UNITED STATES v. DAVID D. LEWIS

District of Columbia
Court of Appeals
No. 13-CO-1456
SEP 29 2016
UNITED STATES,
Appellant,

v. CF2-10190-13

DAVID D. LEWIS,
Appellee.

On Appeal from the Superior Court of the District of Columbia
Criminal Division

BEFORE: WASHINGTON, Chief Judge; and GLICKMAN, FISHER, BLACKBURNE-
RIGSBY, THOMPSON, BECKWITH, EASTERLY and MCLEESE, Associate Judges.

OPINION FOR THE COURT BY MCLEESE, Associate Judge; joined by
GLICKMAN, FISHER, BLACKBURNE-RIGSBY and THOMPSON, Associate Judges.

DISSENTING OPINION BY BECKWITH, Associate Judge; joined by
WASHINGTON, Chief Judge; and EASTERLY, Associate Judge.

JUDGMENT

This case came to be heard on the transcript of record and the briefs filed,
and was argued by counsel. On consideration whereof, and as set forth in the opinion
filed this date, it is now hereby

ORDERED and ADJUDGED that the matter on appeal is reversed and
remanded for further proceedings.

For the Court:

Dated: September 29, 2016.

Opinion by Associate Judge Roy W. McLeese.

Dissenting opinion by Associate Judge Corinne Beckwith.
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. 13-CO-1456
9/29/16
UNITED STATES, APPELLANT,

V.

DAVID D. LEWIS, APPELLEE.

Appeal from the Superior Court
of the District of Columbia
(CF2-10190-13)

(Hon. Robert I. Richter, Trial Judge)

(Argued En Banc June 30, 2015 Decided September 29, 2016)

David B. Goodhand, Assistant United States Attorney, with whom Ronald
C. Machen Jr., United States Attorney at the time the brief was filed, and Elizabeth
Trosman and Christian Natiello, Assistant United States Attorneys, were on the
brief, for appellant.

Joshua Deahl, Public Defender Service, with whom James Klein and Samia
Fam, Public Defender Service, were on the brief, for appellee.

Before WASHINGTON, Chief Judge, and GLICKMAN, FISHER, BLACKBURNE-
RIGSBY, THOMPSON, BECKWITH, EASTERLY, and MCLEESE, Associate Judges.

Opinion for the court by Associate Judge MCLEESE, joined by GLICKMAN,
FISHER, BLACKBURNE-RIGSBY, and THOMPSON, Associate Judges.

Dissenting opinion by Associate Judge BECKWITH, joined by WASHINGTON,
Chief Judge, and EASTERLY, Associate Judge, at page 38.
 2

MCLEESE, Associate Judge: The United States seeks review of the trial

court‘s pretrial order suppressing evidence as fruit of an unlawful search under the

Fourth Amendment. D.C. Code § 23-104 (a)(1) (2012 Repl.). We reverse and

remand for further proceedings.

I.

The evidence presented at the suppression hearing indicated the following.

At approximately 2 a.m. on June 16, 2013, United States Park Police Officer

Michael Alto was driving on Ingraham Street NW when he saw appellee David

Lewis driving a car with a non-functioning headlight. While following the car,

Officer Alto determined that the car was registered to Mr. Lewis and that Mr.

Lewis had a suspended license. Mr. Lewis drove a couple of blocks, pulled over

into a parking spot, and started to get out of the car. Officer Alto pulled up and

asked Mr. Lewis for his license and registration. Mr. Lewis opened the car door,

enabling Officer Alto to see an open bottle of Patrón tequila in the center-console

cup holder. The bottle was half full. Mr. Lewis‘s passenger, Brittney Gibbs, said

that the bottle was hers. Officer Alto told Ms. Gibbs to bring him the bottle, which

Ms. Gibbs did by walking around the car and handing the bottle to Officer Alto.
 3

After confirming that Mr. Lewis‘s license had been suspended, Officer Alto

put Mr. Lewis in handcuffs. Officer Alto did not smell any alcohol coming from

Mr. Lewis or Ms. Gibbs. A second officer, Officer Brown, arrived on the scene,

and Officer Alto asked her to search the car for additional open containers of

alcohol. In Officer Alto‘s experience, ―the majority of times when there is a

tequila or liquor type of beverage in a vehicle, they‘ll be drinking through cups.‖

Similarly, in Officer Brown‘s experience, people very rarely drink directly out of

Patrón bottles and instead usually use cups. The officers decided to search the car,

both for additional evidence of the offense of possession of an open container of

alcohol (POCA) and because it was possible that Ms. Gibbs could have been

permitted to drive the car away afterward, and the officers therefore wanted to

make sure that there was no additional alcohol or other contraband in the vehicle.

At the time Officer Brown arrived, Ms. Gibbs was outside the vehicle,

saying that she needed to go to the bathroom. Officer Brown told Ms. Gibbs that

she could not leave. Officer Brown opened the driver-side door and smelled

marijuana. Officer Brown found a cup containing liquid that smelled like alcohol

on the floor of the front passenger seat, and Ms. Gibbs said that the cup was hers.
 4

Officer Brown also found a loaded handgun and a box of ammunition in a bag on

the back seat. Finally, Officer Brown found a cigarette containing a green plant-

like substance in the passenger-side door.

According to Officer Brown, Ms. Gibbs was not under arrest at the time the

search of the car began. Once Officer Brown found the gun, she placed Ms. Gibbs

in handcuffs. A subsequent search revealed a bag of marijuana in Ms. Gibbs‘s bra.

Ms. Gibbs was arrested for POCA and possession of marijuana. Mr. Lewis was

arrested for carrying a pistol without a license, possession of an unregistered

firearm, possession of unregistered ammunition, and operating a vehicle with a

suspended license.

The trial court granted Mr. Lewis‘s motion to suppress evidence of the gun,

the ammunition, and the marijuana, concluding that the officers did not have

reasonable, articulable suspicion to search the vehicle for evidence of POCA. A

division of this court concluded to the contrary that the officers had reasonable,

articulable suspicion that there was evidence of POCA in the car. United States v.
 5

Nash, 100 A.3d 157, 164-65 (D.C. 2014).1 The division further concluded that the

search of the car was lawful as incident to Ms. Gibbs‘s arrest, even though the

officers did not place Ms. Gibbs under arrest until after the search and it was not

clear whether the officers had at the time of the search intended to arrest Ms.

Gibbs. Id. at 165-68. The en banc court granted rehearing limited to the question

whether the search was lawful as incident to Ms. Gibbs‘s arrest. United States v.

Lewis, 107 A.3d 603 (D.C. 2015) (en banc).

II.

When reviewing a trial court‘s denial of a motion to suppress, we ―view the

evidence in the light most favorable to the prevailing party.‖ Bennett v. United

States, 26 A.3d 745, 751 (D.C. 2011) (internal quotation marks omitted). We draw

all reasonable inferences in favor of upholding the trial court‘s ruling. Milline v.

United States, 856 A.2d 616, 618 (D.C. 2004). We review the trial court‘s legal

conclusions de novo. United States v. Taylor, 49 A.3d 818, 819 (D.C. 2012).

1
Before the division, Mr. Lewis‘s case was consolidated with a factually
unrelated case, United States v. Nash, No. 13-CO-1299. United States v. Nash,
100 A.3d at 159-60. The division affirmed the trial court‘s suppression order in
Mr. Nash‘s case. Id. at 162-64. The current proceeding before the en banc court
involves only Mr. Lewis‘s case.
 6

―A search conducted without a warrant is per se unreasonable under the

Fourth Amendment unless it falls within a few specific and well-established

exceptions.‖ Taylor, 49 A.3d at 821 (internal quotation marks omitted). Under

one such exception, police officers may conduct a warrantless search of a vehicle,

incident to an arrest, if they have reasonable, articulable suspicion to believe that

the vehicle contains evidence of the offense of arrest. Arizona v. Gant, 556 U.S.

332, 351 (2009). The issue before this court en banc is whether such a search,

which we will refer to as a Gant evidence search, is lawful if (a) the search

precedes the arrest for the offense at issue; and (b) it is unclear whether the officers

intended to arrest the suspect before conducting the search. Although the issue has

not yet arisen with much frequency, as far as we are aware every court to have

considered the issue has upheld the validity of such searches. See, e.g., State v.

Fizovic, 770 S.E.2d 717, 720-22 (N.C. Ct. App. 2015) (Gant evidence search of

suspect‘s car was lawful even though suspect was not arrested until after search

and officer did not intend to arrest at time of search). The same issue has arisen in

the context of other types of search incident to arrest, and the overwhelming

weight of authority upholds the legality of such searches even when conducted

before an arrest and in the absence of evidence that the officers subjectively

intended to arrest the defendant at the time of the search. See, e.g., State v. J.J.,

143 So. 3d 1050, 1052 (Fla. Dist. Ct. App. 2014) (per curiam) (upholding search as
 7

lawful incident to arrest where search preceded arrest and where officer explained

search as based on officer safety; no indication that officer intended to arrest at

time of search); People v. Nguyen, 854 N.W.2d 223, 232-34 (Mich. Ct. App. 2014)

(police lawfully searched suspect‘s person incident to arrest, even though search

preceded arrest and officers did not believe they had probable cause to arrest at

time of search), appeal denied, 863 N.W.2d 327 (Mich. 2015); State v. Sykes, 695

N.W.2d 277, 282-87 (Wis. 2005); Moffitt v. State, 817 N.E.2d 239, 246 (Ind. Ct.

App. 2004); United States v. Anchondo, 156 F.3d 1043, 1045 (10th Cir. 1998) (―In

order to be a legitimate ‗search incident to arrest,‘ the search need not take place

after the arrest. A warrantless search preceding an arrest is a legitimate ‗search

incident to arrest‘ as long as (1) a legitimate basis for the arrest existed before the

search, and (2) the arrest followed shortly after the search. Whether or not the

officer intended to actually arrest the defendant at the time of the search is

immaterial to this two-part inquiry.‖) (citations omitted). But see People v. Reid,

26 N.E.3d 237, 239-40 (N.Y. 2014) (search of suspect‘s person was not lawful

incident to arrest because, at time of search, suspect had not been arrested and

officer had no intent to arrest suspect).

We hold that a Gant evidence search is lawful if (a) the police have probable

cause to arrest the suspect for an offense; (b) the suspect recently occupied a
 8

vehicle; (c) the police have reasonable, articulable suspicion to believe that the

vehicle contains evidence of the offense; (d) at the time of the search, the police

have not released the suspect or issued the suspect a citation for the offense; and

(e) the suspect‘s formal arrest for the offense follows quickly on the heels of the

search.2

A.

We turn first to whether a Gant evidence search may precede the arrest of

the suspect. The Supreme Court did not decide that question in Gant. In Rawlings

v. Kentucky, 448 U.S. 98 (1980), however, the Supreme Court addressed the same

question in connection with the search of a defendant‘s person incident to arrest,

stating that ―[w]here the formal arrest follow[s] quickly on the heels of the

challenged search of [a suspect‘s] person, we do not believe it particularly

2
By upholding the legality of Gant evidence searches in the circumstances
stated in text, we do not mean to imply that all other Gant evidence searches would
necessarily be unlawful. For example, we express no view on the question whether
an otherwise lawful Gant evidence search would be rendered unlawful if the
suspect fled from the police after the search and before the police placed the
suspect under arrest. On a separate point, the dissent asserts that we decide the
case on a basis ―which neither party meaningfully briefed.‖ To the contrary, the
parties extensively briefed the issues we decide.
 9

important that the search preceded the arrest rather than vice versa.‖ 448 U.S. at

111. Although Rawlings refers to both ―formal arrest‖ and ―arrest,‖ id., the lower

courts, including this court, have consistently understood the rule announced in

Rawlings to apply without regard to any distinction between formal arrest and

arrest. See, e.g., Millet v. United States, 977 A.2d 932, 935 (D.C. 2009) (―A search

incident to arrest may precede the actual arrest if probable cause exists,

independent of the search, to justify the arrest, and if the arrest follows quickly on

the heels of the search.‖) (internal quotation marks omitted); United States v.

Powell, 376 U.S. App. D.C. 30, 32-33, 483 F.3d 836, 838-39 (2007) (en banc)

(―where police had probable cause to arrest before search, it was of no import that

the search came before the actual arrest‖) (internal quotation marks omitted; citing

cases).

