Morales v. United States

Notice: This opinion is subject to formal revision before publication in the
Atlantic and Maryland Reporters. Users are requested to notify the Clerk of the
Court of any formal errors so that corrections may be made before the bound
volumes go to press.

DISTRICT OF COLUMBIA COURT OF APPEALS

No. 18-CF-734

CESAR MORALES, APPELLANT,

V.

UNITED STATES, APPELLEE.

Appeal from the Superior Court
of the District of Columbia
(CF3-15439-17)

(Hon. Jennifer A. Di Toro, Trial Judge)

(Submitted April 23, 2020 Decided April 8, 2021)

Richard S. Stolker was on the brief for appellant.

Jessie K. Liu, United States Attorney at the time the case was submitted, and
Elizabeth Trosman, Monica Trigoso, and Steven B. Snyder, Assistant United States
Attorneys, were on the brief for appellee.

Before EASTERLY and DEAHL, Associate Judges, and FISHER, Senior Judge. *

Opinion of the court by Associate Judge DEAHL.

Opinion by Senior Judge FISHER, dissenting in part, at page 49.

*
Judge Fisher was an Associate Judge of the court at the time of submission.
His status changed to Senior Judge on August 23, 2020.
 2

DEAHL, Associate Judge: This appeal raises a challenge to the government

eliciting an in-court identification. The salient facts are: (1) a police officer caught

a two-to-three second glimpse of a fleeing suspect’s face more than four months

before trial, (2) the officer did not identify the defendant before trial—no show-up,

lineup, photographic array, or any other type of pretrial identification procedure was

conducted—but (3) on the cusp of trial, the prosecuting attorney handed the officer

a single mugshot of the defendant to study “in preparation for this case.” With that

backdrop, defendant Cesar Morales moved to preclude the government from

attempting to elicit an in-court identification from the officer given the suggestivity

of the pretrial mugshot display. The trial court denied the motion and permitted the

government to elicit an in-court identification. Although describing it as “a close

call,” the court concluded that the mugshot display was not “unduly suggestive.” It

further reasoned that any suggestivity would not impact the reliability of an in-court

identification.

We disagree with the court’s assessment both as to the suggestivity of the

mugshot display and as to the reliability of the officer’s in-court identification. First,

suggestivity is not a close call. Asking a witness on the eve of trial to inspect a

mugshot of the defendant—a stranger whom he had seen only fleetingly and had
 3

never previously identified—is beyond suggestive. This was no mere hint that the

mugshot was of the suspected culprit. It was an inescapable communication to that

effect. We have aptly described single mugshot displays as among “the ‘most

suggestive’ and therefore the ‘most objectionable method[s] of pretrial

identification.’” Patterson v. United States, 384 A.2d 663, 666 (D.C. 1978) (quoting

United States v. Dailey, 524 F.2d 911, 914 (8th Cir. 1975)). The government does

not now dispute that the pretrial mugshot display was unduly suggestive.

Second, having determined the pretrial mugshot display was unduly

suggestive, whether an in-court identification should have been permitted depends

on whether it was nonetheless reliable despite the “corrupting effect of the

suggestive” procedure. Manson v. Brathwaite, 432 U.S. 98, 114 (1977). That is

typically true when, for example, the witness had previously identified or was well-

acquainted with the defendant before being exposed to the suggestive identification

procedure. See, e.g., United States v. Brannon, 404 A.2d 926, 928 (D.C. 1979) (prior

identification); Green v. United States, 580 A.2d 1325, 1327 (D.C. 1990) (named

defendant beforehand and was familiar with him). No similar indicia of reliability

are present here. There was no previous identification; the officer and Mr. Morales

were complete strangers; and the officer’s opportunity to observe the suspect was

fleeting and occurred more than four months before trial. We cannot say, under
 4

those circumstances, that the in-court identification was reliable and untainted by the

suggestivity of the mugshot display.

It was constitutional error to permit the officer to make an in-court

identification of Mr. Morales. The government makes no argument the error was

harmless, and so we reverse Mr. Morales’ convictions.

I.

A jury convicted Mr. Morales of assault with a dangerous weapon and related

weapons charges 1 for a shooting that occurred on August 25, 2017. Surveillance

footage from that afternoon showed two men chasing another into the Columbia

Heights Metro at 14th and Irving Streets NW. As the man being chased started

1
He was convicted of assault with a dangerous weapon, D.C. Code § 22-402
(2020 Supp.), possession of a firearm during the commission of a crime of violence,
D.C. Code § 22-4504(b), unlawful possession of a firearm, D.C. Code § 22-
4503(a)(1), possession of an unregistered firearm, D.C. Code § 7-2502.01(a) (2018
Repl.), and unlawful possession of ammunition, D.C. Code § 7-2506.01(3). He was
sentenced to a total of five years in prison. He also entered a post-trial guilty plea to
one charge of escape from an institution, D.C. Code § 22-2601(a)(1), the lone
conviction we do not reverse because he waived his right to appeal it, as explained
further below. He received an additional ten-month sentence—though the sentence
was suspended—for that offense.
 5

down the escalator, one of his pursuers fired a gun at him from close range. The

shooter’s target was unscathed; he left the scene and was never located. It was broad

daylight and dozens of bystanders immediately fled the area in response to the shot.

Right after the shooting, a police dispatcher broadcast a lookout for “possibly a

Hispanic male” (the suspected shooter) and “a black male with dreads” (his

companion). Officers Doran Gunnells and Alicia Weber were in the immediate

vicinity and, after speaking briefly with witnesses, mounted their bicycles and began

canvassing the area.

Within minutes and about three blocks from the shooting, Officers Gunnells

and Weber saw two men half a block ahead of them who matched the lookout

description. The officers followed them with caution, as they presumed one or both

to be armed. As the two men approached the intersection of 16th Street and Park

Road NW, Officer Gunnells, from a distance of ten to twenty-five feet, 2 directed

them to show their hands. The Hispanic man “glance[d] back real quick,” allowing

2
While seated in the witness stand during direct examination, Officer
Gunnells was asked how close he got to the individual. He estimated the distance
as “[f]rom here to that podium,” which the trial judge estimated as twenty-five feet
without objection. The trial judge later clarified for counsel, and then the jury, that
the distance was closer to twenty rather than twenty-five feet. On appeal, the
government suggests that based on the video evidence the distance was, in fact,
closer to ten feet. We do not resolve this factual discrepancy because our holding
is not affected by it.
 6

Officer Gunnells to see his face from the front or profile for “[t]wo or three seconds,”

and then fled north on 16th Street.

After about five seconds of running, the suspect paused and reached into his

waistband, prompting Officer Gunnells to point his gun at the man while repeating

“Let me see your hands!” Undeterred and non-compliant, the man ran again and

Officer Gunnells gave chase on foot. The suspect then threw a silver object into a

nearby yard, and upon recognizing it as a gun, Officer Gunnells directed Officer

Weber (following close behind him) to stay with it as he continued the pursuit. The

chase continued for another block or two, but Officer Gunnells relented when the

suspect jumped over a gate into someone’s backyard. He then promptly broadcast

an updated lookout for the man, describing him as “a Hispanic male, a beanie on his

head, tattoos on his face and neck, and a jacket tied around his waist.” Officer

Gunnells estimated at trial that the total time he viewed the suspect, from the moment

he first saw him to the moment he lost sight of him after jumping the fence, was

“approximately sixty seconds.” During those sixty seconds, he saw the suspect

primarily from behind, and indicated the only view he had of the man’s face came

during the two-to-three second backward glance preceding the flight.
 7

Six days later, Mr. Morales was arrested. According to evidence introduced

at the preliminary hearing, but not presented to the jury, officers received “an

anonymous tip” from somebody who identified Mr. Morales as the shooter, leading

to further investigation and then his arrest. In the months that followed, Officer

Gunnells was never asked to view a photographic array, witness a lineup, or

otherwise confirm the identity of Mr. Morales as the person he chased after the

shooting.

On December 21, 2017—approximately four months after the shooting and

Officer Gunnells’ failed attempt to apprehend the suspect—the Assistant United

States Attorney handling the trial met with Officer Gunnells in preparation for his

testimony. When Officer Gunnells informed her that he could not describe the

tattoos he observed on the suspect’s face and neck, she showed him a mugshot of

Mr. Morales and the two of them “started looking at the tattoos more closely.” It

was not until after this mugshot display that, in her words, she “realized, oh, I’m

probably going to want to ask [Officer Gunnells] for an in-court identification.” And

so, about two weeks later and two days before the start of trial, she disclosed the

mugshot display to defense counsel.
 8

Defense counsel promptly filed a motion to suppress the out-of-court

identification and any subsequent in-court identification by Officer Gunnells. He

argued the pretrial mugshot display was unduly suggestive and gave rise to a

“substantial likelihood of misidentification.” As a result, defense counsel argued the

government should be precluded from eliciting an in-court identification from

Officer Gunnells and from introducing any account of the pretrial procedure itself.

The government responded that it did not intend to elicit any details about the pretrial

mugshot display, but that it did expect to ask Officer Gunnells for an in-court

identification. In defense of its decision to show Mr. Morales’ mugshot to its

witness, the government explained the officer “could have viewed this picture on his

own” outside the prosecutor’s presence, and in any event, “a person sitting next to

defense counsel”—as is typical of in-court identifications—“is always suggestive”

but that “doesn’t affect the reliability” of the in-court identification. The trial court

denied the suppression motion. It reasoned that, while it was “a close call with

regard to suggestivity,” the mugshot display was not unduly suggestive and, in any

event, it did not “impact[] the reliability of the identification.”

The government’s case at trial consisted largely of video footage, physical

evidence, and Officer Gunnells’ in-court identification. It showed the jury

surveillance footage of the shooting itself—captured by a security camera at a nearby
 9

apartment complex—and surveillance footage from inside the Columbia Heights

Metro Station, which showed bystanders reacting to the shooting and the suspected

victim fleeing through the station’s corridors. The government also introduced

Officer Gunnells’ body-worn camera footage, and after doing that, asked him if the

individual he had chased was in the courtroom. He responded affirmatively and

identified Mr. Morales over defense counsel’s repeated objection. The government

also introduced Officer Weber’s body-worn camera footage. Notably, it did not

attempt to elicit an in-court identification from her, despite her proximity to Officer

Gunnells when the suspect turned to look at the officers and her testimony that she

was “maybe a car length” away from the suspect when she got a “side view” of his

face.

