PUBLISHED

                      UNITED STATES COURT OF APPEALS
                          FOR THE FOURTH CIRCUIT


                                      No. 17-7416


UNITED STATES OF AMERICA,

                    Plaintiff - Appellee,

             v.

ROGER DALE CHARLES, II,

                    Defendant - Appellant.


Appeal from the United States District Court for the Western District of North Carolina,
at Bryson City. Martin K. Reidinger, District Judge. (2:04-cr-00027-MR-DLH-1; 1:16-
cv-00136-MR)


Argued: May 7, 2019                                             Decided: July 26, 2019


Before WILKINSON and NIEMEYER, Circuit Judges, and DUNCAN, Senior Circuit
Judge.


Affirmed and remanded by published opinion. Judge Niemeyer wrote the opinion, in
which Judge Wilkinson and Senior Judge Duncan joined.


ARGUED: Ann Loraine Hester, FEDERAL DEFENDERS OF WESTERN NORTH
CAROLINA, INC., Charlotte, North Carolina, for Appellant. Anthony Joseph Enright,
OFFICE OF THE UNITED STATES ATTORNEY, Charlotte, North Carolina, for
Appellee. ON BRIEF: Anthony Martinez, Federal Public Defender, FEDERAL
DEFENDERS OF WESTERN NORTH CAROLINA, INC., Charlotte, North Carolina,
for Appellant. R. Andrew Murray, United States Attorney, OFFICE OF THE UNITED
STATES ATTORNEY, Charlotte, North Carolina, for Appellee.




                                       2
NIEMEYER, Circuit Judge:

      The “concurrent sentence doctrine” authorizes a court to leave the validity of one

concurrent sentence unreviewed when another is valid and carries the same or greater

duration of punishment so long as there is no substantial possibility that the unreviewed

sentence will adversely affect the defendant or, stated otherwise, so long as it can be

foreseen with reasonable certainty that the defendant will suffer no adverse collateral

consequences by leaving it unreviewed.      In this case, we find that this standard is

satisfied when the only potential harm to the defendant is grounded on unrealistic

speculation.

      In 2005, Roger Charles, II, was convicted of possession of more than 50 grams of

cocaine base with intent to distribute, in violation of 21 U.S.C. § 841(a)(1), and

possession of a firearm by a felon, in violation of 18 U.S.C. § 922(g)(1). On his drug-

trafficking offense, he was sentenced to 360 months’ imprisonment and 10 years of

supervised release, and on his firearm offense he was sentenced to 360 months’

imprisonment and 3 years of supervised release, with the two sentences to be served

concurrently. Both sentences were enhanced by prior felony convictions. On the drug-

trafficking offense, the prior convictions served as the basis for finding Charles to be a

career offender under U.S.S.G. § 4B1.1(a), and on the firearm offense, they served as a

basis for finding Charles to be an armed career criminal under 18 U.S.C. § 924(e) (the

Armed Career Criminal Act (“ACCA”)).

      Ten years later, after the Supreme Court handed down its decision in Johnson v.

United States, 

(2015), which narrowed the category of offenses that can

                                            3
be used to enhance sentences under ACCA, Charles filed a motion under 28 U.S.C.

§ 2255 to reduce the sentences for both of his convictions. The district court found

Charles’s sentence for drug trafficking to be valid and declined to review the sentence for

his firearm conviction, applying the concurrent sentence doctrine. The court concluded

that leaving the firearm sentence unreviewed would cause Charles no “realistic potential

adverse collateral consequence.”

       On appeal, Charles contends that the district court erred in applying the concurrent

sentence doctrine because he would indeed be exposed to the possibility of a collateral

consequence if his firearm sentence were not reviewed and reduced.             Because the

collateral consequence posited by Charles rests on unrealistic speculation, however, we

affirm the district court’s application of the concurrent sentence doctrine in the

circumstances presented to it.

       Nonetheless, after the parties filed their briefs in this appeal, Congress enacted the

First Step Act of 2018, and Charles now claims that leaving his firearm sentence

unreviewed will also have the consequence of denying him relief under that Act. Given

this development, we remand to allow the district court, in the first instance, to consider

Charles’s argument that he is eligible for a reduction of his drug-trafficking sentence

under the First Step Act.


