Shelton v. United States

majority opinion

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. 24-CF-0668

DEMANN SHELTON, APPELLANT,

UNITED STATES, APPELLEE.

Appeal from the Superior Court

of the District of Columbia

(2019-CF2-016024)

(Lynn Leibovitz, Judge)

(Argued February 25, 2026 Decided September 10, 2026)

Areeba Jibril, Public Defender Service, with whom Jaclyn Frankfurt and Shilpa S. Satoskar, Public Defender Service, were on the briefs, for appellant.

R. Allen Darby, Assistant United States Attorney, with whom Jeanine Ferris Pirro, United States Attorney, and Chrisellen R. Kolb, Bryan H. Han, and Luke Albi, Assistant United States Attorneys, were on the brief, for appellee.

Before BLACKBURNE-RIGSBY, Chief Judge, and EASTERLY and SHANKER,

Associate Judges.

Opinion for the court by Associate Judge SHANKER.

Dissenting opinion by Associate Judge EASTERLY at page 35.

SHANKER, Associate Judge: In December 2019, police officers from the

Metropolitan Police Department (MPD) arrested appellant Demann Shelton along

with two other men. After observing that the car the men were in was illegally

parked, officers approached and saw the man in the front passenger seat, Antonio

Smith, acting in an erratic manner, moving his arms frantically all around him. Mr.

Smith subsequently attempted to flee from police, causing the officers to suspect that

he possessed and was attempting to hide contraband. After detaining the three men,

officers searched the car and found a black jacket, among several other items, behind

the front passenger seat. Officers discovered a gun in one of the jacket pockets and

also found drugs and drug paraphernalia in the jacket. Mr. Shelton had been sitting

in the back of the car and there was no dispute that the jacket belonged to him.

The government charged Mr. Shelton with, among other counts, multiple

offenses related to his alleged possession of the gun and accompanying ammunition

found in the jacket. At trial, the government asked the jury to convict Mr. Shelton

on the theory that he was the sole possessor of the gun and drugs. Mr. Shelton’s

defense theory was that Mr. Smith put the gun and drugs in Mr. Shelton’s jacket just

before he fled.

Following closing arguments, the trial court instructed the jury on actual and

constructive possession. After deliberating for about eight hours, the jury sent a note

asking, “Can two or more people have joint possession of an item?” After conferring

with the parties, the court determined that there was sufficient evidence to give a

supplemental jury instruction on the legal concept known as joint constructive

possession. The court’s response to the jury explained that the law recognizes joint

constructive possession but admonished that, in order to convict Mr. Shelton, the

jury had to find that the government proved all the elements of constructive

possession as to him notwithstanding any role that Mr. Smith played with respect to

the items in the jacket. The jury returned a guilty verdict on Mr. Shelton’s gunrelated charges about an hour later.

Mr. Shelton raises one issue on appeal: whether the trial court reversibly erred

when, in responding to the jury note, it stated that the law recognizes that two or

more people can have property in their joint constructive possession. We see no error

in the content of the response to the jury note. Accordingly, we affirm Mr. Shelton’s

firearms convictions.

I. Background

A. Mr. Shelton’s Arrest

On an afternoon in December 2019, MPD officers arrested Mr. Shelton along

with Mr. Smith and Marquis Payne. Before their arrest, the three men had been in a

parked car together, with Mr. Payne in the driver’s seat, Mr. Smith in the front

passenger seat, and Mr. Shelton in the seat behind the driver. Officers patrolling in

the Northwest quadrant of the District observed that the car appeared to be illegally

parked. As officers approached, Mr. Smith saw them and began acting in an erratic

and suspicious manner, moving his arms frantically all around the front passenger

seat. Officers did not, however, see him put anything in the car’s glove box, nor did

they recall him reaching behind his seat.

Mr. Smith’s “furtive movements” indicated to the officers that he may have

been attempting to hide contraband, which prompted them to try and remove him

from the car. After ignoring the officers’ repeated commands, Mr. Smith eventually

opened the car door, pushed aside an officer, and fled from the scene. Following a

brief chase, officers apprehended Mr. Smith and recovered multiple small “zip” bags

containing a “white rock like substance” believed to be narcotics. Mr. Shelton, on

the other hand, did not make any furtive movements or suspicious gestures as

officers approached the vehicle. As Mr. Smith fled, Mr. Shelton stayed with officers

near the car.

Upon searching the vehicle, officers found a black jacket, among several other

items, behind the front passenger seat and removed it from the car. Officers

discovered a gun in one of the jacket pockets. 1 The jacket also contained 52 small

Officers also found a gun in the car’s glove box. The government did not and does not contend that that gun belonged to Mr. Shelton, and the trial court

“zip” bags containing a white rock-like substance, a sandwich bag with 7.8 grams of

additional suspected narcotics, 20 suboxone strips, and a digital scale lacking a

battery. In addition, officers found in the jacket Mr. Shelton’s photo identification

and bank card as well as pieces of mail addressed to him. Officers did not see Mr.

Shelton wearing, holding, touching, or putting anything inside the jacket. After

arresting Mr. Shelton, officers searched him and found $2,442 in cash.

B. Procedural History

Mr. Shelton was charged with multiple drug and gun offenses. After the

government dismissed certain charges and charge enhancements, Mr. Shelton

proceeded to a jury trial on charges of attempted unlawful possession with intent to

distribute a controlled substance (PWID), in violation of D.C. Code

§ 48-904.01(a)(1) (two counts); unlawful possession of a firearm (prior conviction)

(FIP), in violation of D.C. Code § 22-4503; carrying a pistol without a license

(CPWL), in violation of D.C. Code § 22-4504(a)(1); possession of an unregistered

firearm (UF), in violation of D.C. Code § 7-2502.01(a); unlawful possession of

ammunition (UA), in violation of D.C. Code § 7-2506.01(a)(3); and possession of

drug paraphernalia (PDP), in violation of D.C. Code § 48-1103(a).

instructed the jury that there was “no evidence that Mr. Shelton ever possessed” that gun.

During closing arguments in Mr. Shelton’s trial, the government contended

that Mr. Shelton was the sole possessor of the gun and drugs found in the jacket,

asserting that he “was sitting in the back seat . . . by himself” with “a black

jacket . . . that [was] chock-full of evidence” and that he “had the power and intent

to control the items” in the jacket. The government also argued that it did not “make

sense” and was “not logical” to think that Mr. Smith could have placed the gun and

drugs into the jacket from his front seat position. It concluded by asserting that if the

jury “believe[d] beyond a reasonable doubt that this jacket belongs to [Mr. Shelton],

then beyond a reasonable doubt everything inside it is his.” Mr. Shelton, although

acknowledging that the jacket was his, asserted that his presence in the back seat

near the jacket did not prove that he had constructive possession of the gun or the

drugs. He pointed to Mr. Smith’s frantic and panicked movements at the sight of

police and contrasted them with his own cooperative behavior to argue that it was

more likely that Mr. Smith placed the gun and drugs into the jacket in an attempt to

conceal them.

Following closing arguments, the trial court instructed the jury, including on

actual and constructive possession. With respect to constructive possession, the court

explained that “a person may exercise control over property not in his or her physical

possession if that person has both the power and the intent at a given time to control

the property,” and that “[m]ere presence near something or mere knowledge of its

location . . . is not enough to show possession.”

The next day, the jury sent a note asking, “Can two or more people have joint

possession of an item?” and adding that it had reached a verdict on Mr. Shelton’s

PDP charge but was “deadlocked on all other counts.” After the court asked the

parties for their thoughts, the government asked the court to instruct the jury on the

legal doctrine of joint constructive possession. Mr. Shelton objected to the

instruction because the government had not made a joint possession argument at trial

and he had therefore been unable to counter that theory. The court dismissed the jury

and reconvened the parties the following morning for further discussion on how to

respond to the jury note.

In the resumed discussion, Mr. Shelton argued again that the government had

presented no evidence of joint constructive possession and, in fact, had claimed the

opposite: that Mr. Smith had nothing to do with the gun in the jacket. The trial court

was unconvinced and stated that, even though neither party had advanced a joint

constructive possession argument at trial, evidence presented at trial supported that

legal theory. The court cited Mr. Shelton’s argument, supported by evidence elicited

during cross-examination, that Mr. Smith was the source of the firearm in Mr.

Shelton’s jacket. It then determined that if the jury was “making potentially factual

findings that the gun may well have been deposited in the jacket by [Mr. Smith] or

handed back to [Mr. Shelton] by [Mr. Smith],” then the jury was “permitted to infer

that [Mr. Shelton] played a role in accepting the gun, in placing it in the back in the

jacket pocket where it was located as if [Mr. Smith] couldn’t have done the whole

thing.”

The court subsequently instructed the jury:

I’m now going to respond to the question you asked me

about joint possession. The law recognizes the possibility

that two or more individuals can jointly have property in

their constructive possession. Two or more persons have

property in their joint constructive possession when they

each have both the power and the intent at a given time to

control the property.

I instruct you that if you consider joint constructive

possession, in order for you to find the defendant

possessed an item, notwithstanding the role of the frontseat passenger, the Government must still prove beyond a

reasonable doubt all of the elements of each offense and

that the defendant voluntarily and on purpose had the

power and intent to control the item.

I further instruct you that mere presence near something or

mere knowledge of its location, however, is not enough to

show possession. To prove possession of an item against

the defendant in this case, the Government must prove

beyond a reasonable doubt that he had either actual or

constructive possession of it.

You must consider this instruction along with all the

instructions I’ve given you, and you must not give greater

weight to this instruction than to the other instructions you

have been given.

In other words, you must consider all of the instructions as

a whole and give equal weight to this and to all the others.

A little over an hour after receiving the instruction, the jury notified the court

that it had reached a verdict. The jury acquitted Mr. Shelton of both attempted PWID

counts but convicted him of the gun-related charges (and it had already delivered a

guilty verdict on the PDP charge). This appeal, which challenges only the firearms

convictions, followed.

II. Analysis

Mr. Shelton asserts that the trial court reversibly erred in informing the jury

about the legal concept of joint constructive possession because, according to him,

there was insufficient evidence to support his conviction under that legal theory and

the court’s response allowed the jury to convict him despite the lack of evidence that

he “was working together with someone else to jointly possess the gun.”

We see no error. Responding directly to the jury’s specific question, the court

correctly explained that, under the law, two or more people can jointly constructively

possess an item, while elucidating that this requires that “each” person had the power

and intent to control the property. The court then essentially reiterated its initial

constructive possession instruction, telling the jury that it could not convict Mr.

Shelton unless the government proved beyond a reasonable doubt all of the elements

of each offense as to him, including that Mr. Shelton voluntarily and purposefully

had the power and intent to control the items. The court added that the jury had to

find as much “notwithstanding the role of” Mr. Smith; that mere presence near

something or mere knowledge of its location is not enough to show possession; and

that to prove possession by Mr. Shelton, the government had to prove beyond a

reasonable doubt that he had either actual or constructive possession of it.

If the jury heeded these instructions—which we must presume it did—there

is no possibility that it convicted Mr. Shelton without finding that he constructively

possessed the gun, regardless of Mr. Smith’s conduct—and Mr. Shelton concedes

that the evidence was sufficient to support a finding that he constructively possessed

the gun. And if the jury found that Mr. Shelton and Mr. Smith both constructively

possessed the gun—which is, after all, what joint constructive possession is—that

did not serve to expand the basis for liability as to Mr. Shelton; it merely expanded

the number of individuals who constructively possessed the gun.

A. Standard of Review

“When the jury explains specific difficulties, the trial court should clear them

away with concrete accuracy. When a jury sends a note which demonstrates that it

is confused, the trial court must not allow that confusion to persist; it must respond

appropriately.” Parker v. United States, 333 A.3d 1162, 1189 (D.C. 2025) (citation

modified); see (Anthony) Smith v. United States, 306 A.3d 67, 73 (D.C. 2023)

(assessing whether “the instruction adequately addressed the jury’s apparent

confusion and adequately informed the jury of the relevant legal principles”). “A

party is entitled to a jury instruction upon the theory of the case if there is sufficient

evidence to support it.” Washington v. United States, 111 A.3d 16, 23 (D.C. 2015)

(citation modified). The sufficiency of the evidence to support a jury instruction is a

question of law, “as to which our review is de novo.” Peyton v. United States, 278

A.3d 713, 723 (D.C. 2022).

