PRECEDENTIAL
UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT
________________
No. 16-3547
________________
UNITED STATES OF AMERICA
v.
CARLTON WILLIAMS,
Appellant
________________
On Appeal from the United States District Court
for the Western District of Pennsylvania
(D. C. Criminal No. 2-14-cr-00030-001)
District Judge: Honorable Terrence F. McVerry
________________
Argued on May 24, 2017
Before: HARDIMAN, ROTH and FISHER, Circuit Judges
(Opinion filed: August 1, 2018)
Kimberly R. Brunson, Esq. (ARGUED)
Office of Federal Public Defender
1001 Liberty Avenue
1500 Liberty Center
Pittsburgh, PA 15222
Counsel for Appellant
Donovan J. Cocas, Esq. (ARGUED)
Rebecca R. Haywood, Esq.
Michael L. Ivory, Esq.
Office of United States Attorney
700 Grant Street
Suite 4000
Pittsburgh, PA 15219
Counsel for Appellee
________________
OPINION
________________
ROTH, Circuit Judge
During an investigation, federal law enforcement
officials learned that Carlton Williams was involved in the
distribution of heroin. The investigation involved
surveillance of Williams’s activity, which eventually led to a
stop of his car. During the traffic stop, law enforcement
officials conducted a search of Williams’s car and its
contents. As they expected, the officials discovered drugs
2
during the search. Williams subsequently pleaded guilty to
possession of heroin with intent to distribute in violation of
federal drug laws. Williams now appeals the denial of his
suppression motion and application of the United States
Sentencing Guidelines’ career offender enhancement.
Finding no merit in either claim, we will affirm Williams’s
conviction and sentence.
I.
A. Factual Background
The underlying facts are uncontested. During an
investigation that began as early as November 2012, a Drug
Enforcement Administration task force officer learned that
Williams bought heroin in Detroit, Michigan, which he
packaged and sold in Pittsburgh, Pennsylvania. The officer
subsequently placed a GPS tracker on Williams’s car and
monitored his movements for approximately one month. On
January 11, 2013, data from the GPS tracker indicated that
Williams’s car was driven to Detroit. Suspecting that
Williams drove his car to Detroit to retrieve heroin, the task
force officer organized a plan to have Williams’s car stopped
upon its return to Pennsylvania. Pennsylvania State Police
trooper Michael Volk effectuated the traffic stop.
Later that same evening, Trooper Volk observed
Williams’s car speeding and stopped it. The trooper issued a
citation for the traffic violation and told Williams that he was
free to go. Before Williams left, however, Trooper Volk
asked Williams for consent to search his car. Williams
agreed and signed a consent to search form labeled “Waiver
of Rights and Consent to Search.” The parties do not dispute
3
that Williams knowingly, intelligently, and voluntarily
consented to the search of his car, its contents, and his person.
Trooper Volk, with the help of other troopers,
commenced a search of Williams’s car that lasted for
approximately seventy-one minutes. The troopers searched
every part of the car, including its passenger compartment,
trunk, and undercarriage. Unable to locate any narcotics,
Trooper Volk requested the assistance of a narcotics-detection
dog. Shortly thereafter, Trooper Volk updated another
trooper on the progress of the search and indicated that “[the
search] was going to take awhile [because] he hadn’t found
[the heroin], but the K-9 was on its way coming from a
distance.”1
Williams eventually became less patient and told
Trooper Volk “you searched my car three times, now you
hold me up and I have to go.”2 According to Williams, he
made this statement in only “a regular tone of voice that he
expected Trooper Volk to hear but [the trooper] was at a
distance and there was a lot of noise from the turnpike traffic
and the wind.”3 Other than Williams’s own testimony, there
was no evidence that Trooper Volk heard his alleged protest.
The District Court, as a result, found Williams’s testimony
“only credible to a degree.”4
The troopers continued their search despite Williams’s
irritation. As the search continued, Williams requested five
1
United States v. Williams, 2015 WL 5602617, at *6 n.5
(W.D. Pa. 2015).
2
Id. at *6.
3
Id.
4
Id.
4
items from his car, including his two cellular phones. One of
the troopers retrieved Williams’s cellular phones and
attempted to search them before handing them over to
Williams. The trooper was able to read the text messages
contained on only one of the devices because the other device
was password-protected. The trooper who read Williams’s
text messages told Trooper Volk that the messages suggested
that Williams had “something.”5 When Williams was
confronted about the text messages, he warned the officers
that they could not search his phone without a warrant.
The search of the car continued. After fifty-one
minutes, the troopers had not discovered any drugs. They
began to disassemble Williams’s sound system speakers.
Williams objected that the troopers were not permitted to
search his speakers without a warrant. Trooper Volk told
Williams to “relax,” to which Williams replied, “I’ve been
out here half an hour, man.”6 Upon Williams’s protest,
Trooper Volk reassembled the car’s speakers but otherwise
continued searching the vehicle. Soon after, and seventy-one
minutes into the search, Trooper Volk discovered thirty-nine
grams of heroin in a sleeve covering the car’s parking brake
lever. Williams was immediately arrested.
B. Procedural History
Williams was charged with possession of heroin with
intent to distribute, in violation of 21 U.S.C. §§ 841(a)(1) and
841(b)(1)(C). He filed a number of pretrial motions,
including a motion to suppress the evidence seized from his
5
App. 222.
6
Williams, 2015 WL 5602617, at *6.
5
car. Following a two-day hearing and the submission of post-
hearing briefing, the District Court denied Williams’s
suppression motion, because it concluded that Williams had
voluntarily consented to the search and had not unequivocally
withdrawn his consent during the search.
