United States v. Taison McCollum

PUBLISHED

UNITED STATES COURT OF APPEALS
FOR THE FOURTH CIRCUIT

No. 17-4296

UNITED STATES OF AMERICA,

Plaintiff – Appellee,

v.

TAISON MCCOLLUM,

Defendant – Appellant.

Appeal from the United States District Court for the Western District of North Carolina,
at Charlotte. Max O. Cogburn, Jr., District Judge. (3:16-cr-00199-MOC-DCK-1)

Argued: January 25, 2018 Decided: March 20, 2018

Before WILKINSON, TRAXLER, and DUNCAN, Circuit Judges.

Vacated and remanded by published opinion. Judge Duncan wrote the opinion, in which
Judge Traxler joined. Judge Traxler wrote a concurring opinion. Judge Wilkinson wrote
a dissenting opinion.

ARGUED: Joshua B. Carpenter, FEDERAL DEFENDERS OF WESTERN NORTH
CAROLINA, INC., Asheville, North Carolina, for Appellant. Amy Elizabeth Ray,
OFFICE OF THE UNITED STATES ATTORNEY, Asheville, North Carolina, for
Appellee. ON BRIEF: Anthony Martinez, Federal Public Defender, Caleb H. Newman,
FEDERAL PUBLIC DEFENDER WESTERN DISTRICT OF NORTH CAROLINA,
Charlotte, North Carolina, for Appellant. Jill Westmoreland Rose, United States
Attorney, OFFICE OF THE UNITED STATES ATTORNEY, Charlotte, North Carolina,
for Appellee.
DUNCAN, Circuit Judge:

Appellant Taison McCollum pleaded guilty in federal court to possession of a

firearm by a convicted felon in violation of 18 U.S.C. § 922(g). At his sentencing, the

district court applied a sentencing enhancement under § 2K2.1(a)(4)(A) of the Sentencing

Guidelines based on McCollum’s prior conviction under 18 U.S.C. § 1959(a)(5) for

conspiracy to commit murder in aid of racketeering. McCollum argues that the district

court erred because conspiracy under § 1959(a)(5) does not require an overt act and is

therefore broader than generic conspiracy. As we explain below, we are constrained by

the Guidelines text and precedent to agree. The dissent, on the other hand, achieves the

result it prefers by overlooking inconvenient Supreme Court and circuit precedent

interpreting the Guidelines language at issue. 1 As we are reluctant to do so, we vacate

McCollum’s sentence and remand for resentencing.

I.

McCollum pleaded guilty in the Western District of North Carolina to possession

of a firearm by a convicted felon in violation of 18 U.S.C. § 922(g). The presentence

investigation report asserted that McCollum had two prior convictions that qualified as

crimes of violence under § 2K2.1, the Guidelines provision that establishes the base

1
Indeed, the dissent acknowledges that the majority’s “error” is that it
“mechanically applies” the Supreme Court’s direction to compare the elements of crimes,
not their labels. Yet it simultaneously refuses to apply the controlling circuit precedent
that obliges us to analyze inchoate crimes like conspiracy and their objects separately,
and no amount of handwringing, hyperbole, or misdirection can obscure that fact.

2
offense level for a felon in possession of a firearm: a New Jersey conviction for

aggravated manslaughter, and a conviction under 18 U.S.C. § 1959(a)(5) for conspiracy

to commit murder in aid of racketeering. These two convictions supported an

enhancement that increased McCollum’s base offense level from fourteen to twenty-four.

The district court sustained McCollum’s objection to classification of his New Jersey

conviction as a crime of violence but held that McCollum’s § 1959(a)(5) conviction was

properly classified as such. Accordingly, the district court concluded McCollum had one

prior conviction that qualified as a crime of violence under § 2K2.1 and that McCollum’s

base offense level was twenty.

II.

McCollum argues that his enhanced sentence is unlawful because conspiracy to

commit murder in aid of racketeering is not a “crime of violence” since it does not

require an overt act, while conspiracy under the Guidelines does.

We review de novo whether a prior conviction qualifies as a crime of violence

under the Guidelines. United States v. Salmons, 873 F.3d 446, 448 (4th Cir. 2017). We

conclude that § 1959(a)(5) is not categorically a crime of violence because conspiracy

under that provision is, in fact, broader than generic conspiracy, and precedent directs

3
that we consider the inchoate crime of conspiracy and its object independently. We

therefore remand for resentencing. 2

A.

Under § 2K2.1 of the Guidelines, the base offense level for a § 922(g) conviction

is twenty if the defendant has a prior “felony conviction of either a crime of violence or a

controlled substance offense.” U.S.S.G. § 2K2.1(a)(4)(A). The felon-in-possession

Guideline defines a “crime of violence” via cross-reference “to the career-offender

guideline, U.S.S.G. § 4B1.2.” United States v. Shell, 789 F.3d 335, 340 (4th Cir. 2015);

U.S.S.G. § 2K2.1 cmt. n.1. Accordingly, for purposes of the felon-in-possession

Guideline, a crime of violence includes “any offense under federal or state law” that:

(1) has as an element the use, attempted use, or threatened use of physical
force against the person of another, or

(2) is murder, voluntary manslaughter, kidnapping, aggravated assault, a
forcible sex offense, robbery, arson, extortion, or the use or unlawful
possession of a firearm described in 26 U.S.C. § 5845(a) or explosive
material as defined in 18 U.S.C. § 841(c).

