her conclusions about Smith

eventually came from her own dreams and independent recollection.

Based on the suppression hearing and following the parties’ subsequent briefing, the trial

court denied Smith’s motion to suppress. The trial court found that Johnson’s presentation of

Smith’s photo was unproblematic. Specifically, the trial court claimed that because Smith was not

a stranger to Tolliver, Johnson’s identification procedures were not suggestive, as “[i]t would be

logical for an investigator to approach the potential identification of a suspect known to the witness

in this manner.” Judgment Entry, R. 7-1, Page ID #167. On appeal, the Ohio appellate court

recognized that this holding was erroneous and properly found that the identification procedure

4 During the December 9, 2015 conversation at which her mother was present, Tolliver stated that she had briefly met Smith through a mutual friend, and that Smith had driven her to Cleveland once to see her brother. Therefore, she did not have to reveal her drug dealing activities in order to state that she recognized Smith and remembered that Smith was her attacker. Yet Tolliver maintained that she did not remember the day of the incident.

6 No. 23-3604, Smith v. Davis

used by Johnson was impermissibly suggestive and unnecessary. Indeed, “Johnson made repeated

disparaging statements about Smith and even said that Smith’s DNA was found in Tolliver’s

bathroom sink mixed with her blood before Tolliver identified him as her attacker.” State v. Smith,

2018-Ohio-4799, 2018 WL 6313398, at *4 (Ohio Ct. App. Dec. 3, 2018). Based on this finding,

the Ohio Court of Appeals attempted to salvage the trial court’s patently incorrect conclusion by

analyzing whether the identification was nonetheless reliable under “the totality of the

circumstances.” Id. at *5–*6.

In evaluating the reliability of Tolliver’s identification, the Ohio appellate court highlighted

that “[a] strong showing of reliability can arise from the fact that a victim knew the perpetrator of

a crime before the crime was committed.” Id. at *6 (quoting State v. Huff, 763 N.E.2d 695, 702

(Ohio Ct. App. 2001)). The court noted that most of the Neil v. Biggers, 409 U.S. 188 (1972)

factors are “better suited for addressing the reliability of the identification of an unknown suspect,”

and that, by contrast, Smith was known to Tolliver. Id. The court then went on to credit that

Tolliver stated in February 2016 that she was 100% certain that Smith was the person who attacked

her. Id. The court also recognized that Tolliver “recalled Smith coming to her door, letting him

in, and him striking her with a hammer.” Id. Based on this incomplete evaluation of the Biggers

factors coupled with Tolliver’s prior awareness of Smith, the Ohio Court of Appeals held that

“notwithstanding Johnson’s improper statements about Smith to Tolliver before she ‘recalled’ the

day in question, her identification of him is nevertheless reliable.” Id.

Judge Grendell of the Ohio Court of Appeals dissented, reasoning that the improperly

suggestive identification procedure resulted in a completely unreliable identification. Id. at *11–

*13. Tolliver failed to identify Smith on multiple occasions and could eventually do so only after

being pressured by Johnson. Id. at *12. Presumably, as a result of having suffered major brain

7 No. 23-3604, Smith v. Davis

trauma, Tolliver was highly susceptible to believing and adopting Johnson’s theories. Id. Further,

Tolliver only became “certain” about the perpetrator of the crime many months after the incident

occurred. Id. Overall, “[t]he victim was entirely inconsistent in her recollection of events.” Id.

Judge Grendell therefore would have held that the totality of the circumstances test in Biggers

weighed heavily in favor of suppression. Id.

After the Ohio appellate court’s ruling, Smith timely appealed to the Ohio Supreme Court,

which declined to exercise jurisdiction. On February 26, 2020, Smith applied for a writ of habeas

corpus, asserting seven grounds for relief. After reviewing and adopting the recommendations of

the assigned magistrate judge, the district court denied the petition and granted Smith a certificate

of appealability related to his claim that the trial court erred in denying his motion to suppress

Tolliver’s identification.

II. STANDARD OF REVIEW

A district court’s denial of a habeas petition is reviewed de novo. Mitchell v. MacLaren,

933 F.3d 526, 531 (6th Cir. 2019). Because Smith’s suppression claim was adjudicated on the

merits in the state court proceedings, the Antiterrorism and Effective Death Penalty Act of 1996

(“AEDPA”) governs this Court’s review. In accordance with 28 U.S.C. § 2254(d), a state court

adjudication on the merits cannot be overturned unless it “(1) resulted in a decision that was

contrary to, or involved an unreasonable application of, clearly established Federal law, as

determined by the Supreme Court of the United States; or (2) resulted in a decision that was based

on an unreasonable determination of the facts in light of the evidence presented in the State court

proceeding.” Haliym v. Mitchell, 492 F.3d 680, 690 (6th Cir. 2007); see also Williams v. Taylor,

529 U.S. 362, 412 (2000). Because Smith does not meaningfully argue that the Ohio Court of

Appeals based its reliability analysis on an unreasonable determination of the facts, we may

8 No. 23-3604, Smith v. Davis

properly focus our analysis on § 2254(d)(1)—whether the Ohio Court of Appeals decision was

contrary to or unreasonably applied Supreme Court precedent.

A state court decision is “contrary to” established Supreme Court precedent “if the state

court arrives at a conclusion opposite to that reached by the Supreme Court on a question of law

or if the state court decides a case differently than the Supreme Court has on a set of materially

indistinguishable facts.” Haliym, 492 F.3d at 690 (alterations omitted) (quoting Williams, 529

U.S. at 412–13). And a state court decision may be considered to have “unreasonably appli[ed]”

federal law where the decision identifies the correct governing legal principle but “unreasonably

applie[s] that principle to the facts of the prisoner’s case.” Id. (quoting Williams, 529 U.S. at 413).

The state court’s ruling must be such an “extreme malfunction” that no reasonable jurist could

deem it consistent with Supreme Court precedent. Harrington v. Richter, 562 U.S. 86, 102–03

(2011). As applicable to this case, Supreme Court precedent mandates that courts weigh the

reliability of an identification against the suggestiveness of law enforcement’s identification

procedure. See Biggers, 409 U.S. at 198–99 (delineating the two-part framework governing

identifications); Manson v. Brathwaite, 432 U.S. 98, 116 (1977) (directing courts to balance the

suggestive procedure against the totality of the circumstances related to reliability); Perry v. New

Hampshire, 565 U.S. 228, 239 (2012) (holding that, where the unduly suggestive procedure

outweighs the indicia of reliability, the identification must be suppressed). Based on these clearly

established standards, the Ohio courts unreasonably applied federal law by failing to weigh the

lack of reliability of Tolliver’s identification against the highly suggestive and coercive procedure.

III. ANALYSIS

Smith argues that the Ohio courts misapplied constitutional law surrounding suggestive

identifications and that Tolliver’s unreliable identification should have been suppressed.