The Supreme Court in Rawlings did not explain its reasons for holding that a

search incident to arrest may permissibly precede arrest, but the lower courts have

identified at least three. First, permitting search to precede arrest will in some

cases benefit innocent suspects, because if the results of the search negate probable

cause or persuade the officer not to arrest, the suspect will be spared the greater

intrusion and collateral consequences of an arrest. See, e.g., State v. Overby, 590

N.W.2d 703, 706 (N.D. 1999) (―[I]f the person searched is innocent and the search
 10

convinces the officer that his reasonable belief to the contrary is erroneous, it is to

the advantage of the person searched not to be arrested.‖) (internal quotation marks

omitted). Second, in cases in which arrest is inevitable, whether search precedes or

follows arrest does not affect the degree of the intrusion on the suspect. Id. Third,

courts are reluctant to micromanage the precise order in which officers who have

probable cause to arrest conduct searches and arrests, particularly given the safety

and other tactical considerations that can be involved. See, e.g., Anderson v. State,

553 A.2d 1296, 1301-02 (Md. Ct. Spec. App. 1989).

Lower courts, including this court, have applied Rawlings broadly,

upholding searches that preceded arrest in cases involving Gant evidence searches,

searches to protect officer safety or prevent destruction of evidence, searches of a

suspect‘s person, and searches of a vehicle. See, e.g., Powell, 376 U.S. App. D.C.

at 33, 483 F.3d at 839 (upholding search of car incident to arrest where suspect had

not been arrested at time of search; ―Indeed, every circuit that has considered the

question-save one-has concluded that a search incident to arrest may precede the

arrest.‖) (citing cases); State v. Smith, 266 P.3d 1220, 1224 (Idaho Ct. App. 2011)

(upholding Gant evidence search of car where suspect had not been arrested at

time of search); Adams v. State, 815 So. 2d 578, 582 (Ala. 2001) (upholding search

of person incident to arrest where suspect had not been arrested at time of search
 11

but probable cause existed and arrest was ―sufficiently contemporaneous‖);

Minnick v. United States, 607 A.2d 519, 525 n.11 (D.C. 1992) (search of purse

taken from car lawful as search incident to arrest even though search preceded

arrest; citing Rawlings); cf., e.g., Waters v. United States, 311 A.2d 835, 836 (D.C.

1973) (―It is well settled, however, that a search or seizure may precede an arrest,

provided the officer at that point already has probable cause to arrest the possessor

of the items in question.‖).

Mr. Lewis does not appear to dispute that a Gant evidence search can be

lawful even if the search is conducted before arrest, as long as an arrest is under

way. Some of Mr. Lewis‘s arguments, however, seem to imply more broadly that

a completed arrest must precede a Gant evidence search. In any event, we hold

that, as the Supreme Court concluded in Rawlings, it is sufficient that the arrest

follows quickly after the search.

Most broadly, Mr. Lewis relies on the Supreme Court‘s statements that ―[i]t

is the fact of the lawful arrest that establishes the authority to search‖ incident to

arrest. E.g., United States v. Robinson, 414 U.S. 218, 235 (1973). For several

reasons, we do not understand this language to establish a rule that the arrest must

precede a search incident to arrest. First, in the cases relied upon by Mr. Lewis the
 12

Supreme Court was not actually addressing the question whether a lawful search

incident to arrest must follow the arrest. See, e.g., Robinson, 414 U.S. at 234-35

(holding that police may lawfully conduct full search of arrestee‘s person incident

to arrest, even in absence of case-specific basis for search). Both the Supreme

Court and this court have cautioned against treating such language as a dispositive

holding. See, e.g., United States v. Debruhl, 38 A.3d 293, 298 (D.C. 2012) (―We

have stressed, however, that stare decisis is never properly invoked unless in the

decision put forward as precedent the judicial mind has been applied to and passed

upon the precise question.‖) (internal quotation marks omitted); Porter v. United

States, 37 A.3d 251, 265 n.20 (D.C. 2012) (―Questions which merely lurk in the

record, neither brought to the attention of the court nor ruled upon, are not to be

considered as having been so decided as to constitute precedents.‖) (quoting

Webster v. Fall, 266 U.S. 507, 511 (1925)). Second, as previously noted, Mr.

Lewis himself appears to concede that, despite the language upon which he relies,

a completed arrest need not precede a search incident to arrest, as long as an arrest

is under way. Third, interpreting that language to require an arrest to precede the

search would contradict Rawlings.3

3
For similar reasons, we are unpersuaded by Mr. Lewis‘s reliance on language
in Justice Scalia‘s concurring opinion in Thornton v. United States, 541 U.S. 615
(2004), stating that permitting searches for evidence incident to arrest is justified
because the ―fact of prior lawful arrest‖ distinguishes arrestees from society at
(continued…)
 13

On the last point, Mr. Lewis argues that the Supreme Court‘s statement in

Rawlings is dictum. It is true, as Mr. Lewis points out, that the defendant in

Rawlings did not argue in the Supreme Court that a lawful search incident to arrest

must follow arrest. Brief and Reply Brief for Petitioner, Rawlings v. Kentucky, 448

U.S. 98 (1980) (No. 79-5146), 1980 WL 339599, 339603. Nevertheless, the

Supreme Court‘s statement in Rawlings is now deeply entrenched in the law. It

has been cited, and treated as a holding, in many lower-court decisions, including

several decisions of this court. See, e.g., Minnick, 607 A.2d at 525 n.11; United

States v. Montgomery, 377 F.3d 582, 586 (6th Cir. 2004) (―[A]s the Supreme Court

held in [Rawlings], the search-incident-to-a-lawful-arrest rule also permits an

officer to conduct a full search of an arrestee‘s person before he is placed under

lawful custodial arrest as long as the formal arrest follows quickly on the heels of

the challenged search of his person and the fruits of that search are not necessary to

sustain probable cause to arrest him.‖) (brackets, ellipses, and internal quotation

(…continued)
large and distinguishes searches for evidence of the crime of arrest from general
rummaging. 541 U.S. at 630. Moreover, although the opinion for the Court in
Gant did adopt the general approach advocated by Justice Scalia in his concurring
opinion in Thornton, see Gant, 556 U.S. at 343-44, the opinion for the Court in
Gant did not use the word ―prior‖ in stating its holding. Id. On that point, we take
our guidance from the opinion for the Court in Gant rather than the concurrence in
Thornton.
 14

marks omitted). We have located no case in which a court has treated the

statement as dictum. Moreover, well before the decision in Rawlings, the courts in

this jurisdiction had adopted the principle that ―a search or seizure may precede an

arrest, provided the officer at that point already has probable cause to arrest the

possessor of the items in question.‖ Waters, 311 A.2d at 836 (citing Bailey v.

United States, 128 U.S. App. D.C. 354, 357, 389 F.2d 305, 307 (1967)). Taken

together, these considerations preclude this court from disregarding the principle

announced by the Supreme Court in Rawlings. Cf., e.g., State v. Rose, 19 A.3d

985, 1012 n.21 (N.J. 2011) (―Without exploring the intricate distinctions between

dictum and language necessary to decision, we conclude that we must recognize

the clear, direct, explicit, and unqualified statement of the Supreme Court.‖)

(quoting Public Serv. Co. v. General Elec. Co., 315 F.2d 306, 310 n.6 (10th Cir.

1963)); Majette v. New London Hous. Auth., No. X094CV0450000090S, 2005 WL

3112738, *5 (Conn. Super. Ct. Nov. 3, 2005) (although plaintiff argued that

language from decision of Connecticut Supreme Court was dictum, ―the ‗dictum‘

has become more or less entrenched as black letter law‖); Tony Andreski, Inc. v.

Ski Brule, Inc., 475 N.W.2d 469, 473 (Mich. Ct. App. 1991) (Griffin, J.,

concurring) (―While one could argue that the statements in [an earlier decision of

the Michigan Supreme Court] were mere dicta, the doctrine is too firmly

entrenched to be overruled by this intermediate appellate court.‖).
 15

Somewhat more narrowly, Mr. Lewis argues that the Supreme Court‘s

statement in Rawlings must be understood in context. Specifically, Mr. Lewis

argues that an arrest was under way at the time of the search in Rawlings, and that

Rawlings thus should not be read to authorize searches incident to arrest unless an

arrest is at least under way. We conclude otherwise. Although it is not entirely

clear what Mr. Lewis means by the term ―under way,‖ Mr. Lewis appeared to take

the position at oral argument that an arrest is under way as long as, at the time of

the search, the officers intend to transport the suspect to the police station for the

purpose of charging the suspect with a crime, even if the officers have not

communicated their intent in any way. On that view, the question whether an

arrest was under way collapses into the question whether the officers subjectively

intended to place the suspect under arrest. For reasons that we explain infra,

however, we conclude that such an inquiry into officers‘ subjective intent is

foreclosed by controlling decisions of the Supreme Court.

Mr. Lewis also suggests in a footnote that the temporal flexibility provided

by Rawlings should not be extended to Gant evidence searches. We disagree. As

we have already noted, every court of which we are aware to have addressed the

question has applied Rawlings to Gant evidence searches. See, e.g., Smith, 266
 16

P.3d at 1224. Moreover, each of the justifications for the Rawlings rule applies to

Gant evidence searches: permitting such searches to precede arrest in some

instances will benefit suspects, because the results of the search may lead to release

rather than the arrest that otherwise would have occurred; as to suspects who are

going to be both searched and arrested, the order of those events does not affect the

degree of the intrusion; and tactical considerations may lead officers to reasonably

prefer to conduct a Gant evidence search before completing an arrest.

In sum, we hold that a Gant evidence search can be lawful even if the search

precedes arrest.

B.

We turn next to the question whether the search in this case was lawful even

though it is unclear whether, at the time of the search, the officers intended to

arrest Ms. Gibbs. Based on controlling Supreme Court authority, we conclude that

the search was lawful.

The Supreme Court‘s ―Fourth Amendment cases have repeatedly rejected a

subjective approach. Indeed, [the Supreme Court has] never held, outside limited
 17

contexts such as an inventory search or administrative inspection, that an officer‘s

motive invalidates objectively justifiable behavior under the Fourth Amendment.‖

Fernandez v. California, 134 S. Ct. 1126, 1134 (2014) (citation, ellipses, and

internal quotation marks omitted); see also, e.g., Ashcroft v. Al-Kidd, 131 S. Ct.

2074, 2080-81, 2083 (2011) (Supreme Court has ―rejected every request to

examine subjective intent‖ in Fourth Amendment setting, with exception of (a)

special-needs searches, (b) administrative searches, and (c) searches conducted in

absence of individualized suspicion); Brigham City v. Stuart, 547 U.S. 398, 404

(2006) (―An action is reasonable under the Fourth Amendment, regardless of the

individual officer‘s state of mind, as long as the circumstances viewed objectively,

justify the action. . . . The officer‘s subjective motivation is irrelevant.‖)

(brackets, emphasis, and internal quotation marks omitted); Devenpeck v. Alford,

543 U.S. 146, 154 (2004) (―As we have repeatedly explained, the fact that the

officer does not have the state of mind which is hypothecated by the reasons which

provide the legal justification for the officer‘s action does not invalidate the action

taken as long as the circumstances, viewed objectively, justify that action. The

Fourth Amendment‘s concern with ‗reasonableness‘ allows certain actions to be

taken in certain circumstances, whatever the subjective intent. Evenhanded law

enforcement is best achieved by the application of objective standards of conduct,
 18

rather than standards that depend upon the subjective state of mind of the officer.‖)

(brackets, citations, and internal quotation marks omitted).

Gant evidence searches rest on particularized suspicion, because they

require not only probable cause to arrest but also reasonable, articulable suspicion

to believe that the vehicle to be searched contains evidence of the offense of arrest.

Gant, 556 U.S. at 351. Under controlling Supreme Court law, the legality of such

searches thus must be determined based on the objective circumstances, not on

whether at the time of the search the officers subjectively intended to arrest the

suspect. Cf. Peters v. New York, 392 U.S. 40, 66-68 (1968) (upholding search as

lawful search incident to arrest; no direct evidence that officers subjectively

intended to arrest suspect at time of search); id. at 68-69 (Douglas, J., concurring);

id. at 70 (Fortas, J., concurring); id. at 79 (Black J., concurring).

We are not persuaded by Mr. Lewis‘s arguments to the contrary. First, Mr.