The government also admitted a silver .22 caliber semiautomatic handgun

recovered from a front yard along the path of the chase on 16th Street. A firearms

expert opined unequivocally, and without objection, 3 that a shell casing found on the

3
But see Gardner v. United States, 140 A.3d 1172, 1177 (D.C. 2016) (“[I]n
this jurisdiction a firearms and toolmark expert may not give an unqualified opinion
. . . that based on ballistics pattern comparison matching a fatal shot was fired from
one firearm, to the exclusion of all other firearms.”); Williams v. United States, 210
A.3d 734, 743 (D.C. 2019) (“[I]t is error to allow an examiner” to provide “opinion
testimony that unqualifiedly connects a specific bullet to a specific gun.”).
 10

steps of the Columbia Heights Metro Station had been fired from that same handgun.

A DNA analyst also testified that a swab of the gun showed a mixture of three DNA

profiles, including Mr. Morales’ (opining that the odds of the profile originating

from another person were 1 in 41.9 quadrillion).

The defense presented only Mr. Morales’ testimony. He acknowledged

running from Officer Gunnells on 16th Street, but maintained that he had not fired

the gun at the Columbia Heights Metro Station. He testified that as he was leaving

a convenience store on August 25, 2017, he was approached by a nervous and

sweating man named José René, whom Mr. Morales knew as a gang member

associated with MS-13. Mr. René directed Mr. Morales to switch clothes with him,

take the gun he was holding, walk away with his companion—a black man with

dreads 4—and hide the gun somewhere. Mr. Morales testified that he did what Mr.

René asked because he was scared. He also acknowledged that he had previously

told detectives that he was the shooter at the Metro station, explaining that he did so

because he feared Mr. René would retaliate against him and his family had he told

the truth. Mr. Morales was also impeached with prior convictions for assault with a

4
The “black male with dreads,” as he was described in the police lookout, was
never apprehended and never identified.
 11

dangerous weapon, possession of a firearm during a crime of violence, possession

of an unregistered firearm, unlawful possession of a firearm or destructive device,

and unlawful possession of ammunition.

The jury found Mr. Morales guilty on all counts. Following trial, Mr. Morales

entered a guilty plea to one count of escape from an institution as he had absconded

from a halfway house before the shooting at issue here. That plea included a waiver

of his right to appeal the escape conviction. He now brings this timely appeal.

II.

Mr. Morales argues that the government violated his Fifth Amendment due

process rights when it elicited an in-court identification from Officer Gunnells. In

his view, the pretrial mugshot display was “so suggestive as to undermine the

reliability and admissibility of [his] in-court identification.” We agree.

More than fifty years ago, the Supreme Court held in a trilogy of cases decided

on the same day that improper pretrial identification procedures affect not merely

the weight, but the very admissibility of a subsequent in-court identification. United

States v. Wade, 388 U.S. 218, 233–38 (1967); Gilbert v. California, 388 U.S. 263,
 12

272 (1967); Stovall v. Denno, 388 U.S. 293, 301 (1967). That was a significant

change to the doctrinal landscape. Before this trilogy, any doubts about an

identification’s reliability were considered exclusively “a matter of weight and

credibility to be explored on witness examination and argued to the jury.” Patterson

v. United States, 384 A.2d 663, 667 n.6 (D.C. 1978). The trilogy’s cornerstone was

Wade which, unlike this case, concerned the Sixth Amendment right to counsel at

an in-person identification procedure. Despite its Sixth Amendment grounding,

Wade provides critical backdrop for the Fifth Amendment due process questions

raised here, so we detail its holding.

Wade held that a post-indictment lineup, during which a witness is asked if

they can identify a suspect from among some number of individuals, was a “critical

stage of the prosecution” at which the defendant was “as much entitled to” counsel

“as at the trial itself.” 388 U.S. at 237 (quoting Powell v. State of Alabama, 287 U.S.

45, 57 (1932)). In doing so, the Supreme Court acknowledged the “well-known”

“vagaries” of eyewitness identifications and explained that a “major factor

contributing to the high incidence of miscarriage of justice from mistaken

identification has been the degree of suggestion inherent in the manner in which the

prosecution presents the suspect to witnesses for pretrial identification.” Wade, 388

U.S. at 228. It also emphasized that “the influence of improper suggestion upon
 13

identifying witnesses probably accounts for more miscarriages of justice than any

single factor—perhaps it is responsible for more such errors than all other factors

combined.” Id. at 229 (citation and internal quotation marks omitted). The

observations made in Wade reflect a simple, but critical, point: Government-

arranged pretrial identification procedures are “peculiarly riddled with innumerable

dangers and variable factors which might seriously, even crucially, derogate from a

fair trial.” Id. at 228.

The Court in Wade further reasoned that, when the right to counsel at a lineup

is violated, it is an insufficient remedy to simply exclude evidence of the lineup

identification itself. Id. at 240. Such a “rule limited solely to the exclusion of

testimony concerning identification at the lineup itself, without regard to

admissibility of the courtroom identification, would render the right to counsel [at a

lineup] an empty one.” 5 Id. Still, the Court did not adopt a per se rule excluding in-

court identifications that follow such Sixth Amendment violations either. Id.

Instead, Wade held that the government may elicit a subsequent in-court

5
That is because once a witness has identified a suspect before trial “he is not
likely to go back on his word later on, so that in practice the issue of identity may
(in the absence of other relevant evidence) for all practical purposes be determined
there and then, before the trial.” Wade, 388 U.S. at 229 (citation and internal
quotations omitted).
 14

identification only if it “had an independent source,” id. at 240–42, which is to say,

if the government can “establish by clear and convincing evidence that the in-court

identifications were based upon observations of the suspect other than” the

constitutionally deficient pretrial procedure. Id. at 240; see also Gilbert, 388 U.S. at

272. It then noted a host of factors relevant to that inquiry, including the witness’s

opportunity to observe the suspect at the scene, the accuracy of any pre-identification

description, and the lapse of time between the act and the identification procedure.

Wade, 388 U.S. at 241 & n.33.

The following year, the Supreme Court confronted a similar question, more

closely approximating the one before us now, under the Fifth Amendment’s Due

Process Clause: When is a pretrial photographic array—similar to a lineup, but with

pictures—so suggestive as to require preclusion of a subsequent in-court

identification? See Simmons v. United States, 390 U.S. 377, 382–85 (1968). It held

that “convictions based on eyewitness identification at trial following a pretrial

identification by photograph will be set aside” as violating the Due Process Clause

“only if the photographic identification procedure was so impermissibly suggestive
 15

as to give rise to a very substantial likelihood of irreparable misidentification.” Id.

at 384 (emphasis added). 6

The Supreme Court later refined that Simmons inquiry, bifurcating it into two

distinct steps. Courts must first ask whether a pretrial identification procedure is

impermissibly suggestive. Neil v. Biggers, 409 U.S. 188, 197 (1972); see also

Manson v. Brathwaite, 432 U.S. 98, 106 (1977). If it is, then courts must proceed to

the second step of determining whether “the totality of circumstances” nonetheless

demonstrate the in-court identification to be “reliable,” and thus permissible, despite

that suggestivity. Biggers, 409 U.S. at 199; Brathwaite, 432 U.S. at 106. This

second, reliability prong of the analysis is effectively the same as the one posed in

Wade regarding whether the identification stemmed from an “independent source.”

Patterson, 384 A.2d at 666 n.2 (“[T]he factors which are relevant to the

determinations—indeed the very determinations themselves—are essentially the

same” under the Sixth Amendment’s “independent source” and the Fifth

Amendment’s “reliability” inquiries); cf. Clemons v. United States, 408 F.2d 1230,

1237 (D.C. Cir. 1968) (en banc) (asking, under the Due Process Clause post-

6
The petitioner in Simmons did not argue that he had a Sixth Amendment
right to counsel at the pretrial photographic array, though the Supreme Court in
United States v. Ash, 413 U.S. 300, 321 (1973) later rejected that argument.
 16

Simmons, “whether the in-court identification is still admissible because it has an

independent source”).

We now evaluate Mr. Morales’ due process challenge to his in-court

identification under this legal framework. First, we ask whether the pretrial mugshot

display was “so impermissibly suggestive as to give rise to a very substantial

likelihood of irreparable misidentification.” Long v. United States, 156 A.3d 698,

707 (D.C. 2017) (quoting Biggers, 409 U.S. at 197). Because it was, we next

consider whether the government has carried its burden of “establishing that the

identification was nonetheless reliable viewed in the totality of the circumstances.”

Long, 156 A.3d at 707 (citing Kaliku v. United States, 994 A.2d 765, 781–82 (D.C.

2010)); accord Forte v. United States, 856 A.2d 567, 573 (D.C. 2004). We think

not. As to this second prong we stress that our reliability review “may not include

the consideration of evidence of the defendant’s guilt external to the identification.”

Long, 156 A.3d at 708; see also Brathwaite, 432 U.S. at 116 (noting that other

incriminating evidence “plays no part in [the] analysis”). Disregard for that principle

was at the heart of both the trial court’s error and the government’s unavailing

defense of it on appeal, as we explain further below.
 17

A. Suggestivity

1. The High Degree of Suggestivity

The pretrial procedure used here was so impermissibly suggestive that it gave

rise to a substantial likelihood of misidentification. The mugshot display barely

qualifies as an identification procedure at all. As the prosecutor acknowledged at

trial, it was only “for lack of a better term” that she described this as an identification

procedure. The single mugshot display is better characterized as an unmistakable

(even if unintended) communication to Officer Gunnells that Mr. Morales—the man

who had been arrested six days after the shooting, had never been identified by the

officer, and was now going to trial—was the same man the officer had chased from

the scene of the shooting. The government does not dispute that this was an

impermissibly suggestive procedure, but neither does it expressly concede the point,

assuming it only “arguendo.” So we elaborate further.