9 No. 23-3604, Smith v. Davis

The admission of evidence derived from a suggestive identification procedure violates a

defendant’s right to due process if the confrontation leading to the identification was “so

unnecessarily suggestive and conducive to irreparable mistaken identification that [the defendant]

was denied due process of law.” Stovall v. Denno, 388 U.S. 293, 302 (1967); see also Foster v.

California, 394 U.S. 440, 442 (1969). To ascertain whether an identification should be suppressed,

a court utilizes a two-step evaluation. Mills v. Cason, 572 F.3d 246, 251 (6th Cir. 2009). In the

first step, it evaluates “whether the procedure was unduly suggestive.” Id. If the identification

procedure is deemed unduly suggestive, then, in the second step, it “uses five factors set out in

Neil v. Biggers to evaluate ‘the totality of the circumstances to determine whether the identification

was nevertheless reliable.’” Id. (quoting Ledbetter v. Edwards, 35 F.3d 1062, 1071 (6th Cir.

1994)).

A. Johnson’s Unnecessarily Suggestive Confrontation Procedures

The first step involves determining whether the procedure the police used was unduly

suggestive. As explained above, neither party disputes that Johnson’s confrontation procedures—

which involved the use of an unnecessarily suggestive single photograph display—were

impermissible. Indeed, a single photo display “has been widely condemned,” Stovall, 388 U.S. at

302, because it “present[s] greater risks of mistaken identification than a lineup,” Moore v. Illinois,

434 U.S. 220, 229 (1977). The Supreme Court has routinely and repeatedly cast doubt upon the

showing of a single photo of a single suspect to a victim of a crime because of the high likelihood

of misidentification. See Manson, 432 U.S. at 109 (agreeing that the police procedure was

suggestive and unnecessary because only one photo was used in the absence of exigent

circumstances); Simmons v. United States, 390 U.S. 377, 383 (1968) (explaining that the chance

of misidentification is heightened where the police display to the witness only a single photograph

10 No. 23-3604, Smith v. Davis

of the individual or state that they have other evidence that the individual pictured committed the

crime); Biggers, 409 U.S. at 199–200 (holding that the police’s use of a showup was unnecessarily

suggestive and evaluating the reliability of the victim’s identification); Biggers v. Tennessee, 390

U.S. 404, 407 (1968) (Douglas, J., dissenting) (“[S]howing a suspect singly to a victim is pregnant

with prejudice. The message is clear: the police suspect this man.”).

In this case, merely referring to Johnson’s procedures as “impermissibly suggestive” is a

gross understatement, given the manipulative nature of Johnson’s tactics that “includ[ed] repeated

attempts to paint Smith in a negative light and describe him as the attacker.” Smith, 2018 WL

6313398, at *12 (Grendell, J., dissenting); see also Mag. J. R&R, R. 26, Page ID #2986

(“Unquestionably, such statements are likely to influence any victim—but especially one who

exhibited memory loss after a head injury—to identify Mr. Smith as the attacker.”). And it is not

as if this coercive procedure was the police’s only option. As Judge Grendell’s dissent pointed

out, “it is entirely unclear why the police in this case could not have followed the proper procedure

and, at the least, presented a photo array.” Smith, 2018 WL 6313398, at *13 (Grendell, J.,

dissenting).

Accordingly, on one side of the balancing scale is a pre-trial identification procedure that

vastly exceeds what the Supreme Court has previously viewed as “unnecessarily suggestive.” This

was not merely the suggestive presentation of a single photo, Manson, 432 U.S. at 108; an improper

lineup where the main suspect stuck out like a sore thumb, Foster, 394 U.S. at 443; or an

unnecessary “showup,” Biggers, 409 U.S. at 195.5 Instead, this was a suggestive single photo

5 “Showups” involve the police displaying a single suspect to a witness. Unlike lineups or photo arrays, “showups are inherently suggestive and are criticized by scholars and courts because displaying a single person to a witness both suggests to the witness that the person is guilty and fails to test whether the witness can independently identify the person.” J.P. Christian Milde, Bare

11 No. 23-3604, Smith v. Davis

display coupled with an unbridled attempt to disparage a suspect and to persuade the victim to

identify him. Johnson did not merely “suggest” that Smith was the perpetrator—he explicitly

advised Tolliver that Smith brutalized her with a hammer, pitted Tolliver against Smith, and

exaggerated every piece of preliminary evidence he had uncovered purporting to tie Smith to the

crime. Accordingly, Supreme Court precedent demands that only an extremely strong showing of

reliability could overcome Johnson’s aberrant, highly corruptive procedure. Perry, 565 U.S. at

239 (“Where the ‘indicators of [a witness’] ability to make an accurate identification’ are

‘outweighed by the corrupting effect’ of law enforcement suggestion, the identification should be

suppressed.” (quoting Manson, 432 U.S. at 116)).

B. The Reliability of Tolliver’s Identification

Because Smith easily succeeds on the first step, this Court’s analysis is principally focused

on reviewing the state appellate court’s application of the second step of the evaluation—

determining whether Tolliver’s identification of Smith was otherwise reliable. When an

identification was unduly suggestive, a court turns to the five factors set out in Neil v. Biggers to

evaluate “the totality of the circumstances to determine whether the identification was nevertheless

reliable.” Ledbetter, 35 F.3d at 1071. The five Biggers factors include:

(1) the opportunity of the witness to 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 by the witness at the confrontation; and (5) the length of time between the crime and the confrontation.

Necessity: Simplifying the Standard for Admitting Showup Identifications, 60 BOS. COLL. L. REV. 1771, 1774 (June 2019).

12 No. 23-3604, Smith v. Davis

409 U.S. at 199–200.6 This totality of the circumstances test should be applied on a case-by-case

basis. Id. at 196; see also Manson, 432 U.S. at 110 (noting that evaluating the totality of the

circumstances “limit[s] the societal costs imposed by a sanction that excludes relevant evidence

from consideration and evaluation by the trier of fact”); Perry 565 U.S. at 239.

In Biggers, the Supreme Court held that an unnecessarily suggestive showup resulting in

an identification of the defendant was nevertheless reliable and admissible because: (1) the victim

spent up to half an hour with her assailant; (2) her original description of the perpetrator’s general

features corresponded with the defendant’s features; (3) she had no doubt that defendant was the

person who raped her; and (4) she testified that the defendant’s face was something she could

never forget. Biggers, 409 U.S. at 200–01. Using these factors, the Supreme Court held that the

several indicators of the witness’ ability to reliably identify the defendant outweighed the

corrupting effect of the showup. And where the strong showing of reliability prevailed, the

Supreme Court held that the admission of the pre-trial identification did not violate the defendant’s

due process rights.