U.S.S.G. § 4B1.2(a). Further, a “crime of violence” includes “the offenses of aiding and

abetting, conspiring, and attempting to commit such offenses.” Id. cmt. n.1.

2
McCollum also seeks relief under the alternative theory that the Sentencing
Commission cannot use the commentary to expand the definition of “crime of violence”
to include conspiracy, since it would be “flatly inconsistent with the text [of § 4B1.2] . . .
to treat the offense as a crime of violence based solely on the commentary.” See Reply
Br. at 17. Because we hold that McCollum is entitled to resentencing under his first
theory, we do not address McCollum’s alternative theory.

4
 Courts generally follow a well-established procedure to determine whether a

defendant’s prior conviction qualifies as an enumerated crime of violence under the

Guidelines. A defendant may not receive an enhanced sentence merely because the label

attached to his crime of conviction is listed in the enumerated offense clause. Instead, an

enhanced sentence is lawful only if the prior conviction necessarily establishes that the

defendant “has been found guilty of all the elements” of the enumerated offense. See

Taylor v. United States, 495 U.S. 575, 599 (1990).

In Taylor, the Supreme Court instructed that, with regard to the undefined

enumerated offenses, courts must look to the “generic, contemporary meaning” of the

crime, which will typically correspond to the “sense in which the term is now used in the

criminal code of most states,” id. at 598, rather than the term’s common law meaning, see

id. at 594. 3 Taylor instructed that we apply this categorical approach to ensure that the

elements of the crime of conviction are no broader than those of the generic enumerated

offense. See 495 U.S. at 602. Taylor’s categorical approach “serves as the cornerstone

of our analysis [of] whether a prior offense qualifies as a ‘crime of violence’ under

Section 4B1.2(a).” Carthorne, 726 F.3d at 511. “The point of the categorical inquiry is

not to determine whether the defendant’s conduct could support a conviction for a crime

of violence, but to determine whether the defendant was in fact convicted of a crime that

3
Taylor interpreted the Armed Career Criminal Act, but precedents interpreting
“crime of violence” under the Guidelines are interchangeable with precedents interpreting
“violent felony” under the Armed Career Criminal Act. United States v. Carthorne, 726
F.3d 503, 511 n.6 (4th Cir. 2013).

5
qualifies as a crime of violence.” United States v. Cabrera-Umanzor, 728 F.3d 347, 350

(4th Cir. 2013).

When evaluating a defendant’s prior conviction for an inchoate offense listed in

the commentary to § 4B1.2(a), “two sets of elements are at issue: the elements of [the

inchoate crime] and the elements of the underlying . . . offense.” United States v. Dozier,

848 F.3d 180, 185–86 (4th Cir. 2017). 4 Both the inchoate crime and the underlying

offense are subject to Taylor’s categorical approach. An enhanced sentence may follow a

conviction for an inchoate crime only if the defendant’s conviction necessarily

establishes that he was found guilty of a crime whose elements encompass both the

generic inchoate crime and the generic underlying crime. Id.

McCollum argues that § 1959(a)(5) does not support an enhanced sentence

because it does not require an overt act, while generic conspiracy does. The government

disagrees. It argues as a threshold matter that we should not employ the categorical

approach to evaluate a predicate federal crime because we have never previously done so.

Alternatively, the government continues, even if the categorical approach could be

applied to some predicate federal crimes, it should not be applied to § 1959(a)(5) because

it would be unreasonable to believe the Commission intended to exclude obviously

violent crimes like conspiracies to commit murder in aid of racketeering from its

definition of “crime of violence.” Moreover, the government contends that generic

4
Dozier only considered the inchoate crime of attempt under the commentary to
§ 4B1.2. But because “aiding and abetting, conspiring, and attempting” crimes of
violence are listed together in the same sentence, Dozier’s holding logically extends to all
the inchoate crimes listed in that commentary provision.

6
conspiracy, like common law conspiracy, does not require proof of an overt act. We

address these issues below.

B.

We first consider whether the categorical approach applies to federal crimes like

conspiracy to commit murder in aid of racketeering. We conclude that it does for two

reasons. First, although the Commission has not expressly addressed the issue, the text of

§ 4B1.2 and other Commission publications strongly suggests that it does. Second, we

find no textual or analytical basis in our precedent for distinguishing the treatment of

state and federal statutes.

1.

Our starting point is the Guidelines text. The fact that under § 4B1.2(a) a “crime

of violence” includes qualifying offenses “under federal or state law,” U.S.S.G.