                                              I

       After Charles was convicted of drug trafficking and the illegal possession of a

firearm, the Probation Officer prepared a presentence report finding that Charles was a


                                             4
career offender under U.S.S.G. § 4B1.1(a) and, for purposes of sentencing on the firearm

conviction, an armed career criminal under ACCA, 18 U.S.C. § 924(e). The report based

these findings on Charles’s prior Florida felony convictions for (1) battery of a law

enforcement officer, (2) armed burglary with a deadly weapon, (3) escape, and

(4) resisting an officer with violence.   As a career offender, the presentence report

concluded that Charles was subject to an advisory sentencing range of 360 months’ to life

imprisonment. And as an armed career criminal, Charles was subject to a mandatory

minimum sentence for his firearm conviction of 15 years’ imprisonment and a maximum

of life imprisonment, see 18 U.S.C. § 924(e)(1), which was an enhancement of what

would otherwise have been a maximum sentence of 10 years’ imprisonment, see 

      The district court adopted the presentence report and classified Charles as both a

career offender and an armed career criminal. It then sentenced him to 360 months’

imprisonment and 10 years of supervised release on the drug-trafficking offense and 360

months’ imprisonment and 3 years of supervised release on the firearm offense, with both

sentences to run concurrently.

      Charles’s convictions and sentences were affirmed on direct appeal. See United

States v. Charles, 195 F. App’x 133 (4th Cir. 2006) (per curiam).

      Ten years later, in May 2016, Charles filed a motion under 28 U.S.C. § 2255 to

vacate his sentences based on the Supreme Court’s decision in 

, which held that the “residual clause” used to define a sentence-enhancing “violent

felony” in ACCA was impermissibly vague and therefore that “imposing an increased

                                            5
sentence under [ACCA’s] residual clause . . . violate[d] the Constitution’s guarantee of

due process,” and Welch v. United States, 

, 1265 (2016), which made

Johnson retroactively applicable to cases on collateral review. In his § 2255 motion,

Charles argued that because three of his four prior convictions qualified as predicate

convictions only under ACCA’s residual clause, they were illegally used to enhance his

firearm sentence. And by the force of Johnson’s reasoning, he also argued that those

same offenses could not be used to designate him as a career offender under the

Sentencing Guidelines for purposes of enhancing his sentence for drug trafficking.

      The district court placed Charles’s § 2255 motion in abeyance pending the

Supreme Court’s decision in Beckles v. United States, which ultimately held that “the

advisory Guidelines are not subject to vagueness challenges under the Due Process

Clause,” 

, 890 (2017).

      Based on Beckles, the government filed a motion to dismiss Charles’s § 2255

motion. It noted first that after Beckles, it was clear that Charles’s challenge to his

career-offender designation had no merit and that his 360-month sentence on the drug-

trafficking offense was therefore valid. Second, it argued with regard to his firearm

sentence that even if Charles had been improperly classified as an armed career criminal

under ACCA, the court need not address Charles’s challenge to that sentence because

Charles would still have to serve the valid 360-month sentence on his drug-trafficking

conviction and a decision not to review the ACCA enhancement on his firearm

conviction would have no adverse collateral consequence.



                                            6
      In response to the government’s motion, Charles contended that he had been

“sentenced to an unconstitutional, over-the-maximum sentence for his [firearm]

conviction” and that “[t]he concurrent sentence doctrine [was] inapplicable” because the

government had failed to show that he would “suffer no harm” were his ACCA-enhanced

sentence allowed to stand.    He argued that “leaving [his] ACCA sentence in place

[would] carry collateral consequences” by subjecting him to the possibility of a greater

sentence for violating the terms of his supervised release. Specifically, he explained that

because he was sentenced for his firearm conviction under ACCA, the conviction is

classified as a Class A felony, rather than as a Class C felony, and should his supervised

release be revoked, this difference in classification could subject him to an additional 3

years of imprisonment for the supervised release violation. See 18 U.S.C. §§ 3559(a),

