Assessing the Impact of Johnson v. United States on the Void-for- Vagueness Doctrine

By Carissa Hessick
University of North Carolina School of Law
Oct 24, 2016

Johnson v. United States, 135 S. Ct. 2551 (2015), held that the so-called “residual clause” of the Armed Career Criminal Act (“ACCA”) was unconstitutionally vague.  Johnson generated a large amount of litigation in the federal courts.  Less than a year after it was decided, the Supreme Court decided another Johnson case, Welch v. United States, 136 S. Ct. 1257 (2016), which held that the rule in Johnson should be applied retroactively to those defendants whose convictions and sentences have already become final.  The Supreme Court has also agreed to hear two new Johnson cases in the 2016 Term.

Johnson raised important constitutional doubts about federal statutes that employ the so-called “categorical approach” to classifying criminal conduct, as well as doubts about certain Federal Sentencing Guidelines.  This short essay describes Johnson and explores the Johnson-related issues that the Court will hear this Term.  

Johnson and the Vagueness Doctrine

Johnson made three changes to the vagueness doctrine.  First, Johnson held that a law may be unconstitutionally vague even if there are some cases that could be easily resolved under the statutory language.  135 S. Ct. at 2560-61.  Some of the Court’s previous cases had said that a statute was void for vagueness only if it was vague in all applications.  See, e.g., Hoffman Estates v. Flipside, Hoffman Estates, Inc., 455 U.S. 489, 494-95 (1982).  But the Johnson Court rejected this language, stating that it was inconsistent with the Court’s holdings in prior cases.  In particular, the Court pointed to United States v. L. Cohen Grocery Co., 255 U.S. 81 (1921), which held that a law prohibiting “unjust or unreasonable rates” for grocers was unconstitutionally vague, and Coates v. Cincinnati, 402 U.S. 611 (1971), which struck down a law prohibiting people on sidewalks from conducting themselves “in a manner annoying to persons passing by.”  The Court noted that these holdings were incompatible with a rule that “a vague provision is constitutional merely because there is some conduct that clearly falls within the provision’s grasp.”  135 S. Ct. at 2561.  As the Court noted, charging a thousand dollars for a pound of sugar is clearly “unjust or unreasonable,” and spitting in someone’s face on a sidewalk is obviously annoying.  Id.

Second, Johnson made clear that the vagueness doctrine applies not only to statutes that criminalize conduct, but also to statutes that set punishments.  In its brief, the government argued that the vagueness doctrine need not be applied with the same force in cases involving sentencing statutes, as opposed to statutes that define criminal conduct.  This argument found some support in Chapman v. United States, 500 U.S. 453, 467-68 (1991).  Justice Alito echoed this reasoning in his dissent.  But it did not persuade the majority.  After reciting the basic underpinnings of the vagueness doctrine, the majority opinion simply stated: “These principles apply not only to statutes defining elements of crimes, but also to statutes fixing sentences.”  135 S. Ct. at 2557.

 Third, Johnson suggested that the so-called categorical approach to interpreting certain federal statutes raises particular vagueness problems.  The categorical approach—which courts have used to interpret not only the ACCA, but also other federal statutes—requires judges to assess some feature of the defendant’s crime based on the “ordinary case” of an individual committing the crime, rather than the facts surrounding a particular defendant’s crime.  Under the residual clause of the ACCA, for example, a judge must determine whether a defendant had been convicted of a crime that ordinarily “involves conduct that presents a serious potential risk of physical injury to another.”  That is why, in assessing Samuel Johnson’s state law conviction for possession of a short-barreled shotgun, the district court judge in Johnson did not ask about the manner in which Johnson possessed his shotgun.  Whether Johnson was brandishing the shotgun during drug deals or whether he kept it locked in his basement played no role in the judge’s determination whether Johnson’s possession of the short-barreled shotgun “present[ed] a serious potential risk of physical injury to another.”  Instead, the judge had to decide whether the “ordinary case” of possessing a short-barreled shotgun posed such a risk.