§ 4B1.2(a), suggests that the two are to be treated similarly. And treating federal

predicates similarly necessarily implicates the categorical approach. It is beyond debate

that we utilize the Taylor construct and the categorical approach to determine whether a

prior state conviction constitutes a crime of violence. See, e.g., Taylor, 495 U.S. at 590;

Dozier, 848 F.3d at 183; Shell, 789 F.3d at 338; Carthorne, 726 F.3d at 512–13. 5

5
We have previously remarked that the categorical approach requires us to
“determine the elements of the state-law offense in question and compare them to the
generic definition.” See, e.g., United States v. Mungro, 754 F.3d 267, 269 (4th Cir. 2014)
(Continued)
7
 Notably, too, the Commission’s 2016 Supplement to Appendix C explains that

courts employ the categorical approach when applying the enumerated offense clause,

and it offers no exception for federal crimes: “The ‘enumerated offense clause’ identifies

specific offenses that qualify as crimes of violence. In applying this clause, courts

compare the elements of the predicate offense of conviction with the elements of the

enumerated offense in its ‘generic, contemporary definition.’” U.S.S.G. Supp. to App. C,

amend. 798 (2016), at 129. If the Commission had intended for courts to apply the

categorical approach only to state crimes, it could easily have made that clear. When the

Commission wants to single out federal laws, it can--and does--do so explicitly. For

example, the text of § 4B1.2(a)(2) includes as crimes of violence “unlawful possession of

a firearm described in 26 U.S.C. § 5845(a) or explosive material as defined in 18 U.S.C.

§ 841(c).” See U.S.S.G. § 4B1.2(a)(2) (emphases added).

Therefore, we find that the Guidelines text supports the application of Taylor’s

categorical approach to predicate crimes under both state and federal law.

2.

We next consider our precedent. The Supreme Court has not spoken directly to

the point, but we are nevertheless bound by Taylor’s mandate: the categorical approach

(emphasis added); see also Perez-Perez, 737 F.3d 950, 952 (4th Cir. 2013). But these
references to a prior state conviction are an accident of circumstance: those cases
involved state statutes. The parties have identified no case in which we decided--or even
assumed--that the analysis could not apply to a federal statute.

8
aims to realize the Commission’s purpose that enhanced sentences follow convictions for

“crimes having certain specified elements,” not crimes having certain names. See Taylor,

495 U.S. at 588; Dozier, 848 F.3d at 185–86.

The government’s argument that we should abandon the categorical approach

when considering a prior federal conviction whose label is listed in the Guidelines does

not effectuate this mandate. Both state governments and the federal government can

write statutes with elements that deviate from those of the corresponding generic crime,

and our precedent provides us with no basis for treating the two differently.

Furthermore, the government’s position is undermined by the fact that there is no

single federal definition of conspiracy that we can assume the Commission intended to

adopt when it included conspiracy in the commentary to § 4B1.2. The general federal

conspiracy statute requires as an element an overt act. See 18 U.S.C. § 371. At least

fifteen federal conspiracy provisions for nonviolent crimes require an overt act, while at

least ninety-nine do not. See United States v. Pascacio-Rodriguez, 749 F.3d 353, 360

(5th Cir. 2014) (surveying statutes). Eight federal conspiracy provisions for violent

crimes require an overt act, while forty-three do not. Id. at 360–62 (same).

Our review leads us to conclude that the principles expressed by the Supreme

Court and in our prior decisions constrain us to apply the categorical approach to

9
McCollum’s prior conviction under § 1959(a)(5). 6 Our precedent offers no basis for

analyzing the laws of different sovereigns under different standards. 7

6
The dissent complains that our approach is in conflict with that of three other
circuits and warns that our erroneous framework “has serious consequences.” See infra
at 19–22. However, other circuits have employed our approach in similar circumstances.
For example, the Tenth Circuit recently analyzed whether Hobbs Act robbery is a crime
of violence under § 4B1.2(a)(2) and stated that “[o]ur enumerated clause analysis starts
with defining Hobbs Act robbery. We next define generic robbery, and then compare the
two terms to determine whether the minimum conduct needed to constitute Hobbs Act
robbery fits within generic robbery.” United States v. O’Connor, 874 F.3d 1147, 1154–
55 (10th Cir. 2017); United States v. Martinez-Cruz, 836 F.3d 1305, 1314 (10th Cir.
2016) (applying categorical approach analysis to determine whether a conviction for
possession with intent to distribute marijuana in violation of 21 U.S.C. § 846 was an
aggravated felony conviction under the illegal reentry Guideline); see also United States
v. Moldanado, 636 F. App’x 807, 812 (2d Cir. 2016) (analyzing whether a conviction
under 21 U.S.C. § 843(b) could be a predicate crime under § 4B1.1(a), and concluding
that “[b]ecause § 843(b) contains additional elements not included in the definition of
‘controlled substance offense[,]’ [it] is broader than the generic version laid out in the
Guidelines”); but see United States v. Rivera-Constantino, 798 F.3d 900, 902–03 (9th
Cir. 2015); United States v. Sanbria-Bueno, 549 F. App’x 434, 438–39 (6th Cir. 2013);
Pascacio-Rodriguez, 749 F.3d at 367. In light of this circuit split, we think the dissent’s
apocalyptic predictions are exaggerated.
7
The dissent concludes that the generic-definition framework is unnecessary
because the text, structure, and context of the Guideline make it clear that a conspiracy to
commit murder is a crime of violence. In doing so, the dissent makes the issue quite
clear by simply waving a magic wand and eliding everything with which it disagrees,
including Taylor’s mandate that we consider the elements of a crime, not its label. No
clearer example of this tunnel vision can be found than in its failure to apply Dozier, a
case interpreting the application of the Guidelines to an inchoate crime. That the
Supreme Court is moving away from the constraints of Taylor may be devoutly to be
wished. However, as the dissent seems reluctant to acknowledge, our role is to follow,
not lead.