Prior to Williams’s sentencing, the United States
Probation Office prepared a Presentence Investigation Report
(PSR), which the District Court adopted without change. The
sentencing range calculation included U.S.S.G. § 4B1.1’s
career offender enhancement because the District Court
concluded that Williams had two prior convictions for
controlled substance offenses: a 2007 conviction for
possession with intent to distribute heroin and a 1998
conviction under 18 U.S.C. § 1962(c) and (d) of the
Racketeer Influenced and Corrupt Organizations Act (RICO).
Williams admitted to various predicate acts forming the basis
for his § 1962 RICO conviction, all of which were for
possession with intent to distribute either crack cocaine or
heroin in violation of 21 U.S.C. §§ 841(a)(1), 841(b)(1)(C),
841(b)(1)(B)(iii), and 846.7 As a result of the career offender
enhancement, Williams faced a Guidelines sentencing range
of 210-262 months. On May 11, 2016, Williams entered a
conditional guilty plea, preserving his right to appeal the
denial of his suppression motion and the application of the
Guidelines’ career offender designation. Williams was
sentenced to, inter alia, a term of 160 months’ imprisonment.
This appeal followed.
Williams appeals both the denial of his suppression
motion and the District Court’s application of the Guidelines’
7
Supp. App. 19-25.
6
career offender designation. The District Court had
jurisdiction pursuant to 18 U.S.C. § 3231. We exercise
appellate jurisdiction under 28 U.S.C. § 1291 and 18 U.S.C. §
3742(a).8
II.
“We review the District Court’s denial of a motion to
suppress for clear error as to the underlying factual findings,
and we exercise plenary review of its application of the law to
those facts.”9 “‘A finding is clearly erroneous when although
there is evidence to support it, the reviewing court on the
entire evidence is left with the definite and firm conviction
that a mistake has been committed.’”10 Therefore, “‘[i]f the
[D]istrict [C]ourt’s account of the evidence is plausible in
light of the record viewed in its entirety,’ we will not reverse
it even if, as the trier of fact, we would have weighed the
evidence differently.”11
A. The District Court Properly Denied Williams’s
Motion to Suppress
With respect to his suppression motion, Williams
claims that the District Court erred in denying his suppression
8
United States v. Johnson, 587 F.3d 203, 207 (3d Cir. 2009).
9
United States v. Bansal, 663 F.3d 634, 651-52 (3d Cir.
2011) (citing United States v. Perez, 280 F.3d 318, 336 (3d
Cir. 2002)).
10
United States v. Price, 558 F.3d 270, 276-77 (3d Cir. 2009)
(quoting United States v. Pelullo, 173 F.3d 131, 135 (3d Cir.
1999)).
11
Price, 558 F.3d at 277 (quoting Anderson v. City of
Bessemer City, 470 U.S. 564, 573-74 (1985)).
7
motion because he properly withdrew his consent to the
search or was improperly prevented from doing so.
It is well settled that the Fourth Amendment protects
suspects from unreasonable searches.12 “[A] search
conducted without a warrant issued upon probable cause is
[presumptively] unreasonable . . . subject only to a few
specifically established and well-delineated exceptions.”13 A
search conducted with consent is one such “established
exception.”14 The appellant concedes that the search here
began as a consensual one. He contends, however, that the
search ceased to be so when he withdrew his consent or was
prevented from doing so. Before reaching the issue of
whether Williams withdrew his consent in this case, we must
first determine whether the Fourth Amendment allows the
subject of a consensual search to terminate the search by
withdrawing his consent. Neither this Court nor the
Supreme Court has expressly established that the subject of a
consensual search may withdraw consent that he has
voluntarily given. The Supreme Court, however, has
recognized that a person may “delimit as he chooses the
scope of the search to which he consents.”15 In so holding,
the Court has instructed that the standard for measuring the
limitations placed on a consensual search “is that of objective
reasonableness.”16 Thus, in determining the legal bounds of a
12
U.S. Const. Amend. IV.
13
Schneckloth v. Bustamonte, 412 U.S. 218, 219 (1973)
(internal quotation marks omitted) (citing Katz v. United
States, 389 U.S. 347, 357 (1967)).
14
Id. at 219 (citations omitted).
15
Florida v. Jimeno, 500 U.S. 248, 252 (1991).
16
Id. at 251 (internal quotation marks omitted).
8
consensual search, we must determine “what would the
typical reasonable person have understood by the exchange
between the officer and the suspect.”17 Relying on Florida v.
Jimeno’s recognition that a consensual search may be
restricted by individuals, our sister circuits that have
considered whether individuals may withdraw consent to
search have unanimously answered in the affirmative.18
Today, we join them.
Although the Supreme Court has not itself expressly
held that the subject of a consensual search may terminate the
search by withdrawing his consent, considerable support for
such a proposition is easily found in its Fourth Amendment
17
See Jimeno, 500 U.S. at 251 (applying a reasonable person
standard for determining the scope of consent).
18
See United States v. Dyer, 784 F.2d 812, 816 (7th Cir.
1986) (“Clearly a person may limit or withdraw his consent to
a search, and the police must honor such limitations.”);
Painter v. Robertson, 185 F.3d 557, 567 (6th Cir. 1999)
(“[T]he consenting party may limit the scope of that search,
and hence at any moment may retract his consent”); United
States v. Sanders, 424 F.3d 768, 774 (8th Cir. 2005) (“Once
given, consent to search may be withdrawn[.]”); United States
v. McWeeney, 454 F.3d 1030, 1034 (9th Cir. 2006) (“A
suspect is free, however, after initially giving consent, to
delimit or withdraw his or her consent at anytime.”); see also
United States v. Pelle, No. 05-407, 2006 WL 436920, at *4
(D.N.J. Feb. 17, 2006) (“The courts which have decided the
issue, however, have unanimously answered that question in
the affirmative, generally holding that any such withdrawal
must be supported by unambiguous acts or unequivocal
statements.”) (collecting cases).