3583(e)(3). While Charles acknowledged that he also had a concurrent 10-year term of

supervised release for his drug-trafficking conviction, which is a Class A felony, he

argued that he still could “suffer potential adverse consequences from the erroneous

classification of his [firearm conviction] as a Class A felony, because a district court

[would have] discretion to impose consecutive prison terms upon revoking concurrent

terms of supervised release.” (Emphasis added). Based on this possibility, Charles

argued that “the government cannot show that [he] [would] suffer no harm by leaving the

unconstitutional ACCA sentence in place.”

       By an opinion and order dated September 1, 2017, the district court granted the

government’s motion to dismiss Charles’s § 2255 motion. First, the court noted that

Charles’s challenge to his career offender status had been foreclosed by Beckles and

                                            7
therefore that his 360-month sentence on the drug-trafficking offense was valid. And

second, invoking the concurrent sentence doctrine, the court declined to decide whether

Charles’s 360-month term of imprisonment for his firearm conviction was invalid under

Johnson because Charles had been validly sentenced to the same term for his drug-

trafficking offense and the two terms of imprisonment were imposed concurrently. As

for Charles’s argument that he could face adverse collateral consequences should he later

commit a serious violation of his supervised release, the court explained that “[w]hile the

scenario posed by [Charles] [was] technically possible, the factors that would have to line

up to support such a result make it implausible.” It noted:

       The maximum Guidelines range for the most serious violation of
       supervised release calls for a term of incarceration of 51 to 63 months. See
       U.S.S.G. § 7B1.4(a). In order for that Guidelines range to apply, [Charles]
       would have to commit a Grade A violation while on supervised release; and
       that Grade A violation would have to be so serious that the Court would
       have to upwardly vary in order to impose a sentence of more than 84
       months, i.e., an upward variance of at least one-third over the top of the
       recommended Guidelines range.

The court concluded that “[s]uch a hypothetical scenario is simply too speculative to be

considered a realistic potential adverse collateral consequence requiring review of

[Charles’s] sentence on [the firearm conviction].” It accordingly denied Charles’s § 2255

motion and granted the government’s motion to dismiss it, but it also issued a certificate

of appealability with respect to the sentence on Charles’s firearm offense.


                                            II

       When sentenced, Charles was given a sentence of 360 months’ imprisonment and

10 years of supervised release on his drug-trafficking offense and a concurrent sentence

                                             8
of 360 months’ imprisonment and 3 years of supervised release on his firearm offense.

Relying on Johnson, he challenged the length of his firearm sentence. The district court

declined to address his challenge because the sentence for his drug-trafficking offense

was valid and, under the concurrent sentence doctrine, it need not address the claimed

error in the concurring sentence, as any error was harmless.

       Charles contends that the district court erred in taking that position, arguing that he

could be adversely affected by his invalid firearm sentence. He hypothesizes that after he

serves his 360-month sentence, he could commit a serious violation of his supervised

release, thereby implicating the supervised release on two Class A convictions, and that

the court could then stack the two maximum 5-year revocation sentences by having them

run consecutively, for a total term of 10 years. By contrast, if the sentence for his firearm

conviction were corrected, as he argues it should be, the conviction would be reduced

from a Class A to a Class C felony, see 18 U.S.C. § 3559(a), and the maximum sentence

for violating the terms of supervised release on the conviction would then be 2 years, see

Thus, even if the court were to stack the sentences, he could receive at

most 5 years for violating the supervised release on his Class A felony (drug trafficking)

and only 2 years for his firearm sentence (a Class C felony), for a total of 7 years. The 7

years would thus be 3 years less than the 10 years that would be possible if his firearm

offense were not reviewed.

       Agreeing with the government, the district court concluded that the harm that

Charles posited was too speculative and remote to undermine the proper harmlessness

analysis.   It said, “[s]uch a hypothetical scenario is simply too speculative to be

                                              9
considered a realistic potential adverse collateral consequence requiring review of

[Charles’s] sentence on [his firearm conviction].”