In (Anthony) Smith, we noted some ambiguity in our case law about the

standard of review of a trial court’s response to a jury note. 306 A.3d at 71 (noting

that we have said both that the decision on what further instructions, if any, to give

in response to a jury question lies within the trial court’s discretion and also that we

review de novo whether a challenged jury instruction adequately stated the law, and

therefore assuming without deciding that the trial court’s response to a jury note was

subject to de novo review). Contrast Parker, 333 A.3d at 1189 (“Absent an abuse of

discretion, we will not reverse the trial court’s decision [on what further instructions,

if any, to give in response to a jury question] on appeal.”), with Evans v. United

States, 304 A.3d 211, 219 (D.C. 2023) (“If the objection was preserved, we review

the propriety of the trial court’s response to the jury note de novo.”).

We think the review of a response to a jury note is analogous to the review of

jury instructions, and the discrepancy in our case law can be explained on that basis.

In Alleyne v. United States, we recognized a similar lack of clarity in our cases about

the standard of review for jury instructions, and we explained that, “while we review

the trial court’s decision to give an instruction for abuse of discretion, the accuracy

of an instruction itself is a legal question that we review de novo, if the legal question

was preserved.” 327 A.3d 472, 483 n.7 (D.C. 2024) (citation modified); see Fleming

v. United States, 224 A.3d 213, 219 (D.C. 2020) (en banc) (“Although our

terminology has not always been entirely clear on this point, we review de novo

whether challenged jury instructions adequately state the law.”). Moreover, a court

“by definition abuses its discretion when it makes an error of law.” Williams v.

United States, 345 A.3d 570, 579 (D.C. 2025) (citation modified).

Accordingly, we review a trial court’s decision regarding whether and how to

respond to a jury note for an abuse of discretion, while reviewing the legal accuracy

of the response and any further instructions de novo. Lucas v. United States, 240

A.3d 328, 343 (D.C. 2020) (“While we review the trial court’s decision on what, if

any, response to give to a jury’s question for abuse of discretion[,] the accuracy of

the instruction itself is a legal question that we review de novo.”).

B. Joint Constructive Possession

We begin by explaining what joint constructive possession is and what it is

not. “To establish constructive possession, the government must prove beyond a

reasonable doubt that the accused (1) knew that the contraband was present, and

(2) had the ability and intent to exercise dominion or control over the contraband.”

Sanders v. United States, 330 A.3d 1013, 1032 (D.C. 2025); accord Rivas v. United

States, 783 A.2d 125, 129 (D.C. 2001) (en banc). It is well established that, “as a

general matter, knowledge of and proximity to contraband are insufficient to

establish the intent necessary to prove constructive possession.” Sanders, 330 A.3d

at 1032-33.

“Constructive possession may be sole or joint and may be proven by direct or

circumstantial evidence.” Id. at 1032 (quoting Bruce v. United States, 305 A.3d 381,

393 (D.C. 2023)); see also Rivas, 783 A.2d at 129; United States v. Benjamin, 711

F.3d 371, 378 (3d Cir. 2013) (“[U]nder [a] theory of joint possession, dominion and

control need not be exclusive but may be shared with others.” (citation modified)).

But even where multiple people are involved, the government must still show that

each person had the ability and intent to exercise dominion and control over an item

to succeed under the joint constructive possession doctrine; and, for that reason, if

the evidence shows constructive possession by one person, the fact that another

person also had constructive possession does not preclude liability. See Carter v.

United States, 957 A.2d 9, 16 (D.C. 2008) (“[T]he possibility that Tucker, too, may

have had the ability to exercise control over the gun in the car . . . does not

undermine an inference that Carter had the requisite intent to control the gun.

‘Constructive possession may be sole or joint. . . .’” (quoting Rivas, 783 A.2d at

129)); Moore v. United States, 927 A.2d 1040, 1051 (D.C. 2007) (concluding that

“[i]t is of little moment that much the same evidence showed constructive possession

by” two people because “[p]ossession can be joint” and the evidence may justify

attributing the illegal items to both people beyond a reasonable doubt); Criminal Jury

Instructions for the District of Columbia, No. 3.104 §§ B & C (5th ed. 2023) (setting

forth the same language for sole actual or constructive possession and joint actual or

constructive possessive jury instructions, except that the joint possession instruction

adds: “In addition, the law recognizes the possibility that two or more individuals

can jointly have property in their constructive possession. Two or more persons have

property in their joint constructive possession when they each have both the power

and the intent at a given time to control the property.”).

Numerous cases from other jurisdictions support this conception of joint

constructive possession. See, e.g., United States v. Ford, 22 F.4th 687, 693 (7th Cir.

2022) (“Constructive possession can be joint or sole. . . . In cases of joint

constructive possession, the focus is on whether the government can establish a

substantial connection between the defendant and the contraband.”); United States

v. Willis, 14 F.4th 170, 182 (2d Cir. 2021) (“The fact that others may have also

possessed the keys, drugs, and weapons does not preclude a finding that [the

defendant] did so as well.”); Hopson v. Horton, 838 F. App’x 147, 154-55 (6th Cir.

2020) (“Joint possession is a type of either actual or constructive possession; the

essential inquiry is still whether the defendant has control over the weapon.”);

United States v. Wright, 739 F.3d 1160, 1169 (8th Cir. 2014) (stating that

“constructive possession may be joint and need not be exclusive” and concluding

that the evidence as to the defendant individually was sufficient to show that he

possessed narcotics, “whether exclusively or jointly” (citation modified)); Benjamin,

711 F.3d at 379 (“The fact that Esprit had actual possession of the gun and may not

have always been with Benjamin . . . does not preclude Benjamin’s constructive

possession of the same weapon.”); United States v. Duval, 496 F.3d 64, 71, 78 (1st

Cir. 2007) (affirming felon-in-possession convictions despite appellant’s challenge

to instruction that stated that “[f]or purposes of determining a defendant’s guilt, joint

possession is not different from sole possession”); United States v. Garth, 188 F.3d

99, 113 (3d Cir. 1999) (concluding that when “the doctrine of joint constructive

possession applies[,]” the “involvement of multiple actors does not negate the

requirement that [the appellant] knowingly have both the power and intent to

exercise dominion and control” of the contraband); United States v. Perlaza, 818

F.2d 1354, 1362 (7th Cir. 1987) (holding that the trial court’s instruction that the

jury must “consider [the defendants] separately” when deliberating on a theory of

joint constructive possession correctly stated the law because it made clear that each

defendant must individually satisfy the elements of constructive possession to be

convicted); United States v. Davis, 562 F.2d 681, 689 & n.9 (D.C. Cir. 1977)

(holding that the trial court’s jury instruction correctly expressed the law of

possession when it stated that “[y]ou may find that the requirement of possession is

satisfied only if you find beyond a reasonable doubt that the defendant had actual or

constructive possession either alone or jointly with others”).

Thus, because even in the case of multiple possible possessors the elements

of constructive possession must be proved as to each particular defendant, joint

constructive possession neither lowers the bar for a determination of constructive

possession as to an individual defendant nor permits a jury to find constructive

possession based on the conduct of others. It is not a theory of conspiratorial or

vicarious liability—and because it is not such a theory, it does not require proof of

concerted action. See Ortiz v. Dubois, 19 F.3d 708, 713 (1st Cir. 1994) (“[J]oint

possession . . . does not require proof of concerted action or shared state of mind,

and thus, does not offer a basis for vicarious liability for the criminal acts of

others.”).

To be sure, we have made clear that in shared-space cases or cases involving

multiple individuals in proximity to contraband, “there must be something more in

the totality of the circumstances that—together with proximity and knowledge—

establishes that the accused meant to exercise dominion or control over the

narcotics.” Rivas, 783 A.2d at 130; see Ramirez v. United States, 49 A.3d 1246, 1249

(D.C. 2012) (“Constructive possession requires more than mere presence of the

accused on the premises, or simply his proximity to the drugs. There must be

something more in the totality of the circumstances that establishes that the accused

meant to exercise dominion or control over the narcotics.” (citation modified)). That

“something more” can be—and often is—evidence of a joint criminal venture. See

Moore, 927 A.2d at 1051 (noting that “in some cases involving the presence of more

than one person on the premises or even shared occupancy and control, common

possession of contraband is not indicated, and the evidence does not justify

attributing the illegal items to one party rather than another beyond a reasonable

doubt,” and that in such cases, evidence of “a concert of illegal action” involving the

contraband can “reinforce[ ] the inference of joint possession” (citation modified));

Wheeler v. United States, 494 A.2d 170, 173 (D.C. 1985) (“We have held that

proximity may establish a supportable case of constructive possession of narcotics

when colored by evidence linking the accused to an ongoing criminal operation of

which that possession is a part.”); id. (evidence of “concert of illegal action” can

“dispel any fear that the ‘constructive possession’ doctrine has cast too wide a net”

in shared-space cases).

We have never, however, said that the “something more” must in all cases be

concerted action. To the contrary, in Rivas, we made clear that it need not be

concerted action, explaining that the additional evidence necessary to prove

constructive possession in a shared-space case is “comparatively minimal” and could

be “a furtive gesture indicating an attempt to access, hide[,] or dispose of the

object[;] flight or other evidence of consciousness of guilt[;] evidence of

participation in an ongoing criminal venture involving the contraband[;] an

inculpatory statement[;] evidence of prior possession of the item[;] actual possession

of paraphernalia relating to the use or sale of the contraband[;] control of the area or

container in which the contraband is found[;] or the like.” 783 A.2d at 137; see id.

at 128 (stating that the “something more” can be “a word or deed, a relationship or

other probative factor”). The Rivas list was in the disjunctive, and we further

observed that it was “not exhaustive.” Id. at 137 n.13.

This makes sense. If the evidence shows constructive possession by multiple

individuals, there is no reason they should not each face criminal liability for a

possessory offense even if they were not involved in a joint criminal venture.

Accordingly, we have consistently found the evidence sufficient to support

constructive possession by an individual defendant where others might also have

constructively possessed the item, even in the absence of any indication of concerted

action. See, e.g., See Johnson v. United States, 290 A.3d 500, 516 (D.C. 2023)

(“Even when the accused shared the premises with others, additional probative

evidence may suffice to establish constructive possession by the accused, as where

the contraband was found in plain view, in the defendant’s own bedroom or other

personal space, or in proximity to the defendant’s personal items such as mail or

personal papers, photographs, and identification cards.” (citation modified)); Dorsey

v. United States, 154 A.3d 106, 112-13 (D.C. 2017) (even though the defendant did

not live in the apartment and others were present when police entered it, the evidence

was sufficient to show the defendant’s constructive possession based on his close

proximity to the location of the gun, his evasive conduct, indications that he

purposefully went to the gun’s location when officers approached, and a DNA match

to the defendant from the gun); (Tamara) Smith v. United States, 55 A.3d 884, 888

(D.C. 2012) (observing that “something more in the totality of the circumstances”

was needed because the defendant shared the apartment, and concluding that “two

significant factors” (not including concerted action) “constitute[d] ‘something

more’”: (1) a backpack containing the gun was “conspicuous” in the defendant’s

bedroom and adjacent to the defendant’s bed and (2) it was undisputed that the

defendant had been alone in the bedroom prior to the search); Ramirez, 49 A.3d at

1249-50 (observing that “something more” was needed because the defendant shared

an apartment with another individual, and concluding that “the government met its

obligation to show ‘something more in the totality of the circumstances’” because

the defendant was near the drugs, appeared to have been attempting to dispose of

them, and acknowledged that he lived in the apartment in a certain bedroom, and in

that bedroom was the defendant’s property); Moore, 927 A.2d at 1050-51 (relying

on the facts that the defendant had a key to the apartment, the defendant admitted to

living there, and the items were found in the apartment’s only bedroom in plain view

next to the defendant’s personal papers, and holding that, “[t]aken together, these

facts allowed a jury to infer beyond a reasonable doubt that [the defendant] knew the

drugs and gun were in his apartment, and that he had the ability and intent to exercise

dominion and control over them,” before noting that evidence of concerted action

“reinforced” that inference).