Lewis argues that Gant evidence searches necessarily require inquiry into officers‘

subjective intent, because the legality of such searches depends on the offense of

arrest, which in turn depends on the subjective decision of a particular officer. The

Supreme Court did not explain in Gant how courts should determine the offense of

arrest for purposes of assessing the legality of a Gant evidence search. We need
 19

not delve into that issue in the present case, because Mr. Lewis does not dispute

that there was a prompt arrest for POCA. We do note, however, that it is not clear

that the inquiry must be subjective in character. Cf., e.g., Devenpeck, 543 U.S. at

154 (in determining whether there was probable cause to support arrest, courts are

not limited to subjective grounds stated by officer). We thus do not understand

Gant to have implicitly required a subjective inquiry into whether the officers

intend to arrest at the time they conduct a Gant evidence search.

Second, contrary to Mr. Lewis‘s contention, Knowles v. Iowa, 525 U.S. 113

(1998), is consistent with an objective approach. In Knowles, a police officer

stopped Mr. Knowles for speeding. 525 U.S. at 114. The officer issued Mr.

Knowles a citation even though the officer could have arrested Mr. Knowles under

Iowa law. Id. The officer then searched Mr. Knowles‘s car, recovered a bag of

marijuana and a pipe, arrested Mr. Knowles, and charged Mr. Knowles with drug

offenses. Id. The Supreme Court ruled that the officer had conducted an illegal

―search incident to citation‖ rather than a search incident to arrest. Id. at 116-19.

Knowles turned on two objective circumstances. First, at the time of the search at

issue, the suspect had been issued a citation rather than arrested. Second, ―no

further evidence of excessive speed was going to be found either on the person of

the offender or in the offender‘s car.‖ Id. at 118. The Knowles Court never
 20

mentioned the officer‘s subjective intent and in no way suggested that the Court

was adopting a novel exception to its general rule against consideration of

subjective intent in determining the Fourth Amendment reasonableness of searches

based on particularized suspicion.

We recognize that, under an objective approach, officers who have probable

cause to arrest for a minor crime for which they would not ordinarily make an

arrest might be tempted to conduct a Gant evidence search and then decide,

depending on the results of the search, whether in fact to arrest. On the other hand,

under the approach advocated by Mr. Lewis, officers who want to conduct a Gant

evidence search might be tempted to arrest suspects for petty crimes, when they

would not otherwise have done so, in order to permit such a search. It thus is

unclear that the approach advocated by Mr. Lewis would be more protective of

suspects‘ interests. In any event, the Supreme Court has held that comparable

―concerns about improper motives and pretext do not justify subjective inquiries‖

in the context of searches and seizures resting on particularized suspicion. Al-

Kidd, 131 S. Ct. at 2082. For example, the Court held in Al-Kidd that federal

officials had lawfully detained terrorism suspects under the federal material-

witness statute even if those officials did not intend to call the suspects as

witnesses, as long as there was an objective basis for the detention. 131 S. Ct. at
 21

2082-84. And in Brigham City, the Court held that officers may enter a home, the

most protected area under the Fourth Amendment, if they have a reasonable belief

that someone is in danger inside, even if the officers‘ subjective intent is to gather

evidence or arrest suspects rather than to provide aid. 547 U.S. at 405-06.

In sum, we hold that the legality of a Gant evidence search does not depend

on whether the officers intended to arrest the suspect at the time of the search at

issue.

C.

Mr. Lewis makes four additional arguments, which we address in turn.

First, Mr. Lewis relies on the Supreme Court‘s statements that ―an incident search

may not precede an arrest and serve as part of its justification.‖ Sibron v. New

York, 392 U.S. 40, 63 (1968). Under the approach we adopt, however, the search

does not provide any part of the legal justification for the arrest. Rather, the arrest

must be justified by preexisting probable cause.

Second, Mr. Lewis argues that a search cannot be incident to an arrest if the

search is the cause of the arrest. See State v. Funkhouser, 782 A.2d 387, 409 (Md.
 22

Ct. Spec. App. 2001) (decision to arrest cannot be a ―consequence of what was

found in the search‖). The Supreme Court, however, has never suggested such a

principle. Moreover, such a principle would be inconsistent with the objective

approach that the Supreme Court has required in contexts involving particularized

suspicion.

Third, Mr. Lewis argues that, because a search must be justified at its

inception, it is impermissible for subsequent events to affect the lawfulness of a

search. It is true that, under the approach we adopt, the admissibility of evidence

obtained during a Gant evidence search may depend on events that take place after

the search. But that is not unique to the current setting. For example, if officers

executing a search warrant act within the warrant at first but then flagrantly exceed

the scope of the warrant, all of the evidence seized may be subject to suppression.

Cf., e.g., In re 650 Fifth Ave. & Related Props., No. 14-2027, 2016 WL 3913403,

*22 n.32 (2d Cir. July 20, 2016) (―[W]hen items outside the scope of a valid

warrant are seized, the normal remedy is suppression and return of those items, not

invalidation of the entire search, unless it is shown that those executing the warrant

acted in flagrant disregard of the warrant‘s terms.‖) (internal quotation marks

omitted); State v. Rindfleisch, 857 N.W.2d 456, 465 (Wis. Ct. App. 2014) (same).

Thus, evidence obtained during a course of conduct that is lawful at its inception
 23

can become inadmissible based on subsequent events. That is the situation with

Gant evidence searches. They are lawful at their inception if supported by

probable cause to arrest and reasonable, articulable suspicion that evidence of the

offense of arrest is in the vehicle being searched, but the evidence obtained in such

searches may become inadmissible based on subsequent events.

Finally, Mr. Lewis argues that his approach would give clearer guidance to

law enforcement than the approach we adopt. We disagree. Mr. Lewis‘s approach

raises many questions. At the most basic level, it is unclear when an arrest should

be viewed as under way or what it would mean to require that the officers intend to

arrest the suspect. On the latter topic, for example, it is unclear whether under Mr.

Lewis‘s approach a Gant evidence search would be lawful if the police plan to

transport the suspect to the station to then be released on citation. See generally

D.C. Code § 23-584 (b) (2016 Cum. Supp.) (authorizing certain officials to grant

citation release to arrestees after appearance at law-enforcement agency). To take

another example relating to a separate topic, it is unclear how Mr. Lewis‘s

approach would be applied if different officers on the scene had different intents.

In light of these and other similar questions, we conclude that Mr. Lewis‘s

approach would not provide a significantly clearer framework for police than the

objective approach that we adopt.
 24

III.

A.

A central theme of the dissent is that permitting a search incident to arrest to

precede formal arrest would contradict ―the traditional requirement of a completed

formal arrest to justify a search.‖ Post at 72. The Supreme Court, however, has

held that the search-incident-to-arrest exception contains no such requirement.

Rawlings, 448 U.S. at 111 (―Where the formal arrest follow[s] quickly on the heels

of the challenged search of [a suspect‘s] person, we do not believe it particularly

important that the search preceded the arrest rather than vice versa.‖).

Notwithstanding the dissent‘s reservations about Rawlings, we are bound by

Rawlings‘s holding. See generally, e.g., Mercer v. United States, 864 A.2d 110,

114 n.4 (D.C. 2004).

As part of its implicit critique of Rawlings, the dissent states that the

―search-incident-to-arrest exception that existed at common law‖ was limited to
 25

instances in which formal arrest preceded the search. Post at 49. That does not

appear to be the case. The dissent cites no common-law authority holding or

explicitly stating that a lawful search incident to arrest cannot precede the formal

arrest. Id. The Supreme Court has pointed out that the historical underpinnings of

the search-incident-to-arrest doctrine are ―sparse.‖ United States v. Robinson, 414

U.S. 218, 230 (1973). But the Court in Rawlings relied on a number of decisions

holding that a search incident to arrest can lawfully precede the arrest. Rawlings,

448 U.S. at 111 (citing, e.g., United States v. Brown, 150 U.S. App. D.C. 113, 114-

15, 463 F.2d 949, 950-51 (1972) (per curiam) (approving search incident to arrest

where officer, who had probable cause to believe defendant was in possession of

narcotics, introduced himself and then searched defendant; no indication officer

had seized defendant before search or intended to arrest defendant before search;

―Even though a suspect has not formally been placed under arrest, a search of his

person can be justified as incident to an arrest if an arrest is made immediately

after the search, and if, at the time of the search, there was probable cause to

arrest.‖)). Although one can also find contrary decisions, the line of authority on

which Rawlings relied reaches back nearly a hundred years. See, e.g., United

States v. Gorman, 355 F.2d 151, 160 (2d Cir. 1965) (Friendly, J.) (dicta) (―We do

not understand just what values would be served by a rule that would force the

police to impose a justifiable restraint on the person as a condition to making a
 26

search which, if fruitless, might cause them to decide against it; on the other hand,

if the search does lead them to make an arrest for which reasonable cause

previously existed, the search would seem ‗incident to arrest,‘ in any normal use of

language, and the dilemma of seeking to justify the arrest by the search and at the

same time to justify the search by the arrest is obviously not presented.‖) (citations

and internal quotation marks omitted); People v. Simon, 290 P.2d 531, 533 (Cal.

1955) (Traynor, J.) (in case where officer stopped defendant, searched defendant,

found marijuana, and arrested defendant, court concluded that ―search [was] not

unlawful merely because it precede[d] rather than follow[ed] the arrest‖); State v.

McDaniel, 237 P. 373, 376 (Or. 1925) (―It is urged that the arrest followed, and did

not precede, the search. In our opinion it is immaterial whether the arrest preceded

or followed the search, if such acts were practically simultaneous, and if, in fact,

the defendant was guilty of committing a crime in the presence of the officers for

which he might have been arrested. In many instances it is dangerous for an

officer to go through the formality of stating that the accused is under arrest, and

the law does not require him to do so. It is oftentimes safer to act first and talk

afterward.‖).

The dissent also attempts to marginalize Rawlings by describing Rawlings as

presenting ―rare circumstances‖ or ―unusual facts.‖ Post at 54-55 & notes 5-6, 62.
 27

To the contrary, the issue presented in Rawlings arises with great frequency. Many

published appellate decisions apply Rawlings to uphold searches conducted

incident to, but before, arrest. See supra at 9-10. Although the dissent could be

read to suggest that Gant indicated that lawful searches incident to arrest that

precede formal arrest will be ―rare,‖ post at 55 note 6, in fact, Gant used the word

―rare‖ to describe one particular type of search incident to arrest: a justifiable

search of a vehicle, to protect officer safety, occurring before an officer can

adequately secure the suspect. 556 U.S. at 343 n.4. Nothing in Gant suggests that

it would be unusual for lawful searches incident to arrest generally, or Gant

evidence searches specifically, to precede formal arrest.

B.

Although the dissent somewhat tentatively suggests that Rawlings is

inapplicable to Gant evidence searches, post at 59-60, the dissent ultimately

concludes more narrowly that, at a minimum, an arrest must be under way for a

Gant evidence search to be lawful. Post at 56-59. We disagree.
 28

The dissent correctly notes that Justice Scalia‘s concurrence in Thornton v.

United States, 541 U.S. 615 (2004), spoke in terms of searches incident to arrests

that had already occurred. Post at 48-53. According to the dissent, this language

reflects a considered judgment that evidentiary searches incident to arrest should

be limited in conformity with the common-law rule that formal arrest must precede

a search incident to arrest. Id. Moreover, the dissent reasons, the Supreme Court

in Gant adopted the reasoning of Justice Scalia‘s concurrence in Thornton. Id.

Therefore, the dissent suggests, Gant should be understood as implicitly precluding

Gant evidence searches that precede formal arrest. Id. Our analysis differs from

that of the dissent in a number of respects.