In the long term, it may be easier for defendants to prevail on vagueness claims because the Court has made clear that a law need not be “vague in all of its applications” in order to violate the Due Process Clause.  In the short term, federal courts must now decide whether other federal laws that use the categorical approach are also void for vagueness.  Those laws include: 18 U.S.C. § 16(b), 18 U.S.C. § 924(c), and 18 U.S.C. § 3559(c), as well as several Federal Sentencing Guidelines.  (For an interesting argument that Johnson also render’s California’s second degree felony murder rule unconstitutional, see

Johnson and Other Federal Statutes

There are several federal statutes that include language that is similar, but not identical to, the ACCA residual clause.  In the wake of Johnson, the lower federal courts have struggled with whether the language in these other statutes is unconstitutionally vague.  The Supreme Court recently granted certiorari in a case that involves one of those statutes – 18 U.S.C. § 16(b).  There are at least two other statutes – 18 U.S.C. § 924(c) and 18 U.S.C. § 3559(c) – that may also be unconstitutionally vague.

18 U.S.C. § 16(b)

In Lynch v. Dimaya, No. 15-1498, the Supreme Court will address whether 18 U.S.C. § 16(b), is unconstitutionally vague.  Section 16(b) is a general definitional statute, which is incorporated into several other federal laws, including immigration laws.  Section 16(b) defines the term “crime of violence” to include “any other offense that is a felony and that, by its nature, involves a substantial risk that physical force against the person or property of another may be used in the course of committing the offense.”  Like the ACCA, § 16(b) identifies other ways that a crime could be a crime of violence, but this language acts as a residual clause.  And courts have used the categorical approach in deciding whether a defendant’s conduct qualifies as a “crime of violence” under § 16(b).  See, e.g., Leocal v. Ashcroft, 543 U.S. 1, 10–11 (2004).

In the wake of Johnson, the federal courts split over whether § 16(b) is unconstitutionally vague.  Several courts have held that this language is unconstitutionally vague.  See Golicov v. Lynch, -- F.3d --, 2016 WL 4988012 (10th Cir. Sept. 19, 2016); Shuti v. Lynch, 828 F.3d 440 (6th Cir. 2016); Dimaya v. Lynch, 803 F.3d 1110 (9th Cir. 2015), cert. granted, -- U.S.L.W. – (U.S. Sept. 29, 2016) (No. 15-1498); United States v. Vivas-Ceja, 808 F.3d 719 (7th Cir. 2015); see also In re Hubbard, 825 F.3d 225 (4th Cir. June 8, 2016) (finding that Johnson “might” render § 16(b) unconstitutionally vague).  But other courts disagree.  The Fifth Circuit, for example, concluded that this language is not unconstitutionally vague.  See United States v. Gonzalez-Longoria, 831 F.3d 670 (5th Cir. 2016) (en banc).  And the Second Circuit suggested in a recent panel opinion that this is the better approach.   United States v. Hill, 832 F.3d 135, 149-50 (2d Cir. 2016).

In Lynch v. Dimaya, the Supreme Court has taken up not only the question whether the residual clause of § 16(b) is unconstitutionally vague, but also how the vagueness doctrine applies to non-criminal laws.  Indeed, whether the vagueness doctrine applies with the same force in the immigration context as it does in the criminal context is a threshold question in Dimaya.  That is because Dimaya is a removal case.  The respondent in Dimaya is a non-citizen who was convicted of a crime.  The Immigration and Naturalization Act provides for the removal (i.e., deportation) of non-citizens who commit certain crimes, including crimes of violence as defined by § 16(b).  Immigration cases are civil cases, and the Supreme Court has long held that the vagueness doctrine does not apply with equal force to civil statutes.  Consequently, the U.S. government is arguing in Dimaya that the vagueness doctrine should not apply with equal force in removal proceedings.

There are two complications this argument about the relative strength of the vagueness doctrine in the immigration context.  The first is that a Supreme Court case from decades ago, Jordan v. De George, 341 U.S. 223 (1951), suggested that vagueness may apply with the same force in deportation proceedings.  (Removal is the modern term for deportation.)  The second complication is that § 16(b) is a criminal statute.  Congress elected to reference a criminal statute, rather than writing a new definition for which crimes are deportable.  So if the Court accepts the government’s argument that the vagueness doctrine should not apply with the same force in the immigration context, then we might have a situation where a statute is unconstitutionally vague in some contexts but not in others.  That seems like an odd outcome.