10
 C.

Having decided to apply the categorical approach to McCollum’s conviction for

conspiracy to commit murder in aid of racketeering, we now apply that analysis to

determine whether § 1959(a)(5) is a crime of violence under § 4B1.2. As we explain

below, because McCollum’s conviction under § 1959(a)(5) does not establish that he was

actually found guilty of all of the elements of generic conspiracy, we conclude that it is

not.

Our analysis under the categorical approach has four parts. See Perez-Perez, 737

F.3d at 952. First, we determine the relevant offense of comparison. See id. As we have

noted, when determining whether an inchoate crime is a crime of violence, “two sets of

elements are at issue: the elements of [the inchoate crime] and the elements of the

underlying . . . offense.” Dozier, 848 F.3d at 185. Since McCollum concedes that

murder under § 1959(a)(5) qualifies as a crime of violence, see Oral Argument at 5:00–

5:30, our analysis turns on the conspiracy component of his conviction. 8

8
We note that, despite the dissent’s protests, the object of the conspiracy does not
change the elements of the conspiracy offense. While murder and extortion are
enumerated as crimes of violence, the fact that murder may be more violent than
extortion does not mean that a conspiracy to commit murder is more likely to constitute a
crime of violence than a conspiracy to commit extortion. The Guidelines simply state
that “‘[c]rime of violence’ and ‘controlled substance offense’ include the offenses of
aiding and abetting, conspiring, and attempting to commit such offenses.” U.S.S.G.
§ 4B1.2 cmt. n.1. Nothing in the commentary even hints at the possibility that
“conspiracy” has a different meaning depending on the object of the conspiracy, and
Dozier, which interpreted “attempt” under the same provision, analyzed the inchoate
crime in isolation from the object crime. See 848 F.3d at 185. Thus, we analyze the
elements of the conspiracy portion of § 1959(a)(5) without placing a thumb on the scale
to account for the fact that the object of the conspiracy is murder.

11
 Second, we determine the elements of generic conspiracy. See Perez-Perez, 737

F.3d at 952. Because the Guidelines do not define conspiracy, “it should be understood

to refer to ‘the generic, contemporary meaning’ of the crime.” See Rangel-Castaneda,

709 F.3d 373, 376 (4th Cir. 2013) (quoting Taylor, 495 U.S. at 598); U.S.S.G. Supp. to

App. C, amend. 798 (2016), at 129. This meaning will typically correspond to the “sense

in which the term is now used in the criminal codes of most States.” Taylor, 495 U.S. at

598. In Rangel-Castaneda we found that the definition adopted by thirty-two states, the

federal government, and the District of Columbia established a “broad consensus”

sufficient to establish the generic, contemporary definition of a crime. 709 F.3d at 377–

78.

As the parties have noted, the Ninth Circuit recently surveyed conspiracy statutes,

and found that thirty-six states, the District of Columbia, Guam, Puerto Rico, and the

Virgin Islands define conspiracy to require an overt act. See United States v. Garcia-

Santana, 774 F.3d 528, 534–35 (9th Cir. 2014). Additionally, the general federal

conspiracy statute requires an overt act. See 18 U.S.C. § 371. The Ninth Circuit thus

held that “an overt act is an element of the generic definition of conspiracy.” Garcia-

Santana, 774 F.3d at 537. The Tenth Circuit agreed. See Martinez-Cruz, 836 F.3d at

1314. 9

9
The Fifth Circuit considered whether the specific conspiracy, that is, conspiracy
to commit murder, usually includes an overt act requirement. See Pascacio-Rodriguez,
749 F.3d at 364. But our decision in Dozier forecloses that result here. See Dozier, 848
F.3d at 185–86.

12
 So do we. The fact that more than thirty-two states require an overt act is

sufficient to establish the contemporary definition of conspiracy as such. See Rangel-

Castaneda, 709 F.3d at 377–78 (definition adopted by thirty-two states established a

“broad consensus”). 10

The government argues that Etienne v. Lynch, 813 F.3d 135 (4th Cir. 2015),

established that generic conspiracy lacks an overt act requirement, but the comparison is

inapposite. In Etienne, we had to decide whether the definition of conspiracy under the

Immigration and Naturalization Act (the “INA”) adopted the common law definition,

which did not require an overt act. We relied on the “settled principle of statutory

construction that, absent contrary indications, Congress intends to adopt the common law

definition of statutory terms,” id. at 143 (quoting United States v. Shabani, 513 U.S. 10,