First, as we have already noted, supra at 24-26, there does not appear to

have been a settled common-law rule that the formal arrest must precede a search

incident to arrest. Second, the defendant in Thornton had been placed under

formal arrest before the search at issue. 541 U.S. at 618. Justice Scalia thus had

no occasion to address whether evidence searches incident to arrest may, in

conformity with Rawlings, precede the formal arrest. Although Justice Scalia

expressed his conclusion in terms reflecting the circumstances of Thornton, that

cannot reasonably be understood to reflect a considered intent to implicitly carve

out an exception to Rawlings. The dissent itself recognizes the applicable general
 29

principle. Post at 57-58 (quoting Armour & Co. v. Wantock, 323 U.S. 126, 132-33

(1944) (―It is timely again to remind counsel that words of our opinions are to be

read in the light of the facts of the case under discussion. To keep opinions within

reasonable bounds precludes writing into them every limitation or variation which

might be suggested by the circumstances of cases not before the Court. General

expressions transposed to other facts are often misleading.‖)). Third, Gant does

not itself impose a requirement that Gant evidence searches must follow the formal

arrest. In Gant too the defendant had been formally arrested before the search at

issue, 556 U.S. at 336, so the Court had no occasion to consider the applicability of

Rawlings. But the Supreme Court‘s language in Gant suggests that the Court in

Gant understood and intended that Rawlings would continue to permit searches

incident to arrest to be conducted before the formal arrest. For example, in

explaining the basis for Gant evidence searches, the Court states that in some cases

―the offense of arrest‖ -- not, as the dissent would have it, the fact of arrest -- ―will

supply a basis for searching the passenger compartment of an arrestee‘s vehicle

and any containers therein.‖ Gant, 556 U.S. at 344; see also id. at 335

(―[C]ircumstances unique to the automobile context justify a search incident to

arrest when it is reasonable to believe that evidence of the offense of arrest might

be found in the vehicle.‖); id. at 343 & n.4 (indicating that searches incident to

arrest for safety purposes may be conducted before officers ―fully effectuate an
 30

arrest‖). Fourth, we are doubtful that as a lower court we could appropriately infer

that Gant implicitly intended to carve out an exception to the holding of Rawlings

that searches incident to arrest may precede the formal arrest. See, e.g., Agostini v.

Felton, 521 U.S. 203, 237 (1997) (rejecting proposition ―that other courts should

ever conclude that [the Supreme Court‘s] more recent cases have, by implication,

overruled an earlier precedent‖). Fifth, as we have already noted, supra at 6, every

court of which we are aware to have decided the issue has held that Rawlings

applies to Gant evidence searches.

In light of the foregoing considerations, it is not surprising that the dissent

itself does not appear to rest on the idea that Gant implicitly carved out an

exception to Rawlings. Rather, the dissent primarily takes the view that at a

minimum an arrest must be under way before a Gant evidence search may lawfully

be conducted. Post at 56-59. That view does not withstand analysis.

First, the Supreme Court in Rawlings suggested no such limitation. To the

contrary, the Court stated its holding more broadly: it suffices if ―formal arrest

follow[s] quickly on the heels of the challenged search.‖ 448 U.S. at 111. Second,

the overwhelming weight of authority interprets Rawlings to permit searches
 31

incident to arrest to precede the formal arrest, without any indication that arrest

must in some sense be under way at the time of search. Supra at 6-7, 9-11, 13; see

also, e.g., Brown, 150 U.S. App. D.C. at 114-15, 463 F.2d at 950-51 (approving

search incident to arrest where officer, who had probable cause to believe

defendant was in possession of narcotics, introduced himself and then searched

defendant; no indication officer had seized defendant before search or intended to

arrest defendant before search). In contrast, there is virtually no support in the case

law for the dissent‘s novel limitation on Rawlings. Third, the concept of an arrest

being ―under way‖ is remarkably opaque. As previously noted, supra at 15,

counsel for Mr. Lewis indicated at oral argument that an arrest is under way as

long as the officers subjectively intend to make an arrest. On that view, the term

―under way‖ is simply another way of requiring a subjective intent to arrest. The

dissent does not explain whether it agrees with Mr. Lewis that subjective intent to

arrest suffices, or whether instead the dissent believes that officers must take some

unspecified action objectively manifesting their intent to arrest before an arrest will

be deemed under way. Post at 55-57. Given these fundamental uncertainties, we

are not persuaded by the dissent‘s vague assurance that the term ―under way‖ is not

ambiguous. Id. at 58. Relatedly, it is difficult to understand the dissent‘s assertion

that Ms. Gibbs‘s arrest in this case was ―not under way—under any definition of

the term—when the police searched Mr. Lewis‘s car.‖ Id. at 59. To the contrary,
 32

at the time of the search, Ms. Gibbs had been seized, the police had probable cause

to arrest her, and she was arrested promptly after the search. Her arrest thus can be

understood to have been ―under way‖ in an ordinary sense of the term. Merriam-

Webster’s Collegiate Dictionary 1364 (11th ed. 2012) (defining ―under way,‖ inter

alia, as ―in progress‖ or ―afoot‖). Finally, the dissent‘s conclusion that an arrest

was not under way in this case necessarily rests on the view that a Gant evidence

search is unlawful unless the police subjectively intend to arrest the defendant at

the time of the search. As we have already explained, supra at 16-21, controlling

Supreme Court authority forecloses such an approach. The search of the car in this

case was based on particularized suspicion that the car contained evidence of a

crime. The Supreme Court has repeatedly and unequivocally held that such

searches may not be invalidated on the basis of officers‘ subjective intent. Id. at

16-17 (citing cases). The dissent‘s response to this controlling authority is to

suggest that there may be an exception permitting consideration of subjective

intent where ―the government wishes to deviate from objective, existing warrant

exceptions.‖ Post at 72. For the reasons already stated, this case does not involve

a ―deviat[ion] from objective, existing warrant exceptions.‖ In any event, the

Supreme Court could not have been clearer: searches based on particularized

suspicion may not be invalidated based on the officer‘s subjective intent. Given
 33

our obligation to follow the holdings of the Supreme Court, we see no room for the

dissent‘s approach.

C.

According to the dissent, upholding the legality of the search in this case

will ―eviscerate[] the limits‖ imposed in Gant on searches incident to arrest, will

permit ―rummaging at will‖ during searches ―untethered in every respect from an

actual arrest,‖ ―asks almost nothing of police officers before they conduct a search

of a car incident to arrest,‖ and ―invites‖ discriminatory law enforcement. Post at

47, 41 (brackets omitted), 65, 69. These would be very serious objections if they

were well founded, but they are not well founded.

The Supreme Court in Gant precluded a particular type of search incident to

arrest: searches justified neither by a need to protect officer safety nor by a

particularized reason to believe that there would be evidence of crime in the area

searched. E.g., 556 U.S. at 344 (―Neither the possibility of access [to a weapon]

nor the likelihood of discovering offense-related evidence authorized the search in
 34

this case.‖). The search in the present case differs critically from the type of search

precluded in Gant, because the search of the car in the present case rested on

particularized reason to believe that the car contained evidence of the POCA

offense. Upholding the legality of the search in the present case in no way

undermines Gant‘s rejection of suspicionless searches. For the same reason,

upholding the search in this case will not permit officers to ―rummag[e] at will.‖

Post at 41. The Supreme Court used the term ―rummage‖ in Gant to refer to

searches not based on particularized suspicion. Id. at 345 (permitting search

―when there is no basis for believing evidence of the offense might be found in the

vehicle‖ ―implicates the central concern underlying the Fourth Amendment—the

concern about giving police officers unbridled discretion to rummage at will

among a person‘s private effects‖); see also, e.g., Merriam-Webster’s Collegiate

Dictionary 1089 (11th ed. 2012) (defining ―rummage,‖ inter alia, as ―to engage in

an undirected or haphazard search‖). Because it was based on reason to believe

that the car contained evidence of POCA, the search in this case cannot reasonably

be described as involving ―rummaging at will.‖ Post at 41 (brackets omitted).

The search in this case occurred after Ms. Gibbs had been seized and at a

time when the police had probable cause to arrest Ms. Gibbs for POCA. The

search rested on particularized reason to believe that the car would contain
 35

evidence of POCA. After the search, Ms. Gibbs was promptly arrested for POCA.

Far from being ―untethered in every respect from an actual arrest,‖ post at 41, the

search was very closely tethered to Ms. Gibbs‘s arrest. And far from ―ask[ing]

almost nothing of police officers before they conduct a search of a car incident to

arrest,‖ post at 65, Gant evidence searches, as we understand them, require both

probable cause to arrest for an offense and particularized reason to believe that the

vehicle that is searched will contain evidence of that offense.

We turn finally to the dissent‘s view that upholding the legality of the search

in this case will invite discriminatory law enforcement. Post at 68-69. As

previously noted, many courts have upheld searches incident to arrest where the

search preceded the formal arrest and where officers were not shown to have

intended to arrest at the time of the search. Supra at 6-7. Nevertheless, the dissent

has pointed to no evidence that such rulings have contributed to abusive or

discriminatory law enforcement. Although the dissent cites materials describing

discriminatory law-enforcement tactics, including the use of racial profiling and

―baseless[]‖ stops, post at 68 note 14, the dissent does not identify any material

linking those problems to the particular issue in this case.
 36

We share the dissent‘s more general concerns about discriminatory law

enforcement. Post at 68-69. But the dissent fails to explain how the approach it

favors would materially reduce the incentive to conduct searches of suspects whom

officers would otherwise not be inclined to arrest. It is true, as the dissent states,

that ―arrests consume law enforcement resources.‖ Post at 64. Because it is

unclear what steps the dissent would require officers to take before searches would

be permitted, supra at 35, it is unclear what practical consequences the dissent‘s

approach would have. In any event, officers who wanted to conduct searches

might well decide to devote the necessary additional resources. As we and other

courts have noted, supra at 8-9, 29, the approach advocated by the dissent might

well be worse for suspects in some circumstances than an approach that permits

searches that precede formal arrest. In any event, whatever the possible policy

implications of the various approaches, we conclude that binding Supreme Court

doctrine precludes the approach advocated by the dissent.

In sum, the dissent in our view rests on a novel and vague limitation of the

Supreme Court‘s holdings in Rawlings and Gant. Moreover, the dissent‘s

approach is incompatible with the Supreme Court‘s repeated holdings that searches

based on particularized suspicion may not be invalidated based on officers‘
 37

subjective motivation or intent. We therefore respectfully disagree with the

dissent.

IV.

To reiterate, we hold that, under the applicable Supreme Court decisions, a

Gant evidence search is lawful if (a) the police have probable cause to arrest the

suspect for an offense, Maryland v. King, 133 S. Ct. 1958, 1970 (2013)

(―[P]robable cause provides legal justification for arresting [a suspect], and for a

brief period of detention to take the administrative steps incident to arrest. . . . The

validity of the search of a person incident to a lawful arrest [is settled].‖) (internal

quotation marks omitted); (b) the suspect recently occupied a vehicle, Gant, 556

U.S. at 343; (c) the police have reasonable, articulable suspicion to believe that the

vehicle contains evidence of the offense, id.; (d) at the time of the search, the

police have not released the suspect or issued the suspect a citation for the offense,

Knowles, 525 U.S. at 118-19; and (e) the suspect‘s formal arrest for the offense

follows quickly on the heels of the search, Rawlings, 448 U.S. at 111.

In this case, Mr. Lewis does not dispute that the police had probable cause to

arrest Ms. Gibbs for POCA. Ms. Gibbs was a recent occupant of the car. The
 38

division has already ruled that the police had reasonable, articulable suspicion to

believe that the car contained evidence of POCA, and the en banc court left that

ruling undisturbed. Nash, 100 A.3d at 164-65; Lewis, 107 A.3d at 603. Ms. Gibbs

had not been released or issued a citation at the time of the search. Finally, Mr.

Lewis does not dispute that officers made a prompt formal arrest of Ms. Gibbs for

POCA. We therefore reverse the trial court‘s order of suppression and remand for

further proceedings.

So ordered.

BECKWITH, Associate Judge, with whom WASHINGTON, Chief Judge, and

EASTERLY, Associate Judge, join, dissenting: By the time Officer Brown arrived at

the scene on Ingraham Street, Officer Alto had already detained appellant David

Lewis for driving with a suspended license, and Brittney Gibbs, Mr. Lewis‘s

passenger, had already handed over and claimed ownership of the half-full bottle

of tequila Officer Alto spotted in the car‘s center console. Though the officers had

probable cause to arrest Ms. Gibbs for possession of an open container, they did

not arrest her. They searched the car instead. Officer Alto was considering letting

Ms. Gibbs drive the car away, so he wanted Officer Brown ―to check the vehicle
 39

for other open containers of alcohol‖ and to ―make sure there‘s no other

contraband in the vehicle.‖

While Ms. Gibbs sat unrestrained on the police cruiser‘s bumper, Officer

Brown opened the front passenger door and began her search. After she smelled

marijuana, saw a plastic cup with alcohol in it on the floor, and found a marijuana

cigarette in the passenger door handle, she kept searching. It was not until Officer

Brown unzipped a Nike bag in the back seat and discovered a handgun that she

decided to arrest Ms. Gibbs, who was still unrestrained and ―standing outside the

car.‖ Officer Brown then conducted a true search incident to Ms. Gibbs‘s actual

arrest, recovering ―a bag of marijuana that she had stuffed in her bra.‖

The search of Mr. Lewis‘s car was not a search incident to arrest as this

court or the Supreme Court has ever conceived of that exception to the warrant

requirement.