In addition to the threshold question about the strength of the vagueness doctrine in the immigration context, Dimaya also squarely raises the question whether the ruling in Johnson also indicates that § 16(b) is unconstitutionally vague.  The courts that have said § 16(b) is not vague have made both textual and practical arguments.  The first textual argument that the courts have made is that the ACCA spoke in terms of a “risk of physical injury” while § 16(b) speaks in terms of a risk that “physical force . . . may be used.”  Those courts that have said § 16(b) is not vague have said that the language in § 16(b) is more definite than the language in the ACCA.

It is difficult to understand why this is so.  The language from § 16(b) may well be narrower than the language from the ACCA.  By speaking about the use of physical force, § 16(b) seems to include only intentional or maybe reckless conduct.  In contrast, by speaking in terms of the risk of injury, the ACCA arguably includes situations where injury could result from accidental or negligent conduct.  To be clear, it is not certain that § 16(b) is more narrow.  The ACCA speaks in terms of the risk of physical injury to another person, whereas § 16(b) speaks in terms of the risk that physical force will be used “against the person or property of another.”  It is certainly possible that force against property occurs so frequently that it compensates for the fact that § 16(b) appears to exclude negligent or accidental conduct.  But even if § 16(b) is narrower, that does not mean it is more definite.  Courts have to assess in the abstract the risk that physical injury will occur and the risk that physical force will be used.  The language from § 16(b) offers no more guidance than the language from the ACCA about how to make that assessment.

The second textual argument that courts have made is much stronger.  Section 16(b) specifically says that courts must assess the risk that physical force will be used “in the course of committing the offense.”  The ACCA contains no similar limitation.  And the fact that the ACCA did not contain such a limitation seemed to matter to the Johnson Court.  SeeJohnson, 135 S. Ct. at 2559 (“When deciding whether unlawful possession of a short-barreled shotgun is a violent felony, do we confine our attention to the risk that the shotgun will go off by accident while in someone's possession? Or do we also consider the possibility that the person possessing the shotgun will later use it to commit a crime?”).

The final textual argument focuses on the list of enumerated crimes in the ACCA—burglary, arson, extortion, and crimes that involve the use of explosives.  Section 16(b) does not include any similar list.  The Johnson Court noted that the ACCA residual clause appeared to incorporate the amount of risk posed by this list of crimes into the qualitative standard.  The qualitative standard is preceded by the word “otherwise,” which suggests that the enumerated crimes also “present[] a serious potential risk of physical injury.”  But, as the Johnson Court noted, the risk these various enumerated crimes pose is far from clear, and certainly not uniform.  135 S. Ct. at 2558.  While arson and crimes involving explosives presumably create a large risk of physical injury, it is far from clear what risk is created by extortion or burglary.  This confusing list, according to the Johnson majority, made it more difficult to interpret the qualitative standard in the residual clause.  

Because § 16(b) does not include a list of enumerated crimes, several courts have said that it is not as vague as the residual clause of the ACCA.  This textual argument makes sense.  After all, the Johnson Court certainly discussed the ACCA’s enumerated crimes in explaining why the ACCA residual clause was unconstitutionally vague.  But it is far from clear whether the list of enumerated crimes was necessary to the reasoning in Johnson.  Indeed, there are portions of the opinion that suggest it is the combination of a qualitative standard with the categorical approach that made the ACCA residual clause vague.  E.g., 135 S. Ct. at 2557 (“Two features of the residual clause conspire to make it unconstitutionally vague. In the first place, the residual clause leaves grave uncertainty about how to estimate the risk posed by a crime. It ties the judicial assessment of risk to a judicially imagined “ordinary case” of a crime, not to real-world facts or statutory elements. How does one go about deciding what kind of conduct the “ordinary case” of a crime involves?”).  

In addition to the textual differences between the ACCA and § 16(b), courts have also relied on the fact that § 16(b) has not led to the same confusion and disagreement in the courts as did the ACCA in concluding that § 16(b) is not unconstitutionally vague.  These courts have noted that the meaning of the ACCA residual clause resulted in multiple circuit splits, which in turn necessitated multiple Supreme Court rulings on how to interpret the ACCA.