13 (1994)) (emphasis added), to conclude that conspiracy under the INA did not require

an overt act, id. at 145. Here, by contrast, we interpret Guidelines written by the

Sentencing Commission, which requires us to consider a crime’s contemporary meaning,

not its common law meaning--a distinction that the dissent fails to appreciate. See

U.S.S.G. Supp. to App. C, amend. 798 (2016), at 129. Etienne in fact acknowledges that

10
The dissent asserts that our approach, which relies on the conspiracy definition
adopted by most states, does not give effect to the intent of the federal Sentencing
Commission because most federal conspiracies do not require an overt act. But we
presume the Commission is aware of precedent when they write the Guidelines, see Miles
v. Apex Marine Corp., 498 U.S. 19, 32 (1990), and the Supreme Court has instructed that
determination of the generic, contemporary definition of crimes requires a consideration
of state definitions. The dissent points to no indication that the Commission did not have
this instruction in mind when it chose to leave “conspiring” undefined. Should the
Commission intend another position we hope that it will make that clear.

13
“the prevailing contemporary meaning of [conspiracy] . . . requires proof of an overt act.”

813 F.3d at 142.

The third and fourth parts of our categorical approach analysis are straightforward.

Third, we compare the elements of the crime of conviction to those in the generic

definition of the enumerated offense. Perez-Perez, 737 F.3d at 952. If the elements of

the prior crime correspond in substance to those of the enumerated offense, the prior

crime is a crime of violence. Id. If they do not, the fourth step requires that we decide

whether the scope of conduct criminalized by the statute “is categorically overbroad

when compared to the generic definition of the Guideline crime.” Id. at 952–53. A

statute whose elements criminalize a broader range of conduct than the Guidelines crime

“is not categorically a crime of violence.” Id. at 953.

The government does not dispute that conspiracy under § 1959(a)(5) does not

require an overt act. See Appellee’s Br. at 17 (citing United States v. Orena, 32 F.3d 704,

714 (2d Cir. 1994)). Because § 1959(a)(5) does not require an overt act, it criminalizes a

broader range of conduct than that covered by generic conspiracy. McCollum’s

§ 1959(a)(5) conviction therefore cannot support his enhanced sentence because it is not

categorically a crime of violence.

14
 III.

Our analysis is cabined by the text of the Guidelines and by precedent from the

Supreme Court and our prior decisions. For the foregoing reasons, the judgment of the

district court is

VACATED AND REMANDED.

15
TRAXLER, Circuit Judge, concurring:

The dissent ends with the dramatic lament, “Heaven help us.” Frankly, I would be

satisfied if Congress or the Supreme Court would help us. The law in this area, which

Judge Duncan faithfully follows, leads to some seemingly odd results with which I do not

think any of us are particularly happy. But until help comes from some higher level in

the form of substantive changes, this decision, in my judgment, is what the law requires.

16
WILKINSON, Circuit Judge, dissenting:

The majority holds that the federal predicate offense of conspiracy to commit

murder in aid of racketeering activity, 18 U.S.C. § 1959(a)(5), is not a crime of violence

under U.S.S.G. § 2K2.1. To reach this conclusion, it ignores the plain text of the federal

statute and unmistakable textual command of the Guidelines provisions and commentary,

which state in so many words that conspiracy to commit murder is a crime of violence.

When the text is this plain, there is no reason to get into the ambiguous weeds of generic

definitions, which can no more overcome a plain textual command than amorphous

legislative history can overcome a clear statutory mandate.

Even if one were to follow the elusive road to a generic definition the majority

maps out, the conclusion would be the same. Neither the common law nor the great

majority of federal conspiracy offenses requires an overt act. And state criminal

definitions cannot be allowed to rework the textual clarity of the United States

Sentencing Guidelines. Dual sovereignty runs in both directions. Just as there are limits

to federal control over state criminal systems, so too are there limits to state control over

federal criminal law.

The majority’s error is serious. In committing it, the majority ignores,

misinterprets, or fails to follow Supreme Court and circuit precedent at every turn. In

ignoring the Guidelines’ plain text and in misapplying the basic generic-definition

analysis, the majority has arrived at the deeply unfortunate conclusion that participation

in a conspiracy to commit the enumerated violent crime of murder is somehow not a

crime of violence. Properly applied, the categorical approach can ultimately coincide

17
with common sense. Applied as the majority has done here, it reaches a result that no one

inside or outside of Congress and the Sentencing Commission will ever understand. True,

there may be some beings in some other planetary system who can make sense of it all,

but for earthlings, not so much.

I.

This case is easily resolved by the plain language of the Guidelines and their

commentary. This court “must follow the clear, unambiguous language of a particular

guideline unless there is a manifestation of contrary intent.” United States v. Achiekwelu,

112 F.3d 747, 755 (4th Cir. 1997). Because the commentary “is akin to an agency’s

interpretation of its own legislative rules . . . , it must be given ‘controlling weight unless

it is plainly erroneous or inconsistent with’” the relevant guideline. Stinson v. United

States, 508 U.S. 36, 45 (1993) (quoting Bowles v. Seminole Rock & Sand Co., 325 U.S.