9
jurisprudence. The Court recognized in Walter v. United
States,19 and later in Jimeno,20 that a consensual search
satisfies the mandates of the Constitution only if conducted
within the boundaries of the consent given. This recognition
establishes that it is the subject of a consensual search who
decides the terms of the search. Although Walter and Jimeno
expressly consider only a party’s right to limit the particular
things officials may search, nothing in those opinions
suggests that consent, which waives Fourth Amendment
rights, cannot otherwise be narrowed, qualified, or
withdrawn. That a party may terminate a search by
withdrawing his consent is a corollary of the recognition that
the subject of a consensual search determines the parameters
of that search.
Moreover, recognition of a party’s right to take away
the consent that he or she has conferred advances society’s
interest in promoting consensual searches. The Supreme
Court has acknowledged that consensual searches are
important because they promote the effective enforcement of
criminal laws.21 This is particularly true where there is lack
of probable cause to arrest or search because, in such
19
447 U.S. 649, 656 (1980) (“When an official search is
properly authorized—whether by consent or by the issuance
of a valid warrant—the scope of the search is limited by the
terms of its authorization.”).
20
500 U.S. at 252.
21
Schneckloth, 412 U.S. at 228 (“[A] search pursuant to
consent may result in considerably less inconvenience for the
subject of the search, and, properly conducted, is a
constitutionally permissible and wholly legitimate aspect of
effective police activity.”).
10
situations, “a search authorized by a valid consent may be the
only means of obtaining important and reliable evidence.”22
Moreover, a rule restricting the ability to withdraw consent
would likely discourage people from consenting to searches
when they otherwise might have done so. In the present case,
for example, Williams voluntarily authorized the troopers to
conduct a search. He then admonished the troopers that the
search of his speakers and electronic devices was not within
the bounds of his authorization. As a result, the troopers
reassembled the speakers and ceased examining the phone
that was not password-protected. However, “where a suspect
does not withdraw his valid consent to a search for illegal
substances before they are discovered, the consent remains
valid and the substances are admissible as evidence.”23
Turning to the merits of this case, we must decide
whether Williams actually withdrew his consent. As the
parties note, “the ultimate touchstone of the Fourth
Amendment is ‘reasonableness.’”24 Thus, in determining
whether suspects have withdrawn their consent to a search,
courts have been guided by how a reasonable person would
have understood the exchange between law enforcement
22
Id. at 227 (citation omitted).
23
Dyer, 784 F.2d at 816.
24
Brigham City v. Stuart, 547 U.S. 398, 403 (2006) (citations
omitted).
11
officers and suspects.25 Courts agree that a reasonable person
would not understand certain equivocal acts or statements to
convey a suspect’s desire to withdraw consent that he has
voluntarily conferred.26 Ambiguous acts and statements do
not ordinarily lend themselves to a conclusive determination
of whether consent has been withdrawn. Once it has been
established that a suspect has voluntarily consented to a
search, it is his burden to demonstrate that he has withdrawn
that consent by pointing to an act or statement that an
objective viewer would understand as an expression of his
desire to no longer be searched.
With these principles in mind, we hold that the
circumstances here do not demonstrate that Williams
withdrew his consent to the troopers’ search of his car.
Williams knew how to express the absence of consent to
search. As the record demonstrates, Williams told the
troopers that they did not have consent to search his speakers
25
See e.g., United States v. Martel-Martines, 988 F.2d 855,
858 (8th Cir. 1993); see also Jimeno 500 U.S. at 251 (“The
standard for measuring the scope of a suspect’s consent under
the Fourth Amendment is that of ‘objective’ reasonableness—
what would the typical reasonable person have understood by
the exchange between the officer and the suspect?” (citations
omitted)).
26
See, e.g., United States v. $304,980.00 in U.S. Currency,
732 F.3d 812, 820 (7th Cir. 2013) (“[P]olice officers do not
act unreasonably by failing to halt their search every time a
consenting suspect equivocates.”); Martel-Martines, 988 F.2d
at 858 (requiring “‘unequivocal act or statement of
withdrawal’” (quoting United States v. Alfaro, 935 F.2d 64,
67 (5th Cir. 1991)).
12
or his cellular phones. The search of those areas then
stopped.
Williams also argues that he conveyed withdrawal of
his consent to search the car when he complained that he had
been standing “out [there] half an hour” and after he told
officer Volk “you searched my car three times [and] y’all got
me on the side of this road in the middle of the winter holding
me up and I got to go.”27 The District Court held that
Williams’s comments only “constituted manifestations of
irritation” and not statements indicating that he was
withdrawing the consent he had conferred.28 We agree.
Although defendants need not use a special set of words to
withdraw consent, they must do more than express
unhappiness about the search to which they consented.
Other courts have reached the same conclusion when
presented with similar facts. For example, the Eighth Circuit
Court of Appeals in United States v. Gray held that a suspect
had not withdrawn consent simply by objecting that the
search was “ridiculous” and that he was “ready to go.”29 The
court held that such statements amounted only to
“expressions of impatience.”30 The court warned that
“protests about the length of time the search was taking
without any specific request to leave did not under the
circumstances” amount to a withdrawal of consent.31
27
Williams, 2015 WL 5602617, at *6; App. 196, 198-99.
28
Williams, 2015 WL 5602617 at *9.
29
369 F.3d 1024, 1026 (8th Cir. 2004).
30
Id.
31
U.S. v. Sanders, 424 F.3d 768, 774 (8th Cir. 2005) (citing
Gray, 369 F.3d at 1026).
13
Similarly, Williams’s statements here were expressions of
frustration. Williams falls short of meeting his burden of
proof to establish that his consent was withdrawn.