How often courts disagree about the meaning of § 16(b) is hardly a good proxy for whether § 16(b) is vague.  That is because the frequency of circuit splits and Supreme Court certiorari grants depend on a number of factors other than statutory clarity.  And several of those factors are under the control of the government.  The government controls, for example, how often it brings charges (or takes other actions) under a particular statute, and how often it relies on various portions of those statutes.  For example, the government may have chosen to bring far more cases under the residual clause of the ACCA than under the residual clause of § 16(b).  A smaller volume of cases is likely to lead to fewer disagreements over the meaning of a statute.  The government also controls the circumstances under which it brings various charges.  If the government decided to bring § 16(b) charges in only a few types of cases with similar facts, but meanwhile decided to bring ACCA charges in many different circumstances, then courts were more likely to disagree in ACCA cases.  The same is true if the government decides to bring charges that closely track the statutory language or if the government decides to bring charges in more marginal cases.  As these examples demonstrate, the government’s strategic decisions can have a significant effect on whether circuit splits emerge and on whether the Supreme Court grants cert in cases involving a particular statute.  And those strategic decisions have nothing to do with whether a statute is sufficiently clear.

What is more, in the brief in opposition to cert in Lynch v. Dimaya, counsel for Dimaya identified a number of apparent circuit splits involving the meaning of § 16(b).  These splits were not identified in the courts of appeals decisions involving the vagueness of § 16(b).  So perhaps the meaning of § 16(b) is more divisive than the government or the courts of appeals originally believed.  That these cases were identified for the first time at the cert stage may also suggest that the defense-side lawyering we will see in Dimaya will be especially good.

Although Dimaya appears to have highly competent counsel, it is far from clear that the Supreme Court will decide that § 16(b) is unconstitutionally vague.  That is because only six Justices joined the majority opinion in Johnson.  Justice Alito dissented and Justices Kennedy and Thomas concurred in the judgment because they did not think that the ACCA was unconstitutionally vague.  Justice Scalia authored the majority opinion in Johnson, and his death means that there are only five justices remaining from the Johnson majority.  If any of those five agrees with the government on the threshold issue about the vagueness doctrine in the immigration context or on the various textual reasons that § 16(b) is distinguishable from the ACCA, then Dimaya will not be able to secure a ruling that § 16(b) is unconstitutionally vague.  And if the Court splits 4-4 on the vagueness issue, then the circuit split on this issue will persist.

18 U.S.C. § 924(c)

Interestingly, while the circuits have split over whether § 16(b) is unconstitutionally vague, there is no similar split over 18 U.S.C. § 924(c).  Section 924(c) sets mandatory minimum sentences for certain crimes involving weapons.  It contains language that is identical to the language in § 16(b), defining “crime of violence” to include “an offense that is a felony and . . . that by its nature, involves a substantial risk that physical force against the person or property of another may be used in the course of committing the offense.”  The Second Circuit has said that this language is not unconstitutionally vague.  United States v. Hill, 832 F.3d 135 (2d Cir. 2016).  So has the Eighth Circuit.  United States v. Prickett, -- F.3d --, 2016 WL 5799691 (8th Cir. Oct. 5, 2016).

To date, no federal court of appeals has held that this language is unconstitutionally vague.  (Some district court judges have held that it is vague.  E.g., United States v. Smith, 2016 WL 2901661, at *5–6 (D. Nev. May 18, 2016); United States v. Edmundson, 153 F. Supp. 3d 857 (D. Md. 2015).)  Importantly, both the Ninth Circuit and the Seventh Circuit, which held that § 16(b) is unconstitutionally vague have not yet decided whether § 924(c) is vague.  But the Sixth Circuit, which has held that § 16(b) is unconstitutionally vague, see Shuti v. Lynch, 828 F.3d 440 (6th Cir. 2016), has also held that § 924(c) is not vague, see United States v. Taylor, 814 F.3d 340 (6th Cir. 2016).  And the explanation that the Sixth Circuit has offered for these two apparently conflicting opinions is not entirely satisfactory.

The Sixth Circuit first addressed the potential vagueness of this language in United States v. Taylor, which was a § 924(c) case.  Taylor cited the same arguments against finding § 924(c) vague that some courts have accepted in the § 16(b) context.  814 F.3d at 375-79.  But when the Sixth Circuit was later confronted with those same arguments about § 16(b) in Shuti v. Lynch, it did not find those arguments persuasive.