Asking a witness whether they can identify a standalone suspect—rather than

picking him out from a lineup or photo array—is among the most suggestive and

objectionable forms of pretrial identification because of the strong implication that

the individual has been singled out as the suspect by law enforcement. We have said
 18

so before, noting that “single-photo displays are inherently suggestive” and “have

been designated the ‘most suggestive’ and therefore the ‘most objectionable method

of pretrial identification.’” Patterson, 384 A.2d at 666 (D.C. 1978) (quoting United

States v. Dailey, 524 F.2d 911, 914 (8th Cir. 1975)). 7

Yet the fact that this was a single-mugshot display only begins to capture the

degree of suggestivity here. This mugshot display is substantially more suggestive

than most fitting that description. Officer Gunnells was not, in fact, asked whether

7
We are in good company on that point. See, e.g., Stovall, 388 U.S. at 302
(“The practice of showing suspects singly to persons for the purpose of
identification, and not as part of a lineup, has been widely condemned.”); Simmons,
390 U.S. at 383 (recognizing that the danger of misidentification “will be increased
if the police display to the witness only the picture of a single individual who
generally resembles the person he saw”); United States v. Sanchez, 988 F.2d 1384,
1389 (5th Cir. 1993) (“[E]xhibiting a single photograph for identification purposes
is impermissibly suggestive.”); United States v. Murdock, 928 F.2d 293, 297 (8th
Cir. 1991) (“[S]ingle photograph arrays are considered impermissibly suggestive in
the Eighth Circuit.”); United States v. Milhollan, 599 F.2d 518, 523 (3d Cir. 1979)
(procedure “whereby the Chief of Police showed [the witness] a single photograph,
undoubtedly was suggestive”); Mysholowsky v. New York, 535 F.2d 194, 197 (2d
Cir. 1976) (“We have consistently condemned the exhibition of a single photograph
as a suggestive practice, and, where no extenuating circumstances justify the
procedure, as an unnecessarily suggestive one.”); Dailey, 524 F.2d at 914 (court had
“no trouble concluding that showing [defendant’s] photographs to the witness on the
day he was to testify was impermissibly suggestive” because “[b]y showing only
[defendant’s] photograph to the witness, government counsel in effect said ‘This is
the man’”).
 19

the mugshot was of the same man whose face he had glimpsed four months prior.

In the prosecutor’s account, she asked Officer Gunnells if he “remembered any

specific tattoos” of the man he had chased, and when he said he did not, she handed

him Mr. Morales’ mugshot and the two of them “started looking at the tattoos more

closely.” 8 She treated as a foregone conclusion that Mr. Morales was the same man

Officer Gunnells had chased, and that he would identify him as such, though he had

never done so before. While an inquiry about whether a witness can identify a

suspect from a single photograph is exceedingly suggestive, it is even more so when

the answer is presumed rather than sought after. Worse yet, this was no mere suspect

in a case police were still investigating—as is typical of photographic arrays—where

the witness might reasonably feel at liberty to say investigators were focused on the

wrong person. This was a defendant who Officer Gunnells understood was

8
There is no indication that Officer Gunnells made an identification when he
was shown Mr. Morales’ mugshot. Although there was no pretrial identification,
strictly speaking, that is of no import. The threat to due process lies in the procedure
that is utilized by the government and its propensity for exerting an undue and
suggestive influence on the witness. See, e.g., Wade, 388 U.S. at 233 (stressing
“[t]he potential for improper influence”). Whether Officer Gunnells confirmed
before trial that the mugshot was of the man he had chased, or did so for the first
time at trial—as appears to be the case—is of no consequence under this first factor.
If anything, the lack of an identification when first presented with the mugshot
lessens any assurance of reliability under the second factor of this test discussed
below.
 20

proceeding to trial in the near future. The government does not contest that, under

these circumstances, the mugshot was unquestionably and exceedingly suggestive.

2. The Suggestivity Was Not Necessary

The government also does not argue the suggestivity was necessary under the

circumstances. See Perry v. New Hampshire, 565 U.S. 228, 238–39 (2012) (“[D]ue

process concerns arise only when law enforcement officers use an identification

procedure that is both suggestive and unnecessary.”); see also Biggers, 409 U.S. at

198 (“Suggestive confrontations are disapproved because they increase the

likelihood of misidentification, and unnecessarily suggestive ones are condemned

for the further reason that the increased chance of misidentification is gratuitous.”).

Nor could it, as there was no necessity here. The government had every

opportunity to conduct an appropriate and timely pretrial photo array or a lineup

procedure with Officer Gunnells had it wished to do so. The prosecutor’s decision

to instead use a single mugshot display on the eve of trial was not born of necessity,

but—by her own telling—of disregard. The prosecutor explained that she “was just

preparing for trial” and only after the mugshot display did she “realize[], oh, I’m

probably going to want to ask for an in-court identification.” In other words, she
 21

was not thinking about eliciting an in-court identification, despite defense counsel’s

uncontradicted “understanding that . . . there are no identifications,” expressed at a

hearing the same morning as the mugshot display. 9

The identification procedure used here contravenes the Metropolitan Police

Department’s own General Orders on pretrial identification procedures, which

instruct that “[t]he display of a single photograph” should not be used under any

circumstances except “when the perpetrator is known to the witness in a context

other than the commission of the offense.” See MPD General Order 304.7,

Procedures for Obtaining Pretrial Eyewitness Identification (Apr. 18, 2013). That

was not the case here. And as to the timing of the procedure, there was no reason

to delay it until the eve of trial. Delays cause memories to fade and leave them open

to all manner of corrupting effects. The accuracy of a witness’s memory “is likely

9
Taking the prosecutor’s explanation at face value, we fail to see any
appropriate purpose served by giving Officer Gunnells the mugshot to study. Even
had she not intended to elicit an in-court identification, Officer Gunnells might have
spontaneously offered one, or he might have provided a description of the suspect
he chased with newly recalled details about his tattoos drawn directly from the
mugshot itself. Perhaps there’s some good reason that escapes us, but no matter.
Whether the suggestion was purposeful or accidental is of no consequence. United
States v. Rogers, 387 F.3d 925, 937 (7th Cir. 2004) (“It is irrelevant that police
unintentionally” introduced the suggestion; it matters only that “[t]he circumstances
were . . . unduly suggestive”) (citing Brathwaite, 432 U.S. 112–14).
 22

to be greater when retrieval occurs closer to the time of the witnessed events.” Sims

v. United States, 213 A.3d 1260, 1273 n.22 (D.C. 2019) (quoting Nat’l Research

Council, Identifying the Culprit: Assessing Eyewitness Identification 65 (2014)); In

re L.D.O., 400 A.2d 1055, 1057 (D.C. 1979) (“[E]xtra-judicial identification

typically occurs quite soon after the crime when the witness’ memory is fresher and

less likely to have been led afield by incidental events.”). This eleventh-hour

mugshot display was both unnecessarily suggestive and unjustifiably delayed.

3. This Case Is Not of the “Rare Breed” Noted in Patterson

Having found the mugshot display impermissibly suggestive, we would

ordinarily now move onto the reliability part of the analysis. But the government

raises one impediment to doing so. Relying on Patterson, it argues that we should

treat this as one of a “rare breed” of cases where an impermissibly suggestive

procedure nonetheless does not risk irreparable misidentification, a point Patterson

expressly includes in this first, suggestivity prong of the due process analysis. 384

A.2d at 666 (“[W]e do not leave the first stage of the inquiry.”). It is not so clear

why the government thinks that, but suffice it to say Patterson lends no support to

the argument.
 23

In Patterson, a witness had identified his assailants to responding officers at

the scene of, and mere minutes after, the crime. Id. While that same witness was

later exposed to a highly suggestive pretrial identification procedure, we detected no

due process violation in permitting him to make an in-court identification,

concluding: “[W]hen there has been an unequivocal, unsuggested, and otherwise

constitutionally acceptable identification, subsequent identifications”—even those

“refreshed” by lone mugshots—“are not conducive to irreparable misidentification,

in violation of due process.” 10 Id. at 667. At the outset, it is odd that Patterson

thought itself to be engaging in a suggestivity rather than a reliability inquiry. After

all, Patterson’s bottom-line conclusion was that the mugshot displays in that case

“could not have created a ‘very substantial likelihood of irreparable

misidentification,’” 384 A.2d at 666 (quoting Simmons, 390 U.S. at 384), which is

precisely what the reliability inquiry probes. 11 It seems to have wedged a threshold

10
As we explained in Patterson: “[T]he first identifications occurred when
[the witness] sighted the suspects (who had robbed him merely an hour and a half
before) and pointed them out to the officers. He led the police to appellants; the
police did not bring them to [him]. Thus, the initial identifications to the authorities
unmistakably were based on observation without suggestion.” 384 A.2d at 666.
11
See Perry, 565 U.S. at 720 (“[T]he trial judge must screen the evidence for
reliability pretrial. If there is ‘a very substantial likelihood of irreparable
misidentification,’ the judge must disallow presentation of the evidence at trial.”)
(quoting Simmons, 390 U.S. at 384). As apparent support for its view that it was
 24

reliability inquiry into the suggestivity analysis. 12 But how we frame the point

conceptually is of no consequence here.

The critical point in Patterson was that the same witness who had participated

in a suggestive identification procedure had previously identified the suspects on the

scene. That is not the case here. While the government stresses that another person

(who never testified) identified Mr. Morales as the shooter—providing the initial

basis for his arrest—Officer Gunnells did not identify Mr. Morales prior to the

suggestive mugshot display. So Patterson is of no help to the government. We have

never extended Patterson’s reasoning beyond the narrow class of cases where the

engaged in a step-one suggestivity analysis, Patterson cited United States v. Hines,
455 F.2d 1317, 1330–31 (D.C. Cir. 1971). But Hines was engaged in the step-two
reliability analysis when it concluded that the pretrial identification procedure was
not “likely to lead to a misidentification,” 455 F.2d at 1330. The Supreme Court had
simply not yet coined the second prong of the due process analysis as “reliability”
when Hines was decided. When the Court later did so, it treated reliability as
synonymous with probing the likelihood of irreparable misidentification. See
Brathwaite, 432 U.S. at 116 (concluding, in assessing reliability, “we cannot say that
under all the circumstances of this case there is ‘a very substantial likelihood of
irreparable misidentification’”) (quoting Simmons, 390 U.S. at 384).
12
Other courts treat the fact of a prior unsuggested identification as sounding
in the reliability analysis, which is where we think it fits more comfortably. See,
e.g., Hopkins v. State, 721 A.2d 231, 241 (Md. 1998) (noting “previous
identification” as providing assurance that in-court identification was “sufficiently
reliable,” despite any intervening “suggestiv[ity]”).
 25

same witness made a pre-suggestivity identification. See, e.g., United States v.

Brannon, 404 A.2d 926, 928 (D.C. 1979) (same witness); Redmond v. United States,

829 A.2d 229, 234 (D.C. 2003) (same witness); Scott v. United States, 619 A.2d 917,

930 (D.C. 1993) (same witness). Nor would it make any sense to do so. Such an

expansion would mean a single identification by one witness would immunize all

identifications that follow—from any witness, no matter how suggestive the

procedures used to obtain them—from constitutional scrutiny. Neither precedent

nor reason supports that unprincipled result.