The search was not incident to Mr. Lewis‘s arrest: he was handcuffed at the

time of the search and the police had no reason to think they would find evidence
 40

of the offense of driving with a suspended license in the car. See Chimel v.

California, 395 U.S. 752, 763 (1969) (identifying the rationales justifying the

search-incident-to-arrest exception—disarming the suspect and securing evidence

related to the offense of arrest that the suspect might destroy); Arizona v. Gant, 556

U.S. 332, 343 (2009) (holding that ―circumstances unique to the vehicle context

justify a search incident to a lawful arrest when it is ‗reasonable to believe

evidence relevant to the crime of arrest might be found in the vehicle‘‖) (citation

omitted).

Nor was the search somehow incident to Ms. Gibbs‘s eventual arrest. If

anything, it was the officers‘ decision not to arrest Ms. Gibbs that gave rise to their

need to search Mr. Lewis‘s car, because, as Officer Brown testified, Ms. Gibbs

―wasn‘t under arrest and the car wasn‘t impounded,‖ so ―she could have driven

away‖ with contraband in the car. The search had nothing to do with disarming an

arrestee or preventing her from destroying evidence related to an offense of arrest,

see Chimel, 395 U.S. at 763, and it was only after the search turned up a gun that

Officer Brown decided to arrest Ms. Gibbs after all.
 41

Such an investigative search, untethered in every respect from an actual

arrest, runs contrary to the basic Fourth Amendment principle that ―conducting a

Chimel search is not the Government‘s right; it is an exception—justified by

necessity—to a rule that would otherwise render the search unlawful.‖ Thornton v.

United States, 541 U.S. 615, 627 (2004) (Scalia, J., concurring). Officer Brown‘s

wait-and-see approach to Ms. Gibbs‘s arrest involved the type of ―rummag[ing] at

will‖ that the Supreme Court has sought to protect against, see Chimel, 395 U.S. at

767, and this search of Mr. Lewis‘s car, ―conducted outside the judicial process,

without prior approval by judge or magistrate,‖ was not reasonable under the

Fourth Amendment, Katz v. United States, 389 U.S. 347, 357 (1967).

And yet the majority opinion upholds Officer Brown‘s investigative pre-

arrest search because, in its view, it is a ―Gant evidence search‖ that satisfies the

majority‘s new five-part test governing the constitutionality of vehicle searches

incident to arrest. Ante at 7-8. Under this test, a Gant evidence search is ―incident

to arrest‖ and thus constitutionally permissible even where the search precedes the

arrest for the offense at issue and even where it is clear the officers did not intend

to arrest the suspect. Ante at 7-8. Because Ms. Gibbs was ultimately arrested, the

majority concludes, Officer Brown‘s search of the car was a permissible search
 42

incident to arrest, even though it was doubtful Officer Brown intended to arrest

Ms. Gibbs before she found the gun.

The majority justifies this reading of Gant—which neither party fairly

anticipated would be the basis for resolving this case and which neither party

meaningfully briefed1—largely on the ground that it has not found a case holding

to the contrary. Ante at 6. That no court has held to the contrary, however, cannot

justify crafting a new Gant rule that flouts the very premise of Gant and that

transforms the search-incident-to-arrest exception into a search-incident-to-

probable-cause-to-arrest exception to the warrant requirement. Such an exception,

neither specifically established nor well delineated, cannot be squared with

Supreme Court precedent or the Fourth Amendment.

I.

1
In their briefs as in their respective statements of the issue presented for
review, the parties focus on whether a Fourth Amendment search incident to arrest
requires the government to prove more than just probable cause to arrest at the
time of the search.
 43

Arizona v. Gant is the beginning and end of the majority‘s analysis—

specifically, the exception to the warrant requirement Gant recognizes in

circumstances where officers stop a vehicle, arrest its occupant, and search the car

for evidence of the offense of arrest. Given its reliance on Gant, then, the majority

opinion is striking in two ways: (1) in the extent to which the opinion, by

upholding what amounts to an investigatory search that cannot be justified by the

rationales underlying the exception to the warrant requirement it is invoking,

repeats the same kind of mistake the Supreme Court in Gant sought to stop lower

courts from making, and (2) in the extent to which the majority must stray from the

very precepts of Gant‘s vehicle exception to apply it to the facts of this case.

In Arizona v. Gant, 556 U.S. 332 (2009), the Supreme Court set out to

address a problem in Fourth Amendment doctrine that stemmed from its decision

in New York v. Belton, 453 U.S. 454 (1981). In Belton, the Court applied Chimel

to the automobile context and held that when the police have lawfully arrested a

recent occupant of a car they may search the passenger compartment ―as a

contemporaneous incident of that arrest.‖ Id. at 460. Applying the principles of

Chimel, the Court predicated its decision on the ―generalization‖ that articles in the

passenger compartment ―are in fact generally, even if not inevitably, within ‗the

area into which an arrestee might reach in order to grab a weapon or evidentiary
 44

items.‘‖ Id. (quoting Chimel, 395 U.S. at 763). Although the Court in Belton

cautioned that its holding ―in no way alter[ed] the fundamental principles

established in the Chimel case regarding the basic scope of searches incident to

lawful custodial arrests,‖ id. at n.3, the Court in Gant acknowledged that in

practice Belton searches had come to exceed their permissible scope under Chimel

and were being conducted solely for investigative purposes, Gant, 556 U.S. at

341–43. On this expansive reading of Belton, vehicle searches would be

authorized incident to every recent occupant‘s arrest even if the passenger

compartment were not in the arrestee‘s reach at the time of the search.2 Id. at 343.

The lower courts, in construing Belton so broadly, were ―treat[ing] the ability to

search a vehicle incident to the arrest of a recent occupant as a police entitlement

rather than as an exception justified by the twin rationales of Chimel.‖ Id. at 342

(quoting Thornton, 541 U.S. at 624 (O‘Connor, J., concurring in part)).

In seeking to curtail these investigative searches under Belton, the Supreme

Court in Gant began by reaffirming the axiom that a warrantless search is ―per se

unreasonable‖ absent justification under one of the ―few specifically established

2
As Justice Scalia observed in his Thornton concurrence, in some cases
lower courts had upheld searches under Belton ―even when . . . the handcuffed
arrestee has already left the scene.‖ Thornton, 541 U.S. at 628.
 45

and well-delineated exceptions‖ to the warrant requirement. Id. at 338 (quoting

Katz v. United States, 389 U.S. 347, 357 (1967)). One such exception, established

by Chimel, is for searches of a suspect that are incident to the suspect‘s arrest and

that are intended to ensure that the suspect does not have the ability to access

weapons or destroy evidence. Id. (citing Chimel, 395 U.S. at 763). The Gant

Court held that this Chimel exception, which had been impermissibly broadened

for vehicle searches under Belton, authorized the search of a car incident to a

recent occupant‘s arrest ―only when the arrestee is unsecured and within reaching

distance of the passenger compartment at the time of the search.‖ Id. at 343. With

this holding, the Court made clear that Chimel cannot be construed to authorize the

investigative vehicle searches then being conducted under Belton. Id. at 347

(―Construing Belton broadly to allow vehicle searches incident to any arrest would

serve no purpose except to provide a police entitlement, and it is anathema to the

Fourth Amendment to permit a warrantless search on that basis.‖). The Court also

held that separate and apart from Chimel‘s two traditional justifications for the

warrant exception, ―circumstances unique to the vehicle context justify a search

incident to a lawful arrest when it is ‗reasonable to believe evidence relevant to the

crime of arrest might be found in the vehicle.‘‖ Id. at 343 (quoting Thornton, 541

U.S. at 632 (Scalia, J., concurring)).
 46

The Gant Court then applied Chimel and the new vehicle exception to the

facts of the case, in which Mr. Gant, after getting out of a parked car, was arrested

for driving with a suspended license, handcuffed, and placed in the back seat of a

police cruiser before officers searched his car and found a gun and cocaine. Id. at

336. In the Court‘s view, the officers‘ search was unreasonable. Id. at 344. The

two Chimel rationales for a search incident to arrest did not apply because Mr.

Gant was handcuffed and secured in the back of a police car. Id. The rationale

underlying the vehicle exception also did not apply, as there was no chance of

finding evidence of the ―crime of arrest‖—driving without a license—in Mr.

Gant‘s car. Id.

No one here disputes that the police had probable cause to make an arrest for

possession of an open container of alcohol (POCA). It is also undisputed that at

the time of the search the police had not arrested Ms. Gibbs for POCA or for any

other crime. Yet unlike in Gant itself, which explicitly allows only a search of a

car for evidence ―incident to a lawful arrest‖ for a ―crime of arrest,‖ 556 U.S. at

343 (emphasis added), the majority concludes that a Gant car search has a broader

investigatory rationale that is implicated simply by probable cause to believe the

suspect has committed an arrestable offense, not by the fact of arrest itself.
 47

At the outset, the argument for expanding the Gant vehicle exception to

searches incident to probable cause to arrest forgets the fundamental context in

which Gant arose. The Supreme Court in Gant was intent on reining in the purely

investigative searches that had been occurring under Belton, and to that end the

Court expressly rejected the notion of a ―police entitlement‖ to search a car

whenever a recent occupant has been arrested. Gant, 556 U.S. at 342 (citation

omitted). By permitting a search of the vehicle incident to probable cause to arrest

as part of a Gant vehicle search, the majority interprets Gant as authorizing law

enforcement to conduct a broad investigatory search of the vehicle before making a

decision about whether to arrest. This reading of Gant eviscerates the limits the

Court sought to impose on Belton car searches, and in fact would give the police

more latitude to search than they had under lower courts‘ pre-Gant reading of

Belton, which authorized a search incident to arrest only if there had been ―a lawful

custodial arrest‖ and the search was ―a contemporaneous incident of that arrest.‖

Belton, 453 U.S. at 460. We cannot fairly read Gant as professing to scale back

Belton‘s investigatory searches while at the same time authorizing a search-

incident-to-probable-cause-to-arrest exception that by definition invites such

investigatory searches.
 48

The majority‘s reading of Gant also ignores the principles underlying the

vehicle exception, including, in particular, the exception‘s textual origins in Justice

Scalia‘s concurrence in Thornton. Although this exception ―d[id] not follow from

Chimel,‖ Gant, 556 U.S. at 343, the Court made clear that it was still a category of

search incident to arrest, one whose rationale—securing offense-related evidence

that might be found in the vehicle—is informed by the same concerns as those that

inform the evidence-preservation rationale underlying Chimel. Like the traditional

Chimel exception, then, this Gant vehicle search was in no way intended to

authorize the type of open-ended investigative searches of which the Gant Court

expressly disapproved in curbing Belton‘s excesses. Rather, the Gant search was

conceived as a narrow exception that would allow the police to search a car for

additional evidence of the crime for which the arrestee will be charged. This

search was designed only to address ―circumstances unique to the vehicle

context‖—a car that could be driven away from the scene with relevant evidence

still inside. See id.; Wyoming v. Houghton, 526 U.S. 295, 304 (1999) (noting that

―in all car-search cases, the ‗ready mobility‘ of an automobile creates a risk that the

evidence or contraband will be permanently lost while a warrant is obtained‖).

The language in Gant adopting this vehicle exception was taken from Justice

Scalia‘s concurrence in Thornton, 541 U.S. at 632. In Justice Scalia‘s view, even
 49

in the absence of Chimel-like concerns, officers should be able to search a vehicle

for evidence of a crime of arrest. On the facts of Thornton, where the police

arrested Mr. Thornton on narcotics charges after finding marijuana and cocaine on

his body in a lawful patdown, officers should also have been able to search his car

for drugs, even though he was handcuffed and seated in a patrol car, with no access

to weapons and no way of destroying any evidence.

The reasons Justice Scalia gave for this position had everything to do with

the fact of arrest. He made clear that the position was not novel or radical, but

merely a reflection of the search-incident-to-arrest exception that existed at

common law. Thornton, 541 U.S. at 629. He cited pre-Chimel cases from 1914

through 1950 in which the Court had approved of searches for evidence ―relevant

to the crime for which the suspect had been arrested.‖ Id. In every one of these

cases, a suspect had been formally arrested.3 Justice Scalia then quoted at length

an 1872 common law treatise discussing the reasons for allowing such searches,

calling such reasoning ―typical‖:

3
These cases include United States v. Rabinowitz, 339 U.S. 56, 58, 60–64
(1950); Harris v. United States, 331 U.S. 145, 148, 151–52 (1947); Marron v.
United States, 275 U.S. 192, 198–99 (1927); Agnello v. United States, 269 U.S. 20,
30–31 (1925); Weeks v. United States, 232 U.S. 383, 392 (1914).
 50

The officer who arrests a man on a criminal charge
should consider the nature of the charge; and, if he finds
about the prisoner‘s person, or otherwise in his
possession, either goods or moneys which there is reason
to believe are connected with the supposed crime as its
fruits, or as the instruments with which it was committed,
or as directly furnishing evidence relating to the
transaction, he may take the same, and hold them to be
disposed of as the court may direct.