Williams alternatively contends that, even if he did not
withdraw his consent to the search, the evidence should be
suppressed because the “coercive” nature of the search
prevented him from revoking consent.
The Fourth Amendment requires that consent not be
coerced.32 The question of whether Williams’s consent was
at any point the product of coercion is “a question of fact
determined from the totality of the circumstances.”33 In
assessing the voluntariness of a suspect’s consent, we
consider “the age, education, and intelligence of the subject;
whether the subject was advised of his or her constitutional
rights; the length of the encounter; the repetition or duration
of the questioning; and the use of physical punishment.”34
Our analysis “must accord the district court’s conclusion that
[Williams]’s consent was [voluntary] great deference, unless
our examination of the record shows that the district court
committed clear error.”35 Thus, the District Court’s finding
that Williams’s consent was voluntary will not be overturned
unless it is “(1) completely devoid of minimum evidentiary
32
Schneckloth, 412 U.S. at 228.
33
United States v. Antoon, 933 F.2d 200, 203 (3d Cir. 1991).
34
Price, 558 F.3d at 278 (citing Schneckloth, 412 U.S. at 226;
United States v. Kim, 27 F.3d 947, 955 (3d Cir. 1994)).
35
Antoon, 933 F.2d at 204 (citation omitted).
14
support displaying some hue of credibility, or (2) bears no
rational relationship to the supportive evidentiary data.”36
Our assessment of the totality of the circumstances
precludes us from concluding that the District Court
committed clear error. As the District Court noted,
Williams’s interaction with the troopers was not hostile. The
troopers neither made threats nor showed force. No restraints
were employed at the time of the search. The District Court’s
finding that Williams exhibited his ability to intelligently
delimit the scope of the search is supported by the record.
Accordingly, the District Court did not err in finding that,
throughout the entire encounter, Williams’s grant of consent
was not the product of coercion.
B. The District Court Properly Applied the
Guidelines’ Career Offender Enhancement
Williams next appeals his career offender designation,
arguing that his 1998 RICO conviction—predicated on his
distribution of heroin and crack cocaine—was not a requisite
“controlled substance offense” under the Sentencing
Guidelines. We disagree.
Under the Sentencing Guidelines, a defendant must be
sentenced as a “career offender” if: (1) he was at least
eighteen years old when he committed the instant offense of
conviction; (2) the instant offense is a felony crime of
violence or controlled substance offense; and (3) he has at
36
Id. (quoting Krasnov v. Dinan, 465 F.2d 1298, 1302 (3d
Cir.1972)).
15
least two prior felony convictions for a crime of violence or
controlled substance offense.37
In this case, there is no dispute that the instant
offense—possession with intent to distribute heroin in
violation of §§ 841(a)(1) and 841(b)(1)(C)—is a controlled
substance offense. Nor is there any doubt that Williams was
at least eighteen at the time. The parties agree that
Williams’s 2007 conviction for possession with intent to
distribute heroin supplies one of the two required prior felony
convictions. The 1998 RICO conviction, we now hold,
supplies the second.
Ordinarily, to determine whether a prior conviction
qualifies as a crime of violence or controlled substance
offense, we apply a categorical approach.38 We consider only
the elements of the crime of conviction and assess whether
they fall within the bounds of a crime of violence or
controlled substance offense, as defined under the
Guidelines.39 To avoid the “practical difficulties and
potential unfairness” inherent in “determining the precise
facts underlying a defendant’s [prior] conviction,” which may
have occurred years or decades ago, we do not excavate or
dissect the underlying factual record.40
37
U.S.S.G. § 4B1.1(a).
38
Taylor v. United States, 495 U.S. 575, 588 (1990).
39
See, e.g., United States v. Chapman, 866 F.3d 129, 133 (3d
Cir. 2017).
40
United States v. Robinson, 844 F.3d 137, 142 (3d Cir.
2016).
16
There is an exception, however. When a crime is
defined with alternative elements, we may review a limited
set of documents—including the indictment and plea
colloquy, among others—but only to determine which version
of the statute formed the basis of the prior conviction.41 Such
a statute is termed “divisible” and this approach—a more
record-invasive variant of the categorical approach—is called
the “modified categorical approach.”
RICO, in particular Section 1962(c), is one such
divisible statute. That statutory subsection, the basis for
Williams’s 1998 RICO conviction, proscribes “conduct[ing] .
. . [an] enterprise’s affairs through a pattern of racketeering
activity or collection of unlawful debt.” It proscribes two
alternative forms of conduct: either racketeering activity or
the collection of unlawful debt. That fork in the statute has
even more branches. “Racketeering activity,” a statutory
phrase without independent meaning, has “constituent parts”
or alternative “elements” that need to be proven beyond a
reasonable doubt to sustain a conviction.42 Under RICO,
those elements are known as “predicate acts” and include
certain violations of federal law, including “fraud connected
with a case under title 11,” or “fraud in the sale of securities,”
or “the felonious manufacture, importation, receiving,
concealment, buying, selling, or otherwise dealing in a
41
Descamps v. United States, 570 U.S. 254, 261-62 (2013)
(reiterating that, for purposes of this inquiry, we may not
examine a defendant’s prior conduct).
42
Mathis v. United States, 136 S. Ct. 2243, 2248 (2016); see
also 3d Cir. Model Crim. Jury Instr. 6.18.1962C-6 (2018)
(establishing that the government must prove predicate acts
beyond a reasonable doubt to sustain a conviction).
17
controlled substance or listed chemical.”43 Without
consulting the record, we would not know which of these
multiple alternatives yielded Williams’s prior RICO
conviction.