B. Reliability

Having determined the mugshot display was unduly suggestive, we next ask

whether the in-court identification was nonetheless permissible because it was

sufficiently reliable. Brathwaite, 432 U.S. at 114 (“[R]eliability is the linchpin in

determining the admissibility of identification testimony.”).

This requires weighing the independent (and non-suggestive) basis for the

identification against “the corrupting effect of the suggestiv[ity]” itself. Id. That

analysis is similar to the independent source question posed in Wade. Patterson,

384 A.2d at 666 n.2; cf. Clemons, 408 F.2d at 1237. Our aim under both inquiries is
 26

to determine whether the identification stems from a source independent of the

suggestivity, here, Officer Gunnells’ prior chase of the suspect. To make this

determination, we examine five factors outlined in Biggers, though this list is not

exhaustive: 13

[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.

409 U.S. at 199–200.

13
See Haliym v. Mitchell, 492 F.3d 680, 706–07 (6th Cir. 2007) (finding
Biggers factors non-exclusive and deeming the witness’ four prior encounters with
the suspect as bolstering the ensuing identification’s reliability); see also Brathwaite,
432 U.S. at 113 (stressing that reliability depends upon “totality of the
circumstances”). Apart from being non-exhaustive, this list is also likely over-
inclusive, as some of these factors—outlined nearly fifty years ago—have since been
debunked as having no positive correlation to an identification’s reliability. See
State v. Henderson, 27 A.3d 872, 877, 918 (N.J. 2011) (detailing scientific research
in the decades since Biggers and Brathwaite that “offers convincing proof that the
current test for evaluating the trustworthiness of eyewitness identifications should
be revised”). It is an interesting question whether precedent requires us to give
weight to each of these five factors or, alternatively, if we need only “consider[]”
them, Biggers, 409 U.S. at 199, giving each the weight the intervening decades of
empirical research has proven them to be due (even when it is none). But that
question does not matter here where strict fidelity to the Biggers factors leads to the
conclusion that this identification was not reliable.
 27

Upon review of these factors, we cannot say that Officer Gunnells’ in-court

identification was reliable and free of the taint from the mugshot display’s

extraordinary suggestivity. The circumstances strongly suggest that Officer

Gunnells’ in-court identification stemmed from his review of Mr. Morales’ mugshot

only two weeks before trial rather than any independent recollection of the individual

he observed for two to three seconds more than four months earlier.

1. Opportunity to View and Lapse in Time Before Identification

The first, and what is often described as the most important factor in

determining an identification’s reliability, 14 is Officer Gunnells’ opportunity to

observe the suspect at the scene. It was minimal. His entire interaction with the

suspect lasted about a minute, but for most of that time he was chasing the suspect

from behind with no view of his face. More pertinent was the small portion of that

14
See Young v. Conway, 698 F.3d 69, 80 (2d Cir. 2012) (describing witness’s
“prior opportunity to observe the alleged criminal act” as the “most important factor”
of reliability inquiry); People v. Macklin, 125 N.E.3d 1246, 1261 (Ill. App. Ct. 2019)
(Hyman, J., dissenting) (“[C]ourts repeatedly describe a victim’s opportunity to
observe an offender as ‘the most important’ of the Biggers factors.”) (collecting
cases); Commonwealth v. Milburn, 191 A.3d 891, 900 (Pa. Super. 2018) (“The most
important factor in addressing the reliability of an identification is the witness’s
opportunity to observe the perpetrator at the time of the crime.”) (citation omitted).
 28

time in which he viewed the suspect’s face, which was two to three seconds from a

distance of ten to twenty-five feet. See supra note 2. Just as suggestivity is at its

apex here, this case approaches the nadir of a witness’s opportunity to observe. We

have rejected as unreliable identifications where the opportunity to view was

similarly minimal. See Crawley v. United States, 320 A.2d 309, 310, 312 (D.C.

1974) (finding identification unreliable based on “brief” opportunity to view where

witness observed suspect for sixty to ninety seconds and saw suspect only from

behind except when he turned around to admonish witness not to come closer).

When we have found identifications reliable despite similarly brief

opportunities to observe, it has been because the fifth factor—the length of time

between the crime and the confrontation—weighed heavily in the government’s

favor. See, e.g., Holt v. United States, 675 A.2d 474, 482–83 (D.C. 1996) (citing

“uniquely powerful indicia of reliability” where officer observed suspect from a few

car lengths away for approximately one minute while chasing him and made

identification a little over half an hour later). 15 The longer the gap in time between

15
The first and fifth Biggers factors typically drive the reliability analysis, and
have been described as the “only two of the five” factors that scientific research
supports as “clearly related to accuracy in the way that the Court had assumed,”
while the remaining factors “have received mixed support by researchers.” John C.
Brigham et al., Disputed Eyewitness Identification Evidence: Important Legal and
 29

the sighting and the identification procedure, the better the opportunity to observe

must be to provide any assurance of reliability. Here, this fifth Biggers factor

counsels strongly against admissibility, given Officer Gunnells’ minimal

opportunity to observe the suspect and the four-month gap between the chase and

the attempted identification. Biggers, 409 U.S. at 201 (“a lapse of seven months . . .

would be a seriously negative factor in most cases”). 16

The first and fifth Biggers factors, when weighed against the suggestivity of

the government’s mugshot display, just about foreclose any possibility that any in-

Scientific Issues, 36 CT. REV. 12, 17 (1999); see also Benn v. United States, 978 A.2d
1257, 1267–71 (D.C. 2009) (discussing scientific studies regarding lack of
correlation between witness confidence and accuracy).
16
That is in stark contrast to Nichols v. United States, 343 A.2d 336 (D.C.
1975)—relied upon by our dissenting colleague as involving a similarly poor
opportunity to observe the suspect head-on, post at n.3—where a show-up was done
at the scene minutes after the offense. The identifications at issue in Nichols also
relied heavily on the items of clothing worn by the suspects and far more detailed
descriptions, so that the witness’s relative inability to see the suspects’ faces was
less devastating to those identifications than it is to the one we have here. Id. at 345
(noting that the witness identified the suspect’s “uniform, rather than [the suspect]
himself”); id. at 344 (witness’ description of other suspect included “the man’s
general facial features, the cut of his hair, the shape of his head, his race, his height,
and his ‘very broad shouldered’ build” along with his “dark pants [and] dark pullover
shirt”). As discussed further below, while Officer Gunnells described two items of
clothing worn by the suspect he chased, that lends no support to the reliability of his
in-court identification of Mr. Morales in different attire four months later.
 30

court identification would be sufficiently reliable to permit. A glancing view of a

suspect does not provide a reliable basis for an identification made more than four

months later. Particularly not when there has been an intervening and highly

suggestive mugshot display. Brathwaite, 432 U.S. at 114 (courts must weigh

“corrupting effect of the suggestive” procedure against factors of reliability).

2. The Remaining Biggers Factors

While the first and fifth factors are likely fatal to any argument that Officer

Gunnells’ identification was sufficiently reliable, the remaining factors do not help

the government either. See generally United States v. Brown, 700 A.2d 760, 761

(D.C. 1997) (defendant’s burden to establish suggestivity, and upon that showing, it

is the government’s “burden” to show “the identification was reliable nonetheless”).

Officer Gunnells’ degree of attention, the certainty of his identification, and the

accuracy of his prior description offer little assurance that he could reliably identify

the man he chased more than four months later.

As to degree of attention, we do not doubt that Officer Gunnells was paying

close attention to the suspects as he approached, but the importance of his focus is

significantly diminished by his minimal opportunity to observe the two suspects.
 31

This second Biggers factor is closely tied to the first. While a lack of focus might

not be detrimental where the opportunity to observe stretches on for extended

periods, no degree of focus can provide much assurance of reliability when the

opportunity to observe is limited to a few seconds. See Raheem v. Kelly, 257 F.3d

122, 138 (2d Cir. 2001) (granting habeas relief due to unreliable identification when

witnesses observed suspects for “15 minutes,” and “frequently glanced at the[m]”).

That is particularly true here where Officer Gunnells’ principal aim was not to study

the suspects for the purpose of providing later descriptions of them; 17 his goal was

to apprehend them himself. He approached with caution given that they were

suspects in a shooting, and his focus was divided not only between two suspects, but

between the fleeing suspect’s face and his waistline, because he “was fidgeting

around” as if to draw a weapon.

As for Officer Gunnells’ certainty, he was never asked to describe his level of

certainty so we do not know what it was and thus do not give this factor any weight.

17
In Brathwaite, by contrast, the Court noted that the undercover narcotics
officer “could be expected to pay scrupulous attention to detail” when “he knew that
subsequently he would have to find and arrest his vendor.” 432 U.S. at 115. The
situation here was quite different, where Officer Gunnells was approaching two men
whom he believed to be armed and dangerous, and was trying to apprehend them in
the moment.
 32

Officer Gunnells was asked on the stand if he saw the individual he had chased in

the courtroom, to which he responded “[y]es ma’am” and identified Mr. Morales.

He was then asked whether Mr. Morales looked as he had on the day Officer

Gunnells chased him, to which he responded, “[w]ithout the beanie, he looks the

same,” with no mention of his confidence or certainty. We do not think Officer

Gunnells’ seeming lack of hesitation or equivocation amounts to a display of

certainty where he was not presented with any alternative suspects (as in an array or

lineup) to consider. And even had he professed absolute certainty, we would still

give little or no weight to this factor where the certainty was expressed only after the

suggestive mugshot display, and would seem to be a byproduct of it. While the

“relationship between the witness’[s] stated confidence and accuracy of

identifications may be greater at the moment” of an initial timely identification,

“[e]vidence indicates that self-reported confidence at the time of trial is not a reliable

predictor of eyewitness accuracy.” Nat’l Research Council, Identifying the Culprit:

Assessing Eyewitness Identification 108 (2014). 18 Particularly not when the

certainty can be explained by the high degree of suggestivity itself. “The irony” in

weighing this factor under such circumstances “is that the more suggestive the

18
Available at www.nap.edu/catalog/18891/identifying-the-culprit-assessing-
eyewitness-identification https://perma.cc/8AM9-AFR3.
 33

procedure, the greater the chance eyewitnesses will seem confident.” State v.

Henderson, 27 A.3d 872, 918 (N.J. 2011).