6 Scholars and lawyers alike have identified the need for the Supreme Court to revisit these five factors, as only two are scientifically correlated to a witness’ accuracy in identifying the perpetrator—how well the witness was able to view the perpetrator and the length of time between the initial observation and the identification procedure. Indeed, a witness’ level of confidence largely does not correlate with reliability at all. See, e.g., Milde, supra note 5, at 1776 n.25; Suzannah B. Gambell, The Need to Revisit the Neil v. Biggers Factors, 6 WYO. L. REV. 189, 202 (2006); Haliym, 492 F.3d at 705 n.15 (collecting various articles and studies that illustrate “empirical evidence on eyewitness identification undercuts the hypothesis that there is a strong correlation between certainty and accuracy”); Perry, 565 U.S. at 263–65 (Sotomayor, J., dissenting) (collecting studies that show that eyewitness misidentification is the “single greatest cause of wrongful convictions in this country”).

13 No. 23-3604, Smith v. Davis

A faithful application of the Biggers factors to this case demands a different outcome.

Applying the Biggers factors here, as the state courts largely failed to do, the factors

overwhelmingly illustrate that Tolliver’s identification should have been suppressed. The first

factor—the opportunity of the witness to view the defendant during the crime—clearly favors

suppression. Unlike Biggers, in which the victim had ample opportunity to view her assailant (up

to thirty minutes), Tolliver testified that she “was so busy” that she simply pushed at the handle of

the screen door, immediately turned away to get her coat and shoes, and was hit by a hammer.

Pl.’s Br. In Opposition to Suppression, R. 7-1, Page ID #128, 132. She turned back around to

reach for her Taser, but was hit again and blacked out. At most, she had a few seconds to view

her assailant through the obstruction of a screen door which, under previous identification cases,

is not enough to confer reliability. Cf. Manson, 432 U.S. at 114 (finding that the first factor favored

a reliability finding where the witness had two to three minutes to view assailant and opened the

door for him twice, looking him directly in the eyes and having a short conversation); Simmons,

390 U.S. at 383 (noting that improper employment of photographs by police may cause witnesses

to err in identifying criminals, particularly where the witness “obtained only a brief glimpse of a

criminal, or [saw] him under poor conditions”); Haliym, 492 F.3d at 705 (witness viewed the

suspect at a close range for the duration of the attack and testified that he saw his mother stabbed

thirty times); Howard v. Bouchard, 405 F.3d 459, 472 (6th Cir. 2005) (witness had three

opportunities to view the perpetrator, and one of those opportunities was for more than a minute);

Mills, 572 F.3d at 251 (witness had considerable time to observe the perpetrator where the events

lasted more than an hour).

The second factor—the witness’ degree of attention—favors suppression for the same

reasons. Cf. Ojile v. Smith, 779 F. App’x 288, 294 (6th Cir. 2019) (noting the second Biggers

14 No. 23-3604, Smith v. Davis

factor weighed against reliability because the witness only viewed the perpetrator during a single

brief observation while under the influence of stress or excitement). Tolliver was not focused on

the assailant behind the screen door, but rather she was “busy” and focused on gathering her coat

and shoes. In fact, Tolliver told Johnson that she “did not pay attention to if [the person at the

door] pulled up or not.” Pl.’s Br. In Opposition to Suppression, R. 7-1, Page ID #132. And after

the assailant hit her, she was concentrating on obtaining a self defense weapon, her Taser on the

counter. In contrast to other cases in which witnesses faced and focused on their assailants, the

nature of Tolliver’s assault illustrates that she could not pay the requisite attention to weigh in

favor of reliability. Cf. United States v. Crews, 445 U.S. 463, 473 n.18 (1980) (attaching particular

significance to the fact that the victim viewed her assailant at close range for a period of five to

ten minutes under excellent lighting conditions and “with no distractions”); Biggers, 409 U.S. at

200 (noting that victim faced her rapist “directly and intimately,” as “[s]he was no casual observer,

but rather the victim of one of the most personally humiliating of all crimes”).

The third factor—the accuracy of the witness’ prior description of the defendant—also

weighs against reliability. We have previously held that “[t]he purpose of looking to prior

identifications is to find an indicium of reliability. If [the witness] failed to describe [the suspect]

before being presented with his photo in a suggestive manner, that fact should not be ignored, but

rather cuts in favor of [defendant’s] argument that [the witness] identified him merely because of

the suggestive photo line up.” United States v. Crozier, 259 F.3d 503, 512 (6th Cir. 2001). The

lack of any description prior to the confrontation therefore supports Smith’s argument that Tolliver

identified him merely because of the suggestive photo. At best, this factor is inapplicable here,

where Tolliver never provided a description of her assailant and initially claimed to not remember

the attack at all. Cf. Haliym, 492 F.3d at 705 (“The third factor [is] inapposite. Nothing in the

15 No. 23-3604, Smith v. Davis

record suggests that [the witness] provided a description of the suspect to the police prior to the

line-up.”).

The fourth factor—the level of certainty shown by the witness at the pretrial

identification—can cut both ways. Even after being told that Smith committed the crime, Tolliver

continued to express hesitancy and could not discern her dreams from reality. The Supreme Court

has required an evaluation of the “level of certainty demonstrated by the witness at the

confrontation.” Biggers, 409 U.S. at 199; see also United States v. Washam, 468 F. App’x 568,

577–78 (6th Cir. 2012) (Clay, J., dissenting) (explaining that the district court mishandled the

fourth factor “when it allowed the [witness’] changed testimony to trump their initial

identifications,” as the Biggers factor pertains to the level of certainty at the time of confrontation);

Crozier, 259 F.3d at 512 (finding an indicium of reliability where the witnesses picked the

defendant as the robber “within five seconds of viewing the photo line-up”). Measuring

confidence from the time of confrontation is imperative, as suggestive remarks from law

enforcement consistently inflate eyewitness certainty.7 The record reflects that Tolliver exhibited

an extremely low level of certainty during Johnson’s initial confrontation, as she failed to identify

Smith as her attacker, and switched from stating “who is that” to admitting that she had briefly met

Smith through a mutual friend.

7 Gary L. Wells & Deah S. Quinlivan, Suggestive Eyewitness Identification Procedures & the Supreme Court’s Reliability Test in Light of Eyewitness Science: 30 Years Later, 33 LAW & HUM. BEHAV. 1, 12 (Feb. 2009) (“The problem with using eyewitness certainty as a second-prong reliability factor in Manson-type situations is that it has already been determined (under the first prong) that a suggestive procedure was used with this eyewitness. As with view and attention, we know that confirmatory suggestive remarks from the lineup administrator consistently inflate eyewitness certainty for eyewitnesses who are in fact mistaken.” (collecting studies)).