Fortunately, because Section 1962(c) is divisible, we
may consult select portions of the record under the modified
categorical approach to make that determination. The
superseding indictment and Williams’s 1998 plea colloquy
are illuminating. They reveal that Williams pleaded guilty to
a RICO violation under Section 1962(c) and five underlying
RICO predicate acts.44 All five of those predicate acts of
racketeering were violations of 21 U.S.C. § 841(a)(1)—or
conspiracy to commit such a violation under 21 U.S.C. §
846.45 Specifically, he admitted to “manufactur[ing],
distribut[ing], or dispens[ing], or possess[ing] with intent to
manufacture, distribute, or dispense” heroin or crack
cocaine.46 Without probing the record further or examining
Williams’s prior conduct, we now know that Williams’s prior
RICO conviction necessarily implicated only a limited
portion of 18 U.S.C. § 1961(1)(D), namely, only the
“felonious manufacture,” or “recei[pt],” or “buying, selling,
or otherwise dealing in a controlled substance.”47 This
limited and non-fact-intensive review of the record unmasks
the specific version of the RICO statute under which
Williams was convicted: “conduct[ing] . . . [an] enterprise’s
affairs” through “a pattern of racketeering activity” by
43
18 U.S.C. § 1961(1)(D).
44
App. 324.
45
Supp. App. 19-25.
46
21 U.S.C. § 841(a)(1).
47
18 U.S.C. § 1961(1)(D)
18
“felonious[ly] manufactur[ing],” or “receiving,” or “buying,
selling, or otherwise dealing in a controlled substance or
listed chemical.”48
The final step in this analysis is to assess whether the
offense of conviction—as decoded by this selective review of
the record—sweeps any more broadly than the relevant
generic offense,49 in this case a “controlled substance
offense” as defined in the Guidelines. Section 4B1.2 of the
Guidelines defines a “controlled substance offense” as “the
manufacture, import, export, distribution, or dispensing of a
controlled substance (or a counterfeit substance) or the
possession of a controlled substance (or a counterfeit
substance) with intent to manufacture, import, export,
distribute, or dispense.”50 The specific version of RICO
implicated by Williams’s prior conviction encompasses only
the “felonious manufacture,” or “recei[pt],” or “buying,
selling, or otherwise dealing in a controlled substance”: It is
categorically a subset of the Guidelines’ definition of a
“controlled substance offense.” For that reason, Williams’s
prior RICO conviction was a “controlled substance offense”
under the Guidelines.
Because both his 2007 heroin distribution conviction
and his 1998 RICO conviction were prior felony convictions
for controlled substance offenses, the District Court correctly
applied the career offender enhancement to Williams.
48
Id. §§ 1961(1)(D), 1962(c).
49
Mathis, 136 S. Ct. at 2249.
50
U.S.S.G. § 4B1.2(b).
19
III.
For the foregoing reasons, we will affirm the judgment
of the District Court.
20
HARDIMAN, Circuit Judge, concurring in part and
concurring in the judgment.
I agree with my colleagues that the District Court did
not err when it denied Williams’s motion to suppress evidence.
I also agree that Williams is—as the District Court found—a
career offender under § 4B1.1 of the United States Sentencing
Guidelines (2015) (USSG). As to that second issue, I concur in
the judgment only because I cannot subscribe to the Majority’s
modified categorical approach, which I believe misapplies the
Supreme Court’s decisions in Taylor v. United States, 495 U.S.
575 (1990), and Mathis v. United States, 136 S. Ct. 2243
(2016). But because a proper application of the modified
categorical approach would yield absurd results in cases
involving RICO predicate offenses, I am convinced that the
Supreme Court would not apply it here. Accordingly, I agree
with my colleagues that Williams is a career offender.
At the outset, it’s important to note that the Supreme
Court has not yet applied Taylor (or Mathis) in a case involving
a RICO predicate offense. And although some of our sister
courts have adjudicated cases involving the interplay between
RICO and the § 4B1.1 career offender guideline, they have not
settled on a consistent mode of analysis. For example, the
Ninth Circuit placed “the focus of the inquiry . . . on
the conduct for which [the defendant] was convicted” without
mentioning the categorical approach or citing Taylor. United
States v. Scott, 642 F.3d 791, 801 (9th Cir. 2011) (per curiam).
The First Circuit has taken a different tack, explaining that in
determining whether a RICO conviction counts toward the
career offender enhancement, courts should “in fidelity to
Taylor principles . . . merely assess the nature and object of the
racketeering activity as described in the indictment and fleshed
1
out in the jury instructions.” United States v. Winter, 22 F.3d
15, 19–21 (1st Cir. 1994). Like the First Circuit, a panel of the
Eleventh Circuit professed fealty to Taylor, but it looked to
“the facts to which [the defendant] stipulated” in comparing
the conduct underlying the defendant’s prior racketeering
conviction with the definition of a “controlled substance
offense” under USSG § 4B1.2(b). United States v. Rosquete,
208 F. App’x 737, 739–41 (11th Cir. 2006) (per curiam).
Here, my colleagues have chosen to follow the path
marked by the Supreme Court in Taylor and Mathis. And if the
Taylor/Mathis framework applies to this case, the Majority is
quite right that the relevant statute (18 U.S.C. § 1962(c)) is
divisible, which requires application of the modified
categorical approach.
But the modified categorical approach yields a result
contrary to the one the Majority reaches. Section 1961(1)(D),
which specifies the type of racketeering activity Williams was
engaged in, is not “categorically a subset of the Guidelines’
definition of a ‘controlled substance offense.’” Maj. Op. 19.