Contrary to the dissent’s characterization, the above explication of joint

constructive possession neither is “groundbreaking” nor effects a “change” in the

law. Post at 36. As our reliance on, among other binding decisions, Rivas, Carter,

and Moore, as well as on persuasive authority from other jurisdictions, demonstrates,

our view is amply supported. Indeed, in Carter, we said that “the possibility that

Tucker, too, may have had the ability to exercise control over the gun in the car does

not undermine an inference that Carter had the requisite intent to control the gun.”

957 A.2d at 16 (citation modified). Replacing the names “Tucker” and “Carter” in

that quotation with, respectively, “Smith” and “Shelton” resolves this case. 2 By

contrast, the dissent’s assertion that joint constructive possession is an independent

theory of, essentially, conspiratorial liability, involving a different showing of intent

and different evidence, is entirely unsupported. The dissent does not point to a single

case where the evidence was sufficient to show the defendant’s sole constructive

possession but liability was precluded because someone else also satisfied the

requirements of constructive possession and there was no evidence that the

defendant and that individual had acted in concert.

The dissent’s attempt to characterize Rivas as such a case reflects a

fundamental misunderstanding of our analysis in that case. Post at 44-46. In Rivas,

the question was whether there was sufficient evidence of Mr. Rivas’s own

In his brief on appeal, Mr. Carter argued that Mr. Tucker possessed the firearm and “there was no evidence presented that Mr. Carter and Mr. Tucker planned a crime, conferred during the offense, or shared proceeds of the crime afterwards.” Carter v. United States, Brief of Appellant, 2007 WL 5844065, at *15-16 (citation modified). We were “satisfied” that the evidence was sufficient to show that Mr. Carter had constructive possession of the firearm and ammunition based on his own conduct, and we then rejected the notion that Mr. Tucker’s constructive possession bore on Mr. Carter’s intent to exercise dominion and control over the gun, without mentioning any concerted action by the two men. Carter, 957 A.2d at 16. Carter controls, but all the dissent can say about it is that it is “simply a sufficiency of the evidence case.” Post at 67.

constructive possession of narcotics; that is not the case here, where both Mr.

Shelton and our dissenting colleague recognize that there was sufficient evidence

that he constructively possessed the gun and drugs in his jacket. Post at 38. In Rivas,

we rejected the government’s attempt to impute liability to Mr. Rivas based on the

much clearer constructive possession of Mr. Melgar, the owner and driver of the car

in which the drugs were found. 783 A.2d at 128, 135. Here, by contrast, joint

constructive possession was not applied to extend liability from Mr. Smith to Mr.

Shelton. 3 The question, rather, was whether, where Mr. Shelton constructively

possessed the gun and drugs, Mr. Smith’s possible additional constructive

possession cast doubt on Mr. Shelton’s constructive possession. Rivas would be

helpful to the dissent’s position if there the evidence had been sufficient to show Mr.

Rivas’s constructive possession and we said that, nonetheless, the fact of Mr.

Melgar’s constructive possession required reversal of Mr. Rivas’s conviction

because the government had not proven that the two had acted together.

The dissent asserts that we have “no way of knowing this,” post at 64 n.11, but we do: the trial court instructed the jury that it could not do so, and we assume that jurors follow the court’s instructions. See Foote v. United States, 108 A.3d 1227, 1239 (D.C. 2015) (noting the “crucial assumption underlying our constitutional system of trial by jury that jurors carefully follow instructions” (citation modified)). If the dissent means that we cannot be 100% positive that the jury did not ignore its instructions, that can be said in any case.

The dissent’s assertion that two people cannot exercise dominion and control

over an item unless they do so under some sort of agreement “to preclude control by

the rest of the world,” post at 42, is likewise demonstrably incorrect. It takes little

imagination to picture an individual, Jones, who secrets his drugs in a desk drawer

in his shared apartment, oblivious that his roommate, Williams, is aware and

partakes of the stash. Jones had constructive possession and Williams had

constructive possession. It would make little sense to say—and no authority

suggests—that Jones cannot be found to have constructively possessed the drugs

because Williams also did and the two were not acting together “to preclude control

by the rest of the world.” 4

Ultimately, what the dissent appears to not understand is that, even after the

jury note, no one was seeking to establish Mr. Shelton’s constructive possession

based on Mr. Smith’s constructive possession. The issue—as framed by the trial

court’s response to the note—was whether, if Mr. Shelton’s constructive possession

In this hypothetical, at trial Jones could very well assert a defense that the drugs belonged to Williams. If the jury sent a note asking whether Williams’s dominion and control over the drugs meant that Jones could not be guilty unless the two were working in concert, surely the correct response would be “no.”

was established on the basis of his own conduct, it mattered whether Mr. Smith also

had constructive possession. Settled authority makes clear that it did not. 5

C. The Trial Court’s Response to the Jury Note

In light of the above correct understanding of the concept of joint constructive

possession, we conclude that the trial court neither abused its discretion nor legally

erred in responding to the jury in the way that it did.

The court first answered the jury’s specific question by correctly affirming

that “two or more individuals can jointly have property in their constructive

possession.” Mr. Shelton does not argue that this was legally inaccurate. Nor does

he dispute that there was sufficient evidence that someone else possessed the gun at

some point; indeed, that was his defense at trial and he maintains on appeal that “Mr.

Smith’s actions, in particular, suggested that he could have placed the gun in the

jacket, either by dropping the gun into the jacket in the back seat or by hiding it in

For this reason, our dissenting colleague’s ominous prophecy that our decision here “breaks down the guardrails we have erected to protect against guilt by proximity and association,” post at 70, is unfounded. It is, has always been, and remains the case that to conclude that a defendant jointly constructively possessed an item, the jury must find that the defendant individually satisfied the elements of constructive possession and cannot rely on the defendant’s mere knowledge of and proximity to contraband. Rivas, 783 A.2d at 128, 130, 137-38. Nothing in this majority opinion suggests, as the dissent contends, that a jury may convict a defendant of constructive possession “simply because there is another person present.” Post at 70.

the jacket before dropping the jacket in the back seat.” 6 And, as the jury appeared

unclear about the legal concept of constructive possession, which was at issue with

respect to both the gun and narcotics charges Mr. Shelton was facing, the court

properly sought to address that confusion. See Johnson, 290 A.3d at 519 (“Where a

jury has demonstrated confusion, the trial judge may not allow that confusion to

continue, but must make an appropriate and effective response.” (citation

modified)). The trial court’s answer that the law recognizes joint constructive

possession was correct, as it would have been improper to suggest that the jury could

not find Mr. Shelton guilty of constructive possession on the ground that someone

else also constructively possessed the contraband, see Carter, 957 A.2d at 16, or to

decline to answer the question and thereby allow the jury’s confusion to continue.

The trial court then sought to eliminate from the jurors’ minds the possibility

that Mr. Shelton could be guilty simply because “the gun [was] shoved into his lap”

or he “had to accept it but didn’t really intend to exercise dominion and control”—

Nor does the dissent argue that the evidence was insufficient to show Mr. Smith’s constructive possession of the items in the jacket, and in any event, that would not be a claim belonging to Mr. Shelton, nor would it undermine his own constructive possession. The dissent argues that “there was no evidence that the two men interacted, much less that they possessed or hid the gun together.” Post at 36. Contrary to the dissent’s assertion that we do not “not answer the evidentiary question the parties have briefed,” post at 36, our answer is that no such evidence was required.

that is, that he could be guilty based on mere proximity or vicarious liability. To that

end, it inserted the following language into the standard jury instruction for joint

constructive possession (which already makes clear that “each” person must have

both the power and the intent at a given time to control the property and that “[m]ere

presence near something or mere knowledge of location . . . is not enough to show

possession”):

I instruct you that if you consider joint constructive

possession, in order for you to find the defendant

possessed an item, notwithstanding the role of the frontseat passenger, the Government must still prove beyond a

reasonable doubt all of the elements of each offense and

that the defendant voluntarily and on purpose had the

power and intent to control the item. (Emphasis added.)

The trial court thus accurately instructed the jury that it had to find that the

government proved beyond a reasonable doubt that Mr. Shelton satisfied the

elements of constructive possession, and it eliminated the risk that the jury would

misunderstand that joint constructive possession allowed it to convict Mr. Shelton

based on the conduct of Mr. Smith. Because “[w]e ordinarily presume that the jury

understands and obeys the trial judge’s instructions[,]” Holloway v. United States,

25 A.3d 898, 903 (D.C. 2011), we must presume that the jury did not convict Mr.

Shelton of constructive possession without finding that he voluntarily and

purposefully had the power and intent to control the gun. The court’s response

accordingly did not permit the jury to find Mr. Shelton guilty on an unproven or

invalid theory. 7

Our dissenting colleague repeatedly declares as established fact that the jury was “unpersuaded” by the government’s sole constructive possession theory and then “swiftly” found Mr. Shelton guilty after the court’s response to the jury note. Post at 35, 38; see post at 55-57. But again, if we assume, as we must, that the jury followed its instructions, there is no possibility that it found Mr. Shelton guilty without finding that he individually satisfied all the elements of constructive possession. See Allen v. United States, 603 A.2d 1219, 1224 (D.C. 1992) (en banc) (“It would be absurd to upset a verdict upon a speculation that the jury did not do their duty and follow the instructions of the court.” (citation modified)).

In any event, we have no idea why the jury sent its note or how the court’s response bore on the jury’s deliberations. See Owens v. United States, 497 A.2d 1086, 1093 (D.C. 1985) (“We cannot indulge in speculation about the jury’s thought processes, and thus we must disregard appellant’s conjectures as to how the jury may have arrived at its verdict.” (citation modified)). In this regard, we observe that the jury note did not specifically ask whether two people can jointly possess a gun, but asked whether two or more people can “have joint possession of an item[.]” When it sent the note, the jury was “deadlocked” on both the attempted PWID and the gunrelated counts, and after the court’s response, which referred to property or items more generally, the jury acquitted Mr. Shelton of the attempted PWID counts. It is entirely possible that the court’s instruction—that joint constructive possession is a thing but that the government still had to prove beyond a reasonable doubt that Mr. Shelton voluntarily and on purpose had the power and intent to control “the item” “notwithstanding the role of the front-seat passenger”—led to the jury’s acquittal of Mr. Shelton on the attempted PWID charges. This is all the more plausible because (1) a separate gun was found in the car’s glove box, in front of Mr. Smith’s seat, suggesting that he placed his gun there and the gun in the jacket was Mr. Shelton’s, and (2) the drugs found in the jacket were similar in nature and packaging to the drugs found on Mr. Smith, suggesting that the jury might have thought that he stashed what he quickly could in the jacket. We do not know that this is true, but our dissenting colleague certainly does not know that it is not true. We further note that, contrary to the dissent’s characterization, post at 58, this is not a point about harmlessness; the point is that the dissent is engaging in pure speculation when it

Mr. Shelton’s arguments to the contrary are unavailing. He contends that the

trial court’s confirmation that joint constructive possession is a legal doctrine was

improper because the government’s theory was that he solely constructively

possessed the gun; although he argued and elicited evidence (on cross-examination)

that Mr. Smith placed the gun in the jacket, that was directed only at creating

reasonable doubt about his guilt and therefore could not support an instruction on

joint constructive possession. 8 He is wrong. “A defendant may open the door to a

joint-possession instruction through his own evidence and arguments.” United States

v. Norris, 21 F.4th 188, 194 (1st Cir. 2021); see id. (“Norris also attempted to pin

the gun and drugs on another occupant of the residence searched. . . . In short, the

posits that the jury convicted Mr. Shelton of the gun-related charges because of the trial court’s answer to the jury note.