Id. at 630 (citing 1 J. Bishop, Criminal Procedure § 211, p. 127 (2d ed. 1872)).

―There is nothing irrational,‖ Justice Scalia added, ―about broader police authority

to search for evidence when and where the perpetrator of the crime is lawfully

arrested,‖ and it is this ―fact of prior lawful arrest‖ that ―distinguishes the arrestee

from society at large, and distinguishes a search for evidence of his crime from

general rummaging.‖ Id. He even appeared to suggest that the police might in

some sense be obligated to search for such evidence relevant to a charge of arrest.

See id. at 632 (quoting Smith v. Jerome, 93 N.Y.S. 202, 203 (Sup. Ct. 1905) (―This

right and duty of search and seizure extend, however, only to articles which furnish

evidence against the accused.‖ (emphasis added))).

Even as he endorsed these searches, Justice Scalia acknowledged that an

evidence-gathering justification for a search is ―far less compelling‖ than a Chimel

rationale, where ―officer safety or imminent evidence concealment or destruction is
 51

at issue‖ and where ―officers should not have to make fine judgments in the heat of

the moment.‖ Id. at 632. Justice Scalia therefore indicated that the exception

should be carefully circumscribed to apply only to ―relevant‖ evidence against the

accused, and that the search of Mr. Thornton‘s car was legal because Mr. Thornton

―was lawfully arrested for a drug offense‖ and the police had reason ―to believe

that further contraband or similar evidence relevant to the crime for which he had

been arrested might be found in the vehicle from which he had just alighted and

which was still within his vicinity at the time of arrest.‖ Id. (emphasis added).

The reasoning of Justice Scalia‘s concurrence, upon which Gant exclusively

relied in adopting an evidence-gathering search-incident-to-arrest exception, for

the most part does not apply in the absence of a formal arrest. First, while a person

who has been formally arrested undoubtedly has a lowered expectation of privacy

in his body and effects, see, e.g., Maryland v. King, 133 S. Ct. 1958, 1978 (2013),

the government cites no case, and we are aware of none, suggesting that a person

the police have probable cause to believe committed an offense, but who has not

yet been subject to arrest for that offense, has a lowered expectation of privacy on
 52

that basis alone.4 Second, Justice Scalia‘s concurrence—and the common law

exception it resurrects—makes reference to the right and duty of officers to collect

relevant evidence against the accused. This right and duty arise presumably

because formal arrest reflects an anticipation of formal charges and a formal

criminal proceeding against the suspect. The very concepts of ―relevant evidence‖

and the ―accused‖ presuppose a criminal proceeding of some kind. See, e.g., Am.

Tobacco Co. v. United States, 328 U.S. 781, 787 (1946) (―The verdict in a criminal

case is sustained only when there is ‗relevant evidence from which the jury could

properly find or infer, beyond a reasonable doubt,‘ that the accused is guilty.‖

(quoting Mortensen v. United States, 322 U.S. 369, 374 (1944))); Wheeler v.

United States, 930 A.2d 232, 249 (D.C. 2007) (in cases ―where the evidence

requires careful weighing, the need for unfettered jury adjudication is at its zenith,

and requires that each juror have considered all relevant evidence and be firmly

4
In a footnote, the majority asserts that unlike Justice Scalia‘s statement in
Thornton that the ―fact of prior lawful arrest distinguishes the arrestee from society
at large,‖ Gant made reference only to searches as being incident to ―lawful
arrest,‖ omitting the word ―prior.‖ Ante at 12 note 3. But Gant in no way
purported to augment Justice Scalia‘s reasoning, much less to radically alter it by
authorizing an evidence search incident to arrest that precedes the arrest. Gant‘s
discussion of Justice Scalia‘s approach was brief and unequivocal; it adopted the
approach and moved on. Reading Gant‘s omission of ―prior‖ in this way also
appears at odds with the Gant Court‘s purpose of narrowing the search-incident-to-
arrest exception to address the problems Belton engendered. If the Court had also
intended to broaden the exception by authorizing a search without the predicate of
an arrest, it would have said so.
 53

convinced that there is no reasonable doubt as to the accused‘s guilt‖). The duty to

collect relevant evidence against a person who officers have reason to believe will

be formally accused of a crime has no meaningful application in a scenario in

which officers objectively have probable cause to arrest but have chosen not to do

so. Any argument that the common law would accommodate the type of search

conducted in Mr. Lewis‘s case is in no way suggested by Gant, by the concurrence

in Thornton, or by the cases and treatises Justice Scalia relied upon. To the

contrary, these authorities all suggest that so novel an argument is at odds with the

rationales underlying the common law exception.

At bottom, the majority‘s argument under Gant—that a search of a car

incident to probable cause to arrest is justified because the police might arrest the

person—appears no stronger than the argument the Supreme Court rejected in

Gant: that a search of a car incident to arrest under Chimel is justified even when

the arrestee is already secured because the police might have conducted the search

earlier. Both arguments assume that, ―one way or another, the search must take

place.‖ Thornton, 541 U.S. at 627. Yet both arguments overlook the fact that

searching without a warrant ―is not the Government‘s right; it is an exception—

justified by necessity—to a rule that would otherwise render the search unlawful.‖

Id.
 54

II.

The majority justifies allowing pre-arrest Gant vehicle searches incident to

arrest on two main grounds: (1) that no Supreme Court precedent explicitly states

the opposite—that pre-arrest Gant searches are impermissible—and (2) that

requiring a formal arrest to be either completed or under way would contradict the

Supreme Court‘s opinion in Rawlings v. Kentucky, 448 U.S. 98 (1980). Ante at 9,

12–13. But these justifications in turn get the warrant requirement backwards and

derive from a strained reading of Rawlings that disregards Supreme Court

precedent predicating the search-incident-to-arrest exception—whether under

Chimel or Gant—on a formal arrest or perhaps, in rare circumstances like in

Rawlings, an imminent and inevitable arrest.5

5
Contesting the notion that Rawlings presents ―rare circumstances,‖ the
majority argues that ―the issue presented in Rawlings arises with great frequency‖
and that ―[m]any published appellate decisions apply Rawlings to uphold searches
conducted incident to, but before, arrest.‖ Ante at 27. But the fact that lower
courts have cited Rawlings in trying to justify a search-incident-to-probable-cause-
to-arrest rule does not mean that the facts of Rawlings—where officers were in the
process of arresting the suspect but had not yet formally done so—arise with great
frequency.
 55

That no Supreme Court case explicitly holds that pre-arrest Gant searches

are unconstitutional is true as far as it goes. But the Fourth Amendment does not

give the police unfettered power to conduct searches and seizures unless judges tell

them not to, and this court is not creating policing policy on a blank slate. Rather,

we start from the premise that a warrantless search is per se unreasonable, subject

only to exceptions that have been defined clearly by the Supreme Court. See

Coolidge v. New Hampshire, 403 U.S. 443, 455 (1971) (general warrant

requirement is ―subject only to a few specifically established and well delineated

exceptions . . . [that] are jealously and carefully drawn‖ (internal quotation marks

and citations omitted)). With the exception of Rawlings, the Supreme Court‘s

cases on the search-incident-to-arrest exceptions of Chimel and Gant involve

circumstances in which a formal arrest was completed and explicitly state the

exception in terms of a search incident to a lawful arrest.6 In the absence of

6
The footnote from Gant that the majority cites in suggesting that the Court
―understood and intended that Rawlings would continue to permit searches
incident to arrest to be conducted before the formal arrest,‖ ante at 29, makes clear,
consistent with Rawlings, that certain pre-arrest searches would be permissible in
the ―rare case‖ where an officer has not ―fully effectuate[d] an arrest.‖ Gant, 556
U.S. at 343 n.4.
 56

precedent establishing a broader exception, the warrantless search of Mr. Lewis‘s

car is presumed unreasonable.7

With respect to the majority‘s second justification for allowing pre-arrest

searches, requiring that a completed or underway arrest precede a Chimel search is

entirely consistent with Rawlings. In Rawlings, six police officers entered a house

with an arrest warrant for a suspect and, not finding the suspect but spotting

marijuana seed, detained the house‘s occupants, including Mr. Rawlings, for forty-

five minutes, while two of the officers left to get a search warrant for the house.

448 U.S. at 100. Later, the occupants were read their Miranda rights8 and a

woman was ordered to empty her purse. Id. at 100-01. When it became clear that

the purse contained drugs, the woman turned to Mr. Rawlings and told him ―to

take what was his.‖ Id. at 101. Mr. Rawlings confessed that the drugs belonged to

him, and the officers searched his person, finding $4,500 in cash and a knife, and

then immediately placed him under formal arrest. Id. In three sentences at the

very end of its opinion, after addressing at length whether Mr. Rawlings had

7
That Gant was decided so recently also blunts the force of the majority‘s
argument that pre-arrest Gant vehicle searches are permissible because the
Supreme Court has never held them unconstitutional.
8
Miranda v. Arizona, 384 U.S. 436 (1966).
 57

standing to challenge the search of his companion‘s purse, the Court rejected Mr.

Rawlings‘s contention that the search of his person was not incident to a valid

arrest, stating that ―[w]here the formal arrest follow[s] quickly on the heels of the

challenged search of petitioner‘s person, we do not believe it particularly important

that the search preceded the arrest rather than vice versa.‖ Id. at 111.

This language in Rawlings reflects a commonsense acknowledgment that

where a formal arrest is under way at the time a suspect‘s wingspan is searched, as

it surely was on the facts of Rawlings, a hypertechnical insistence upon excluding

evidence uncovered in such a search would unnecessarily constrain the discretion

of law enforcement. The majority casts this reading of Rawlings as reflecting an

―implicit critique‖ of the case, ante at 24, but the opposite is true. Rawlings

reaches the right result. The objective circumstances there—Mr. Rawlings was

detained by four police officers for forty-five minutes while two other officers

went to get a warrant—were such that there was never any question that Mr.

Rawlings was going to be arrested at the time he was searched, and given that fact,

the Court sensibly did not believe it ―particularly important‖ that the search

preceded the ―formal arrest,‖ because such an arrest had already been set in

motion. Read in light of these facts, Rawlings cannot be construed as endorsing

pre-arrest Chimel searches where an arrest is not at least under way. See Armour
 58

& Co. v. Wantock, 323 U.S. 126, 132–33 (1944) (―remind[ing] counsel that words

of [the Court‘s] opinions are to be read in the light of the facts of the case under

discussion‖ and that ―[g]eneral expressions transposed to other facts are often

misleading‖).9 And while the majority suggests that it is unclear what such an

―underway‖ arrest means, ante at 15, the meaning is not ambiguous. If the arrest

has not begun, there is no reasonable dispute that it is about to begin. It is

9
In suggesting otherwise, the majority observes that Rawlings ―relied on a
number of decisions holding that a search incident to arrest can lawfully precede
the arrest.‖ Ante at 25. But the cases cited in Rawlings do not endorse a search-
incident-to-probable-cause-to-arrest exception to the warrant requirement, and they
were wrongly decided if they do. The officers involved in those cases were
conducting the searches with a conditional intent to arrest for the offense for which
they had probable cause if the searches were fruitful. See, e.g., United States v.
Brown, 463 F.2d 949, 950 (D.C. Cir. 1972) (where the officer‘s search was clearly
intended to confirm or dispel that the envelope ―protruding from appellant‘s shirt
pocket‖ contained narcotics—as the appellant‘s eyes were ―glassy,‖ he was
behaving suspiciously in an area where the officer had ―frequently observed‖
narcotics transactions, and the envelope was ―of the type in which [the officer] had
found narcotics on previous occasions‖—the court upheld the search even though
the suspect had ―not formally been placed under arrest‖ at the time the officer
seized the envelope); Bailey v. United States, 389 F.2d 305, 307–09 (D.C. Cir.
1967) (holding that where the government conceded that ―in law the arrest of the
car‘s [occupants] took place at the time [the officer] approached the car, perhaps
with gun drawn, and told appellants to sit still and keep their hands in plain sight,‖
that officer could, as a search incident to arrest, confiscate a wallet that he saw on
the car floor, even though he had not yet conducted the formal arrest). For this
reason, and because the circumstances of those cases triggered no concerns about
pretextual searches, the cases also do not control this case. See Wayne R. LaFave,
Search and Seizure § 5.4 (a) (5th ed. 2012) (noting that it is ―particularly
unsettling‖ to interpret Rawlings to permit pre-arrest searches where there is
probable cause to arrest and ―when the offense for which pre-search probable
cause existed was so trivial that it otherwise would likely have been ignored‖).
 59

imminent and inevitable. The concept of an ―underway‖ arrest distills, in a word,

the concern at the end of Rawlings.