       The issue before us thus turns on the proper understanding of the concurrent

sentence doctrine and whether, in the circumstances of this case, it allowed the court to

leave unreviewed Charles’s challenge to his firearm sentence.

       At the outset, we note that the concurrent sentence doctrine rests on the same

rationale underlying harmless-error review — namely, the recognition that to help

promote the overall functioning of our justice system, courts should “conserve judicial

resources by . . . cleans[ing] the judicial process of prejudicial error without becoming

mired in harmless error.” United States v. Hasting, 

, 501 (1983) (citation

omitted); see also Benton v. Maryland, 

, 798–99 (1969) (White, J.,

concurring) (explaining that “[i]n a time of increasingly congested judicial dockets,” the

concurrent sentence doctrine “is not a rule of convenience to the judge, but rather of

fairness to other litigants”).

       In United States v. Truong Dinh Hung, 

(4th Cir. 1980), we

articulated the doctrine as follows:

       The concurrent sentence rule provides that where a defendant receives
       concurrent sentences on plural counts of an indictment and where the
       conviction on one count is found to be good, a reviewing court need not
       pass on the validity of the defendant’s conviction on another count. . . .
       [T]he application of the rule has, though, been restricted to situations where
       there is no substantial possibility that the unreviewed conviction will
       adversely affect the defendant’s right to parole or expose him to a
       substantial risk of adverse collateral consequences.




                                            10


(emphasis added). In that case, we applied the doctrine after considering

whether the unreviewed conviction could affect the defendants’ eligibility for parole;

their place or conditions of confinement; or their “salient factor score[s],” as would be

determined by the Parole Commission. Concluding that the defendants would not suffer

any such adverse collateral consequences, we declined to review the merits of one of the

defendants’ convictions. See 

       Truong Dinh Hung has since remained the leading authority for the application of

the doctrine in our circuit, although our formulation has sometimes varied in our effort to

summarize it. In United States v. Mumford, 

(4th Cir. 1980), we cited

Truong Dinh Hung to rule that, because the defendant in that case “was sentenced to

concurrent jail terms on the first five counts of mail fraud” and because there was “no

substantial possibility” that one of those convictions would “subject him to adverse

collateral consequences,” “under the concurrent sentence doctrine it would not be

necessary to review the validity of [that] conviction.” 

n.7. In United States v.

Webster, 

(4th Cir. 1981), we again relied on Truong Dinh Hung, albeit with

a slightly different formulation of the doctrine. We stated in Webster that “[w]e may

assume, for the purposes of this case, that under the ‘concurrent sentence’ doctrine, it is

permissible to affirm the convictions without investigating whether they were for some

reason erroneous, provided we can foresee with reasonable certainty that no adverse

collateral consequences will redound to the defendant.” 

(emphasis added).

In United States v. Walker, 

(4th Cir. 1982), we relied on Truong Dinh

Hung again to articulate the doctrine, stating that “a reviewing court need not pass upon

                                            11
the validity of [a] defendant’s conviction for additional counts once it has affirmed his

conviction on one count, but only if ‘there is no substantial possibility that the

unreviewed conviction will adversely affect the defendant’s right to parole or expose him

to a substantial risk of adverse collateral consequences.’” 

(emphasis added).

And in United States v. Hill, 

(4th Cir. 1988), we relied on Truong Dinh

Hung to summarize the doctrine, yet more briefly, stating that it provides “that where a

defendant is serving concurrent sentences and one conviction is shown to be valid, the

court may decline to pass upon the validity of the other conviction. It requires a showing

that the defendant will suffer no harm by letting both the valid and unreviewed

convictions stand.” 

In short, while some of our language differs, none of our

cases indicate that one formulation is wrong or even distinct from any other.

      Charles focuses on the different language used in Truong Dinh Hung and Webster

to claim that we should limit application of the rule to those cases in which a court “can

foresee with reasonable certainty” that there will be no adverse collateral consequences,

as was articulated in 

. But when Webster’s formulation is

rightly analyzed as a summary of Truong Dinh Hung’s, there is no meaningful difference

between the two, and Charles’s entreaty provides him with no meaningful help.