On the remaining factor, there was not much to Officer Gunnells’ description

of the suspect he chased. He described the suspect as “[1] a Hispanic male, [2] a

beanie on his head, [3] tattoos on his face and neck, and [4] a jacket tied around his

waist.” Noticeably missing from this lookout are common suspect descriptors such

as age, height, weight, build, and complexion. See, e.g., Cureton v. United States,

386 A.2d 278, 285–86 (D.C. 1978) (eyewitness description revealed “close attention

to detail” where it included assailants’ “race, sex, age, height and weight” and

described “hair style, facial hair, skin color, clothing, and the type of weapon

employed by each”). The second and fourth descriptors included in Officer

Gunnells’ description—both relating to the suspect’s clothing—are irrelevant

because Mr. Morales was not captured until later in the week and there’s no

indication he was wearing the same clothes. 19

19
It is a mistake to say the described beanie and tied-around jacket were
“accurate” descriptions of Mr. Morales, though the dissent counts these “distinctive”
items of apparel as a point in favor of the in-court identification’s reliability. Post
at 59. Mr. Morales was never seen wearing those items of clothing unless, of course,
one starts from the premise that Mr. Morales is the suspect Officer Gunnells chased.
 34

All that remains is a description of “a Hispanic male” with non-descript

“tattoos on his face and neck.” While this accurately describes Mr. Morales, it is

sparse on specifics, and Officer Gunnells admittedly could not describe any of the

tattoos he observed. We would give more weight to “distinctive features of a

particular tattoo” if one had been described, but put little stock in the generic

description provided by Officer Gunnells, given that “[t]attoos have become

ubiquitous in modern society.” Small v. State, 180 A.3d 163, 182 (Md. Ct. Spec.

App. 2018) (quoting State v. Rios, 156 A.3d 18, 44 (Conn. App. Ct. 2017)).

It is true, as the dissent notes, that identifications with “some questionable

feature” are standard “grist for the jury mill.” Post at 60 (quoting Brathwaite, 432

U.S. at 116). But some grist is spoiled and to grind it with the rest would contaminate

the flour. This identification was of that tainted variety. It contained more than just

some questionable feature, but for the reasons already discussed, it was unreliable at

its core.

3. The Irrelevance of Corroborating Evidence

The government also argues that Officer Gunnells’ identification was reliable

because of the existence of other evidence, unrelated to the trustworthiness of
 35

Officer Gunnells’ identification, showing that Mr. Morales was in fact the gunman.

But its argument stems from confusion about what reliability is. Whether a

particular piece of evidence is reliable is distinct from whether it happens to be

accurate, or as we have previously put it, “the reliability inquiry may not include the

consideration of evidence of the defendant’s guilt external to the identification”

itself. Long, 156 A.3d at 708. Reliability measures whether identification testimony

is itself sufficiently trustworthy, which is what each of the Biggers factors speaks to.

And that inquiry is independent of whatever additional evidence of guilt the

government might have. Corroborating evidence might prove the identification to

be correct, but it tells us nothing about its reliability. A stopped watch is not a

reliable indicator of the time even during those two moments per day it happens to

be correct, just as a broken compass is not reliable just because its holder happens to

face in the same direction its needle is stuck.

The Supreme Court made this very point in Brathwaite. After evaluating each

of the five Biggers factors, the Court noted that the strength of other corroborating

evidence unrelated to the quality of the witness’s identification itself “plays no part

in our [reliability] analysis.” Brathwaite, 432 U.S. at 116. That was no unthinking

aside. In his concurrence, Justice Stevens acknowledged that “it is sometimes

difficult to put other evidence of guilt entirely to one side” when assessing reliability,
 36

but commended the Court’s opinion for “carefully avoid[ing] this pitfall and

correctly rel[ying] only on appropriate indicia of the reliability of the identification

itself.” Id. at 118. And he had made that same point previously in United States ex

rel. Kirby v. Sturgis, 510 F.2d 397 (7th Cir. 1975) (Stevens, J.), a case Brathwaite

drew from considerably. In Kirby, then-Judge Stevens stressed that the court was

not “placing any weight on the corroboration resulting from the defendants’

possession of the stolen property,” id. at 404, foreshadowing—and no doubt

informing—Brathwaite’s echo that corroborating evidence plays no part in the

reliability analysis.

We need not dig further into the origins of the principle, because we have

reiterated it more recently. In Long, we explained “that corroborative evidence

external to an identification ‘plays no part’ in the due process analysis of the

reliability of the identification.” 156 A.3d at 708 (quoting Brathwaite, 432 U.S. at

116). To illustrate, we noted that “even where there was irrefutable evidence of the

defendant’s guilt, if an identification were made by a witness who, it transpired, was

not even present at the event, we could hardly term the identification reliable.” Long,

156 A.3d at 708 (quoting Raheem, 257 F.3d at 140). “Put another way, the fact that

an identification is proved accurate by evidence unrelated to the identification itself
 37

does not mean the identification is reliable within the meaning of a proper due

process analysis.” Long, 156 A.3d at 708.

The government cites to one of our cases suggesting, in dicta, a deviation from

this principle: Curtis v. United States, 886 A.2d 92 (D.C. 2005). In Curtis, after

extensive discussion of the suggestivity prong of the analysis, and concluding that

the pretrial procedure was not in fact suggestive, our court passingly—in two

sentences—noted strong corroborative evidence as supporting the reliability of the

identification in any event. Id. at 95. It is an understandable misstep. It is indeed

“sometimes difficult to put other evidence of guilt entirely to one side,” Brathwaite,

432 U.S. at 118 (Stevens, J., concurring). But it is not a holding. “[T]he judicial

mind was not asked to focus upon, and the opinion did not address, the point at

issue,” so it “lends no support to the government’s position.” Alfaro v. United States,

859 A.2d 149, 154 (D.C. 2004) (quotations omitted). And even if it had purported

to be it was inconsistent with the Supreme Court’s opinion in Brathwaite, as we have

since explained in Long.

Because it is so tempting to consider external evidence that is indicative of a

defendant’s guilt when speaking in terms of an identification’s reliability, we

reiterate that the reliability inquiry is akin to asking whether the identification has
 38

an independent source. Patterson, 384 A.2d at 666 n.2 (“[T]he factors which are

relevant to the determinations—indeed the very determinations themselves—are

essentially the same.”). Thinking of the issue in those terms might disabuse courts

and litigants from conflating the identification’s reliability with the weight of

corroborating evidence. Adding further support for this conceptualization is Wade,

where the factors articulated as relevant to the independent source inquiry were

roughly the same as those later articulated in Biggers as relevant to the reliability

inquiry. Wade, 388 U.S. at 241 (independent source factors include “the prior

opportunity to observe the alleged criminal act, the existence of any discrepancy

between any pre-lineup description and the defendant’s actual description . . . and

the lapse of time between the alleged act and the lineup identification”).

Corroborating evidence has a role to play on appeal, to be sure, but its

relevance is limited to whether any error was harmful. It does not inform the

suggestivity or reliability calculus and it is thus as irrelevant to the trial court’s

admissibility determination as it is to our assessment of whether an error occurred.

Raheem, 257 F.2d at 141 (“[E]vidence of record that is unrelated to an identification

but that is supportive of a finding of guilt is properly considered in harmless-error

analysis, not in the due process inquiry of whether the identification has

reliability.”); United States v. Emanuele, 51 F.3d 1123, 1128 (3d Cir. 1995) (“[O]nly
 39

factors relating to the reliability of the identification will be relevant to a due process

analysis. Independent evidence of culpability will not cure a tainted identification

procedure, nor will exculpatory information bar admission of reliable identification

testimony. We will consider other evidence only to determine whether an error, if

present, was harmless.”).

There is good reason for that. That a defendant is dead to rights is not a sound

reason to permit unreliable evidence to be admitted against him.

C. Harm

The admission of Officer Gunnells’ in-court identification was constitutional

error in violation of the Fifth Amendment’s Due Process Clause. We must therefore

reverse Mr. Morales’ convictions unless the government has “prove[d] beyond a

reasonable doubt that the error . . . did not contribute to the verdict obtained.”

Chapman v. California, 386 U.S. 18, 24 (1967). The government has not attempted

to make that showing and so we reverse Mr. Morales’ convictions.

Mr. Morales correctly notes that the government bears the burden of

demonstrating that the improper admission of Officer Gunnells’ in-court
 40

identification was harmless beyond a reasonable doubt. The government, in turn,

makes no argument that the error was harmless under any standard. It does not cite

(or allude to) Chapman, Kotteakos, 20 or any other authority concerning harmless

error. The word harm does not appear in its brief, nor does harmful, harmless, or

any other permutation or synonym. The government has, therefore, waived any

argument that this constitutional error was harmless beyond a reasonable doubt.

Rose v. United States, 629 A.2d 526, 535 (D.C. 1993) (“It is a basic principle of

appellate jurisprudence that points not raised on appeal are deemed to be waived.”);

see also United States v. Sineneng-Smith, --- U.S. ---, 140 S. Ct. 1575, 1579 (2020)

(“[A]s a general rule, our system is designed around the premise that parties

represented by competent counsel know what is best for them, and are responsible

for advancing the facts and argument entitling them to relief.”) (quotation marks

omitted).

Our dissenting colleague nonetheless would exercise discretion to sua sponte

hold this constitutional error harmless. Assuming we even have such discretion, 21

20
Kotteakos v. United States, 328 U.S. 750, 764–65 (1946) (articulating the
test for non-constitutional errors).
21
It appears to be an open question in this jurisdiction whether we have that
discretion when it comes to constitutional errors. We certainly do when it comes to
 41

this is not the case in which to exercise it. Any discretionary authority to sua sponte

address harmless error “must be exercised with restraint,” and “only when

harmlessness is obvious.” Randolph v. United States, 882 A.2d 210, 223 (D.C.

2005). If it is “at least debatable,” “the government should not be relieved of its

waiver.” Id. Harmlessness is “at least debatable” here, rendering a sua sponte harm

analysis improper. Id.