The dissent assumes that the reason the government argued that Mr. Smith could not have deposited the gun and drugs in Mr. Shelton’s jacket was that it thought that only sole, and not joint, possession was a viable scenario. Post at 35. That is not the most logical explanation for the government’s framing. Mr. Shelton’s defense was that Mr. Smith (solely) possessed the gun and drugs, so naturally the government sought to foreclose the possibility that the jury would conclude that if Mr. Smith placed the gun and drugs in the jacket, Mr. Shelton was not guilty. The government did not then pivot from sole to joint constructive possession, but rather took (and now takes) the (correct) position that if Mr. Shelton constructively possessed the gun and drugs, it is “of little moment” whether Mr. Smith also did. Moore, 927 A.2d at 1051.

In this same vein, contrary to the dissent’s claim, post at 57, the trial court’s instruction did not turn an “exculpatory” defense (that Mr. Smith was the sole possessor of the gun) into an “inculpatory” one, because both before and after the instruction, Mr. Shelton’s guilt turned on whether he had had constructive possession of the gun, regardless of whether Mr. Smith also did.

evidence extant and Norris’s own theory of the case made the joint possession

instruction appropriate. It matters not that the government and Norris each argued

for a different sole-possessor theory, because these arguments permitted the jury to

infer that both purported possessors shared the contraband.” (citation modified)); see

also Henry v. United States, 94 A.3d 752, 757 (D.C. 2014) (“The evidence

supporting a requested instruction may be an amalgam of portions of the

government’s evidence and portions of the defense evidence.” (citation modified));

United States v. Driggers, 913 F.3d 655, 657-58 (7th Cir. 2019) (holding that joint

constructive possession instruction was proper even though government’s theory

was that defendant solely possessed the gun); United States v. Howard, 687 F.3d 13,

19 (1st Cir. 2012) (stating that because, “when crafting jury instructions a judge must

consider all of the evidence introduced at trial, in other words, the government’s as

well as the defense’s,” a joint possession instruction was appropriate based on the

defendant’s evidence and arguments (citation modified)); United States v. Carrasco,

257 F.3d 1045, 1050 (9th Cir. 2001) (finding a joint possession jury instruction

appropriate when the defendant testified, and his counsel argued, that the gun he was

charged with possessing belonged to his vehicle passenger). 9

The dissent does not address these cases in asserting, without supporting authority, that “[t]he defense argument that Mr. Smith might have possessed the gun

Mr. Shelton next asserts that, even if there was evidence that Mr. Smith

possessed the gun, it was insufficient evidence to support the court’s response to the

jury because it did not establish that he and Mr. Smith or Mr. Payne “were involved

in some greater joint scheme or conspiracy with respect to the gun.” But as we have

explained, joint constructive possession is not an alternative theory of liability

premised on concert of action. Joint constructive possession simply means that it is

possible for two individuals to constructively possess an item, either together or

independently, if they each had the power to exercise control over the item. See

United States v. Rainone, 816 F.3d 490, 494 (7th Cir. 2016) (“The joint possession

instruction properly informed the jury that Rainone could still have possessed the

gun even if others had also possessed the firearm.”). Both joint constructive

possession as a legal doctrine itself and the court’s response to the jury note here

foreclosed the possibility that the jury could have imputed liability to Mr. Shelton

based on Mr. Smith’s conduct, so the lack of evidence of concerted action is

irrelevant. See United States v. Fairley, 137 F.4th 503, 514-15 (6th Cir. 2025)

(stating that in cases of joint possession, an instruction that “conveys (1) the jury’s

obligation to determine whether the evidence conclusively proved the defendant

did not, on its own, support [a joint constructive possession inference] and thereby open the door to a joint-constructive-possession instruction” and that “an alternative sole possessor is not the same thing as a potential joint possessor.” Post at 53.

guilty and (2) that it should not permit the possible additional criminal liability of

others to influence its decision” is appropriate (citation modified)). 10

Mr. Shelton and our dissenting colleague cite Thomas v. United States, 806

A.2d 626 (D.C. 2002), for the proposition that “the trial court erred in telling the jury

it could convict Mr. Shelton of joint constructive possession of the gun in the jacket

where there was no evidence of a ‘concert of illegal action’ between Mr. Shelton and

Mr. Smith or Mr. Payne.” In Thomas, it was unclear from the evidence whether Mr.

Thomas or other individuals who were standing with him, including another suspect,

Mr. Parker, threw a gun to the ground. Id. at 627-28. At Mr. Thomas’s trial, the court

instructed the jury on actual and constructive possession, and the jury subsequently

sent a note asking whether “more than one person [can] have constructive

possession?” Id. at 628. The court responded to the jury question simply with the

word “yes.” Id. We held that the court’s response was improper because the

government did not present at trial evidence or a theory that Mr. Thomas and Mr.

Parker both had constructive possession of the gun; “there was no evidence

Our dissenting colleague asserts that, in rejecting the view that joint constructive possession requires a joint criminal venture, we are “attacking a strawman.” Post at 62. That is, however, precisely Mr. Shelton’s argument as to why the trial court’s response to the jury note was erroneous; indeed, he explicitly complains that “there were no facts from which the jury could infer a joint criminal enterprise connecting multiple occupants of the car with the firearm.”

indicating how long Mr. Thomas and Mr. Parker were together before the police

arrived or what they had been doing” or any of “the requisite circumstances

indicating a concert of illegal action that obviously tends to dispel the natural fear

that the doctrine of constructive possession is casting too wide a net.” Id. at 629

(citation modified).

The concern in Thomas, as in Rivas, was the possibility that the jury thought

“that Mr. Thomas could be convicted if Mr. Parker, or one of the others, threw down

the gun.” Id. We found that problematic in the absence of something more showing

Mr. Thomas’s ability and intent to exercise dominion and control over the

contraband. Id. We called circumstances indicating a concert of illegal action a

“requisite” something-more factor, but that was because facts showing concerted

action were required under the circumstances in Thomas, where there was nothing

else linking Mr. Thomas with the gun and the jury’s question and the trial court’s

response suggested the possibility that Mr. Thomas was convicted because the jury

thought someone threw the gun down and Mr. Thomas could be guilty regardless of

who it was. We decline to ascribe load-bearing weight to Thomas’s use of the word

“requisite,” especially where doing so would run headlong into Rivas, which Thomas

post-dated.

Thomas does not stand (and has never been cited) for the proposition that to

prove joint constructive possession where multiple people each had the power and

intent to control an item, the government must in all cases also prove concerted

action. Such a requirement would, moreover, make little sense. Take a scenario in

which two individuals share a bedroom to which they have equal and unfettered

access, and a gun is found in the open on a nightstand in the bedroom next to both

individuals’ personal property. Both individuals would have had constructive

possession of the gun, 11 and there is no logic in or legal basis for preventing both

individuals from being guilty of a possessory crime even if they had never engaged

in, as Mr. Shelton puts it, “some greater joint scheme or conspiracy with respect to

the gun.” 12

See Johnson, 290 A.3d at 516 (“Even when the accused shared the premises with others, additional probative evidence may suffice to establish constructive possession by the accused, as where the contraband was found in plain view, in the defendant’s own bedroom or other personal space, or in proximity to the defendant’s personal items such as mail or personal papers, photographs, and identification cards.” (citation modified)); Proctor v. United States, 172 A.3d 396, 403 (D.C. 2017) (“Evidence showing the accused’s control or occupancy of the premises in which the contraband is found may also serve to prove constructive possession.”).

In re R.G., 917 A.2d 643 (D.C. 2007), is not to the contrary. There, the flaw in the government’s constructive-possession position was simple: there was “no evidence in the present record that could fairly be viewed as proving beyond a reasonable doubt that R.G., rather than her companion, exercised dominion and control over the pistol, or even that she exercised joint control with him.” Id. at 650. Here, there was plenty of evidence that Mr. Shelton exercised dominion and control over the gun.

The essence of Mr. Shelton’s and the dissent’s arguments is that the trial

court’s response to the jury note allowed the jury to convict Mr. Shelton of

constructive possession of the gun based on the conduct of Mr. Smith despite the

absence of evidence that the two were in cahoots. That argument rests on a faulty

premise—that joint constructive possession requires concert of action—and it also

ignores the bulk of the trial court’s response, which made abundantly clear that the

jury could not convict Mr. Shelton of constructive possession unless it found that he

had the power and intent to control the gun notwithstanding the role of Mr. Smith

and that mere presence near something or mere knowledge of its location is not

enough to show possession. The court’s answer that the law recognizes joint

constructive possession perhaps allowed the jury to find that Mr. Smith also

constructively possessed the gun. But that is “of little moment” for Mr. Shelton,

Moore, 927 A.2d at 1051, as the jury had to find that he independently constructively

possessed the gun, and he concedes that the evidence was sufficient to support such

a finding.

III. Conclusion

For the foregoing reasons, we affirm Mr. Shelton’s FIP, CPWL, UF, and UA

convictions, as well as his PDP conviction, which he did not challenge on appeal.

So ordered.

EASTERLY, Associate Judge, dissenting: Sole and joint constructive possession

are different legal theories of guilt which require the government to make different

evidentiary showings. At Demann Shelton’s trial, everyone—the government, the

defense, and the court—understood this. The government exclusively pursued the

former theory. From opening statements through summation, it argued that Mr.

Shelton had solely constructively possessed the gun found in the jacket beside him

in the back seat of Marquis Payne’s car, and it dismissed as “not logical” and beyond

permissible inference that the front-seat passenger, Antonio Smith, might have

hidden the gun there, as the defense had argued. The jury appeared unpersuaded by

the government’s sole constructive possession theory, and, after a full day of

deliberation, it asked in a note including a claim of deadlock, “[c]an two or more

people have joint possession of an item?” Over Mr. Shelton’s objection, and

notwithstanding its recognition of the government’s exclusive pursuit at trial of “a

theory of sole possession,” the court concluded that the evidence presented allowed

an inference that Mr. Shelton had “played a role in accepting the gun” from Mr.

Smith, and thus the jury was “permitted to find joint constructive possession.” The

court then so instructed the jury, and the jury swiftly found Mr. Shelton guilty of

various firearm offenses.

On appeal, Mr. Shelton challenges the trial court’s assessment that there was

sufficient evidence in the record for the jury to consider the theory of joint

constructive possession in addition to that of sole constructive possession, arguing

that there was no evidence that the two men interacted, much less that they possessed

or hid the gun together. The government, now a strong proponent of a joint

constructive possession theory, defends the trial court’s ruling. But the majority

opinion does not answer the evidentiary question the parties have briefed.

Instead, the majority opinion opts to change the law. It declares sole and joint

constructive possession to be one and the same. Joint constructive possession only

means the possession need not be exclusive. Thus, so long as there is a foundation

for an instruction on sole constructive possession and one or more other people were

around, the majority opinion reasons, the jury may also consider joint constructive

possession. On this basis, the majority opinion holds that the trial court did not err

in instructing the jury about joint constructive possession.

The majority opinion presents its holding as grounded in black letter law, but

it is anything but. Its erasure of the distinction between the theories of sole and joint

constructive possession is groundbreaking in this jurisdiction and cannot be squared

with this court’s en banc decision in Rivas v. United States, 783 A.2d 125 (D.C.

2001), and its progeny, including Thomas v. United States, 806 A.2d 626 (D.C.

2002). These cases show that, although constructive possession always requires

proof of the same three elements—knowledge of the contraband, the ability to

exercise dominion and control over the contraband, and the intent to do so—the

operation of the ability and intent elements necessarily changes depending on

whether constructive possession is sole or joint. When the government seeks to prove

sole constructive possession, it must show that the defendant has the ability and

intent to exercise sole dominion and control over the contraband. By contrast, when

the government seeks to prove joint constructive possession, it must show that the

defendant has the ability and intent to exercise joint, or shared, dominion and

control. The government may pursue whichever theory it chooses—or both if it has

the evidentiary support—but the majority opinion is wrong to say that proof of joint

constructive possession is no different from that of sole constructive possession.