Under Guidelines § 4B1.2(b), a “controlled substance offense”
encompasses “the manufacture, import, export, distribution, or
dispensing of a controlled substance (or a counterfeit
substance) or the possession of a controlled substance (or a
counterfeit substance) with intent to manufacture, import,
export, distribute, or dispense.” That definition differs from
Williams’s RICO conviction, which involved “the felonious
manufacture, importation, receiving, concealment, buying,
selling, or otherwise dealing in a controlled substance or listed
chemical (as defined in section 102 of the Controlled
Substances Act), punishable under any law of the United
States.” 18 U.S.C. § 1961(1)(D). A comparison of the two
provisions makes clear that Williams’s RICO offense
2
encompasses conduct that § 4B1.2(b) does not cover, such as
“receiving, concealment, buying . . . or otherwise dealing in a
controlled substance.” See id. Because it “sweeps more broadly
than the generic crime,” Williams’s RICO conviction is not a
qualifying offense under the modified categorical approach.
See Descamps v. United States, 570 U.S. 254, 261 (2013).
Would the Supreme Court really conclude that
Williams’s RICO conviction did not constitute a “controlled
substance offense”? I think not. The predicate acts underlying
Williams’s conviction included the distribution of and
possession with intent to distribute: (1) in excess of a kilogram
of heroin; (2) in excess of 50 grams of cocaine base; (3) in
excess of 5 grams of cocaine base; (4) less than 5 grams of
cocaine base; and (5) less than 100 grams of heroin. The
enumeration of these predicate acts plainly establishes that
Williams’s RICO conviction is for a controlled substance
offense.
To hold that it is not defies common sense not only in
this case, but in any RICO case predicated on federal drug
crimes. This is so because in every such case the “element” that
Taylor and Mathis require us to compare to USSG § 4B1.2(b)
will be the same: 18 U.S.C. § 1961(1)(D). See Maj. Op. 18–19.
An application of the modified categorical approach will thus
generate the same nonsensical answer—that a RICO
conviction based on controlled substance offenses is not a
“controlled substance offense”—every time.
I cannot accept that Congress, the United States
Sentencing Commission, or the Supreme Court would endorse
such an absurd result. Accordingly, I would hold that the
approach the Court has articulated in cases like Taylor,
Descamps, and Mathis does not apply here. The categorical
3
approach was developed to ensure that federal defendants who
have committed essentially the same crimes in the past don’t
receive disparate sentences merely because they committed
those prior offenses in different states. See Taylor, 495 U.S. at
591–92. That policy justification has no relevance here, where
the nature of the prior federal conviction is clear on the face of
the docket.
Were the Supreme Court confronted with the question
before us, I think it would not attempt to pound the square peg
of RICO into the round hole of the categorical/modified
categorical approach. It would be especially surprising for the
Court to do so not only because the predicate offense at issue
here is markedly different from the state burglary crimes at
issue in Taylor and Mathis, but also because several Justices
have expressed dissatisfaction with the categorical approach
generally.1 For these reasons, I would not apply the categorical
approach of Taylor and Mathis to the RICO offense at issue
here. Instead, I would hold that Williams’s RICO conviction
1
See, e.g., Mathis, 136 S. Ct. at 2258 (Kennedy, J.,
concurring) (“[T]oday’s decision is a stark illustration of the
arbitrary and inequitable results produced by applying an
elements based approach to this sentencing scheme.”); id. at
2268 (Alito, J., dissenting) (categorical approach “has
increasingly led to results that Congress could not have
intended”); see also Shepard v. United States, 544 U.S. 13, 27
(2005) (Thomas, J., concurring in part and concurring in the
judgment) (criticizing Taylor’s implications for factfinding at
sentencing in light of Apprendi v. New Jersey, 530 U.S. 466
(2000)); id. at 35–36 (O’Connor, J., joined by Kennedy and
Breyer, JJ., dissenting) (“The Court’s overscrupulous regard
for formality leads it not only to an absurd result, but also to a
result that Congress plainly hoped to avoid.”).
4
was, on its face, a controlled substance offense that counted
toward the USSG § 4B1.1 career offender enhancement.
* * *
For the reasons stated, I join the Court’s opinion
regarding the denial of Williams’s motion to suppress, and I
concur in the Court’s judgment that Williams is a career
offender.
5
ROTH, J., Concurring
Although I agree with the reasoning and the
conclusions of the majority opinion, which I in fact wrote, I
write separately because of my concern that the categorical
approach, along with its offspring, the modified categorical
approach, is pushing us into a catechism of inquiry that
renders these approaches ludicrous. The categorical approach
was developed to avoid the “practical difficulties and
potential unfairness” inherent in “determining the precise
facts underlying a defendant’s conviction when those facts
are not plain from the elements of the offense itself.”1
Given the unique structure of RICO, we are able to
determine easily what predicate offenses led to the RICO
conviction. There is no possibility of lack of precision.
RICO’s cross-referential and multi-layered structure makes it
unlike other criminal statutes. Williams pleaded guilty to a
RICO violation under 18 U.S.C. § 1962(c). Section 1962(c)
proscribes participating in the conduct of an interstate
enterprise’s affairs through a “pattern of racketeering
activity,” which RICO defines as “at least two acts of
racketeering activity.”2 Thus, as we recently recognized, in
order to a make out a § 1962(c) RICO violation, the
government must prove beyond a reasonable doubt that the
defendant himself engaged in at least two predicate acts of
racketeering.3 Because a guilty plea requires a defendant to
1
United States v. Robinson, 844 F.3d 137, 142-43 (3d Cir.
2016).
2
18 U.S.C. § 1961(5).
3
United States v. Ferriero, 2017 WL 3319283, at *6 (3d Cir.
Aug. 4, 2017)
1
admit to every element of the offense, a defendant pleading to
a RICO violation must therefore admit to at least two
predicate acts. The mere showing of any two predicate acts
of racketeering alone, however, is not enough to sustain a
1962(c) RICO conviction. The law requires a jury to
unanimously agree on the specific racketeering acts
committed by the defendant and to find all the elements of
those predicate acts. 4 By the terms of the statute, the two
predicate racketeering acts are inextricably linked to the
substantive § 1962(c) violation, and a RICO conviction
cannot stand without them.