      In an effort to describe the proper degree of certainty that harm will not result

from an unreviewed sentence or conviction, we stated in Truong Dinh Hung that there

must be “no substantial possibility” of harm. And in Webster, we stated that there must

be a “reasonable certainty” of no harm. Each articulation of the standard is designed to

limit by a measure the requirement of “no possibility” or absolute “certainty.” Thus, “no

                                            12
possibility” is limited to “no substantial possibility,” and absolute “certainty” of no harm

is limited to “reasonable certainty” of no harm.        The message conveyed by both

formulations is that when we consider future events, there must always be some

circumspection that eliminates from the forward-looking analysis unrealistic speculation

of adverse collateral consequences. See Eason v. United States, 

, 1124 (8th

Cir. 2019) (recognizing that if “highly speculative adverse collateral consequences” were

to bar courts from applying the concurrent sentence doctrine, there would be nothing left

to “this useful rule” (cleaned up)). Thus, when determining the potential consequences of

leaving a sentence unreviewed, we must not engage in unrealistic speculation about the

possibility of future harm.

       In view of intervening Supreme Court decisions, however, we must recognize an

additional limitation to the concurrent sentence doctrine. Because 18 U.S.C. § 3013

imposes a special monetary assessment for each count of conviction, the Court has held

that the concurrent sentence doctrine cannot be applied to avoid reaching a defendant’s

challenge to one of his federal convictions, reasoning that the additional punishment of

the assessment means that the sentences are not truly concurrent. See Ray v. United

States, 

, 737 (1987) (per curiam). Moreover, even beyond the existence of a

special assessment for each of multiple convictions, the Court has indicated that there are

other collateral consequences attaching to unreviewed convictions that limit the

applicability of the concurrent sentence doctrine. In Rutledge v. United States, 

(1996), the Court noted:



                                            13
       The separate conviction, apart from the concurrent sentence, has potential
       adverse collateral consequences that may not be ignored. For example, the
       presence of two convictions on the record may delay the defendant’s
       eligibility for parole or result in an increased sentence under a recidivist
       statute for a future offense. Moreover, the second conviction may be used
       to impeach the defendant’s credibility and certainly carries the societal
       stigma accompanying any criminal conviction.

(cleaned up).

       Yet, while the concurrent sentence doctrine cannot be applied to avoid reviewing

the validity of one of a defendant’s convictions, the doctrine still has continuing force as a

species of harmless-error review where a defendant seeks to challenge the legality of a

sentence that was imposed for a valid conviction, but where the challenged sentence runs

concurrently with a valid sentence of an equal or greater duration. See, e.g., 

; In re Williams, 

, 1356–57 (11th Cir. 2016); United States v.

Harris, 

, 1139 (10th Cir. 2012); United States v. Pierre, 

, 88,

90–91 (1st Cir. 2007).

       In this case, Charles’s conviction and 360-month sentence for drug trafficking is

valid, and he does not suggest otherwise on appeal. Rather, he challenges his concurrent

360-month sentence — not his conviction — for illegally possessing a firearm. Thus,

given Charles’s drug-trafficking sentence, his challenge to his firearm sentence can have

no effect on his total term of imprisonment. Likewise, he would still face 10 years of

supervised release imposed as part of his drug-trafficking sentence, which is longer than

the 3-year term of supervised release that was imposed for his firearm conviction.

       Charles argues nonetheless that under an extremely unrealistic hypothetical he has

posited, he could be adversely affected if his firearm sentence were not reviewed and

                                             14
corrected. Specifically, he describes a situation where, after serving a 30-year term of

imprisonment and being released from prison when he is nearly 60 years old, he would

commit a violation of his supervised release within the first three years of his release —

the period during which he would be serving two concurrent terms of supervised release

on the two sentences. Moreover, his violation of his supervised release would have to be

so serious that the district court would not be satisfied with the Guidelines’ recommended

sentence of 51 to 63 months’ imprisonment for the most serious grade violation, see

U.S.S.G. § 7B1.4(a), but would decide that it needed to vary upward to impose a total

revocation sentence that was greater than 84 months’ imprisonment, which it could only

achieve by stacking near-maximum revocation sentences by imposing them

consecutively. By contrast, were Charles to succeed in reducing his firearm sentence, 84

months’ imprisonment would be the statutory maximum that he could face for a violation

of his supervised release. In short, Charles argues that he is exposed to the risk of serving

up to 120 months’ imprisonment — two 60-month sentences stacked consecutively — for

a future violation of his supervised release, which would be 36 months more than if his

firearm sentence were reviewed and reduced.