410, 414 (1945)).

Because McCollum pled guilty to being a felon in possession of a firearm in

violation of 18 U.S.C. § 922(g), the relevant guideline is § 2K2.1. That guideline

provides for a heightened base offense level if a defendant has previously been convicted

of “a crime of violence.” See U.S.S.G. § 2K2.1. The term “crime of violence” is not

defined in § 2K2.1, but the commentary to that guideline instructs that the term “has the

meaning given . . . in § 4B1.2(a) and Application Note 1 of the Commentary to § 4B1.2.”

Id. § 2K2.1 cmt. n.1.

First and foremost among the crimes of violence enumerated in § 4B1.2(a) is

“murder.” See id. § 4B1.2(a)(2). This is not surprising. Murder is singular. It is the

18
quintessential violent crime. To leave no room for doubt, the commentary to § 4B1.2

instructs that “the offenses of aiding and abetting, conspiring, and attempting to commit”

crimes of violence are also included within the crime of violence definition. Id. § 4B1.2

cmt. n.1.

This clear textual command should by any measure wrap up this case. According

to the text and commentary of the Guidelines, “conspiring . . . to commit” “murder” is a

crime of violence. This language maps perfectly onto the statutory language of 18 U.S.C.

§ 1959(a)(5): “conspiring to commit murder.” If the Commission had wished to exclude

this crime from the crime of violence definition, it would not have used language

identical to that in the statute. Awaiting help from elsewhere, as the concurring opinion

prefers, is like waiting in a bus station for someone who has long since arrived. The

answer is already abundantly clear.

It is also commonsensical. This court has long acknowledged that the statute under

which McCollum was previously convicted “punishes violent crimes.” United States v.

Umaña, 750 F.3d 320, 336 (4th Cir. 2014). Further, “no mechanical exercise can ever

fully supplant . . . common sense and good judgment” when applying the Guidelines.

United States v. Flores-Granados, 783 F.3d 487, 498 (4th Cir. 2015). Courts must be

careful to “not read the Guidelines in a way that makes the Sentencing Commission look

foolish.” United States v. Turner, 998 F.2d 534, 538 (7th Cir. 1993). That is precisely

what the majority has done in suggesting that the Commission meant to exclude from the

definition of crime of violence the crime of conspiracy to commit murder when it used

inclusively the words “conspiring to commit” “murder.”

19
 Instead, the majority mechanically applies the generic-definition methodology

employed in Taylor v. United States, 495 U.S. 575 (1990). But the Supreme Court has

instructed that the Taylor framework is “not required by the categorical approach.”

Esquivel-Quintana v. Sessions, 137 S. Ct. 1562, 1571 n.3 (2017). To be sure, courts may

find this approach useful when it “helps shed light on the common understanding and

meaning of the federal provision being interpreted.” Id. (internal quotation marks

omitted). In Taylor, for instance, the Court emphasized that it was “not readily apparent”

what “Congress intended ‘burglary’ to mean,” in part because the term was defined by

the courts and state criminal codes “in many different ways.” 495 U.S. at 580. In that

context, the Court found the generic-definition analysis to be a useful tool in divining

Congress’s intent. See id. at 598-99. And this court has likewise found Taylor’s generic-

definition approach to be a useful tool in some cases. See, e.g., United States v. Dozier,

848 F.3d 180, 185-87 (4th Cir. 2017) (describing the “unique complexity of general

attempt statutes” for purposes of applying the Guidelines before applying the generic-

definition approach); United States v. Rangel-Castaneda, 709 F.3d 373, 376 (4th Cir.

2013) (noting that “where Congress has not indicated how a prior offense enumerated in

a sentencing enhancement statute is to be interpreted,” courts should consider the

“generic” meaning of the crime). 1

1
Notably, the government did not argue in either Dozier or Rangel-Castaneda that
the generic-definition analysis was unnecessary. In neither case, then, was the court asked
to choose between a conclusion rooted in the Guidelines’ plain meaning and one based on
Taylor’s generic-definition approach.

20
 But when a guideline’s meaning is already patently clear from its text, structure,

and context, there is no need to rely on the Taylor framework. In its reliance on that

framework, the majority is in conflict not only with the Supreme Court’s decision in

Esquivel-Quintana, 137 S. Ct. at 1571 n.3, but also with three other circuits, see United

States v. Rivera-Constantino, 798 F.3d 900, 904 (9th Cir. 2015) (“[W]hen the plain

meaning of a term is readily apparent from the text, context, and structure of the relevant

Guidelines provision and commentary, that meaning is dispositive . . . .”); United States

v. Sanbria-Bueno, 549 F. App’x 434, 438 (6th Cir. 2013) (declining to apply the generic-

definition approach because it “is only applied to interpret the underlying offense where

it is unclear in what sense the term was used by Congress or the Sentencing Commission”

(citations omitted)); United States v. Rodriguez-Escareno, 700 F.3d 751, 754 (5th Cir.