16 No. 23-3604, Smith v. Davis

However, approximately two months later, Tolliver changed her mind and stated that she

was 100% certain that her assailant was Smith. See Manson, 432 U.S. at 115 (finding a high level

of certainty where the witness testified at the time of confrontation that there “is no question

whatsoever” about the identification). But, again, the relevant time interval is the time of

confrontation, not at a subsequent trial or a later hearing. See Biggers, 409 U.S. at 199; Washam,

468 F. App’x at 578 (Clay, J., dissenting) (“A witness’ level of certainty [after the time of

confrontation], especially after the identified individual has been arrested and indicted, is not part

of the Biggers’ analysis.”). Although Tolliver provided an explanation for her prior hesitation and

testified during the suppression hearing with a high level of certainty that would normally support

reliability, such a finding must be one of only tempered strength, as Tolliver exhibited an utter lack

of confidence at the time of confrontation. And this purported confidence should be further

tempered by the fact that this fourth factor is frequently criticized as one that rarely correlates with

the reliability of a witness’ identification.8

Finally, the fifth factor—the length of time between the initial observation and the

identification—slightly favors suppression, as approximately four months passed before Tolliver

could identify Smith as the aggressor.9 See Biggers, 409 U.S. at 201 (expressing concern about

8 “Confidence by the witness in an eyewitness identification has been shown to be of little significance to accuracy.” Gambell, supra note 6, at 202 (collecting studies); see also Perry, 565 U.S. at 264 (Sotomayor, J., dissenting) (recognizing “confidence is a poor gauge of accuracy”). And, what’s more, “witness confidence of an identification increases with time instead of decreasing.” Gambell, supra note 6, at 202. 9 Scientific studies consistently show that recognition may be extremely high immediately following an event, but then fades quickly. For example, one study found that a witness’ recognition of a photograph after two hours was 100%, while after four months deteriorated to 57%. See Elizabeth F. Loftus & James M. Doyle, Eyewitness Testimony: Civil & Criminal § 3-2 (3d ed. 1997). In addition, the passage of time also can create changes in how the witness

17 No. 23-3604, Smith v. Davis

the lapse of seven months between the crime and the confrontation and noting that this “would be

a seriously negative factor in most cases”); Manson, 432 U.S. at 116 (photographic identification

took place only two days later, and the Supreme Court highlighted that “[w]e do not have here the

passage of weeks or months between the crime and the viewing of the photograph”); id. at 131

(Marshall, J., dissenting) (“[T]he fact is that the greatest memory loss occurs within hours after an

event . . . [i]f the time gap is any greater, reliability necessarily decreases.”); United States v.

McCrary, No. 1:18-cr-26, 2018 WL 3575053, at *9 (S.D. Ohio July 25, 2018) (“There was no time

lag of hours, days, or months, to hinder the [witness’] memory and thwart a positive

identification.”); Crozier, 259 F.3d at 512 (agreeing that the length of time cut in favor of

suppression where “[o]ne month had passed between the robbery and the impermissibly suggestive

photographic line up”). Although the Sixth Circuit has upheld similar time periods of multiple

months where other strong indicia of reliability existed, this case clearly does not involve a strong

showing on any of the other Biggers factors to mitigate against the fifth factor’s weak showing.

See, e.g., Howard, 405 F.3d at 473 (finding that three months is not an excessive length of time

where witness provided a prior accurate description, had ample opportunity to view the perpetrator,

exhibited a high degree of attention, and exhibited confidence by immediately picking the

defendant out of a lineup); United States v. Causey, 834 F.2d 1277, 1285 (6th Cir. 1987) (finding

remembers the original event based on information acquired well after the witnessing event has occurred. This phenomenon is referred to as “post-event influence.” For example, one study found that “after witnessing a clean-shaven person commit an act, participant-witnesses who were given information suggesting that he had a moustache incorporated that information into their later descriptions of the person.” Wells & Quinlivan, supra note 7, at 14. Where Johnson engaged in highly influential and overbearing tactics to convince Tolliver that Smith was her attacker, her recollection is unquestionably susceptible to law enforcement’s information affecting her later reports.

18 No. 23-3604, Smith v. Davis

that three months was not an excessive length of time where witness could clearly see the

perpetrator and provided an unequivocal and very accurate description of the defendant).

On balance, no fair-minded jurist could possibly find that the totality of the Biggers factors

cut in favor of admitting Tolliver’s largely unreliable identification. But the Ohio courts swept

this severe weakness under the rug, either by failing to analyze the Supreme Court’s factors

altogether or by cursorily citing Biggers and halfheartedly devoting a couple of sentences to its

analysis. Nonetheless, we must examine all theories that could have supported the state court’s

conclusion. Because the Supreme Court has permitted a review of the “totality of the

circumstances,” the state court is also arguably permitted to evaluate case-specific factors that may

not fit cleanly into the delineated Biggers factors. Cf. Perry, 565 U.S. at 239 n.5. In this case, the

Ohio Court of Appeals relied heavily upon the fact that Tolliver knew Smith before the crime,

explaining that a strong showing of reliability can arise from the fact that a victim knew the

perpetrator of a crime before the crime was committed. Cf. Haliym, 492 F.3d at 706 (noting that

“[w]itnesses are very likely to recognize under any circumstance the people in their lives with

whom they are most familiar, and any prior acquaintance with another person substantially

increases the likelihood of an accurate identification”).

However, the Ohio Court of Appeals failed to recognize that Tolliver had only seen Smith

one single day in the past. Tolliver testified during the suppression hearing that she did not know

Smith well, she only knew him “just as well as [she] knew everybody else.” Tr. Suppression Hr’g,

R. 10-1, Page ID #819. Every case in this Circuit that has allowed a witness’ prior knowledge of

a suspect to contribute to the reliability of the identification involved a significantly closer

relationship than that of Tolliver and Smith. Cf. Moss v. Hofbauer, 286 F.3d 851, 862 (6th Cir.

2002) (finding identification reliable where the witness had previously seen the suspect on a daily

19 No. 23-3604, Smith v. Davis

basis as a neighbor); Crozier, 259 F.3d at 511 n.2 (explaining that familiarity should operate as a

sliding scale factor, because “the more familiar the person, the more reliable the identification”);

Haliym, 492 F.3d at 706 (highlighting that the witness knew the suspect from four previous

occasions); United States v. Beverly, 369 F.3d 516, 539 (6th Cir. 2004) (witness knew suspect

“very well” and “would readily recognize” him, as the suspect was the witness’ husband); United

States v. Borrelli, 621 F.2d 1092, 1095 (10th Cir. 1980) (holding that a trial court did not abuse its

discretion by permitting identification testimony of defendant’s stepfather who recognized the

defendant in a bank surveillance photo); Bryant v. Turner, No. 2:15-cv-02929, 2016 WL 5930919,

at *8 (S.D. Ohio Oct. 12, 2016) (witness could readily identify the suspect, who was his cousin).

The facts in this case are unquestionably distinguishable. In contrast to a showing of a close

relationship, a few weeks prior to the incident, Tolliver asked her friend, Christina Yeager, for a

ride home from a court date. When Yeager arrived to pick up Tolliver, Smith was driving the car.