In any event, Ms. Gibbs‘s arrest was not under way—under any definition of

that term—when the police searched Mr. Lewis‘s car. Officer Alto decided not to

arrest Ms. Gibbs after confiscating the open bottle of tequila, and so Ms. Gibbs

remained sitting unrestrained on the bumper of the police cruiser while Officer

Brown searched the vehicle to see what other contraband she might find. Unlike in

Rawlings, the circumstances here showed that Ms. Gibbs was unlikely to be

arrested before Officer Brown found the gun at the end of her search. Officer

Brown‘s investigatory search of the vehicle, with no arrest of Ms. Gibbs imminent

or inevitable, bears little resemblance to the search incident to an underway arrest

in Rawlings.

Of course, this case involves a Gant vehicle search, and it is not clear that

Rawlings applies at all in this context. For his part, Mr. Lewis argues that while

concerns for officer safety may sometimes justify pre-arrest Chimel searches when

the arrest is under way, those concerns are not present when the police conduct

Gant vehicle searches. It may make sense not to stand on formalities when an
 60

arrest is already under way and there is a real danger that the suspect might gain

access to a weapon or destroy evidence, but the same beat-him-to-the-draw logic

does not apply so clearly to searches of vehicles for evidence relevant to a crime of

arrest. The majority‘s only direct response to this point—that other courts have

applied Rawlings to Gant evidence searches, ante at 15-16—is less than

compelling.

Even if the Rawlings logic does apply to pre-arrest Gant searches, the

resulting rule should then be that Gant vehicle searches incident to arrest are lawful

only where a formal arrest is completed or under way. But the majority rejects this

modest approach and suggests instead that under Rawlings, Gant vehicle searches

should be permissible whenever officers have probable cause to arrest for an

offense and where the vehicle may contain evidence of that offense (along with the

majority‘s other conditions).

Recognizing that the Supreme Court in Rawlings ―did not explain [the]

reasons‖ of its fleeting analysis, the majority looks to other state court decisions

and gleans from them three policy rationales that it says underlie Rawlings and that

support a search-incident-to-probable-cause-to-arrest approach to Gant. Ante at
 61

10, 16. But imported policy rationales cannot create a well-delineated exception to

the warrant requirement where none exists. And in any event, the majority‘s

policy grounds for a pre-arrest Gant exception fall short on the merits.

As its first policy rationale, the majority asserts that a pre-arrest Gant search

based on probable cause might benefit an innocent suspect if it negates such

probable cause and leads to a decision not to arrest. Under a bright-line approach

requiring a formal arrest to justify a search incident to arrest, the argument goes,

the police will simply arrest more people, including some people they may not

have arrested had they searched them first and found nothing incriminating.

But determining how a search-incident-to-probable-cause-to-arrest rule

would affect innocent suspects involves a series of complex empirical questions.

The precarious assumptions that emerge from such an inquiry are not persuasive

grounds for engaging in an otherwise counterintuitive interpretation of Supreme

Court precedent to allow a search incident to arrest in the absence of an arrest.

Indeed, while the majority cites State v. Overby, 590 N.W.2d 703, 706 (N.D.

1999), for the proposition that ―if the person searched is innocent and the search

convinces the officer that his reasonable belief to the contrary is erroneous, it is to
 62

the advantage of the person searched not to be arrested,‖ ante at 10, the Overby

court was itself divided in that regard. Chief Justice VandeWalle, joined by

another justice, wrote that ―[t]he insult remains,‖ adding, ―[m]ore importantly,‖

that

the reality is that an officer who is suspicious may be
subconsciously tempted to conduct a search before
making the arrest with the expectation or hope that the
search will produce such irrefutable evidence of the
commission of a crime that a lack of probable cause to
arrest prior to the search will be overlooked, or such
suspicion as did exist will be viewed more favorably in
light of the evidence discovered in the search if, in fact,
there is evidence discovered.

Id. at 708 (VandeWalle, C.J., concurring specially).

The majority‘s second policy rationale, that a pre-arrest Gant search adds no

further intrusion where an arrest is already ―inevitable,‖ ante at 10, 16, may well be

true. The difficulty, however, is in determining when an arrest is truly inevitable.

Sometimes a set of unusual facts, like those in Rawlings, will objectively indicate

the inevitability of an arrest at the time of a pre-arrest search. An approach that

attempts to extract from Rawlings and Gant an administrable rule—that a search

incident to arrest is legal only in the face of a completed or ―underway‖ arrest—

provides a workable proxy for inevitability. The majority‘s proposed conditions to
 63

the legality of a pre-arrest Gant search, in contrast, are not proxies for the

inevitability of an arrest. The circumstances of the five-factor test—that officers

have probable cause to arrest, that the suspect has just come from a car, that the

police have reason to believe the car might contain evidence of the arrestable

offense, that the police have not released the suspect or issued a citation, and that

the suspect‘s arrest for the offense follows quickly on the heels of the search—in

no way imply that the suspect inevitably would have been arrested for that offense.

Indeed, there are strong indications that the officers had no intent to arrest Mr.

Lewis or Ms. Gibbs for POCA, at least before their pre-arrest search uncovered

evidence of more serious crimes.

As to the majority‘s final policy rationale for pre-arrest Gant searches,

courts do appear ―reluctant to micromanage the precise order in which officers

who have probable cause to arrest conduct searches and arrests‖ in the context of a

Chimel search where, as in Rawlings, the arrest is under way. Ante at 10. And the

majority‘s citation on this point, Anderson v. State, 553 A.2d 1296 (Md. Ct. Spec.

App. 1989), says nothing beyond that unremarkable truism. See id. (noting that it

would be an ―absurdity‖ to require officers in a ―combat situation‖ right before

arrest to ―lose critical seconds,‖ and lose the chance to ―beat[] his opponent to the
 64

draw,‖ in order to ―say the operative words, you are under arrest‖).10 Requiring

that an arrest be completed or under way is entirely consistent with judges‘

concerns about micromanaging police. This approach does not seek to

micromanage the conduct of officers acting in what are often stressful situations,

but rather to force these officers to verify at the front end that their searches are

tethered to an actual arrest supported by probable cause, rather than mere general

rummaging that may or may not lead to an arrest depending on what the officers

find. And because arrests consume law enforcement resources, predicating a

search incident to arrest upon an actual arrest has the added benefit of prompting

police officers to think more critically about whether the offense for which

probable cause exists—and which provides the basis for the search—is serious

enough to justify the costs associated with an arrest. Such an approach could also

foster trust and respect between police officers and community members.11

10
The government argues in its brief that mere probable cause to arrest
creates a combustible arrest situation because the suspect will know he is in
trouble. In circumstances that are devoid of other objective indications of an intent
to arrest, however, it is far from clear that a suspect would realize an officer has
probable cause to arrest him. And that same logic could be used to justify a search
incident to any traffic stop of a vehicle on the theory that reasonable suspicion to
stop the vehicle creates a combustible stop situation.
11
Cf. U.S. Dep‘t of Justice, Investigation of the Baltimore City Police
Department 8 (2016), https://www.justice.gov/opa/file/883366/download
(reporting that ―[i]n some cases, [Baltimore City Police Department] supervisors
(continued…)
 65

The majority‘s test, by contrast, asks almost nothing of police officers before

they conduct a search of a car incident to arrest, and relies instead on a judicial

officer‘s after-the-fact determination of probable cause to ensure the lawfulness of

the search—a determination that will of course be made only if the suspect is

arrested. Requiring an underway or completed arrest—and accordingly a

probable-cause-to-arrest determination—before searching a car provides an

additional safeguard against abuses in one of the few circumstances where we

allow the police to search a vehicle ―without prior approval by judge or magistrate‖

in the form of a search warrant. See Coolidge, 403 U.S. at 454; see also Riley v.

California, 134 S. Ct. 2473, 2482 (2014) (recognizing that ―a warrant ensures that

(…continued)
have ordered officers to specifically target African Americans for stops and
arrests,‖ that ―[t]hese failures [in policing] contribute to the large racial disparities
in BPD‘s enforcement that undermine the community‘s trust in the fairness of the
police,‖ and that ―BPD leadership has acknowledged that this lack of trust inhibits
their ability to forge important community partnerships‖); U.S. Dep‘t of Justice,
Investigation of the Ferguson Police Department 5–6 (2015),
https://www.justice.gov/sites/default/files/opa/press-releases/attachments/2015/
03/04/ferguson_police_department_report.pdf (highlighting ―the lack of trust
between the Ferguson Police Department and a significant portion of Ferguson‘s
residents, especially African Americans,‖ and concluding that this ―distrust . . . is
longstanding and largely attributable to Ferguson‘s approach to law enforcement,‖
which ―results in patterns of unnecessarily aggressive and at times unlawful
policing; reinforces the harm of discriminatory stereotypes; discourages a culture
of accountability; and neglects community engagement‖).
 66

the inferences to support a search are ‗drawn by a neutral and detached magistrate

instead of being judged by the officer engaged in the often competitive enterprise

of ferreting out crime‘‖ (quoting Johnson v. United States, 333 U.S. 10, 14

(1948))).

It is true that the conditions the majority places on pre-arrest searches avert

some of the possible abuses. An officer could not, for example, simply pull over a

motorist for an extralegal reason (such as his race) on the basis of probable cause

for an arrestable but pretextual traffic offense like reckless driving, search his car

for evidence of some other crime, and expect to have the search upheld. But

despite the majority‘s insistence that its test will not lead to abuses because Gant

evidence searches must rest on ―particularized suspicion,‖ ante at 18, 20–22, 32–

36, the majority‘s approach does nothing to stop officers from searching a car if the

arrestable but pretextual offense is something like POCA, where evidence of the

offense may well be in the car, but where officers have no intent to arrest for the

minor offense if they fail to find evidence of other crimes. In such a case, the fact

that there is particularized suspicion to believe the car contains evidence of that

offense provides no comfort to a citizen whom the police want to stop and search
 67

for evidence of other unrelated crimes, using POCA as a pretext.12 See Wayne R.