2012) (“There is no reason to search outside the Guidelines for a definition of

‘conspiracy’ . . . .”). Indeed, the use of the generic-definition framework in plain-meaning

cases would not only be unhelpful but may prove positively detrimental. It “would only

becloud what is clear from the Guideline itself.” Rodriguez-Escareno, 700 F.3d at 754.

And in doing so, it would frustrate the goals of the Commission, introducing uncertainty

and inconsistency where the Commission sought to avoid it and allowing the most violent

cohort of criminal offenders to avoid the enhanced penalties the Commission felt they

deserved.

The majority wishes the Commission had defined conspiracy differently, namely

in the way the majority prefers. But the Commission spoke, mercifully, in plain, non-

Delphic language. To get where it wishes to go, the majority implies qualifiers to the

21
simple term conspiracy; it succumbs to the most ancient of antitextual practices, which is

to add the proverbial gloss.

This case is one in which reliance on the generic-definition framework is wholly

unnecessary. The Guidelines are clear. “[M]urder” is a crime of violence. U.S.S.G. §

4B1.2(a)(2). “[C]onspiring . . . to commit” any crime of violence is itself a crime of

violence. Id. § 4B1.2 cmt. n.1. Therefore, conspiracy to commit murder is a crime of

violence.

Case closed. End of story.

II.

The majority’s decision to undertake the generic-definition analysis, and its

erroneous application of that analysis, has serious consequences. For the majority’s view

will do nothing less than label nonviolent a multitude of state and federal convictions for

conspiracy to commit one of the enumerated violent offenses in § 4B1.2. It would render

nonviolent such crimes as conspiracy to commit genocide, 18 U.S.C. § 1091(d);

conspiracy to use weapons of mass destruction, id. § 2332a; conspiracy to commit arson,

id. § 81; and conspiracy to take hostages, id. § 1203(a). See United States v. Pascacio-

Rodriguez, 749 F.3d 353, 363 n.45 (5th Cir. 2014). The majority would introduce

disuniformity into federal sentencing, as conspiracy convictions would be counted in

different ways, notwithstanding their essential similarity in being “offenses of . . .

conspiring . . . to commit” crimes of violence. U.S.S.G. § 4B1.2 cmt. n.1. The majority

would further require federal courts to tally up state laws even where the federal

Guidelines language could not be more clear. Its errors are thus twofold: one in adopting

22
the generic-definition framework in the first place, and the other in applying it so

improperly. Either error by itself would be serious; together they are egregious.

All this is done in the course of nullifying directly applicable circuit precedent. In

Etienne v. Lynch, 813 F.3d 135 (4th Cir. 2015), this circuit, indeed this very panel, held

that generic conspiracy does not contain an overt act requirement. I had supposed Etienne

to be the law of the circuit until the majority pulled the rug out from under its own

decision.

The majority attempts to distinguish Etienne on the ground that that case involved

the INA whereas this case involves the Guidelines. Maj. Op. at 13-14. Armed with no

support, it asserts that the Guidelines are exempt from the “settled principle of statutory

construction that, absent contrary indications, Congress intends to adopt the common law

definition of statutory terms.” Id. at 12 (quoting Etienne, 813 F.3d at 143). But that is not

the case. In fact, Taylor—the very authority the majority declares itself “bound by”—

began with the common law.

The Court in Taylor inquired into “the generic sense in which [‘burglary’] is now

used in the criminal codes of most States,” 495 U.S. at 598, but only after determining

that there was good reason to depart from “the maxim that a statutory term is generally

presumed to have its common-law meaning,” id. at 592. This was because “the

contemporary understanding of ‘burglary’ ha[d] diverged a long way from its common-

law roots,” and so relying on the common-law definition “would not [have] comport[ed]

with the purposes of the enhancement statute.” Id. at 593. Instead, it would have “come

23
close to nullifying” the enhancement by placing the vast majority of burglary offenses

beyond its scope. Id. at 594.

As in Taylor, the analysis here ought to begin with the common law. As the

Supreme Court has “consistently held[,] . . . the common law understanding of

conspiracy ‘does not make the doing of any act other than the act of conspiring a

condition of liability.’” United States v. Shabani, 513 U.S. 10, 13-14 (1994) (quoting

Nash v. United States, 229 U.S. 373, 378 (1913)). In other words, conspiracy at common

law did not have an overt act requirement.

As this court explained in Etienne, there is no reason to jettison this common-law

definition. First, whereas in Taylor only a handful of states had retained burglary’s

common-law definition, 495 U.S. at 593, today one-third of states retain the common-law

definition of conspiracy, Etienne, 813 F.3d at 144. 2 The contemporary understanding of

conspiracy, then, has more than a “‘little in common’ with conspiracy’s common-law

counterpart.” Id. Second, whereas in Taylor the vast majority of states had “expanded”

the definition of burglary, 495 U.S. at 593, in the case of conspiracy, “those states that

have added the overt act requirement have narrowed the definition,” Etienne, 813 F.3d at

144. Conspiracy’s common-law definition would therefore encompass even those

modified state offenses.