In this situation, I see no need to determine whether §
1962(c) is a divisible statute; I see no need to discuss the
alternative forms of conduct covered by it; I see no need to
decode the offense of conviction or to ask whether it sweeps
more broadly than the relevant generic offense.
Instead, I should be able to look at the RICO
conviction and the integrated predicate acts. The predicate
acts here are drug offenses, violations of 21 U.S.C. §§
841(a)(1) and 846, both controlled substance offenses. Voila,
we can ascertain the precise facts underlying the defendant’s
conviction. The contortions of the modified categorical
approach are needless.
(citing In re Ins. Brokerage Antitrust Litig., 618 F.3d 300, 371
(3d Cir. 2010)).
4
United States v. Brown, 583 F.2d 659 (3d Cir. 1978); see
also Mod. Crim. Jury Instr. 3rd Cir. 6.18.1962C-8 (instructing
that a jury must unanimously agree on the specific
racketeering acts committed by the defendant).
2
I write separately to explain this approach with the
hope a future panel may see that there is no need, in the case
of a RICO conviction, to engage in the catechism of the
modified categorical approach.
3
FISHER, Circuit Judge, concurring.
I join the majority opinion except for Part II.A. There,
the majority rules that consent to a search may be revoked, but
that Carlton Williams did not do so. I would hold that he did. I
am concerned that the majority’s ruling erects obstacles that
will make it difficult, in the future, for individuals to withdraw
consent to police searches. I nevertheless concur in the result
because—taking consent out of the equation—there was
probable cause to search Williams’s car.
I.
I agree with the majority’s ruling, joining other Circuits,
that consent to search may be revoked. A person may “delimit
as he chooses the scope of the search to which he consents,”
and the scope of the consent is measured by asking, “[W]hat
would the typical reasonable person have understood by the
exchange between the officer and the suspect?” Florida v.
Jimeno, 500 U.S. 248, 251-52 (1991).
I also agree with the majority’s recounting of the facts,
which is largely as follows. Trooper Michael Volk pulled
Williams over on the Pennsylvania Turnpike, gave him a
warning, told him he was free to go, and then asked if he would
consent to a search of his car. Williams signed the proffered
consent form. Then, for an hour and ten minutes on a dark
winter night, he stood on the side of the highway with a second
state trooper, Trooper Vresh, while Trooper Volk searched his
car. A third trooper, Trooper Brautigam, also arrived shortly
after the stop began.
Williams testified at the suppression hearing that he
eventually said, “[Y]ou searched my car three times, now you
hold me up and I have to go.” United States v. Williams, No.
2:14–cr–30, 2015 WL 5602617, at *6 (W.D. Pa. 2015). The
1
District Court found this testimony “only credible to a degree”
because it was not audible on the video from the dashboard-
mounted camera in Trooper Volk’s cruiser.1 Id. The District
Court does not explain the impact of its assessment that
Williams’s credibility on this point was less than total. Instead,
the opinion simply goes on to note that “there is no evidence
that Trooper Volk heard Williams’ alleged comments,
assuming such were made.” Id.
Later in the search, Williams asked for his cell phones,
which were in his car. Trooper Brautigam got them and began
to read messages on one of the phones. Williams said that the
officers could not search his phone without a warrant. When
the troopers began to disassemble the stereo speakers in his
trunk, Williams protested, “[Y]ou need a warrant to go through
my speakers.” Id. In the ensuing exchange, Williams said,
“I’ve been out here half an hour, man” (this was an
understatement; it had been more than fifty minutes). Id.
Finally, Trooper Volk discovered heroin hidden in the parking
brake assembly. Id.
I agree with the majority that when Williams said he had
been “out [there] half an hour,” he could have been expressing
impatience rather than withdrawing consent. But I view
differently his earlier statement: “[Y]ou searched my car three
times [and] y’all got me on the side of this road in the middle
1
A review of the video makes clear that the
microphone was on Trooper Volk’s person. It clearly picked
up everything Trooper Volk said, and less clearly picked up
others’ speech when they were speaking with him. It faintly
and sporadically picked up other sounds, including comments
by Williams and the trooper or troopers standing near him.
When Trooper Volk got into his cruiser and closed the door,
the road noise and all other sounds died away.
2
of the winter holding me up and I got to go.” Maj. Op. II.A.
(emphasis added). I believe that a typical reasonable person
would interpret this statement as withdrawing consent to any
further search. See Jimeno, 500 U.S. at 251-52 (asking what
“the typical reasonable person [would] have understood by the
exchange between the officer and the suspect”). Indeed, it
would be difficult to withdraw consent much more clearly,
aside from saying, “Stop,” or “I no longer consent.” As the
majority rightly notes, no one should have to say “a special set
of words to withdraw consent.” Maj. Op. II.A. The majority’s
holding, though, may render anything short of a “special set of
words” insufficient.
My position arguably conflicts with the Eighth Circuit,
which ruled that a defendant’s “indicat[ion] that he needed to
be on his way” did not withdrawal consent to a search. United
States v. Ross, 263 F.3d 844, 846 (8th Cir. 2001). Regardless
of any conflict, however, I believe that the majority disregards
the “reasonableness” that is the “touchstone” of the Fourth
Amendment, Brigham City v. Stuart, 547 U.S. 398, 403 (2006)
(citations omitted), when it requires something more than
Williams’s simple, direct statement that he needed to go.