In finding this error obviously harmless beyond a reasonable doubt, the

dissent relies primarily on Mr. Morales’ own self-incriminating testimony, presented

as part of the defense case. Post at 61–62. Such reliance is misplaced because it

skips over an initial question whether the government’s own case would have even

survived the motion for judgment of acquittal absent the improper identification. See

United States v. Rothbart, 653 F.2d 462, 466–67 (10th Cir. 1981) (constitutional

error not harmless where “defendant’s motion for judgment of acquittal at the end

non-constitutional errors. Randolph v. United States, 882 A.2d 210, 221–23 (D.C.
2005) (recognizing our discretion, to be exercised with restraint, to “disregard any
nonconstitutional error which does not affect a substantial right of the accused”)
(quoting Arnold v. United States, 358 A.2d 335, 341 (D.C. 1976) (en banc)). But
we have found no binding precedent that purports to extend this discretion to, much
less exercised it in, a case involving a constitutional error. But cf. Wilson-Bey v.
United States, 903 A.2d 818, 844 n.45 (D.C. 2006) (en banc) (addressing
constitutional harm argument made for the first time in en banc briefing).
 42

of the government’s evidence may have been successful” but for the erroneously

admitted evidence); United States v. Ferreira, 821 F.2d 1, 7–9 (1st Cir. 1987) (noting

significance of improperly admitted evidence in the trial court’s decision to deny

defendant’s motion for judgment of acquittal as a reason for concluding error was

not harmless). If it would not have, then Mr. Morales’ testimony does not factor into

the harm equation because had he been acquitted at the close of the government’s

case, he never would have had the opportunity to testify. When we consider this

threshold question, we think it far from obvious beyond a reasonable doubt that the

government’s case would have survived the motion for judgment of acquittal absent

Officer Gunnells’ identification. In fact, it is likely that it would not have.

Subtracting Officer Gunnells’ in-court identification, there were only two

pieces of evidence linking Mr. Morales to the shooting when the government rested

its case (i.e., before Mr. Morales testified). One was the surveillance and body-worn

camera footage, neither of which depicted the suspect in particular detail or from a

range that would permit any juror to confidently use it as a basis for making their

own identification. 22 The other was Mr. Morales’ DNA on the handgun, which

22
The government did not suggest otherwise in opposing the motion for
judgment of acquittal. It did argue, quite forcefully, that the person depicted in the
video footage of the shooting was the same as the person depicted in Officer
 43

seems compelling in the abstract, except that there were two other DNA profiles

right alongside it. A one in three chance is hardly the stuff of proof beyond a

reasonable doubt. It was Officer Gunnells’ identification that tied the entire

government case together by providing the lone identification of Mr. Morales at the

scene. Both the trial court and the government understood as much. The trial court

described its denial of Mr. Morales’ motion for judgment of acquittal as a close

call—even with Officer Gunnells’ identification—stating: “I think this case does go

to the jury. I agree that it is close on the issue of whether or not the person with the

gun is this person.” And in opposing the motion for acquittal, the government

stressed Officer Gunnells’ identification multiple times in its otherwise brief

argument, showing how crucial the government considered the identification to be

for purposes of withstanding the motion (“Officer Gunnells identified the person at

16th and Park as the Defendant. . . . Officer Gunnells sees him toss [the gun]”). 23

Gunnells’ body-worn camera footage, stressing that the two images were of men
wearing identical clothing and accompanied by the same companion mere minutes
apart and in close proximity. But as to the all-critical question of who that person
was, the government relied heavily and repeatedly on Officer Gunnells’
identification in order to make that link.
23
While we are focused on the evidence as it existed at the close of the
government’s case-in-chief—independent of the defense case—it is worth noting
that the government also stressed Officer Gunnells’ identification as a principal basis
for convicting Mr. Morales in closing, even after Mr. Morales’ testimony. We have
 44

The dissent contends we must assess the “trial record as a whole” and include

the defense case in our harm assessment, regardless of whether a judgment of

acquittal would have been granted at the close of the government’s case absent

Officer Gunnells’ identification. Post at 63 (quoting United States v. Hasting, 461

U.S. 499, 509 (1983). We disagree. It is true that we generally look at the entire

record when assessing harm, but not always. For instance, when a defendant testifies

“in order to overcome the impact” of illegally admitted evidence, we sometimes

conclude her “testimony was tainted by the same illegality” so that it must be

discounted from the harm analysis. Smith v. United States, 529 A.2d 312, 318 (D.C.

1987) (quoting Harrison v. United States, 392 U.S. 219, 223 (1968)). 24 That is to

say we should not blind ourselves to the effects that constitutional violations have

routinely noted that a “prosecutor’s stress upon the centrality of particular evidence
in closing argument tells a good deal about whether the admission of the evidence
was prejudicial.” Morten v. United States, 856 A.2d 595, 602 (D.C. 2004) (quoting
Allen v. United States, 837 A.2d 917, 923 (D.C. 2003)) (internal quotation marks
and brackets omitted).
24
We tend to apply such analysis in cases concerning ill-gotten confessions,
but the principle extends beyond that discrete realm. See Kauffman v. Sec’y of the
Air Force, 415 F.2d 991, 998–99 (D.C. Cir. 1969) (in the context of a Fourth
Amendment violation, asking “whether appellant’s testimony was impelled by the
need to explain away the [illegally obtained and admitted] notebook”); United States
v. Pelullo, 173 F.3d 131, 134 (3d Cir. 1999) (government required to prove that
defendant’s “decision to testify at the first trial was not caused by the Brady
violation”).
 45

on the course of the trial itself. 25 Under the present circumstances, we will not ignore

the critical fact that the constitutionally infirm evidence supplied the proof needed

to overcome a motion for judgment of acquittal, as the dissent would have us do. 26

That conclusion is bolstered by the fact that the inquiry under Chapman is not

whether the evidence was ultimately overwhelming against a defendant, but instead

whether the government has established “beyond a reasonable doubt that the error

complained of did not contribute to the verdict obtained.” Chapman, 386 U.S. at 24.

25
Imagine a case where a government agent manufactured DNA test results
purporting to conclusively establish that a defendant had held a murder weapon
when, in fact, he never had. If that defendant then falsely “admits” to having held
the weapon in order to offer some innocent explanation for it—believing it a lost
cause to testify truthfully and willing to perjure himself to combat a false murder
charge—why should that render the manufactured DNA evidence harmless?
Improperly admitted evidence may sometimes leave a defendant with “no alternative
but to take the stand in the hope of lessening its damaging impact,” Smith, 529 A.2d
at 318, and discounting such testimony from the harm analysis is often critical to
protecting the underlying constitutional right.
26
The sufficiency case law the dissent relies on for the contrary proposition
goes to a different question. Post at 64. Those cases hold that a defendant waives
her right to appeal a motion for judgment of acquittal at the close of the government’s
case if she presents a defense case, so that any appeal regarding the sufficiency of
the evidence must account for the defense evidence as well. But we are not
concerned with a sufficiency challenge at the moment, but see infra Part III, and that
same waiver analysis does not extend to the question whether an inarguably
preserved error was harmless. It certainly does not “obviously” so extend to warrant
sua sponte harmless error review. Randolph, 882 A.2d at 223.
 46

If evidence supplies the needed proof to overcome a motion for judgment of

acquittal—as is the case with Officer Gunnells’ identification—then it necessarily

contributed to any guilty verdict obtained because a judicially-directed verdict of

“not guilty” would have resulted in its absence. In fact, under these circumstances

we do not think one could even “say, with fair assurance, after pondering all that

happened without stripping the erroneous action from the whole, that the judgment

was not substantially swayed by the error,” Kotteakos, 328 U.S. at 764–65, under

the test for non-constitutional errors. Perhaps that is why the government chose not

to argue in this court that any error was harmless, as it is accustomed to doing when

that argument is a viable one. United States v. Rodriguez, 880 F.3d 1151, 1163 (9th

Cir. 2018) (“If any party in federal litigation is in a superior position to raise harmless

error, it is the United States of America, the most long-standing and frequent litigant

in our federal courts.”).

We thus reverse Mr. Morales’ convictions, save one. The lone exception is

his conviction for escape. Mr. Morales pled guilty to that charge post-trial and

waived his right to appeal it. Given his waiver of appellate rights as to that

conviction, we affirm it.
 47

III.

Mr. Morales argues that the evidence at trial was insufficient to support his

convictions. We disagree. Despite reversing his convictions, we must address this

question because if the evidence against him were indeed insufficient as to any

charge, the Double Jeopardy Clause would bar his retrial. Roberts v. United States,

216 A.3d 870, 882 (D.C. 2019). We conclude that the evidence was plainly

sufficient to support Mr. Morales’ convictions.

“In reviewing claims of insufficient evidence, we review the evidence in the

light most favorable to the government, giving it the benefit of all reasonable

inferences in its favor.” Stroman v. United States, 878 A.2d 1241, 1244 (D.C. 2005).

Reversal is required only “where there is no evidence upon which a reasonable mind

could fairly conclude guilt beyond a reasonable doubt.” Harris v. United States, 668

A.2d 839, 841 (D.C. 1995). We consider all evidence admitted by the trial court,

including erroneously admitted evidence, when conducting a sufficiency review.

See Thomas v. United States, 557 A.2d 599, 600–01 (D.C. 1989) (en banc) (per

curiam). Which is to say we do not strip out Officer Gunnells’ identification

testimony in our sufficiency analysis, but consider it along with the rest of the

evidence admitted at trial.
 48

Mr. Morales argues that there was no evidence tying him to the shooting at

the Columbia Heights Metro Station. He asserts that “[w]hile it is clear from the

video clips . . . that someone fired into the entrance of the Columbia Heights Metro

Station on the date and at the time charged, nothing in those clips established that

[he] was that person.” Instead, Mr. Morales argues, the person in the video could

have been “anyone.” He is wrong about that. The video clips were far from the only

evidence before the jury. The jury also heard evidence that, when first arrested and

questioned in connection to the shooting, Mr. Morales admitted to being the shooter.

They also viewed the body-worn camera footage of Officers Gunnells and Weber

showing Mr. Morales, by his own admission, running from the officers minutes later,

within blocks of the shooting, and wearing the exact same clothing as the shooter.

Mr. Morales counters that his testimony regarding “how he came to be in

possession of a firearm and ammunition”—that he received it from the gunman, who

he then exchanged clothing with, immediately after the shooting—was “unrebutted.”

Hardly. The prosecution presented a detective as a rebuttal witness who called into

question the existence of both José René and the location of the alleged clothing

swap. Even absent this rebuttal testimony, the jury is not required to accept

“unrebutted” testimony as true. One of the jury’s core functions is to weigh the

credibility of trial witnesses. They were free to reject Mr. Morales’ version in its
 49

entirety rather than accepting his exculpating yarn about being an innocent passerby

who happened to come into possession of the firearm, and came to wear the

gunman’s clothes, in the immediate aftermath of the shooting.

IV.

We reverse the trial court’s ruling permitting the in-court identification, vacate

Mr. Morales’ convictions—except for his escape conviction, D.C. Code § 22-

2601(a)(1), which we affirm—and remand the case for further proceedings

consistent with this opinion.

So ordered.