The evidence presented at Mr. Shelton’s trial did not support a theory of joint

constructive possession as it is correctly understood. The government effectively

acknowledged this when it exclusively argued at Mr. Shelton’s trial that he solely

possessed the gun. The trial court only determined that an instruction on joint

constructive possession was warranted because it thought that the trial evidence

supported an inference that Mr. Shelton had “played a role” in helping Mr. Smith

hide the gun. But no such evidence was presented. Thus, per Thomas, the court never

should have told the jury that it could consider the theory of joint constructive

possession. Moreover, this instructional error cannot be deemed harmless under the

standard for nonconstitutional error. Kotteakos v. United States, 328 U.S. 750 (1946).

Although the evidence of sole constructive possession was sufficient to support a

guilty verdict (as Mr. Shelton conceded at oral argument), the record indicates that

the jury was not convinced of Mr. Shelton’s guilt under this theory before it sent its

combination joint-constructive-possession-question-and-deadlock note and that it

only reached a guilty verdict after it was told that it could consider an unsubstantiated

theory of joint constructive possession. Certainly, we have no basis to say that it is

“highly probable the error did not affect the jury’s verdict.” Moore v. United States,

114 A.3d 646, 658 (D.C. 2015) (citation modified); see also Kotteakos, 328 U.S. at

776. Accordingly, this court should reverse Mr. Shelton’s conviction and remand his

case for a new trial.

I. The Law of Constructive Possession

A. The Doctrine’s Purpose and Limits

Constructive possession is a judge-made “doctrine used to broaden the

application of possession-type crimes to situations in which actual physical control

cannot be directly proved.” Wayne R. LaFave, Substantive Criminal Law § 6.1(e)

(3d ed. 2025). This court has long acknowledged the “natural fear” that applications

of the doctrine will “cast[] too wide a net,” Thomas, 806 A.2d at 629 (quoting Brown

v. United States, 546 A.2d 390, 397 (D.C. 1988), and Curry v. United States, 520

A.2d 255, 264 (D.C. 1987)), and thereby punish “mere bystander[s]” for possessing

contraband. Rivas, 783 A.2d at 128; accord In re R.G., 917 A.2d 643, 649 (D.C.

2007) (quoting Rivas); Proctor v. United States, 172 A.3d 396, 403 (D.C. 2017)

(same); Jackson v. United States, 61 A.3d 1218, 1222 (D.C. 2013) (same). After all,

the doctrine is a tool for interpreting statutes that penalize “possession” of various

types of contraband, and courts are not free to employ their interpretive powers to

stretch the meaning of that term beyond any reasonable understanding of what the

legislature could have intended. See Grayson v. AT&T Corp., 15 A.3d 219, 237-38

(D.C. 2011). More particularly, our court cannot act as a legislature and create a

general “legal duty [in the District of Columbia] to distance oneself upon learning

of the presence of a[n illegal] gun on the premises or within one’s immediate

proximity.” Conley v. United States, 79 A.3d 270, 287 (D.C. 2013) 1; see also Rivas,

783 A.2d at 145 (Ruiz, J., concurring) (cautioning that courts must “ensure that in

setting the evidentiary minimum to establish the judge-made elements of

constructive possession as a proxy for actual possession, the jury does not stray from

the offense of . . . possession that the legislature has criminalized, as possession is

normally understood”) (emphasis in original).

The court in Conley acknowledged the lack of such a general duty in the District’s statutes and further recognized that the legislature’s ability to create such a duty would be constrained by due process concerns. Id. at 281-89 (striking down as unconstitutional a statute that attempted to criminally punish people who failed to distance themselves from illegal guns in cars without proof that they knew of an obligation to do so).

To ensure that the doctrine of constructive possession is properly cabined, this

court has defined a three-part test for its application. The government must prove

that the defendant (1) knew the contraband was present, (2) had the ability to

exercise dominion and control over the contraband, and (3) had the intent to do so.

Rivas, 783 A.2d at 129, 129 n.3 (citing Bernard v. United States, 575 A.2d 1191,

1195 (D.C. 1990)); see also id. at 138 n.14 (Ruiz, J., concurring) (explaining that

cases predating Bernard omitted the intent element); Blackmon v. United States, 835

A.2d 1070, 1075 (D.C. 2003) (recognizing that Rivas endorsed the above three-part

test). Some of our more recent cases have described the test as having two parts by

combining the ability and intent-to-control requirements in one element. See, e.g.,

Bruce v. United States, 305 A.3d 381, 393 (D.C. 2023). However enumerated, a

separate assessment of intent is a “critical” component. Rivas, 783 A.2d at 135-38

(concluding that the intent element was not supported by sufficient evidence);

accord In re R.G., 917 A.2d at 647-649 (same); Bolden v. United States, 835 A.2d

532, 536-37 (D.C. 2003) (same); Burnette v. United States, 600 A.2d 1082, 1083

(D.C. 1991) (same). The intent requirement in particular “protect[s] against the

possibility that bystanders may be unwittingly caught up criminally for mere

innocent presence at the scene of illegal activity” in a way that the requisite showing

of ability to exercise dominion and control does not: “‘[s]omeone might have

effective power over [contraband] simply because [it] w[as] located within

reach . . . but if such a person had power over the [contraband] . . . but had no

intention to exercise that power, there might still be no crime.’” Rivas, 783 A.2d at

139 (Ruiz, J., concurring) (quoting United States v. Zavala Maldonado, 23 F.3d 4, 8

(1st Cir. 1994)).

B. Sole and Joint Constructive Possession as Distinct Theories of Guilt

Like actual possession, constructive possession can be sole or joint. Rivas,

783 A.2d at 129. But that distinction does not turn on a simple headcount of potential

defendants. Sole and joint constructive possession are different theories of guilt.

Although both incorporate the same three elements described above, they

necessarily differ in operation and the evidence needed to prove them.

To begin with, it makes no sense to say that sole constructive possession and

joint constructive possession are established by the very same showing of ability and

intent to exercise dominion or control. “Dominion and control” denotes the power

to “guide the destiny” of contraband, see Curry, 520 A.2d at 263, 265 (explaining

that “[d]ominion or control over an object is shown when the accused has some

appreciable ability to guide its destiny”); accord Rivas, 783 A.2d at 128, 137 n.12;

see also In re R.G., 917 A.2d at 647, 651. 2 When a single individual wields this

power, they have the ability and intent to preclude control by the rest of the world.

See Burnette, 600 A.2d at 1084 (explaining that “the concept ‘intent to exercise

dominion and control’” refers to “an intent—inferable from the circumstances—to

assert dominion as against (hypothetical) competing claims to possession by

others”); Possession, Black’s Law Dictionary (12th ed. 2024) (defining “possession”

as “the right under which one may exercise control over something to the exclusion

of all others”). It follows that when a person wields this power with others, i.e.

“jointly,” that person has the ability with those others to preclude control by the rest

of the world and the intent to share possession with those others. Joint Possession,

Black’s Law Dictionary (12th ed. 2024) (defining “joint possession” as

“[p]ossession shared by two or more persons”). To collapse this distinction—to say

that dominion and control operates the same way whether a person wields it singly

or jointly—would strip the concept of any genuine meaning and risks branding mere

bystanders as joint possessors because no showing of an intent to share control is

required. In short, for the concept of “dominion and control” to be recognizable in

Our court has long referred to “dominion and control” and “dominion or control” interchangeably, see, e.g., Rivas, 783 A.2d at 129 (using both constructions). But, as Curry reflects, we understand these paired terms to constitute a hendiadys, conveying a single idea.

any real sense, the elements of sole constructive possession and joint constructive

possession must operate differently.

Relatedly, the type of control—sole or joint—matters evidentiarily. The

government regularly must rely on circumstantial evidence to prove constructive

possession, but the inferences that can be drawn from this evidence differ depending

on whether anyone else possibly possessed the contraband. If, for example, a

defendant is found alone, in a public space, seated next to contraband in plain view

for a period of time, those facts may permit an inference that the defendant has the

ability and intent to exercise dominion and control over the contraband—and to do

so alone—because there is no potentially competing claimant whose presence would

call that inference into doubt. But if others are present during that time with equally

provable knowledge of and access to the same contraband, the inference of sole

control is “considerably weakened” and the possibility that the individual is a mere

bystander is concomitantly strengthened. Rivas, 783 A.2d at 141 (Ruiz, J.,

concurring). In this scenario, the government cannot establish constructive

possession by presenting circumstantial evidence from which a factfinder could infer

that any number of individuals had the ability and intent to control the contraband,

somehow, some way. Doing so would relieve the government of its burden to prove

that a specific, nonbystander defendant in fact had the ability and intent to exercise

a cognizable form of control over the contraband. Instead, in a setting where more

than one person might constructively possess contraband, the government must offer

“something more,” i.e., additional circumstantial evidence that supports either a

theory of sole constructive possession or a theory of joint constructive possession.

Rivas, 783 A.2d at 128, 130. Under the former, the government must present

evidence from which the factfinder can infer the defendant’s ability and intent to

control the contraband alone, notwithstanding others’ comparable access to the

contraband. Under the latter, the government must present evidence from which the

factfinder can infer both the defendant’s ability and intent to control the contraband

“with others” who enjoy comparable access to the contraband—in other words, that

the defendant had the ability and intent to share control with another person or

persons (and to deny control to everyone else). 3 Rivas, 783 A.2d at 128 (emphasis

added).

Rivas’s endorsement of sole and joint constructive possession as distinct

theories of guilt with distinct intent inquiries is reflected in its analysis of Mr. Rivas’s

The catalog of circumstantial evidence that Rivas identifies—e.g., a furtive gesture, flight, participation in an ongoing criminal venture, an inculpatory statement, prior possession, or “the like,” 783 A.2d at 137—helps to determine the intent at issue. A defendant’s furtive gesture toward contraband, for example, may be evidence of an intent to control it alone or jointly with another person depending on whether the surrounding facts connect that gesture to another individual as well as the contraband. Similarly, participation in an ongoing criminal venture may signal an intent to solely possess the contraband if there is no evidence that anyone else nearby was involved in a defendant’s criminal activity, or an intent to jointly possess the contraband with others if there is evidence of others’ involvement.

case. 4 Rejecting a rule that a defendant could be convicted of constructive possession

of contraband simply because he was found with the contraband in the close confines

of a car, the court held that “whether [a defendant] had the necessary intent to control

(individually or with others)” had to be proved by something more than mere

proximity to contraband in plain view. 783 A.2d at 128 (emphasis added). Although

the government theorized that Mr. Rivas, the front seat passenger of the car, “jointly

possessed” with the driver the cocaine found on the console between them, id. at

134-35, the court first indicated that sole constructive possession had not been

proved. Specifically, the court observed that there was no evidence that Mr. Rivas

had ever touched the bag, or had been selling drugs, or had any evidence on him

linking him to the drugs. Id. at 129, see also id. at 134-35. The court then considered

whether there was sufficient evidence of joint constructive possession and

determined that there was not because the evidence of his intent to “share[]

As illustrated by Rivas, the intent element will often be the linchpin in assessing whether constructive possession has been proved because it is often easy to show that a collection of individuals in proximity to contraband have the ability to exercise dominion and control over it, either solely or jointly, especially if the individuals are known to each other. 783 A.2d at 129-30, 135, 137; accord Smith v. United States, 55 A.3d 884, 890 (D.C. 2012) (Schwelb, J., concurring) (identifying as the “only question of substance . . . the final and most challenging element of a constructive possession case, namely” whether the defendant “intended, individually or jointly . . . to exercise dominion and control over the contraband and to guide its destiny”) (emphasis in original).

possession” of the cocaine with the driver or to have “a stake” in the drugs was

insufficient. Id. at 135.