LaFave, Search and Seizure § 5.4 (a) (5th ed. 2012). Under the majority‘s

approach, moreover, an officer who does not intend to arrest but conducts a search

anyway will inevitably look for evidence that will firm up probable cause or justify

an arrest on grounds other than those that formed the basis of the probable cause—

a consequence that, besides defying the rationale of a Gant search, will effectively

lower the bar to probable cause and make investigative searches more common.13

12
The government contends that such concerns are unfounded because there
is no evidence of abuse in the more than 25 years since Rawlings was issued. As
an initial matter, not all courts have construed Rawlings to allow searches incident
to probable cause to arrest. See, e.g., People v. Reid, 26 N.E.3d 237, 239 (N.Y.
2014) (―It is irrelevant that, because probable cause existed, there could have been
an arrest without a search. A search must be incident to an actual arrest, not just to
probable cause that might have led to an arrest, but did not.‖); State v. Funkhouser,
782 A.2d 387, 406 (Md. Ct. Spec. App. 2001) (―It is axiomatic that a search
incident to lawful arrest is absolutely dependent on the fact of an actual arrest.‖).
And even where courts have so construed Rawlings, we are aware of no data from
1980 to the present indicating the absence of such abuses. Indeed, it is far from
clear how such discriminatory pretextual searches would make their way into
published judicial decisions in the first place. These searches will receive judicial
scrutiny where they fortuitously uncover evidence of a more serious crime and thus
lead to an arrest, but will not in the many cases in which the police find nothing of
interest.
13
See supra note 10.
 68

These concerns are by no means hypothetical and carry with them serious

implications for disparate enforcement in policing practices.14 By allowing the

police to conduct a search of a car in the absence of an underway or completed

arrest based on probable cause to believe a specific offense has been committed,

14
See generally, e.g., Complaint, Wilkins v. Maryland State Police, No. 93-
468 (D. Md. Feb. 12, 1993), http://www.clearinghouse.net/chDocs/ public/PN-
MD-0003-0007.pdf. In this class-action lawsuit, Robert Wilkins, an African-
American attorney who now serves as a federal appellate judge, successfully
challenged the racial profiling practices of the Maryland State Police after the
police baselessly stopped and detained him and three relatives and searched their
car on a highway in Maryland because the police believed they fit the state‘s drug
courier profile. The litigation exposed the prevalence of pretextual stops and
ultimately led to a consent decree in which the Maryland State Police agreed to
revise and update its policies and procedures against racial profiling, retrain its
troopers on proper traffic stops, develop a Police-Citizen Advisory Committee to
address ongoing concerns about racial profiling, as well as take other measures to
combat this problem. See Consent Decree, Wilkins v. Maryland State Police, No.
93-468 (D. Md. Apr. 22, 2003), http://www.clearinghouse.net/chDocs/public/ PN-
MD-0003-0012.pdf; see also Baltimore Report, supra note 11, at 7 (noting that
―BPD stopped African-American [pedestrians] three times as often as white
residents after controlling for the population of the area in which the stops
occurred,‖ that ―African Americans accounted for 82 percent of all BPD vehicle
stops, compared to only 60 percent of the driving age population in the City,‖ and
that ―BPD searched African Americans more frequently during pedestrian and
vehicle stops, even though searches of African Americans were less likely to
discover contraband‖); Ferguson Report, supra note 11, at 4 (―Data collected by
the Ferguson Police Department from 2012 to 2014 shows that African Americans
account for 85% of vehicle stops, 90% of citations, and 93% of arrests made by
FPD officers, despite comprising only 67% of Ferguson‘s population. African
Americans are more than twice as likely as white drivers to be searched during
vehicle stops even after controlling for non-race[-]based variables such as the
reason the vehicle stop was initiated, but are found in possession of contraband
26% less often than white drivers, suggesting officers are impermissibly
considering race as a factor when determining whether to search.‖).
 69

the majority‘s rule dispenses with an important mechanism—an arrest—that makes

it harder for officers to target individuals because of race or ethnicity and conduct

wide-ranging investigatory searches of their vehicles even though the officers have

no intent to arrest at the start of the search. An arrest requirement does just the

opposite: it ties permissible police conduct to the fact of an arrest, authorizing a

Gant search of the vehicle only if ―evidence relevant to th[at] crime of arrest‖

might reasonably be found inside. Gant, 556 U.S. at 343. In light of the well-

publicized concerns about racial profiling in some major cities, see, e.g., Floyd v.

City of New York, 959 F. Supp. 2d 540 (S.D.N.Y. 2013), keeping the focus on the

arrest properly mitigates a risk that the majority‘s approach actually invites—that

officers will use an arrestable offense as a pretext to rummage through the cars of

people from marginalized communities in order to look for evidence of other

crimes.

III.

In its final argument in favor of broadly construing Rawlings to implicitly

allow pre-arrest Gant searches, the majority contends that any alternative

approach, such as a rule that Gant searches are permissible only if a formal arrest
 70

is completed or under way, would inevitably require an inquiry into the subjective

motivations of the arresting officers—an inquiry the Supreme Court has

discouraged in cases like Whren v. United States, 517 U.S. 806 (1996). See ante at

16–18. This argument pervades the section of the majority‘s opinion defending its

conclusion that a pre-arrest Gant vehicle search is permissible even where the

officers had no intention of executing a formal arrest.

As a threshold matter, the approach most consistent with Gant and with

Justice Scalia‘s Thornton concurrence—a bright-line rule that Gant vehicle

searches are impermissible absent a completed formal arrest—would require no

inquiry into the subjective intent of the officers. It is solely by virtue of the

majority‘s insistence that the Rawlings logic apply to Gant searches that this

question of subjective inquiry arises at all.

And in any event, the majority misapprehends Supreme Court precedent like

Whren and Ashcroft v. Al-Kidd, 563 U.S. 731 (2011), in insisting that the Court‘s

―repeated[] reject[ion]‖ of ―a subjective approach,‖ ante at 16-17 (quoting

Fernandez v. California, 134 S. Ct. 1126, 1134 (2014)), bolsters the conclusion

that pre-arrest searches are permissible where officers subjectively do not intend to
 71

arrest. As the Supreme Court made clear in Florida v. Jardines, ―those cases

merely hold that a stop or search that is objectively reasonable is not vitiated by

the fact that the officer‘s real reason for making the stop or search has nothing to

do with the validating reason.‖ 133 S. Ct. 1409, 1416 (2013). Thus in Whren, the

Court held that a traffic stop based upon probable cause for a traffic offense, even

where the officer might have an invidious subjective motivation for the stop, is

lawful under the Fourth Amendment. Whren, 517 U.S. at 813 (―Subjective

intentions play no role in ordinary, probable-cause Fourth Amendment analysis.‖);

see also Jardines, 133 S. Ct. at 1416 (―defendant will not be heard to complain that

although he was speeding the officer‘s real reason for the stop was racial

harassment‖); Devenpeck v. Alford, 543 U.S. 146, 153 (2004) (―arresting officer‘s

state of mind (except for the facts that he knows) is irrelevant to the existence of

probable cause‖); Reid, 26 N.E.3d at 240 (―[Whren and Devenpeck] hold that a

stop or arrest is valid where it is supported by the necessary level of suspicion or

probable cause, whatever the actual motive for the officer‘s action.‖).

But there is a difference between a defendant‘s delving into an officer‘s

subjective intent to invalidate a search that is clearly otherwise legal under
 72

objective, existing warrant exceptions15 and a defendant‘s insisting that, if the

government wishes to deviate from objective, existing warrant exceptions, it

should do so only where that subjective intent aligns with the rationales underlying

such existing exceptions. In a Whren-like situation, the government might

reasonably be concerned that a valid traffic stop, objectively justified under well-

established Supreme Court precedent, would be invalidated because of the vagaries

of trial courts‘ credibility assessments about officers‘ subjective intentions in

stopping a motorist. But here, the search of Mr. Lewis‘s vehicle was not

objectively justified under existing and well-established exceptions to the warrant

requirement, and thus the concerns expressed in Whren and its progeny do not

apply with equal force. It is the government, not the defense, that hopes to

persuade the court to deviate from the traditional requirement of a completed

formal arrest to justify a search.16 To do so, the government should have to show

15
See Brigham City v. Stuart, 547 U.S. 398, 403–06 (2006) (officers‘
subjective motivations in entering a house were irrelevant where police were
providing assistance under the emergency aid exception to the warrant
requirement).
16
Counsel for the government actually appeared to concede at oral
argument that the subjective intent of the officer will sometimes be relevant to
determining the legality of a search when he acknowledged that he would not want
the judge to exclude an officer‘s testimony that he had made a definite decision to
arrest.
 73

that the arrest was under way, even if that requires a subjective inquiry into officer

intent.

Any subjective inquiry into an officer‘s intent to arrest, moreover, would

rarely require the type of extensive examination into an officer‘s state of mind that

the Supreme Court has generally guarded against in the Fourth Amendment

context.17 In the vast majority of cases involving searches incident to arrest, that

an officer subjectively intended to arrest a suspect will be readily discernible based

on the objective circumstances of the encounter. See, e.g., Jardines, 133 S. Ct. at

1416–17 (analyzing whether officers had ―a purpose to conduct a search‖ based on

what ―their behavior objectively reveals‖); Devenpeck, 543 U.S. at 154 (arresting

officer‘s ―subjective intent is always determined by objective means‖). Before Mr.

Rawlings was ―placed [] under formal arrest,‖ for example, he was detained for

forty-five minutes, was administered Miranda warnings, and had confessed to
17
Even if it did, at least one case binding on this court examined an
officer‘s subjective intent in the form of his trial testimony in concluding that a
search was unreasonable under the Fourth Amendment. See White v. United
States, 271 F.2d 829, 831 (D.C. Cir. 1959) (―According to the officer‘s own
testimony . . . the search did not turn upon an arrest for vagrancy, and to attribute it
to such an arrest would be to take an unrealistic view of the evidence. The arrest
for vagrancy was incidental to the search, rather than the converse. . . . We have
more than once excluded evidence obtained by a search which in truth was not
incidental to an arrest, but when in fact the arrest was incidental to a search. This
is such a case.‖ (citations omitted)).
 74

possessing drugs found in a companion‘s purse—circumstances objectively

indicating the officers‘ intent to arrest him. See Rawlings, 448 U.S. at 100–01.

The officers‘ intent to arrest was similarly evident in Millet v. United States, 977

A.2d 932 (D.C. 2009), where, by the time the police searched Mr. Millet, officers

had already lawfully recovered a bag of marijuana from his car, had arrested Mr.

Fountain, the driver, for driving under the influence of marijuana, and had ―told

Millet and Fountain that they would both be charged in connection with the

marijuana found in the bag.‖ Id. at 934.18

18
The majority relies heavily on broad statements from cases such as
Devenpeck that subjective intent has no place in a Fourth Amendment setting. And
yet the Supreme Court has expressly endorsed subjective inquiries into an officer‘s
state of mind in certain Fourth Amendment contexts, including inquiries that seem
very similar to the question of an officer‘s intent to arrest. In Murray v. United
States, 487 U.S. 533 (1988), for example, the Court held that under the
independent source doctrine, an unlawful warrantless entry into a warehouse would
not require suppression of evidence obtained during a subsequent warrant-based
search of the warehouse if the government could establish (1) that the warrant was
based on information independent of what the officers saw during the warrantless
entry and (2) that the officers would have sought a warrant even if they had not
previously entered the warehouse. Id. at 541–43; see also id. at 547 (Marshall, J.,
dissenting) (stating that the majority opinion in Murray ―makes the application of
the independent source exception turn entirely on an evaluation of the officers‘
intent‖); United States v. Jadlowe, 628 F.3d 1, 9–10 (1st Cir. 2010) (in
independent-source case where officer testified that he ―would have sought a
warrant even had the officers not seen the bricks of cocaine in the garage,‖ the
court explained that the ―police officers‘ subjective intent to seek a warrant is a
factual determination subject to clear error review‖); United States v. Restrepo, 966
F.2d 964, 972 (5th Cir. 1992) (explaining that ―the core judicial inquiry before the
(continued…)
 75

Of course, discerning an officer‘s subjective intent to arrest may in some

cases prove less clear-cut, and officers may be understandably reluctant to subject

themselves and the validity of their searches and seizures to the uncertainty of trial

courts‘ credibility determinations. In such instances, officers have a

straightforward and objectively verifiable alternative available to them: they can

make the arrest.

IV.

The rule the majority crafts for evaluating the constitutional validity of a

pre-arrest Gant vehicle search is ostensibly narrow, but its five-part test cannot

mask what is, at bottom, a profound departure from the essential moorings of the

Supreme Court‘s well-established exceptions to the warrant requirement for both

Chimel-style searches incident to arrest and Gant vehicle searches alike. The

majority opinion is no mere application of Gant to a new set of facts. On the

(…continued)
district court on remand is a subjective one: whether information gained in the
illegal search prompted the officers to seek a warrant to search [the residence]‖).
 76

contrary, in upholding the pre-arrest search in this case under Gant even in the

absence of any indication the police intended to arrest Ms. Gibbs before the search,

the majority spurns Justice Scalia‘s unmistakable view that such an exception to

the warrant requirement is justified by the ―fact of prior lawful arrest.‖ 541 U.S. at

630. The weakness of the majority‘s holding in this regard is underscored by its

unsupportable and largely unexplained contention that the Gant Court, while

expressly adopting the exception Justice Scalia proposed in Thornton, nevertheless

rejected its analytic roots.

There are undeniable costs to expanding existing exceptions to the warrant

requirement beyond their well-established bounds. Most notably, investigative

searches will become more commonplace. Such searches, as the Supreme Court

has repeatedly emphasized, ―implicate[] the central concern underlying the Fourth

Amendment—the concern about giving police officers unbridled discretion to

rummage at will among a person‘s private effects.‖ Gant, 556 U.S. at 345. With

its new anomalous conception of the Gant vehicle search, the majority opinion

downplays this central concern, discarding a key limiting principle in the

foundation of Gant itself that is designed to address precisely this problem—―the

fact of prior lawful arrest.‖ Thornton, 541 U.S. at 630.
 77

I respectfully dissent.