The trio then dropped Tolliver’s paperwork at her trailer and took Tolliver to Cleveland to purchase

drugs. This single day was the first and only time that Tolliver had met Smith, making their

incipient acquaintanceship, at best, a minimal contributor to the reliability analysis. Our precedent

has never deemed such a tenuous relationship close enough to confer reliability in the Biggers

analysis.

Importantly, in addition to Smith and Tolliver’s relationship being easily distinguishable

from prior case law, every single case that allowed a prior acquaintanceship to color the reliability

analysis also involved a strong showing of the Biggers factors. In other words, the totality of the

circumstances evaluation mandates that at least a majority of the Biggers factors cut in favor of a

finding of reliability, and only then can prior knowledge of the suspect bolster the reliability

evaluation. Cf. Haliym, 492 F.3d at 705 (finding sufficient reliability where, in addition to several

20 No. 23-3604, Smith v. Davis

prior face-to-face meetings, the totality of the circumstances included a strong showing of (1) a

good opportunity to view the suspect, (2) a heightened degree of attention where witness was

fixated upon the commission of the crime and communicated with the attacker, and (3) a short

period of a few days between the confrontation and the identification). But, as stated above, no

such Biggers showing was or could be made here. And one or two brief face-to-face meetings

cannot overcome the overwhelming lack of reliability indicated by a proper weighing of the

Biggers factors.

The Ohio Court of Appeals also credited Tolliver’s testimony that she remembered all

along that Smith was her attacker, as well as her claim that did not identify him in December 2015

because she did not want her mother to know that she was still dealing drugs.10 This explanation

makes little sense, as she eventually told Johnson that Smith had driven her one time to Cleveland

“to see her brother.” Tr. Suppression Hr’g, R. 10-1, Page ID #803–04. In other words, Tolliver

had already offered an innocent explanation for her limited knowledge of Smith that would have

allowed her to identify Smith as her attacker without revealing that she was trafficking drugs. Yet

she continued to deny that she remembered the day in question or that Smith was her assailant.

Even putting the obvious weaknesses of this February 2016 explanation to the side, an eleventh-

hour explanation cannot save the totality of the circumstances evaluation where the vast majority

of the Biggers factors weigh in favor of suppression. Cf. Kyles v. Whitley, 514 U.S. 419, 444

(1995) (noting that adjustments to a witness’ original story “can be fatal to its reliability”).

10 Although the dissent erroneously attempts to highlight and credit Tolliver’s second explanation for her supposed dishonesty—that she did not want to affect her pending drug cases— the district court did not have this information before it when ruling on Smith’s motion to suppress. At the time of the ruling at issue, the district court only heard testimony related to Tolliver’s alleged wariness of her mother’s presence during her conversation with Johnson.

21 No. 23-3604, Smith v. Davis

Finally, it is worth mentioning that the Ohio Court of Appeals’ opinion suffers from a

glaring omission in its totality of the circumstances evaluation—that Tolliver had suffered serious

brain trauma prior to the unnecessarily suggestive confrontation procedure, which likely rendered

her more susceptible to suggestion and negatively affected several of the Biggers prongs. See

Smith, 2018 WL 6313398, at *12 (Grendell, J., dissenting); see also Jerrod Brown, Interview

Considerations – Traumatic Brain Injury, INT’L ASS’N CHIEFS OF POLICE (Nov. 9, 2022)

(explaining that traumatic brain injuries may increase the risk of suggestibility, rendering a witness

more vulnerable to providing the response that they believe the authority figure wants to hear

rather than stating the truth). In addition to her vulnerable mental state, the chance of

misidentification increases further where the police indicate to the witness that they have other

evidence linking the person pictured to the crime. Heightened suggestibility, coupled with

Johnson’s encouragement to deem Smith as the perpetrator, has been scientifically linked to

confabulation—the development of false memories. See, e.g., Demothenes Lorandos, Adult

Suggestibility, 2 LITIG. HANDBOOK FORENSIC MED., PSYCHIATRY, & PSYCH. § 10:12 (May 2023)

(collecting studies illustrating that photos and suggestion can create false memories over time,

including a study that found that “confirmatory feedback increased false memory for forcibly

confabulated events, increased confidence in those false memories, and increased the likelihood

that participants would freely report the confabulated events 1 to 2 months later”). The Supreme

Court has explicitly recognized that, in these circumstances, “the witness thereafter is apt to retain

in his memory the image of the photograph rather than of the person actually seen, reducing the

trustworthiness of subsequent . . . identification.” Simmons, 390 U.S. at 383–84.

In summary, the state court’s cursory reliability analysis hung its hat on the fact that

Tolliver had met Smith on one occasion prior to the assault and, months after the original

22 No. 23-3604, Smith v. Davis

confrontation with his photo where she was unable to identify him, newly proclaimed that she was

100% certain that Smith was her attacker. Beyond this extremely weak showing of reliability, the

state court then failed to balance these minimal showings against the immensely suggestive law

enforcement procedure. Manson, 432 U.S. at 114, 116. Instead of being merely presented with a

photograph, Tolliver was told exactly who committed the crime, what evidence law enforcement

had to support Smith’s arrest, and that the suspect subsequently threatened and disparaged her.

Faced with such powerful suggestions and evidence, it should be no surprise that she identified

Smith with newly discovered confidence as her assailant months later. And the totality of the

circumstances test mandated by the Supreme Court to probe the reliability of Tolliver’s subsequent

identification overwhelmingly points in favor of suppression. Where eyewitness identifications

rank among the most compelling evidence available to a jury and are the leading cause of wrongful

convictions, such an unreliable identification should never have been presented at trial. The state

court’s failure to suppress this identification testimony was an objectively unreasonable

application of Supreme Court precedent, thus clearing the highly deferential bar imposed by

AEDPA. Cf. Wiggins v. Smith, 539 U.S. 510, 527–28 (2003) (where a state court’s decision

reaches a result not supported by Supreme Court precedent, “[t]he requirements of § 2254(d) []

pose no bar to granting petitioner habeas relief”).

C. The Ohio Court’s Unreasonable Application of Supreme Court Precedent

The manifest injustice in this case stems from the Ohio trial court’s erroneous original

decision that Johnson’s interview procedures, which indisputably ran afoul of Supreme Court and

Sixth Circuit precedent, were warranted and not overly suggestive. Following this error, the Ohio

Court of Appeals attempted to cobble together some semblance of a faithful application of the law

regarding the reliability of Tolliver’s identification. Without engaging in any meaningful analysis

23 No. 23-3604, Smith v. Davis

of the Biggers factors—perhaps because the Biggers factors overwhelmingly favored

suppression—and instead relying upon Tolliver’s nascent acquaintanceship with Smith, the Ohio

court completely failed to salvage the trial court’s erroneous reasoning.11 Ultimately, Smith’s due

process claim is not barred by § 2254, inasmuch as the Ohio court unreasonably applied Supreme

Court precedent in evaluating the reliability of Tolliver’s identification. 28 U.S.C. § 2254(d)(1);

Harrington, 562 U.S. at 102–03.