       Such a hypothetical scenario, however, is highly speculative and unrealistic. It

would require: (1) that Charles commit a violation of his supervised release; (2) that his

violation occur in the 3-year window after his release from prison when he would be

serving both release terms; (3) that his violation be extraordinarily serious; (4) that the

court find the Guidelines’ recommendation of 51 to 63 months’ imprisonment for the

most serious violation to be insufficient to punish the violation; and (5) that the court find

                                             15
it necessary to issue a variance sentence that would be at least 22 months longer than the

highest recommended sentence to reach a sentence in excess of 84 months’

imprisonment. Not only is the prospect of all of this happening remote, but, most

significantly, Charles’s hypothetical could only arise if he were to choose to violate the

terms of his supervised release. Thus, his posited adverse effect would be entirely within

his ability to avoid. As the Eighth Circuit recently stated in identical circumstances,

“[The hypothetical] could not occur unless [the defendant] chooses to commit serious

violations of law during his future term of supervised release.          Thus, the adverse

consequences are entirely within [his] control to avoid.         Literal application of such

speculative consequences, resting upon a supposition of [the defendant’s] continued

criminality, would effectively bar the application of the concurrent sentence rule.”

(cleaned up).

       In short, we conclude that in the circumstances presented to the district court, it

did not abuse its discretion by applying the concurrent sentence doctrine and declining to

review Charles’s sentence for illegal possession of a firearm.


                                            III

       Following the parties’ briefing on appeal, however, Congress enacted the First

Step Act of 2018, which, among other things, allows district courts to apply the Fair

Sentencing Act of 2010 retroactively to sentences for certain crimes committed before

August 3, 2010. The Fair Sentencing Act reduced the disparity between sentences for

crack cocaine offenses and powder cocaine offenses. In a letter to this court, Charles


                                            16
noted this development and argued that, as a result, he will suffer a new collateral

consequence if we do not review his firearm sentence because leaving that sentence

unreviewed will effectively deny him the benefit of the First Step Act. As he reasons,

even if he were granted a reduction of his drug-trafficking sentence under the First Step

Act, such a reduction would be meaningless because his firearm sentence of 360 months’

imprisonment would remain in place, unreviewed.

       The government contends that Charles is not eligible for a reduction of his drug-

trafficking sentence under the First Step Act and therefore that its enactment “does not

undercut the district court’s conclusion that no realistic consequence would flow from its

decision not to review” Charles’s firearm sentence.

       We conclude that because the impact of the First Step Act was not considered by

the district court, as the Act was enacted after the court ruled, the better course is to

remand this case to the district court for it to consider the effect of the Act in the first

instance. Should the district court conclude that Charles is eligible for a reduction of his

drug-trafficking sentence under the Act and further that such a reduction would be

warranted, this remand will enable the court to consider the merits of Charles’s § 2255

motion challenging his firearm sentence. Cf. 

(recognizing, in a

case in which the defendant had been sentenced to 15 years for burglary and a concurrent

sentence of 5 years for larceny, that “should the length of the burglary sentence be

reduced to less than five years,” the defendant “would then clearly have a right to have

his larceny conviction reviewed”). If, on the other hand, the court were to decide not to

reduce Charles’s sentence for his drug-trafficking offense under the First Step Act, it

                                            17
could then rely on the concurrent sentence doctrine — as it previously applied the

doctrine and as we affirm here — to dismiss Charles’s § 2255 motion.

                                                      AFFIRMED AND REMANDED




                                          18