Our post-Rivas cases similarly illustrate the distinction between the theories

of sole and joint constructive possession and indicate that, to prevail on the latter

theory, the government must, at a minimum, present proof of a defendant’s intent to

share control of the contraband with another. Thus, for example, in In re R.G., we

explained that the central question in the case was whether the evidence was

sufficient to establish R.G.’s intent to exercise dominion and control over the gun

found in the bedroom where she and her boyfriend slept. 917 A.2d at 647. We then

determined that the circumstantial evidence was insufficient to support a finding that

she, rather than her boyfriend, intended to exercise sole dominion and control over

the pistol, id. at 648-49, and further explained that, “although [she] could

theoretically be in joint constructive possession of the pistol even if the weapon

belonged to [him],” the District had not shown that she had “shared possession” with

him “or that she had a personal stake in [the gun].” Id. at 650. By contrast, in Moore

v. United States, although we acknowledged that “common possession of

contraband” is not always indicated in shared spaces, we concluded that the

government had proved that one defendant had, along with the other two, jointly

constructively possessed drugs found in the apartment where the other two lived

because “the evidence showed a close personal and business relationship [selling

drugs] among the three defendants.” 5 927 A.2d 1040, 1051 (D.C. 2007); see also

Mitchell v. United States, 64 A.3d 154, 157 (D.C. 2013) (explaining that when the

government seeks to prove that a defendant constructively possessed contraband it

“faces the challenging burden of proving, inter alia, that the defendant intended,

individually or jointly with another person or persons, to exercise dominion and

control over the weapons and to guide their destiny”) (emphasis added). 6

In Moore and other cases, we have said that “evidence of a concert of illegal action” supports the “inference of joint [constructive] possession.” 927 A.2d at 1051 (citation modified); see also, e.g., Thomas, 806 A.2d at 629. Although such evidence can help cabin the doctrine, evidence of illegal action is not required in every instance of joint constructive possession. See supra at note 3 (cataloging the types of circumstantial evidence that may inform whether a defendant had the requisite intent to solely or jointly possess contraband); see also In re R.G., 917 A.2d at 650 (requiring proof of “shared possession” or “a personal stake”). And even illegal action by the defendant and another person might not inform whether a defendant is guilty under a theory of joint constructive possession, for example, where two individuals intentionally share a gun for separate criminal aims—one to protect themselves during drug deals, the other to use during armed robberies. In other words, the key is that the evidence must support an inference of an intent to share; the purpose for which the contraband is being shared is a relevant but not limiting factor. See, e.g., Burwell v. United States, 901 A.2d 763, 769 (D.C. 2006) (concluding that driver of vehicle jointly constructively possessed marijuana in glove compartment where the smell of marijuana, its presence on the car’s lighter, and emptied cigar shells, showed that he and his companions had been smoking or were going to smoke it together).

This distinction between sole and joint constructive possession is evident in pre-Rivas decisions, too. Compare Burnette, 600 A.2d at 1085 (concluding that evidence was insufficient to show that backseat passenger in a car intended to exercise dominion and control over a gun under the mat at his feet and also that there was insufficient evidence that he was linked to an ongoing criminal operation with

Of course, just because sole and joint constructive possession are distinct

theories of guilt does not mean that the government is limited to one or the other.

Rather, the government may pursue theories of both sole and joint constructive

possession in the same way that it may seek to convict a defendant of alternative

theories of actual or constructive possession. See, e.g., Taylor v. United States, 267

A.3d 1051, 1059 (D.C. 2022). But, as in all cases where the government wishes the

factfinder to consider alternative legal theories of guilt, the government must be able

to present proof that could support a separate determination of guilt under each

theory’s distinct formulation of the elements of the offense. Jennings v. United

States, 993 A.2d 1077, 1080 (D.C. 2010) (explaining that the need for “legally

the others in the car involving the gun) and Speight v. United States, 599 A.2d 794, 796-98 (D.C. 1991) (concluding that evidence of intent to possess drugs in apartment where the defendant did not live and where his relationship to the drugs and the others present was unproven was insufficient) with McDaniels v. United States, 718 A.2d 530, 532 (D.C. 1998) (concluding that where defendant was the driver of a car which sped away from the police and from which an AK-47 fell or was thrown at the same time the defendant and his companions abandoned the car, “[a] jury could reasonably infer . . . that appellant . . . jointly and constructively possessed it with the other occupants” because the “attempted concealment of this weapon could not have been done without appellant’s active participation”) and Roy v. United States, 652 A.2d 1098, 1107 (D.C. 1995) (where evidence indicated that defendant and companion were “associated together” in an effort to sell a gun, it supported an inference that “the two men had the requisite ability and intention jointly to exercise dominion and control over the weapon and to guide its destiny”). But I rely primarily on post-Rivas decisions to avoid any confusion about whether intent is a standalone element in the test for constructive possession (which some of our older cases fail to acknowledge) and whether the government enjoys a relaxed burden of proof to establish a passenger’s constructive possession of contraband in a vehicle (a proposition rejected by Rivas).

sufficient evidence to support an alternative charge offered by the government” is

an “evidentiary restraint” on instructing the jury about alternative legal theories); see

also, e.g., Tyree v. United States, 942 A.2d 629, 636 (D.C. 2008) (explaining that

“[w]hether to give an aiding and abetting instruction is a matter within the discretion

of the trial court, provided there is evidence sufficient to support a finding of guilt

on this theory”).

We applied this principle to the doctrine of constructive possession in

Thomas, 806 A.2d 626. Recognizing that sole and joint constructive possession are

analytically distinct theories of guilt, we concluded that the government, which had

exclusively pursued “a sole constructive possession or actual possession theory,”

had “presented no evidence supporting [a] joint constructive possession theory.” Id.

at 629. We therefore held that the court had erred by instructing the jury, in response

to its question “whether ‘more than one person [can] have constructive possession

of a thing,’” that it could consider the theory of joint constructive possession

alongside a sole constructive possession theory. Id. (explaining that the jury’s

question and the court’s response contained “an implicit assumption that Mr.

Thomas could be convicted if [one of his companions had] thr[own] down the gun”

but that “[t]his assumption is entirely unsupported by the evidence”). Thomas

controls our resolution of this case.

II. The Inapplicability of Joint Constructive

Possession in Mr. Shelton’s Case

A. The Trial Court’s Error

In the government’s prosecution of Mr. Shelton for various possession-of-afirearm offenses, from start to finish, it exclusively pursued a theory of sole

constructive possession. The government based this argument on its evidence that

the police found Mr. Shelton sitting by himself in the back seat on the driver’s side

of a car and found a jacket on the seat next to him with a loaded gun in one of its

pockets. The jacket contained identification, a bank card, and mail with Mr.

Shelton’s name on it. 7 In making this argument, the government specifically asserted

that Mr. Shelton “still had the power and intent to control” the jacket and its contents

after he set the jacket down at his side, and the jacket and its contents were “not

suddenly in possession of the car as a whole.” Anticipating the defense argument

that the gun belonged to Mr. Smith, not Mr. Shelton, the government argued in

closing that “[a]t no point did [the jury] hear any testimony from officers saying that

The jacket also contained a number of small clear baggies containing what appeared to be crack cocaine and 20 suboxone strips, and the government prosecuted him for possession with intent to distribute both substances. But, as the defense highlighted in closing, the government never presented any evidence that the substance in the baggies had been tested and determined to be cocaine, and suboxone strips are sold by pharmacies in packages of 20 strips, approximately a week’s dose for one person with a prescription (which the government failed to prove Mr. Shelton did not have). The jury acquitted Mr. Shelton of both drug offenses.

Mr. Smith,” the front seat passenger who fled the car, “had turned around behind

him and was trying to conceal something in the back seat when the officers were

approaching”; instead the government stressed that Mr. Smith had his own concerns

and “[h]is actions were focused forward in the glove box area,” where the police

found another gun.

In rebuttal, the government doubled down on its theory of sole constructive

possession after the defense—highlighting evidence about the location of the gun

behind the front-seat passenger, the frantic hand movements of that passenger, Mr.

Smith, upon the approach of the police, and Mr. Smith’s flight from the car (in

contrast to Mr. Shelton, who remained seated)—argued that Mr. Smith might have

hidden the gun in Mr. Shelton’s jacket before he exited the car and ran from the

police. The government told the jury that, because the defense theory that Mr. Smith

had turned around and put his gun in Mr. Shelton’s jacket while also “rifling through

a glove box with a second set of hands” was not “physically possible,” did “n[o]t

make sense,” and was “not logical,” the jury could not “make an inference as to

that.” And the government concluded by telling the jury it should find Mr. Shelton

guilty because the jacket and everything inside it were Mr. Shelton’s.

After summations, the trial court instructed the jury that “a person” may

constructively possess contraband “not in his or her physical possession if that

person has both the power and the intent at a given time to control the property.”

(emphasis added). The jury deliberated for a full day before it sent a note asking

“[c]an two or more people have joint possession of an item[?]” The court indicated

that it was of the view that the jury should not be told that joint constructive

possession was a viable theory unless there was evidence that Mr. Shelton had the

ability and intent to share possession of the gun with either of his companions in the

car: The court distinguished between a jury finding either “that the defendant

essentially accepted the gun” or that he was given a gun that he “d[id]n’t even know

about or want to have dominion and control over,” and indicated that the former

would support a theory of joint constructive possession, but the latter would not. The

court simply thought that evidence had been presented that permitted an inference

that Mr. Shelton had had the ability and intent to share the gun with Mr. Smith.

The trial court reasoned that (1) the defense had asked the jury to infer that

Mr. Smith possessed the gun based on his actions in the front seat, and (2) the fact

that “the gun was found deep inside one of the jacket pockets” supported the

inference that the gun had been “handed back to [Mr. Shelton] by [Mr. Smith]” and

that Mr. Shelton had “played a role in accepting the gun, in placing it in the back in

the jacket pocket where it was located.” Accordingly, the court found “that, on this

record, the jury is permitted to find joint constructive possession.” It then instructed

the jury that “[t]he law recognizes the possibility that two or more individuals can

jointly have property in their constructive possession” and indicated that the jury

could consider this theory in deciding Mr. Shelton’s case. Approximately an hour

later, the jury reached a verdict and found Mr. Shelton guilty of all the charged

possession-of-a-firearm offenses.

The trial court was wrong to conclude that the evidence presented supported

a theory that Mr. Shelton and Mr. Smith had joint, i.e., shared, possession of the gun

found in Mr. Shelton’s jacket. Specifically, there was no evidence to support the

requisite showing that Mr. Shelton had the intent to control the gun “with others,” as

opposed to “individually,” Rivas, 783 A.2d at 128. The defense argument that Mr.

Smith might have possessed the gun did not, on its own, support this inference and

thereby open the door to a joint-constructive-possession instruction; an alternative

sole possessor is not the same thing as a potential joint possessor. See supra at Part

I.B. Moreover, the evidence on which the defense argument was based—Mr. Smith’s

frantic hand gestures while the police approached the car—did not support this

theory of joint constructive possession. As the government correctly argued in

closing, there was no testimony that Mr. Smith turned to the back of the car and

made those gestures. And the court’s inference that Mr. Shelton helped Mr. Smith

conceal the gun in the jacket because of the gun’s location deep inside the jacket

pocket had no evidentiary foundation. No witness testified that the gun was so

positioned. Instead, the court relied entirely on its observation of the body-worn

camera footage of an officer removing the gun from a jacket pocket. But this removal

occurred after the police took the jacket out of the car and put it on the ground, and

there was no evidence about how the police handled the jacket, much less that they

had taken pains to make sure that the items in the jacket did not shift in transport.

In addition to the grounds for a joint-constructive-possession instruction

identified by the trial court, the government argues on appeal that “the evidence of

drug dealing was . . . sufficient to allow the jury to reasonably infer [Mr.] Shelton

and [Mr.] Smith jointly [constructively] possessed the firearm.” The police did

recover drugs from both Mr. Shelton and Mr. Smith, but there was no evidence that

they were engaged in drug dealing (or packaging) together at the time the police

approached them. See Rivas, 783 A.2d at 134 (noting the absence of such evidence

in concluding that the government’s joint constructive possession theory was

unfounded). All the police saw was Mr. Shelton sitting in the car and Mr. Smith

making frantic gestures and then fleeing from the car. Nor did the evidence even

indicate that they possessed drugs from the same source; the alleged crack cocaine

they each possessed appeared to have been packaged differently—Mr. Shelton’s

were packaged in a single knotted clear bag and in clear zip bags, whereas Mr.