The different outcomes in Taylor and Shuti are difficult to reconcile, and the Shuti opinion provides little in the way of assistance.  The Shuti Court claimed that its decision was “wholly consistent” with Taylor, Shuti, 828 F.3d at 449, even though the Taylor Court stated that “§ 16(b) appears identical to § 924(c)(3)(B) in all material respects” and even though Taylor specifically rejected the reasoning from the Ninth and Seventh Circuits finding § 16(b) vague, Taylor, 814 F.3d 340, 379.  The Shuti Court insisted that Taylor “appl[ied] Johnson’s real-world conduct exception to uphold the constitutionality of 18 U.S.C. § 924(c)(3)(B).”  Shuti, 828 F.3d at 450.  But the Taylor Court specifically stated that § 924(c), like the ACCA, “requires the application of a categorical approach, which requires courts to look at the ordinary case of the predicate crime.”  Taylor, 814 F.3d 340, 378.  The Shuti Court also noted that the Taylor decision was handed down before the Supreme Court decided Welch, and that the Welch opinion “made clear” that the “ACCA’s vagueness ‘rests in large part on its operation under the categorical approach.’” Shuti, 828 F.3d at 450 (quoting Welch, 136 S.Ct. at 1262.  One could certainly quibble with whether the Supreme Court meant to provide authoritative guidance about the meaning of its Johnson decision in the Welch case (which dealt only with whether te rule announced in Johnson was retroactive).  But perhaps this language from Shuti indicates that the Sixth Circuit may be willing to reconsider its position on § 924(c).

Of course, if courts were to apply § 924(c) without using the categorical approach, then it would not be unconstitutionally vague under Johnson.  The Johnson Court made clear that it did not question the constitutionality of statutes that require the application of qualitative standards to real world conduct.  Johnson v. United States, 135 S. Ct. 2551, 2561 (2015) (“As a general matter, we do not doubt the constitutionality of laws that call for the application of a qualitative standard such as “substantial risk” to real-world conduct.”).  To be clear, the Shuti Court’s claim that § 924(c) is, in fact, applied to real world conduct is highly questionable.  The courts of appeals that have denied vagueness challenges to § 924(c) have done so while acknowledging that § 924(c) is applied using a categorical approach rather than a facts and circumstances approach.  See United States v. Prickett, No. 15-3486, 2016 WL 5799691, at *1 (8th Cir. Oct. 5, 2016); United States v. Hill, 832 F.3d 135, 148 (2d Cir. 2016); Taylor, 814 F.3d 340, 378.  But courts could apply§ 924(c) to real world conduct, rather than using the categorical approach, in order to avoid the constitutional problems associated with the categorical approach in the wake of Johnson.  Indeed, several lower courts appear to have acknowledged this possibility.  See, e.g., United States v. Checora, 2015 WL 9305672 (D. Utah Dec. 21, 2015) (questioning application of categorical approach in Section 924(c)(3) cases, particularly in the context of deciding a pretrial motion to dismiss); United States v. McDaniels, 2015 WL 7455539, at *2-3 (E.D. Va. Nov. 23, 2015) (refusing to apply categorical approach for pre-trial motion to dismiss and sending determination of crime of violence to the jury as a question of fact).

The possibility of avoiding vagueness problems by jettisoning the categorical approach is possible in § 924(c) cases, but not in § 16(b) cases.  That is because § 924(c) cases involve the question whether a defendant committed a crime of violence in this particular case, whereas ACCA cases and § 16(b) cases (at least § 16(b) cases involving immigration removal proceedings) involve whether a defendant had previously been convicted of a crime of violence.  The Supreme Court has signaled that, if judges applied a qualitative standard to facts of a past crime, then such fact-finding might run afoul of the Sixth Amendment right to a jury trial.  The Court has pointed to this potential Sixth Amendment problem as a reason to retain the categorical approach for the remaining sections of the ACCA.  See Mathis v. United States, 136 S. Ct. 2243, 2252 (2016).  But the categorical approach is not necessary in § 924(c) cases.  In § 924(c) cases, judges could simply instruct juries to decide whether a defendant’s conduct in that particular case “involve[d] a substantial risk that physical force against the person or property of another may be used in the course of committing the offense.”  That would allow federal courts to avoid the constitutional doubt that currently infects § 924(c) cases.