Although “deference and latitude” must be afforded to the state court in accordance with

AEDPA, fair-minded jurists could not disagree that the scant indicia of reliability in this case

simply cannot outweigh the egregious and highly influential identification procedure utilized by

Johnson. In direct contravention of the Supreme Court’s holding in Foster, Johnson’s coercive

and suggestive tactics “made it all but inevitable that [the victim] would identify petitioner whether

or not he was in fact ‘the man.’” Foster, 394 U.S. at 443. And, even worse than the identification

procedure utilized in Foster, the police disparaged Smith to Tolliver, intentionally painting a

picture of a repeat offender who had served time for attempted murder and wanted Tolliver dead.

11 The dissent criticizes us for “fault[ing] the state court for not conducting a detailed, factor-by-factor Biggers analysis” and for providing our own analysis of the factors. Dissent at 32. But rather than rubber stamping the state court’s constitutionally deficient decision under the guise of federalism or comity, AEDPA requires us to determine whether the state court’s conclusion rested upon an objectively unreasonable application of Supreme Court precedent. We can only make such a determination by examining the facts before us in light of the relevant clearly established law. Cf. Miller-El v. Cockrell, 537 U.S. 322, 340 (2003) (“Even in the context of federal habeas, deference does not imply abandonment or abdication of judicial review. . . . A federal court can disagree with a state court’s . . . determination and, when guided by AEDPA, conclude the decision was unreasonable.”); Harrington, 562 U.S. at 91 (“Judges must be vigilant and independent in reviewing petitions for the writ.”).

24 No. 23-3604, Smith v. Davis

While the dissent attempts to brush aside the clear application of Foster based on

immaterial factual differences, “AEDPA does not ‘require state and federal courts to wait for some

nearly identical factual pattern before a legal rule must be applied.’” Panetti v. Quarterman, 551

U.S. 930, 953 (2007) (quoting Carey v. Musladin, 549 U.S. 70, 81 (2006) (Kennedy, J., concurring

in judgment)); Caver v. Straub, 349 F.3d 340, 350 n.7 (6th Cir. 2003) (explaining that, for the

purposes of § 2254(d)(1), a Supreme Court rule “must be applied in the many factually distinct

situations that will come before the lower courts”). Under the dissent’s interpretation of the

AEDPA bar, federal courts could never rectify even the most obvious, egregious constitutional

errors, as a conniving lawyer can always contrive a factual distinction. But AEDPA does not

require perfect congruence, and applying the dissent’s insurmountable test to this case would be

especially perverse, given that the facts of this case vastly exceed the gravity of the constitutional

violation the Supreme Court found in Foster.

When the enormous weight of this improper identification procedure is balanced against

the dubious traces of Tolliver’s reliability, any fair-minded juror would conclude that the scale

crashes to the side demanding suppression. Accordingly, the Ohio courts’ failure to engage in the

balancing test mandated by clearly established Supreme Court precedent resulted in the objectively

unreasonable admission of eyewitness evidence that jurors view as disproportionately compelling.

Manson, 432 U.S. at 114, 116; Perry, 565 U.S. at 239. Prompted by coercive suggestiveness and

riddled with internal inconsistencies, Tolliver’s identification nonetheless served as the key piece

of the government’s evidence—this error is exactly the “extreme malfunction in the state criminal

justice system” that federal habeas corpus is intended to correct. Harrington, 562 U.S. at 102.

25 No. 23-3604, Smith v. Davis

IV. CONCLUSION

Smith was convicted based on an unduly suggestive and coercive single photo

identification procedure intended to elicit a positive identification from a victim who had just

suffered a severe and debilitating head trauma. Ohio courts impermissibly excused this flagrant

violation of Smith’s right to due process. Not even AEDPA deference can insulate such an

unreasonable application of Supreme Court precedent from proper review and reversal. For the

foregoing reasons, we REVERSE the district court’s denial of the writ of habeas corpus, and

REMAND with instructions that the district court issue Smith a writ of habeas corpus unless the

State proceeds, within 180 days, to prosecute Smith in a new trial without utilizing Tolliver’s

identification of Smith, which shall be suppressed and excluded from evidence.

26 No. 23-3604, Smith v. Davis

THAPAR, Circuit Judge, dissenting. David Smith seeks a writ of habeas corpus. He

argues police improperly influenced a witness to identify him as the perpetrator, and he claims the

state trial court should have suppressed the witness’s identification. The Ohio Court of Appeals

held the identification was reliable and rejected Smith’s claim. The majority disagrees with the

merits of that analysis. But under AEDPA, a disagreement with a state court isn’t enough to grant

relief. Indeed, “even serious errors aren’t enough to warrant relief by themselves.” James v.

Corrigan, 85 F.4th 392, 393 (6th Cir. 2023). Rather, we may grant habeas only if the state court

caused an “extreme malfunction” of the justice system by unreasonably applying clearly

established Supreme Court precedent. Harrington v. Richter, 562 U.S. 86, 102 (2011). Here, the

state court’s decision was reasonable, so Smith isn’t entitled to habeas relief.

I.

David Smith wanted to buy crack and heroin from Quortney Tolliver. Tolliver’s supply

had run out. But Smith was persistent. After texting or calling her more than eighty times, the

pair finally agreed that Smith would drive Tolliver to Cleveland so she could restock. In exchange,

Tolliver would give Smith a deal on the drugs. The plan seemed good enough: in fact, Smith had

driven Tolliver to Cleveland in exchange for drugs two weeks earlier.

Things didn’t go as smoothly this time. When Smith was close to Tolliver’s trailer, he

called to confirm she was ready. Tolliver said she was. So Smith walked up to the trailer, and

Tolliver opened the door and went to get her shoes. But as she turned, Smith pulled out a hammer

and hit her in the head. Tolliver collapsed. Smith hit her again, took her drug money, and left her

for dead.

Tolliver was taken to the hospital, where she fell into a weeks-long coma. When she woke

up, police came to interview her. Tolliver still couldn’t speak due to her injury, so she wrote her

27 No. 23-3604, Smith v. Davis

responses, and her mom read them to the police. Officers showed Tolliver twenty-four photos and

asked her to identify the assailant. Smith wasn’t included in the array, and Tolliver didn’t identify

any of the individuals who were. Officers asked if Tolliver remembered the incident. She signaled

no.