Smith’s were in small “yellow zips”—and Mr. Shelton possessed other drugs

(suboxone strips) that Mr. Smith did not. See supra at note 7. Thus, the government’s

reliance on generalized expert opinion testimony about the drugs, money, and

paraphernalia found on the two men is unavailing; it supports an inference only that

each man might have intended to sell whatever substances they each possessed, not

that Mr. Shelton intended to sell drugs jointly with Mr. Smith, much less that he

intended to possess the gun jointly with Mr. Smith to do so. See Rivas, 783 A.2d at

152 (Ruiz, J., concurring) (acknowledging the limitations of such expert testimony

in establishing intent).

Absent an evidentiary foundation, the court’s response to the jury note—

telling the jury that joint constructive possession was a viable theory of guilt—was

error in this case just as it was in Thomas, 806 A.2d at 629. The only theory of guilt

the jury should have been permitted to consider was that of sole constructive

possession. Cf. Tyree, 942 A.2d at 638 (holding that it was error to allow the jury to

consider an aiding and abetting theory that had “no factual underpinning”; instead,

the jury should have been able to decide only whether the defendant “‘was either the

principal or a nonparticipant’” in the murder (quoting Brooks v. United States 599

A.2d 1094, 1100 (D.C. 1991))).

B. Harm

The court’s erroneous instruction to the jury was not harmless under the

Kotteakos standard for nonconstitutional error. Under Kotteakos, the government

bears the burden to provide this court with “fair assurance” that the error did not

“substantially sway” the judgment. 328 U.S. at 765. In other words the government

must prove it is “highly probable the error did not affect the jury’s verdict.” Moore,

114 A.3d at 658 (citation modified); see also Kotteakos, 328 U.S. at 764. “[I]f we

are ‘left in grave doubt’ as to whether” an error affected a verdict, or if “‘the court is

in virtual equipoise as to the harmlessness of the error . . . , [we] should treat the

error . . . as if it affected the verdict.’” Hinton v. United States, 979 A.2d 663, 691

(D.C. 2009) (en banc) (quoting Kotteakos, 328 U.S. at 765 and Fry v. Pliler, 551

U.S. 112, 121 n.3 (2007)).

In Mr. Shelton’s case, the government’s sole constructive possession theory

was unquestionably strong. Nonetheless, the jury appeared to be unpersuaded by it.

After deliberating for a full day, it asked the court about joint constructive possession

at the same time that it said that it was deadlocked on all charges but one (the

unlawful possession of drug paraphernalia charge). The jury’s inquiry about joint

constructive possession thus indicates that it was looking for a different theory on

which to convict Mr. Shelton of gun possession. See Gray v. United States, 79 A.3d

326, 337-38 (D.C. 2013) (court’s inadequate response to jury’s note was not

harmless where it was inferable that jury asked its question because it was

contemplating convicting the defendant on grounds unsupported by the evidence);

Thomas, 806 A.2d at 629 (same). The court’s joint constructive possession

instruction gave the jury that different theory, and, because there was no evidence to

support it, left the jury nowhere to look but Mr. Shelton’s reasonable-doubt argument

that Mr. Smith might have hidden the gun in Mr. Shelton’s jacket. Thus, a defense

argument meant to be exculpatory as to a sole constructive possession theory became

inculpatory as to an unfounded theory of joint constructive possession, especially

because the court never told the jury that it had to find that Mr. Shelton intended to

share possession of the gun with Mr. Smith in order to convict Mr. Shelton of joint

constructive possession. 8 The fact that the jury reached a verdict so quickly after it

was erroneously informed that it could consider a joint constructive possession

theory further supports an inference that this instruction was the basis for its

determination that Mr. Shelton was guilty of all the charged gun-possession offenses.

Evans v. United States, 304 A.3d 211, 231-232 (D.C. 2023) (where the “jury pretty

clearly did not buy [the government’s] account,” and the jury reached a verdict

“within forty-five minutes of receiving the court’s [erroneous] response” to a

question posed in a note, the court “ha[d] no fair assurance that the [trial court’s]

error was harmless”).

The majority contests that the trial court’s joint constructive possession instruction “turn[ed] an ‘exculpatory’ defense (that Mr. Smith was the sole possessor of the gun) into an ‘inculpatory’ one,” but only by effectively restating its positions that sole and joint constructive possession are one and the same and the court’s instruction was not erroneous. Ante at 28 n.8. These, however, are merits claims and do not rebut my point that, if sole and joint constructive possession are different theories requiring different formulations of intent, then the court’s unfounded and deficient joint constructive possession instruction cannot be deemed harmless.

As discussed below, the majority opinion fails to perceive any error in the trial

court’s response to the jury’s note, see infra at Part III, and thus conducts no formal

harmlessness analysis. But to the extent the majority argues that the jury’s decision

to acquit Mr. Shelton of both PWID counts shows that the jury was unaffected by

the court’s joint constructive possession instruction, ante at 27-28 n.7, the majority

misassigns and misunderstands the burden of proof under Kotteakos. Mr. Shelton

does not have the burden to prove that he was in fact harmed by the court’s erroneous

instruction; rather, as explained above, it is the government’s burden to establish that

it is “highly probable” that Mr. Shelton was not so harmed. Where the jury had other

reasons to question whether the government had carried its burden of proof as to the

PWID charges, see supra at note 7, the fact that the jury acquitted Mr. Shelton of

those charges despite the erroneous joint constructive possession charge does not

“eliminate[] [our] doubt about whether [the court’s instructional] error influenced

the jury’s decision” as to Mr. Shelton’s gun possession charge. Hinton, 979 A.2d at

691.

Further, to the extent the majority opinion suggests that any instruction about

joint constructive possession was harmless because the court told the jury that Mr.

Shelton’s guilt of joint constructive possession should be considered

“notwithstanding the role of the front seat passenger,” ante at 26-27, it is likewise

mistaken. In conjunction with an instruction that the jury could consider the alternate

theory of joint constructive possession, this was at best confusing; and, absent a clear

directive to the jury that it had to find that Mr. Shelton intended to possess the gun

with Mr. Smith to convict him based on this alternate theory, this segment of the

court’s response to the jury should not assure us that the court’s error was harmless.

See ante at 10-11 (citing Parker v. United States, 333 A.3d 1162, 1189 (D.C. 2025)

(“When the jury explains specific difficulties, the trial court should clear them away

with concrete accuracy. When a jury sends a note which demonstrates that it is

confused, the trial court must not allow that confusion to persist; it must respond

appropriately.”) (citation modified) and (Anthony) Smith v. United States, 306 A.3d

67, 73 (D.C. 2023) (assessing whether “the instruction adequately addressed the

jury’s apparent confusion and adequately informed the jury of the relevant legal

principles”)).

III. Response to the Majority Opinion’s

Conception of Constructive Possession

The majority opinion sidesteps the question briefed by the parties, namely

whether the trial court had a factual basis to instruct the jury that it could consider

the theory of joint constructive possession alongside that of sole constructive

possession. See Shelton Br. Issue Presented (arguing there was “no evidence” to

support a joint constructive possession instruction); USAO Br. Issues Presented

(arguing a joint constructive possession instruction was proper because “the

evidence sufficiently established” it—or, alternatively, any error was harmless). The

majority declares that no such assessment is necessary, ante at 25 n.6, because joint

constructive possession is not a distinct legal theory of guilt at all—it is nothing more

than a descriptor indicating that constructive possession need not be exclusive to one

person. Ante at Part II.B. Accordingly, the majority asserts that constructive

possession is proved the same way regardless of how many individuals possibly

possess the contraband, i.e., that evidence of a defendant’s sole constructive

possession provides adequate support for a theory of joint constructive possession.

Ante at Part II.B. On this basis, the majority concludes that the trial court did not err.

Ante at Part II.C.

The majority opinion’s erasure of the distinction between the theories of sole

and joint constructive possession is not only conclusory; it is also both

groundbreaking and irreconcilable with this court’s case law.

The majority opinion never explains how joint dominion and control can

meaningfully function within a unitary conception of constructive possession. See

supra Part I (explaining that it cannot). The majority opinion’s citations to bare

statements that constructive possession can be sole or joint, ante at 13—a

proposition I endorse—do not fill this gap or support the proposition that these

theories are one and the same. Nor does the smattering of District of Columbia cases

cited by the majority to illustrate this proposition, ante at 19-20, support this thesis.

These cases say nothing about whether the jury was properly instructed about joint

constructive possession as well as sole constructive possession, the issue presented

here. Instead, they merely show that, in the specific factual scenarios presented, the

evidence (e.g., the DNA on the contraband or its proximity to defendant’s papers or

identification) was sufficient to support the defendant’s guilt of sole constructive

possession—and if the evidence is sufficient as to one theory, there is no need to

parse out and assess sufficiency as to the other. 9

See Johnson v. United States, 290 A.3d 500 (D.C. 2023) (evidence sufficient to show defendant constructively possessed the gun where it was in his bedroom, “in plain view” and “within close reach” next to him while he slept, and there was Instagram footage of him “waving” a gun); Dorsey v. United States, 154 A.3d 106 (D.C. 2017) (evidence sufficient to show defendant constructively possessed the gun with DNA on it that strongly aligned with his own was found in kitchen cabinet immediately after defendant walked out of that room where the other occupants, females and children, were elsewhere in the apartment); Smith, 55 A.3d at 884 (evidence sufficient to show defendant constructively possessed the gun found in a child’s backpack conspicuously located in her bedroom where she had been the sole occupant in the week prior to the search); Ramirez v. United States, 49 A.3d 1246 (D.C. 2012) (evidence sufficient to show defendant constructively possessed the drugs in shared apartment where he was seemingly washing drugs down the kitchen sink as the police arrived and other drugs were found in closet next to where he slept); Carter v. United States, 957 A.2d 9, 16 (D.C. 2008) (evidence was sufficient to show defendant constructively possessed gun in plain view on the left side of the driver seat even though there was another person in the car, because defendant was the driver of the car and this court has held that “an inference of constructive possession may be drawn against the driver of a vehicle in which contraband was found in plain view and in close proximity to the driver’s seat”).

The problem with the majority opinion is not simply that it lacks support for

its analysis, but also that it conflicts with this court’s case law. Our precedent makes

clear that the requisite intent for constructive possession is not always the intent to

exercise exclusive dominion or control over contraband as needed for sole

constructive possession and may instead be the intent to share dominion or control

over contraband as needed for joint constructive possession. See supra Part I.B.

(discussing Rivas, R.G., Moore, and Mitchell). Further, our precedent makes clear

that, in order to instruct the jury on one theory of constructive possession or the

other, the government must present evidence to support that distinct theory. See

supra Part I.B. (discussing Thomas).

With respect to Rivas, the majority opinion acknowledges that our court has

held that, “in shared-space cases or cases involving multiple individuals in proximity

to contraband, ‘there must be something more in the totality of the circumstances

that—together with proximity and knowledge—establishes that the accused meant

to exercise dominion or control over the [contraband].’” Ante at 17 (quoting Rivas,

783 A.2d at 130). But the majority overlooks the fact that Rivas demonstrates that

the “something more” catalog of circumstantial evidence must then be used to

answer the specific question posed by the theory at issue. See supra Part I.B. Instead,

the majority opinion largely focuses on attacking a strawman—namely that proof of

a “joint criminal venture” is always required to satisfy the intent element for joint

constructive possession. Ante at 16, 18, 19 (asserting that joint constructive

possession is not a theory of conspiratorial or vicarious liability). And diverted by

this uncontroverted point,10 the majority never explains why no proof of intent to

work “with others” is needed to prove joint constructive possession, Rivas, 783 A.2d

at 128, or what “more” in the way of circumstantial evidence was presented in this

case to support an inference of Mr. Shelton’s intent to jointly possess the gun with

Mr. Smith (as distinct from the circumstantial evidence supporting an inference of

his intent to solely possess the gun: its presence, not loose in the car, but in the pocket

of his jacket at his side in the back seat, not within easy reach of either front seat

passenger).