2
The Fifth Circuit counted “34 states [that] require an overt act as an element of
all criminal conspiracies” and “13 states [that] do not require an overt act for any
conspiracy offense.” Pascacio-Rodriguez, 749 F.3d at 365. Arizona, New Jersey, and
Utah require an overt act for some crimes. Id. (citing Ariz. Rev. Stat. Ann. § 13-1003(A);
N.J. Stat. Ann. § 2C:5-2(d); Utah Code Ann. § 76-4-201).

24
 The bond between statutory usages and common law, so basic also to American

law, has been too casually broken. It is simply not the case that adopting the common-law

definition of conspiracy would “come close to nullifying” that term in the Guidelines.

Taylor, 495 U.S. at 594. Rather, that is precisely the consequence of the majority’s

addition of an overt act requirement. Its dismissal of the common law, as noted, is no

small matter. It will exclude from the Guidelines’ crime of violence definition the vast

majority of federal conspiracy offenses—including multiple murder-conspiracy

offenses—and roughly one-third of all state conspiracy offenses. This is not what the

Commission intended. Because the common-law definition of conspiracy is neither

“obsolete” nor “inconsistent with the [guideline]’s purpose,” the search for a generic

definition of conspiracy need go no further. Id. at 594.

But further the majority goes, and the more severe its error grows. For proper

application of the multijurisdictional analysis in this case indicates that the contemporary

understanding of conspiracy does not contemplate an overt act requirement. The

multijurisdictional method of divining generic definitions calls upon courts to survey the

overall landscape of criminal laws in the various jurisdictions across the country,

including both the federal government and the states. See Taylor, 495 U.S. at 598-99;

Rangel-Castaneda, 709 F.3d at 377-78 (consulting the laws of the federal government,

the states, and the District of Columbia). The majority, without explanation, narrows the

inquiry to “state definitions,” refusing to give appropriate weight to federal law. Maj. Op.

at 13 n.10. By putting a thumb on the scale, the majority warps the overall balance of

relevant authorities, which in reality cannot be reconciled with its holding.

25
 The lion’s share of federal conspiracy statutes contain no overt act requirement.

As the Fifth Circuit tallied in 2014, 24 federal conspiracy statutes require an overt act

whereas 142 do not. See Pascacio-Rodriguez, 749 F.3d at 360-62. The majority relies

wholly on the general federal conspiracy statute, which does require an overt act. 18

U.S.C. § 371. But that provision covers a wide variety of violent and nonviolent offenses,

and is therefore far broader than the sort of conspiracy the Sentencing Commission was

concerned with here. The great majority of federal conspiracy provisions dealing

specifically “with conspiracies to commit crimes that arguably would be within the

definition of a ‘crime of violence’” contain no such requirement. Pascacio-Rodriguez,

749 F.3d at 362 (noting that 8 such provisions “require an overt act, while 43 do not”

(footnote omitted)). Surely the Commission did not intend to exclude those offenses

when it included conspiracy within the crime of violence definition.

The majority correctly notes that most states have added an overt act requirement

to their own conspiracy statutes. Maj. Op. at 12. But the number of states that retain the

common-law definition is by no means “few.” Taylor, 495 U.S. at 593. Rather, “[o]ne-

third of the states retain the common-law definition [of conspiracy] outright.” Etienne,

813 F.3d at 144. As the majority sees it, this two-thirds cluster of state conspiracy statutes

requiring an overt act “is sufficient to establish the contemporary definition of conspiracy

as such.” Maj. Op. at 13. If those state laws aligned with the federal law, I might be

inclined to agree. See Rangel-Castaneda, 709 F.3d at 377. But here those state laws must

be weighed against the overwhelming consensus among federal conspiracy statutes that

no overt act is necessary.

26
 Even more fundamentally, “allow[ing] a straw-poll of the states [to] determine the

meaning of federal law” would “run contrary to federal supremacy.” Etienne, 813 F.3d at

145 n.5. To hold that the plain meaning of a guideline must change whenever the number

of states shifts leaves federal law in an unstable condition. The majority inexplicably pays

scant attention to federal law in its bean-counting exercise even though the text it claims

to be applying is the United States Sentencing Guidelines. Our government is designed to

accord respect for the distinct provinces of the dual sovereigns that comprise it. The

principle of dual sovereignty reverberates in such varied contexts as the Fifth

Amendment’s guarantee against double jeopardy, see Heath v. Alabama, 474 U.S. 82, 90-

91 (1985), and the anti-commandeering principle, see New York v. United States, 505

U.S. 144, 176 (1992). So too does that principle resonate in cases such as this. While the

criminal laws of the states may help illuminate the definitions of generic federal crimes, it

is ultimately the intent of the federal lawmakers that must control. And in this case, that

intent could not be more clear.

III.

The majority has ignored plain Guidelines text, its own prior precedent, and

elementary common sense. It has embraced the principle that law must of necessity be

counterintuitive, that the straightforward must yield to the convoluted, and that the

obscure must supersede the obvious, as though clouds had been summoned to hide the

sun. And this, finally, is what we have come to: plotting to murder one’s fellow human

beings is not a crime of violence.

Heaven help us.

27