FISHER, Senior Judge, dissenting in part: I of course agree that the evidence

was sufficient, but I cannot accept the majority’s conclusion that we should reverse

appellant’s convictions. In my view, the trial court properly permitted the jury to

hear, and to weigh, Officer Gunnells’ in-court identification. Even assuming that

decision was error, it was harmless beyond a reasonable doubt.
 50

I. Background

When ruling on the admissibility of Officer Gunnells’ in-court identification,

we must consider the totality of the circumstances, viewing the record in the light

most favorable to the trial court’s ruling. Therefore, at the risk of tiring the reader,

I will revisit the facts. Around 12:20 p.m. on August 25, 2017, Officers Doran

Gunnells and Alicia Weber were leaving a restaurant when a security officer

reported that a shooting had just occurred at the Columbia Heights Metro Station.

“[A]lmost at the same time,” the officers received a radio assignment for “shots fired

at the Metro at 14th and Irving.” They were directed to proceed to Hiatt Place, N.W.,

to look for a “black male with dreds and for possibly a Hispanic male.” The officers

rode their bicycles on Hiatt Place toward Park Road, N.W., and soon saw two

individuals “walking in a very hurried manner” on Park Road. Officer Gunnells

asked his dispatcher to repeat the description and the officers approached the two

men. Officer Gunnells said “Hey, let me talk to you for a second” and directed the

men to show their hands.

One of them, a Hispanic man, began “digging in his waistband” and continued

to walk away from the officer, beginning to run when he reached 16th Street. Officer

Gunnells dismounted and gave chase on foot. While fleeing, the man “thrust his
 51

right arm off to the side as if throwing something” and Officer Gunnells saw a “silver

object fly out into the yard”; as he passed the object, he confirmed that it was a gun.

Officer Weber stopped to stay with the gun while Officer Gunnells continued his

pursuit, but the man escaped. During the chase, Gunnells radioed a description of

the man he was pursuing.

Shortly before trial, the prosecutor informed defense counsel that she had

shown Officer Gunnells and Officer Weber a “mug shot” of appellant Morales

during trial preparation. In response, appellant filed a motion to suppress any

identifications by the officers, arguing that the display of the photograph was

“impermissibly suggestive, and the resulting identification is therefore unreliable.”

For reasons I will discuss in more detail below, the motion was denied and Officer

Gunnells identified appellant Morales as the Hispanic male he chased on August 25.

An expert in forensic biology and science testified that some of the DNA recovered

from the gun matched appellant’s DNA.

Appellant testified at trial, admitting (1) that he was the man the officers had

chased and (2) that he had possessed the gun. However, he asserted that he fled

because he was “scared” and that he had not participated in the shooting at the Metro

station. Instead, appellant claimed the shooter — who was depicted in two different
 52

videos recorded by cameras at the Metro station — was a person named José René.

Appellant testified that he was walking to a convenience store when René

approached him, directed appellant to accompany him behind the store, and ordered

appellant to “[t]ake this and hide it” while handing him a gun. Appellant also

testified that René ordered appellant to exchange clothes with him and depart with

René’s companion, and that appellant and René are of similar height and “look

alike.” Appellant stated that he complied because he was “scared” of René, as he

believed that René was a member of the MS-13 gang.

On cross-examination, appellant acknowledged that he had previously told

detectives that he was the shooter at the Metro station and that he had fired one time.

He also admitted that he had previously told the same detectives that he had been

instructed to “do a hit on someone,” which was why he chased the victim toward the

Metro station and then fired at him.
 53

II. Analysis

A. The in-court identification

The pre-trial identification — if that is an accurate description 1 — was not

disclosed to the jury and is not directly at issue. And appellant does not claim that

the circumstances of the in-court identification were themselves impermissibly

suggestive. Instead, appellant contends that Officer Gunnells’ in-court identification

“should be set aside where the pretrial identification procedures were impermissibly

suggestive.”

However, there is no rule requiring per se exclusion. Manson v. Brathwaite,

432 U.S. 98, 111–14 (1977). Instead, we must engage in a balancing process. Even

when undue suggestivity has been shown, if “[w]e cannot say that under all the

1
When discussing the pre-trial procedure, the prosecutor stated that “[a]s far
as the identification — for lack of a better term, we’ll say identification — the
disclosure that I made to counsel was not that they made an identification, but that
. . . I expect the officers to make an in-court identification.” The record does not
indicate that Officer Gunnells made any identification of appellant after being shown
the photograph, but merely discloses that the prosecutor showed the photograph to
the officer. In any event, the procedure was unduly suggestive and must be analyzed
as if an identification was made.
 54

circumstances . . . there is ‘a very substantial likelihood of irreparable

misidentification,’” then “such evidence is for the jury to weigh.” Cureton v. United

States, 386 A.2d 278, 287 (D.C. 1978) (quoting Brathwaite, 432 U.S. at 116). In

other words, “if the indicia of reliability are strong enough to outweigh the

corrupting effect of the police-arranged suggestive circumstances, the identification

evidence ordinarily will be admitted, and the jury will ultimately determine its

worth.” Perry v. New Hampshire, 565 U.S. 228, 232 (2012).

In assessing reliability, we consider various factors, including:

(1) the opportunity of the witness to view the criminal
during the crime, (2) the witness’ degree of attention, (3)
the accuracy of his prior description of the criminal, (4)
the level of certainty demonstrated at the confrontation,
and (5) the time between the crime and the confrontation.

Stewart v. United States, 490 A.2d 619, 623 (D.C. 1985). But we may not consider

“evidence of the defendant’s guilt external to the identification.” Long v. United

States, 156 A.3d 698, 708 (D.C. 2017). “Put another way, the fact that an

identification is proved accurate by evidence unrelated to the identification itself

does not mean the identification is reliable within the meaning of a proper due

process analysis.” Id.
 55

I agree that having the officer view a mug shot of Mr. Morales prior to trial

was unduly suggestive. 2 See Patterson v. United States, 384 A.2d 663, 666 (D.C.

1978) (“single-photo displays are inherently suggestive” and “the most

objectionable method of pretrial identification”). However, the use of a suggestive

pre-trial procedure is not dispositive; we still must review whether the identification

bears sufficient indicia of reliability. Brathwaite, 432 U.S. at 116 (stating that

“identifications arising from single-photograph displays may be viewed in general

with suspicion,” but nonetheless finding the identification reliable and, thus,

admissible).

Based on a proffer of facts and the arguments of counsel, Judge Di Toro made

a preliminary ruling that she would allow the officer to make an in-court

identification of Mr. Morales. “[B]ecause I think the identification is reliable, it can

be admitted.” However, the motion to suppress was denied without prejudice, and

2
The trial judge asked why the officer was shown the mug shot. The
prosecutor explained that while preparing for trial she asked Officer Gunnells about
his broadcast description of the fleeing suspect, which mentioned his tattoos. She
asked if he “remembered any specific tattoos, and he didn’t.” She then showed him
the mug shot and they “started looking at the tattoos more closely.” During cross-
examination, the prosecutor asked appellant Morales about several tattoos on his
face and neck.
 56

the court told defense counsel that “you can come to the bench if you think that an

insufficient foundation has been laid.”

The issue was examined in more detail at trial. Before any in-court

identification was attempted, the trial court heard Officer Gunnells describe his

response to the radio assignment and his interaction with the two suspects. He

focused particularly on “the Hispanic male” who had begun reaching into his

waistband. Officer Gunnells testified that, as he approached, he saw the man from

the side and from the front, and then, as the man began to run, from the back. The

government also played the radio broadcast in which Officer Gunnells described the

man he saw as a “Hispanic male, a beanie on his head, tattoos on his face and neck,

and a jacket tied around his waist.” However, the trial judge did not allow an in-

court identification at that point, saying “I don’t think we’re there yet.”

Officer Gunnells stated that he had appellant in his sight for about sixty

seconds and “saw him from the front” for two or three seconds during the chase.

(The video from his body-worn camera indicates it was a little longer.) He was about

as close as “[f]rom here to that podium.” The parties construe the record differently,

variously asserting that the officer was, at the closest, somewhere from ten to twenty-

five feet away from appellant.
 57

The government then played the footage from Officer Gunnells’ body-worn

camera. As certain portions were played, the officer explained what was happening.

After seeing the video and hearing the narration, Judge Di Toro ruled that Officer

Gunnells could now attempt an in-court identification. The court noted that the

identification would be subject to cross-examination and added that “the jury can

decide at this point . . . what weight to give the identification based on the amount

of time that they’ve now seen he had visually through his camera and whatever else

is elicited.” Officer Gunnells then made an in-court identification, pointing to the

defendant when asked if “the individual that [he] chased . . . on August 25th” was

present.

B. The identification was admissible

Juries regularly consider “evidence with some element of untrustworthiness.”

Brathwaite, 432 U.S. at 116. We do not ask whether the in-court identification was

free from doubt, but whether it was reliable enough to satisfy the requirements of

due process — avoiding a substantial likelihood of irreparable misidentification.

Under the due process clause, “the ‘central question’ is whether ‘the identification

was reliable even though the confrontation procedure was suggestive.’” Cureton,

386 A.3d at 284–85 (quoting Neil v. Biggers, 409 U.S. 188, 199 (1972)). If the
 58

identification satisfies that constitutional requirement, we allow the jury to fulfill its

traditional function of determining how to evaluate the identification along with the

other evidence. See, e.g., Watkins v. Sowders, 449 U.S. 341, 347–49 (1981) (“the

proper evaluation of evidence under the instructions of the trial judge is the very task

our system must assume juries can perform”; “[w]here identification evidence is at

issue . . . no . . . special considerations justify a departure from the presumption that

juries will follow instructions” because “the time-honored process of cross-

examination [is] the device best suited to determine the trustworthiness of

testimonial evidence”). With that principle in mind, I turn to the indicia of reliability

revealed by the testimony about that initial confrontation.

The officer undoubtedly focused closely on the man he was chasing, as he was

a specially trained police officer investigating a nearby shooting. See Brathwaite,

432 U.S. at 115 (emphasizing that the witness was “not a casual or passing observer,”

but “a trained police officer on duty” who “could be expected to pay scrupulous

attention to detail, for he knew that subsequently he would have to find and arrest”

the suspect). Officer Gunnells testified that he “observed” the person for about sixty

seconds, from multiple angles, and from a distance that the government contends

was as close as ten feet away. The jury was able to compare this testimony to the

footage from the body-worn camera. Officer Gunnells’ ability to recognize
 59

appellant was based on both the longer opportunity he had to observe appellant from

various angles and his more abbreviated opportunity to view the appellant head-on. 3

Officer Gunnells broadcast a description of the man he encountered, stating

repeatedly that he saw a “Hispanic male, a beanie on his head, tattoos on his face

and neck, and a jacket tied around his waist.” This description, while not minutely

detailed, was accurate (as the camera footage reveals) and confirms that the officer

was paying close attention to distinctive characteristics of the man he was pursuing.