Similarly, the majority opinion’s assertion that Rivas has no application to this

case because the theory of “joint constructive possession was not applied to extend

As noted above, I do not controvert it, see supra at note 5, and the majority misconstrues my argument when it states that this dissent “assert[s] that two people cannot exercise dominion and control over an item unless they do so under some sort of agreement.” Ante at 23. I have said nothing about two people reaching an “agreement.” (Nor have I said anything about two people “acting together,” ante at 23.) I have solely discussed the need in the context of joint constructive possession to show that an individual defendant has an ability to exercise dominion and control with one or more other people and an intent to share dominion and control with those other people. As for Mr. Shelton, the core of his argument is that the record was “devoid of facts from which a jury could infer that [he] had assisted Mr. Smith . . . with the intent to jointly exercise control over” the gun—an argument about the absence of evidence of an intent to share dominion and control, which could be proved by a concert of illegal action but need not necessarily be proved that way.

liability from Mr. Smith to Mr. Shelton,” ante at 22, ignores the fact that Rivas

recognizes a principle—incompatible with the majority’s holding—that the intent

element of constructive possession varies depending on whether that possession is

sole or joint. 11

The majority opinion’s efforts to distinguish and limit Thomas are no more

persuasive than its analysis of Rivas. First, the majority argues that Thomas is

factually inapposite because in that case “there was nothing else linking Mr. Thomas

with the gun and the jury’s question and the trial court’s response suggested the

possibility that Mr. Thomas was convicted because the jury thought someone threw

the gun down and Mr. Thomas could be guilty regardless of who it was.” Ante at 32.

But the majority opinion misses the point that the court in Thomas held that the jury

could not be instructed about joint constructive possession because there was no

In any event, the majority has no way of knowing that the jury did not apply “joint constructive possession . . . to extend liability from Mr. Smith to Mr. Shelton,” ante at 22; nor can it even say that it is “highly probable” that the jury did not rely on the court’s erroneous joint constructive possession instruction, given that the court’s “notwithstanding” clause did not require the jury to find that Mr. Shelton intended to share dominion and control with Mr. Smith, and so did not eliminate the risk that the jury simply impermissibly inferred Mr. Shelton’s guilt from Mr. Smith’s actions. See supra at Part II.B. And, for this reason, the majority’s citations to cases recognizing the proposition that the jury is generally deemed to follow a trial court’s instructions, see ante at 22 n.3, 27-28 n.7, are unhelpful because the court’s instructions to the jury were wrong. The court both told the jury that it could consider an unfounded theory and then failed to tell the jury what it needed to find to convict Mr. Shelton under this theory.

evidence supporting that theory of guilt. 806 A.2d at 626. The same is true in Mr.

Shelton’s case. The fact that Mr. Shelton’s identification and mail were inside the

jacket located next to him in the back seat provides no support for an inference that

he had an intent to share control of the gun also found inside the jacket with another

person, which is what joint constructive possession demands. See supra Part I.B.

The majority further argues that “Thomas does not stand (and has never been

cited) for the proposition that to prove joint constructive possession where multiple

people each had the power and intent to control an item, the government must in all

cases also prove concerted action.” Ante at 33. But again, the question is whether the

government presented the “requisite” evidence, Thomas, 806 A.2d at 629, of an

intent to share dominion and control of contraband “with others,” Rivas, 783 A.2d at

128, so as to support a joint constructive possession instruction. Moreover, this

understanding of what Thomas requires does not “run headlong into Rivas,” ante at

32; to the contrary, Thomas aligns with and builds on Rivas’s recognition that joint

constructive possession requires an application of the intent element that is distinct

from the intent element for sole constructive possession.

And this court’s decision in In re R.G., 917 A.2d 643, squarely contradicts the

majority’s assertion in its discussion of Thomas that an unspecified type of

constructive possession would necessarily be proved in a hypothetical scenario

where two individuals who share a bedroom have equal and unfettered access to a

gun found in plain view next to both their personal property. Ante at 33. On

comparable facts (the gun in In re R.G. was not in plain view, but we accepted the

trial court’s finding that the defendant knew about it, id. at 648), this court held that

the evidence was legally insufficient to establish that the defendant had the intent to

possess the gun solely or jointly. 917 A.2d at 648-650 (identifying intent as the

pivotal issue and concluding that the government had not proved beyond a

reasonable doubt that the defendant “rather than her companion, exercised dominion

and control over the pistol, or even that she exercised joint control with him”)

(emphasis added). 12

Instead of taking guidance from Thomas, the majority opinion looks to this

court’s later decision in Carter, 957 A.2d 9, asserting that it “resolves this case.”

In a later hypothetical, the majority posits that one person, Jones, has a stash of drugs and the other, Williams, “partakes” of it without seeking permission and then argues that “it would make little sense to say—and no authority suggests—that Jones cannot be found to have constructively possessed the drugs because Williams also did and the two were not acting together ‘to preclude control by the rest of the world.’” Ante at 23. I agree—not because there is no distinction between sole and joint constructive possession, but because there is. Jones could be prosecuted under a sole constructive possession theory even though Williams was surreptitiously skimming from Jones’s stash because Jones intended to have exclusive possession over the drugs. Jones could not be prosecuted under a joint constructive possession theory, however, because under the majority’s own telling he lacked the requisite intent—he had no intent to share his stash with Williams. Rather than illustrating that these two theories of constructive possession are the same, the hypothetical shows that they are different and thereby undercuts the majority’s analysis.

Ante at 21. But Carter does not purport to modify or limit Thomas. In fact, it never

cites to Thomas because, as noted above, it is simply a sufficiency of the evidence

case. See supra at note 9. Thus, it says nothing about Thomas’s holding that a jointconstructive-possession instruction requires its own evidentiary foundation. 13

In the absence of support from this court’s precedent, the majority opinion

relies heavily on “numerous cases from other jurisdictions” to “support [its]

conception of joint constructive possession.” Ante at 14-16, see also ante at 28-29.

But the District’s common law of constructive possession is its own, and our court’s

prior decisions elucidating and limiting the application of our doctrine is binding on

this division of the court. That said, no monolith of case law exists outside of the

District that holds that sole and joint constructive possession are one and the same

and a jury may be instructed that the constructive possession can be joint, i.e.,

nonexclusive, without regard to whether the government presented evidence

supporting an inference that the defendant had the intent to share the contraband with

others. To the contrary, a number of the federal appellate decisions cited by the

If the joint-constructive-possession question had actually been presented in Carter, the evidence would have—unlike in Mr. Shelton’s case—plainly supported an instruction on that theory. Along with gun possession, Mr. Carter was convicted of assault with intent to commit robbery while armed on an aiding and abetting theory. And the evidence supporting that conviction—his service as the gun supplier and getaway driver, 957 A.2d at 16-19—would have amply supported an inference that he intended to share the gun with the individual who committed the robbery. See Rivas, 783 A.2d at 128, 130.

majority opinion accept that sole and joint constructive possession are distinct

theories; they do the work the majority opinion has not done in this case and assess

whether the evidence presented supported an instruction regarding joint constructive

possession. 14 Others acknowledge that proof of intent to share is required for joint

constructive possession, but say nothing about whether a separate jury instruction

on this theory was warranted because they are sufficiency of the evidence cases. 15 A

couple are simply unilluminating with respect to the issues presented in this case. 16

See United States v. Norris, 21 F.4th 188, 194 (1st Cir. 2021); United States v. Howard, 687 F.3d 13, 19 (1st Cir. 2012); United States v. Carrasco, 257 F.3d 1045, 1050 (9th Cir. 2001). The fact that these decisions conclude that a joint constructive possession instruction was warranted carries little weight given the distinct facts presented. See, e.g., Norris, 21 F.4th at 194 (male and female DNA found on gun, in conjunction with other evidence that the male defendant presented showing female housemate possessed the gun, supported a theory of joint constructive possession).

See United States v. Fairley, 137 F.4th 503, 513 (6th Cir. 2025) (recognizing that an “intention . . . to exercise dominion and control . . . through others” is a means of proving constructive possession); United States v. Benjamin, 711 F.3d 371, 379 (3d Cir. 2013) (same); United States v. Garth, 188 F.3d 99, 113 (3d Cir. 1999) (same); see also Hopson v. Horton, 838 F. App’x 147, 154 (6th Cir. 2020) (observing that joint constructive possession is proved under Michigan law if two or more defendants are “acting in concert” (citation modified)).

See United States v. Wright, 739 F.3d 1160 (8th Cir. 2014) (examining only whether evidence of defendant’s knowledge of the contraband was sufficient to support defendant’s conviction on a constructive possession theory); United States v. Duval, 496 F.3d 64 (1st Cir. 2007) (same); United States v. Perlaza, 818 F.2d 1354, 1361-62, 1361 n.3, 1361 n.4 (7th Cir. 1987) (upholding instruction informing jury that it had to make an individual determination that a defendant jointly possessed the contraband and could not find the defendant guilty solely on the basis of the codefendant’s possession).

And the remaining handful, mostly from the Seventh Circuit, provide no persuasive

support for our court to overhaul our law of constructive possession and follow suit

(which we could do only sitting en banc) because they provide little in the way of

analysis and express no concern as we have done about exceeding our judicial role

and casting the net of the doctrine of constructive possession too widely. 17

Apart from the majority’s failure to explain how joint dominion and control

can meaningfully function within a unitary conception of constructive possession,

its lack of support in this court’s case law, and its reliance on mostly inapposite and

nonbinding federal cases, the biggest flaw in the majority opinion is its failure to

contain the doctrine of constructive possession as this court has long committed to

See United States v. Ford, 22 F.4th 687, 693 (7th Cir. 2022) (upholding sentence enhancement based on constructive possession of weapon without discussing the elements of that doctrine); United States v. Driggers, 913 F.3d 655, 657-58 (7th Cir. 2019) (upholding joint constructive possession instruction without discussing the elements of that doctrine); United States v. Rainone, 816 F.3d 490, 494 (7th Cir. 2016) (same); see also Ortiz v. Dubois, 19 F.3d 708, 713 (1st Cir. 1994) (concluding in habeas case that felony-murder conviction based on felony of joint unlawful possession of a firearm in a motor vehicle, which “does not require proof of concerted action or shared state of mind,” was erroneous because the government consequently had not “prove[n] beyond a reasonable doubt that appellant intentionally assisted . . . in the commission of the felony,” as required by Massachusetts law in felony-murder cases); United States v. Davis, 562 F.2d 681, 684 (D.C. Cir. 1977) (concluding evidence of drug possession was sufficient, employing a test for constructive possession that lacked an express intent element); see also id. at 699 (Bazelon, J., dissenting) (critiquing the majority for its approach because it failed to consider intent and “confuse[d] possible toleration with active participation”).

doing. As noted above, over the years we have imposed structure on the doctrine,

clearly defining its elements as well as key terms like “dominion and control,” all

with the aim of protecting individuals we have no reason to believe the legislature

wished to punish from being convicted of possessory crimes. But the majority sets

course in the opposite direction and breaks down the guardrails we have erected to

protect against guilt by proximity and association. Under the majority’s analysis, a

jury that is unable to convict based on a sole constructive possession theory may

nonetheless convict a defendant of joint (only in the sense of nonexclusive)

constructive possession simply because there is another person present—even if the

government has not presented any evidence to support an inference that the

defendant intended to possess contraband with that other person. This is precisely

the wider net we have long warned against, and it invites exactly the outcome we

have said is not allowed, namely, punishing a defendant for the conduct of others,

rather than for his own. Moreover, it pushes the boundaries of our role as judges.

The observations by our colleague, Judge Ruiz, in her concurrence in Rivas, bear

repeating:

If we permit too low an evidentiary threshold for

constructive possession, we w[ill] in essence be setting a

heightened standard of behavior, on pain of criminal

sanction, for individuals faced with close proximity to

contraband in plain view. If knowing proximity to

contraband is to be a crime, it is for the legislature to

declare it. What we must do is ensure that in setting the

evidentiary minimum to establish the judge-made

elements of constructive possession as a proxy for actual

possession, the jury does not stray from the offense of drug

possession that the legislature has criminalized, as

possession is normally understood.

783 A.2d at 145.

* * *

For these reasons, I cannot endorse either the majority opinion’s creation of a

new rule treating sole and joint constructive possession as one and the same or its

application of this new rule to Mr. Shelton’s case. Accordingly, I respectfully dissent.

Shelton v. United States

Authorities cited

Opinion · public domain