A month later, Lieutenant Greg Johnson visited Tolliver. She was in a nursing home,

accompanied by her mother. Johnson introduced himself and told Tolliver, “I think I found out

who did this to you.” R. 10-1, Pg. ID 768. He showed Tolliver a photo of Smith. At first, Tolliver

asked, “Who is that?” Id. at 825. Seconds later, she confirmed that she recognized Smith, that

Smith had given her a ride before, and that he was supposed to take her to Cleveland the day she

was attacked. Johnson then stated that Smith’s DNA was found in Tolliver’s trailer. He also said

Smith is “very violent” and “hoping you’re dead.” Id. at 771–72. Sometime during the interview,

Tolliver told Johnson she dreamt—or remembered, she wasn’t sure which—that Smith hit her with

a hammer.

Weeks later—four-and-a-half months after the attack—Tolliver was sentenced on separate

drug charges. She asked Lieutenant Johnson to meet her immediately after the sentencing. When

Johnson arrived, Tolliver said she remembered the incident and knew Smith attacked her. Johnson

asked whether Tolliver was sure this wasn’t a dream. Tolliver confirmed she was “one hundred

percent sure.” Id. at 798. Why hadn’t she shared this during the earlier interview? According to

Tolliver, she didn’t want her mom to learn she was selling drugs. And she didn’t want her

admission to affect her then-pending drug charges.

Eventually, Ohio charged Smith with attempted murder, aggravated robbery, and other

crimes. Before trial, Smith moved to suppress Tolliver’s identification, arguing Lieutenant

28 No. 23-3604, Smith v. Davis

Johnson was unnecessarily suggestive when interviewing her. The state trial court denied the

motion, and a jury found Smith guilty.

On appeal, Smith argued the court shouldn’t have admitted Tolliver’s identification. The

Ohio Court of Appeals held that Lieutenant Johnson was unnecessarily suggestive when

interviewing Tolliver. State v. Smith, 2018-Ohio-4799, ¶ 37 (Ohio Ct. App. Dec. 3, 2018). But,

applying Supreme Court precedent, the court concluded that Tolliver’s identification was

sufficiently reliable and didn’t need to be suppressed. Id. ¶¶ 38–49 (citing Neil v. Biggers, 409

U.S. 188 (1972)). When Smith raised this claim in a federal habeas petition, the district court

agreed with the Ohio court’s analysis and denied relief. Smith v. Eppinger, No. 20-CV-00438

(JPC), 2023 WL 4071835 (N.D. Ohio June 20, 2023).

II.

When police use “unnecessarily suggestive” tactics to secure a witness identification, due

process limits the government’s ability to introduce that identification at trial. Biggers, 409 U.S.

at 198. Here, the Ohio Court of Appeals held that Lt. Johnson’s procedures were unnecessarily

suggestive. Smith, 2018-Ohio-4799, ¶ 37. But not all unnecessarily suggestive identifications

must be suppressed. Manson v. Brathwaite, 432 U.S. 98, 106 (1977). If an identification has

“sufficient aspects of reliability,” it may be admitted. Id. Indeed, suppressing a reliable

identification—even one that was obtained through unnecessarily suggestive procedures—would

be a “draconian sanction” that “may frustrate rather than promote justice.” Perry v. New

Hampshire, 565 U.S. 228, 239 (2012) (quoting Manson, 432 U.S. at 113). The Supreme Court

has identified five factors—called the Biggers factors—that help determine whether an

identification is reliable. Biggers, 409 U.S. at 199–200. If a witness had ample opportunity to see

the criminal, was paying attention during the crime, and identified the defendant soon after, her

29 No. 23-3604, Smith v. Davis

identification is more reliable. Id. On the other hand, if the witness is uncertain of her testimony,

or if she failed to accurately describe the defendant until after the suggestive interview, the

identification is less reliable. Id. Throughout it all, the “central question” is whether, “under the

‘totality of the circumstances,’” the identification was reliable even though police were unduly

suggestive. Id. at 199; see also Perry, 565 U.S. at 239 & n.5 (2012) (suggesting the five factors

aren’t exhaustive).

The Ohio Court of Appeals applied Biggers and held that Tolliver’s identification was

reliable. Smith, 2018-Ohio-4799, ¶¶ 38–49 (citing Biggers, 409 U.S. at 199). Accordingly, it

rejected Smith’s suppression claim. Id. ¶ 49. The majority holds that the state court’s decision

unreasonably applied Supreme Court precedent. See 28 U.S.C. § 2254(d)(1). In doing so, the

majority misapplies “the deferential standard of review demanded by the Antiterrorism and

Effective Death Penalty Act of 1996 (AEDPA).” Shoop v. Cunningham, 143 S. Ct. 37, 37 (2022)

(Thomas, J., dissenting from denial of certiorari).

A.

Where (as here) the state court applies the correct governing law to the facts, our review is

extremely limited. Shinn v. Kayer, 592 U.S. 111, 118 (2020) (per curiam). To show that the Ohio

Court of Appeals’ decision was an “unreasonable application” of Biggers, Smith must do more

than establish that the state decision was wrong. Harrington, 562 U.S. at 101–03. He must also

show that the decision was such an “extreme malfunction[]” of the justice system that every fair-

minded jurist would conclude it contradicted Biggers. Id. at 102.

Smith doesn’t clear this high bar. The Ohio Court of Appeals reasonably concluded that

Tolliver’s identification was reliable. As the state court emphasized, Tolliver knew Smith before

the attack. Smith, 2018-Ohio-4799, ¶¶ 47, 67. Smith had previously driven her to Cleveland, and

30 No. 23-3604, Smith v. Davis

the pair communicated more than 80 times the day before the attack. Id. That “substantially

increases” the reliability of Tolliver’s identification. Haliym v. Mitchell, 492 F.3d 680, 706 (6th

Cir. 2007). Moreover, Tolliver had ample opportunity to see Smith and was paying attention to

him before the attack. See Biggers, 409 U.S. at 199. Indeed, she was expecting Smith, opened the

door for him, and turned to get her shoes to leave with him. Smith, 2018-Ohio-4799, ¶¶ 30, 48.

Thus, it makes sense that Tolliver was “one hundred percent sure” Smith attacked her. R. 10-1,

Pg. ID 798; see Biggers, 409 U.S. at 199. Finally, although Tolliver failed to identify Smith until

months after the incident—and initially said she didn’t remember the attack—the court reasonably

credited her explanation for the delay and shift: she didn’t want her mom to learn she was selling

Smith drugs, and she didn’t want the testimony to affect her pending drug charges. Smith, 2018-

Ohio-4799, ¶¶ 34, 48, 59; see Biggers, 409 U.S. at 199. Given these facts, it wasn’t an “extreme

malfunction” for the court to conclude that Tolliver’s identification was reliable.

B.

So how does the majority conclude otherwise? By declining to apply AEDPA’s deferential

standard of review. Instead, the majority conducts its own de novo analysis of the Biggers factors

and then states that no fair-minded jurist could disagree with its conclusion. Majority Op. 14–19

(applying the five Biggers factors and then appending a sentence stating that “no fair-minded jurist

could possibly find that the totality of the Biggers factors” supported Tolliver’s identification).