Cf. Brathwaite, 432 U.S. at 115 (in-court identification reliable despite being made

eight months after an officer initially described “the [drug] vendor’s race, his height,

his build, the color and style of his hair, [] the high cheekbone facial feature,” and

“clothing the vendor wore”); Turner v. United States, 622 A.2d 667, 673 (D.C. 1993)

3
See Nichols v. United States, 343 A.2d 336, 344 (D.C. 1975) (despite
allegedly suggestive process, identification was sufficiently reliable to be admitted,
even though identification was based on view from an eighth-floor window, as “a
witness’ close observation of facial characteristics is not necessary” when the
“witness has had sufficient opportunity to observe the figure of a person committing
a crime, and has a vivid impression of such characteristics as height, build, race, hair
style, and clothing”); see also United States v. Wong, 40 F.3d 1347, 1360 (2d Cir.
1994) (despite allegedly suggestive process, identification was nonetheless reliable
when a witness “observed the gunman after she ducked under the table at the
restaurant, staring him in the face for two to three seconds before he turned away,”
as a “fleeting but real good look at [the] assailant [was] sufficient for identification,”
especially when the “degree of attention was very high” because of his possession
of a firearm) (brackets and internal quotation marks omitted).
 60

(even assuming undue suggestivity, identification made after viewing an “intruder”

from “about fifteen feet” was sufficiently reliable when the witness “specifically

noted the person’s clothing, build, height, hair and skin color”).

Officer Gunnells also identified appellant with certainty, adding that

“[w]ithout the beanie, he look[ed] the same” as he did on August 25, 2017. And

though four months elapsed between the confrontation and the in-court

identification, that length of time is not unusual or disqualifying.

This jury was especially well-equipped to assess Officer Gunnells’ testimony

because it was able to compare his characterization of events — such as how close

Officer Gunnells was to the man he chased and for how long and from what angle

the officer viewed the man’s face — with the video from his body-worn camera. In

such a situation, it is particularly appropriate to allow the jury to weigh the evidence,

rather than to exclude it altogether. As the Supreme Court has summarized, “[w]e

are content to rely upon the good sense and judgment of American juries, for

evidence with some element of untrustworthiness is customary grist for the jury mill.

Juries are not so susceptible that they cannot measure intelligently the weight of

identification testimony that has some questionable feature.” Brathwaite, 432 U.S.
 61

at 116. This identification bore sufficient indicia of reliability to allow the jury to

perform its traditional function.

C. Any error was harmless

We have emphasized during the course of these opinions that “reliability” is

not the same thing as accuracy. However, while focusing so closely on vocabulary,

we should not lose sight of the fact that accuracy is our ultimate concern. The two-

step process — addressing undue suggestivity, then reliability — is designed to

prevent misidentification. See Brathwaite, 432 U.S. at 116 (“Surely, we cannot say

that under all the circumstances of this case there is a substantial likelihood of

irreparable misidentification.”) (internal quotation marks omitted). Thus, our

attention properly shifts to corroboration and accuracy when we assess the

harmlessness of any error in admitting the evidence.

Appellant acknowledged that he was the person fleeing from Officer

Gunnells. While this testimony does not enter into our consideration of whether the

identification was sufficiently reliable to be admitted, it did “effectively conced[e]

that [the] identification of him was accurate.” Curtis v. United States, 886 A.2d 92,

95 (D.C. 2005). Both parties agreed that appellant was the man Officer Gunnells
 62

confronted and the undisputed physical evidence corroborated this as well. Thus,

even if we were to assume that the trial court erred in admitting the identification,

we should be confident that “the guilty verdict actually rendered in this trial was

surely unattributable to the error.” Ellis v. United States, 941 A.2d 1042, 1049 (D.C.

2008) (quoting Sullivan v. Louisiana, 508 U.S. 275, 279 (1993) (describing test for

harmless constitutional error)).

At trial, appellant did not contest that Officer Gunnells chased him, but tried

to explain why he was in possession of the firearm, dressed in the same clothes as

the shooter, and fleeing from the police. The essential question the jury resolved

was not whether appellant was the man who possessed the firearm, but whether he

possessed it voluntarily. That question was answered independently of the

identification, making it clear that the guilty verdict was not attributable to the

alleged error. Because identity was conceded by the defense when the defendant

himself testified, any error in admitting the identification would have been harmless

beyond a reasonable doubt. 4

4
The jurors heard and saw the following evidence from which they could
reasonably infer that appellant was the shooter: (1) the shooting took place shortly
before appellant was confronted by police; (2) appellant fled when approached by
police; (3) appellant testified that he was the man who discarded the gun while
fleeing from police; (4) DNA evidence linked appellant to the gun; (5) an expert
 63

My colleagues improperly disregard the evidence presented by the defense.

In United States v. Hasting, the Court explained that a reviewing court must consider

“the trial record as a whole,” 461 U.S. 499, 509 (1983), when assessing

harmlessness. See Delaware v. Van Arsdall, 475 U.S. 673, 681 (1986) (“Since

Chapman, we have repeatedly reaffirmed the principle that an otherwise valid

conviction should not be set aside if the reviewing court may confidently say, on the

whole record, that the constitutional error was harmless beyond a reasonable

doubt.”); White v. United States, 613 A.2d 869, 875 (D.C. 1992) (en banc) (same). 5

testified that the cartridge case recovered from the Metro station was fired by the
gun discarded by appellant; (6) the person depicted in footage from Officer
Gunnells’ body-worn camera looked similar to the person depicted in the videos of
the shooting at the Metro station; (7) the person depicted in both videos was wearing
the same clothes; and (8) though appellant testified at trial that his statement was
untrue, the jury heard that appellant admitted to police that he was the shooter.
Additionally, as the jury was shown each of the videos mentioned above, it had the
ability to compare appellant’s appearance to that of the person depicted in each
video.
5
The decision in Smith, on which my colleagues rely, is not nearly as broad
as they suggest. It recognized that a defendant’s testimony will be considered
tainted, and thus excluded from a harmlessness inquiry, where he testified “in order
to overcome the impact of confessions illegally obtained and . . . improperly
introduced[.]” Smith v. United States, 529 A.2d 312, 318 (D.C. 1987) (quoting
Harrison v. United States, 392 U.S. 219, 223 (1968)). Nothing like that happened
here.
 64

This commitment to considering the whole record is consistent with the

approach of this court and the D.C. Circuit to appeals challenging the sufficiency of

the evidence to support a conviction. United States v. Foster, 783 F.2d 1082 (D.C.

Cir. 1986) (en banc); Hairston v. United States, 497 A.2d 1097, 1104 n.12 (D.C.

1985); Franey v. United States, 382 A.2d 1019 (D.C. 1978). As the court explained

in Foster, the use of “waiver” language in these cases is “a conventional fiction used

to describe and produce the result that the courts will not blind themselves to

incriminating evidence introduced by the defendant who chooses to respond, rather

than to demur, to the government’s case.” 783 F.2d at 1084. This is not a

technicality, but a “policy judgment”—“a defendant demonstrated to be guilty

beyond a reasonable doubt on the basis of all the valid and admissible evidence will

not be set free merely because, had an earlier erroneous ruling been made correctly,

the trial would have ended before sufficient evidence to convict had been

introduced.” Id. at 1084–85. 6

6
Moreover, even if Officer Gunnells had not made an in-court identification,
a trial judge who had listened to his testimony and that of Officer Weber, heard that
appellant’s DNA was on the gun, and seen the videos from the Metro shooting and
the footage from Officer Gunnells’ body-worn camera, properly would have denied
a motion for judgment of acquittal made at the close of the government’s case-in-
chief and allowed the jury to decide the case.
 65

My colleagues also contend that the government has waived any claim of

harmless error. It is true that the word “harmless” does not appear in its brief, but

the government made essentially the same argument by emphasizing that “appellant

himself corroborated Officer Gunnells’ testimony when he acknowledged at trial

that he was the person who fled from Officers Gunnells and Weber, and discarded

the firearm.” To reinforce this point, the government quoted from our decision in

Curtis, where we held that there had been no error in denying the motion to suppress

because the defendant admitted that he punched the victim, “thus effectively

conceding that [the victim’s] identification of him was accurate.” Curtis, 886 A.2d

at 95. 7 This should be enough to refute any claim of waiver. I simply am reading

the government’s brief fairly rather than raising the issue of harmlessness sua sponte.

In any event, this court is commanded by statute to “give judgment after an

examination of the record without regard to errors or defects which do not affect the

substantial rights of the parties[,]” D.C. Code § 11-721(e) (2012 Repl.), and we have

7
While this sentence in Curtis could be construed as a misdirected comment
on the reliability of the identification, it seems to be more appropriately treated as a
recognition that any error was harmless. Cf. Long, 156 A.3d at 708 (explaining that
“the fact that an identification is proved accurate by evidence unrelated to the
identification itself does not mean the identification is reliable within the meaning
of a proper due process analysis”).
 66

“the authority to find trial court error harmless notwithstanding the government’s

failure to claim harmlessness.” Randolph v. United States, 882 A.2d 210, 223 (D.C.

2005). 8 We have concluded that we should exercise this discretionary authority with

restraint, and “only when harmlessness is obvious.” Id. In my judgment, this is such

a case. We should affirm all of appellant’s convictions.

8
The majority asserts that it is an open question whether we have this
authority when it comes to constitutional errors. Ante at note 21. It may be an
“open” question in the sense that we apparently have never disavowed or clarified
the language from Arnold which is quoted in Randolph (and footnote 21). But it is
not a hard question under current law. The comment seems to trace back to (1) a
footnote in United States v. Lee, 489 F.2d 1242, 1246 n.10 (D.C. Cir. 1973) (which
addresses non-constitutional error), and (2) to a time when people doubted that
constitutional errors could ever be deemed harmless. After Chapman v. California,
386 U.S. 18 (1967), and Arizona v. Fulminante, 499 U.S. 279 (1991), we now
recognize that, except for so-called “structural” errors, claims of constitutional error
at trial are reviewable for harmlessness. The comment from Arnold also traces back
to a previous version of 28 U.S.C. § 2111, the statutory harmless error principle,
which referred to “technical errors.” See Hasting, 461 U.S. at 507–09 & n.7. The
current version of 28 U.S.C. § 2111 is essentially the same as our D.C. Code § 11-
721(e).