The majority supports its holding by saying that
Williams knew how to withdraw consent because he “told the
troopers that they did not have consent to search his speakers
or his cellular phones.” Maj. Op. II.A. But Williams did not
say that, and the inaccuracy of the majority’s paraphrase is
consequential. Williams actually said that the troopers needed
a warrant to search his speakers and cell phone. He was wrong,
so his statements show only that he misunderstood Fourth
Amendment law. See California v. Acevedo, 500 U.S. 565, 580
(1991) (generally, search of speakers would be permissible
because warrantless search of an automobile encompasses “the
containers within it”); Williams, 2015 WL 5602617, at *10
3
(“Under existing law at the time, Trooper Brautigam was
authorized to examine Williams’ cell phones.”). As a logical
matter, Williams’s attempt to assert garbled Fourth
Amendment “rights” does not bear on whether he knew how to
withdraw consent to a search. If anything, his comments show
that he lacked an understanding of his rights that would have
permitted him to speak with the level of clarity the majority
seems to want.
The majority further supports its conclusion by noting
that there was no evidence—aside from Williams’s own
testimony—that Trooper Volk heard his comment, “[Y]ou
searched my car three times, now you hold me up and I have
to go.” I do not believe that Trooper Volk’s ability to hear this
comment is relevant. As the majority notes, he was assisted by
two other troopers. The District Court found, more specifically,
that Troopers Vresh and Brautigam arrived partway into the
traffic stop, and that “Trooper Volk . . . directed [Williams] to
wait with Trooper Vresh while Volk searched the vehicle.” Id.
at *5. The record fully supports this finding, to which I defer.
The dash camera video shows Williams walking off camera
with Vresh to wait as directed, and then shows Brautigam
walking on and off camera, clearly going back and forth
between Williams’s car and where Williams was standing with
Vresh. Because the troopers were working as a team, Williams
should have been able to withdraw consent by speaking to any
of them. The District Court made no findings with regard to
whether Troopers Vresh or Brautigam could hear Williams’s
comment, but the record shows that they, not Trooper Volk,
were the ones in a position to hear Williams.
II.
Despite the fact that I believe Williams withdrew his
consent to the search, I would affirm the District Court’s denial
4
of the suppression motion because there was probable cause to
search Williams’s car.
The facts leading up to the Turnpike stop were these. A
confidential informant conducted two controlled buys of
heroin from Williams in November and December 2012. Based
on the way the drug was packaged, Pittsburgh Police task force
Detective Eric Harpster believed the heroin was from Detroit.
The detective obtained a warrant to install a tracking device on
Williams’s car. On the day of the fateful traffic stop in January
2013, Detective Harpster tracked the car as it drove from
Pittsburgh to Detroit, made a very quick turnaround, and drove
back toward Pennsylvania. Detective Harpster contacted
Trooper Volk. He asked Trooper Volk to conduct a traffic stop
and try to find any heroin that might be in Williams’s car.
“The automobile exception to the warrant requirement
permits law enforcement to seize and search an automobile
without a warrant if ‘probable cause exists to believe it contains
contraband.’” United States v. Burton, 288 F.3d 91, 100 (3d
Cir. 2002) (quoting Pennsylvania v. Labron, 518 U.S. 938, 940
(1996)). In Burton, a confidential informant attempting to
make a controlled buy of a small amount of cocaine
unwittingly interrupted what appeared to be a large drug
transaction taking place in a particular house. Id. at 94-95. A
police drug task force, alerted to the large transaction, saw the
defendant leave the house in question. Id. at 95. He put a plastic
bag in the trunk of a car, made a telephone call, and drove
away. Id. We ruled that the police had probable cause to search
the car at that point “[b]ecause [they] observed Burton leave
what they thought to be a drug deal and place the results of that
transaction inside his trunk.” Id. at 100.
“The probable cause inquiry is ‘commonsense,’
‘practical,’ and ‘nontechnical;’ it is based on the totality of the
circumstances and is judged by the standard of ‘reasonable and
5
prudent men.’” United States v. Donahue, 764 F.3d 293, 301
(3d Cir. 2014) (quoting Illinois v. Gates, 462 U.S. 213, 230-31
(1983)). “At bottom,” the probable cause analysis “deal[s] with
probabilities.” Id. (quoting Gates, 462 U.S. at 231). Here, the
probable cause determination is similar to Burton. Williams
had previously sold heroin that appeared to be from Detroit to
a confidential informant. Williams’s car traveled to Detroit on
the day in question, turned around quickly, and started
traveling back. Looking at these facts in a practical and
nontechnical fashion, the probability was that Williams had
picked up drugs in Detroit and was bringing them back to
Pittsburgh. Therefore, the police had probable cause to search
the car regardless of whether Williams consented.
Contrary to Williams’s argument, Detective Harpster
was not required to recount the full probable cause analysis to
Trooper Volk. “[T]he arresting officer need not possess an
encyclopedic knowledge of the facts supporting probable
cause, but can instead rely on an instruction to arrest delivered
by other officers possessing probable cause.” Burton, 288 F.3d
at 99. Nor is Williams persuasive when he implies that
Detective Harpster needed to impart a certain quantum of
information to Trooper Volk; we have held that “[a]n officer
can lawfully act solely on the basis of statements issued by
fellow officers . . . .” Id. at 99 (quoting Rogers v. Powell, 120
F.3d 446, 453 (3d Cir. 1997)). Finally, the situation did not
need to be “dynamic” or “fast-paced” in order to justify
Trooper Volk’s reliance. The situation in Burton was not
particularly dynamic or fast-paced, but even so, we concluded
that the arresting officer could rely on the task force officer’s
instruction. See id. at 94-96.
III.
For these reasons, I concur in the judgment affirming
the denial of the suppression motion. However, I would not
6
reach that result on the basis that Williams consented to the
search (and failed to effectively withdraw his consent). Instead,
I would hold that although he withdrew consent, there was
probable cause to search his car. This analysis is more faithful
to our Fourth Amendment jurisprudence, and in particular, the
reasonableness that is its touchstone.
7