Again and again, the Supreme Court has rejected this exact approach. Shinn, 592 U.S. at 119

(holding that a reviewing court may not “essentially evaluate[] the merits de novo, only tacking on

a perfunctory statement at the end of its analysis asserting that the state court’s decision was

unreasonable” (quoting Sexton v. Beaudreaux, 585 U.S. 961, 968 (2018))); Harrington, 562 U.S.

31 No. 23-3604, Smith v. Davis

at 101–02 (similar). The majority pursues that precise approach today despite the Supreme Court’s

unambiguous prohibition.

The majority then tries to show the state court’s analysis was more than just wrong. But

those efforts come up short.

First, the majority faults the state court for not conducting a detailed, factor-by-factor

Biggers analysis. It says the state court “largely failed” to apply Biggers by “halfheartedly

devoting a couple sentences to its analysis” or by “failing to analyze the Supreme Court’s factors

altogether.” Majority Op. 14, 19. But the Ohio Court of Appeals cited and quoted extensively

from Biggers, laid out its factors, and analyzed its “central question”: whether Tolliver’s

identification was reliable under the totality of the circumstances. Smith, 2018-Ohio-4799, ¶¶ 37–

49 (citing Biggers and considering “the applicable factors”). It “strains credulity” to believe the

court refused to apply a test “it had taken the trouble to recite.” Early v. Packer, 537 U.S. 3, 9

(2002) (per curiam). More fundamentally, the majority errs by targeting the state court’s reasoning

rather than its bottom-line decision. It doesn’t matter how many sentences the state court used to

conduct its Biggers analysis. Rogers v. Mays, 69 F.4th 381, 391–92 (6th Cir. 2023) (en banc),

cert. denied sub nom. Rogers v. Pounds, 144 S. Ct. 830 (2024). We review the state court’s

“decision,” not the comprehensiveness of its reasoning. Id. Indeed, the majority’s analysis

disregards the Supreme Court’s reminder that we “have no authority to impose mandatory opinion-

writing standards on state courts.” Johnson v. Williams, 568 U.S. 289, 300–01 (2013). The Ohio

Court of Appeals didn’t need to offer any explanation for its decision—let alone a multi-paragraph

discussion of each Biggers factor. See Harrington, 562 U.S. at 98.

Second, the majority criticizes the state court for emphasizing that Tolliver had known

Smith before he attacked her. According to the majority, one previous interaction isn’t enough to

32 No. 23-3604, Smith v. Davis

confer reliability. At most, then, we’re left with a disagreement between the state court and the

majority about how to weigh evidence. So who wins? Under AEDPA, the answer is clear: we

defer to the state court when no clearly established Supreme Court precedent stands in the way.

28 U.S.C. § 2254(d)(1); see Harrington, 562 U.S. at 102 (asking whether the arguments supporting

a state court’s decision clearly contradict “the holding in a prior decision of [the Supreme] Court”).

And here, the majority can’t find a single Supreme Court case holding that a court can’t consider

the victim’s previous interaction with her attacker in its reliability analysis. See Majority Op. 19–

20. There is no such case. And it’s impossible to conclude that state court a misapplied a “clearly

established” Supreme Court precedent when no such precedent exists. 28 U.S.C. § 2254(d)(1).

Moreover, our circuit previously held that a victim’s prior acquaintance with her attacker

is relevant to the reliability analysis. Haliym, 492 F.3d at 706 (“[A]ny prior acquaintance with

another person substantially increases the likelihood of an accurate identification.”).1 In assessing

whether Tolliver’s prior acquaintance with Smith made her identification more reliable, the state

court wasn’t unreasonable for taking the same view as our own circuit precedent.

Finally, the majority argues that the state court’s application of Biggers was unreasonable

because it resulted in a decision that contravened Foster v. California. 394 U.S. 440 (1969). But

a fair-minded jurist could conclude that Foster is distinguishable and doesn’t dictate the outcome

1 The majority contends that relying on a victim’s prior knowledge of the suspect is only allowed if a majority of the Biggers factors already support reliability. Majority Op. 20–21. Even if this were an accurate statement of our circuit precedent, which it’s not, the majority points to no Supreme Court case holding as much. On the contrary, Supreme Court precedent allows courts to consider facts that fall outside the Biggers factors—as the majority itself acknowledges. Majority Op. 19 (“Because the Supreme Court has permitted a review of the totality of the circumstances, the state court is also arguably permitted to evaluate case-specific factors that may not fit cleanly into the delineated Biggers factors.” (internal quotation omitted)).

33 No. 23-3604, Smith v. Davis

here. In that case, a witness saw a tall man in a leather jacket rob a bank. 394 U.S. at 441. Police

later asked the witness to identify the robber from a lineup of three people. However, only the

defendant was wearing a leather jacket, and the other two men were half a foot shorter. Id. When

that didn’t work, police allowed the witness to talk one-on-one with the defendant. Id. When that

still didn’t work, police showed the witness a five-person lineup. Of the five participants, only the

defendant had appeared in the earlier lineup. Id. at 441–42. As a result, the witness identified the

defendant as the robber. Id. at 442. But because the officers’ tactics made it “virtually inevitable”

the witness would pick the defendant, the Court suppressed the identification. Id. at 443.

True, as the majority points out, the identification procedures used here were not a model

of police behavior. The police said negative things about Smith to Tolliver and showed Tolliver

Smith’s picture. For its part, the state court transparently acknowledged these procedural defects.

However, as the state court recognized, there is a crucial difference between Smith’s case and

Foster: unlike the victim in Foster, Tolliver had been acquainted with her attacker (Smith) before

the attack occurred. That makes Tolliver’s identification more reliable than the identification in

Foster. See Haliym, 492 F.3d at 706 (“[A]ny prior acquaintance with another person substantially

increases the likelihood of an accurate identification.”); see also United States v. Crozier, 259 F.3d

503, 511 & n.2 (6th Cir. 2001) (similar). The state court rested on this distinction and held that

the identification was reliable because of Tolliver’s prior acquaintance with Smith. Again, merely

disagreeing with that determination isn’t grounds on which we may grant habeas. The point of

AEDPA is to leave all but the most indisputably obvious legal questions to the judgment of state

courts, subject to direct appeal, and out of the hands of federal courts on collateral review.

Entertaining the merits of debatable legal questions on collateral review disregards the respect for

federalism that Congress has mandated. Because the Ohio Court of Appeals’ decision to consider

34 No. 23-3604, Smith v. Davis

prior acquaintance didn’t directly contradict any holding of the Supreme Court, it wasn’t

unreasonable for the Ohio Court of Appeals to conclude Tolliver’s identification was more reliable

than the one in Foster.

* * *

I would affirm the district court’s decision. For that reason, I respectfully dissent.