To be clear, it does not appear that judges are currently instructing juries to apply the qualitative standard in § 924(c) to the facts of the particular case.  District courts appear to employ the categorical approach at trial either by instructing the jury that a particular crime is, as a matter of law, a crime of violence, or by instructing the jury that they must return a conviction on a § 924(c) count if they find a defendant guilty of one of the other charged crimes.  In either scenario, district court judges are engaging in a categorical analysis and treating the “crime of violence” question as a question of law rather than a question of fact for the jury.

If § 924(c) is unconstitutionally vague, it may also affect certain pre-trial detention decisions under 18 U.S.C. § 3142(e)(3)(B).  Section 3142(e)(3)(B) provides that defendants will be detained pending trial, rather than released on bail, in certain circumstances, including if a judicial officer finds that there is probable cause to believe that the person committed a crime under § 924(c).   If § 924(c) is unconstitutionally vague when it sets mandatory minimum sentences, it may also raise constitutional problems in the pre-trial detention context.  

But it is unclear whether the vagueness doctrine applies to statutes regarding pre-trial detention.  Johnson tells us that the vagueness doctrine applies “not only to statutes defining elements of crimes, but also to statutes fixing sentences.”  135 S. Ct. at 2557.  A statute about pre-trial detention, as a formal matter, neither defines a crime nor fixes punishment.  As a general matter, “the void-for-vagueness doctrine requires that a penal statute define the criminal offense with sufficient definiteness that ordinary people can understand what conduct is prohibited and in a manner that does not encourage arbitrary and discriminatory enforcement.” Kolender v. Lawson, 461 U.S. 352, 357 (1983). A vague pre-trial detention statute raises problems of arbitrary and discriminatory enforcement.  But it is unclear whether it raises problems associated with notice.  And so a court could reasonably conclude that a § 3142(e)(3)(B), in so far as it incorporates the residual clause from § 924(c), is not subject to a vagueness attack.

On the other hand, there are reasons to think that the vagueness doctrine does extend to statutes regarding pre-trial detention.  For one thing, the language from Johnson notwithstanding, the Supreme Court has not limited the vagueness doctrine to criminal laws that define crimes and fix punishments.  See, e.g., Gentile v. State Bar of Nevada, 501 U.S. 1030 (1991) (finding Nevada Supreme Court Rule void for vagueness).    For another, although a pre-trial detention statute does not raise the same notice problems as statutes defining crime and fixing punishments, the Supreme Court has explained that “the more important aspect of vagueness doctrine is . . . the requirement that a legislature establish minimal guidelines to govern law enforcement.”  Kolender, 461 U.S. at 357-58 (internal quotation marks omitted).  Indeed, one might read City of Chicago v. Morales, 527 U.S. 41 (1999), as suggesting that the arbitrary and discriminatory enforcement concern, standing alone, is enough to invalidate a vague statute.  (The plurality concluded that the ordinance in question gave insufficient notice, but Justice Kennedy’s concurrence and Justice Breyer’s concurrence appear to have rested on concerns about enforcement.)

18 U.S.C. § 3559(c)

In addition to § 16(b), and § 924(c), there is one additional federal statute that may be unconstitutionally vague after Johnson, 18 U.S.C. § 3559(c).  Section 3559(c) mandates a sentence of life in prison for defendants with certain criminal histories, including for defendants who have been convicted of two or more “serious violent felonies.”  Section 3559(c) defines the term “serious violent felony” to include “any other offense punishable by a maximum term of imprisonment of 10 years or more . . . that, by its nature, involves a substantial risk that physical force against the person of another may be used in the course of committing the offense.”  This language is nearly identical to the language in § 924(c) and § 16(b).  The definition includes the phrase “by its nature,” which federal courts have generally interpreted to mandate the categorical approach.  Consequently, several courts have read this language to require the categorical approach.  E.g., United States v. Echeverria-Gomez, 627 F.3d 971, 975 n.13 (5th Cir. 2010); United States v. Kennedy, 133 F.3d 53, 56–57 (D.C. Cir. 1998).  Despite this language, not all courts have held that § 3559(c) requires the categorical approach.  See, e.g., United States v. Mackovich, 209 F.3d 1227, 1239–40 (10th Cir. 2000).

If the courts are using the categorical approach to apply § 3559(c), then a ruling in Lynch v. Dimaya that § 16(b) is unconstitutionally vague may also render § 3559(c) unconstitutional.  There is a further complication, however.  Section 3559(c) allows defendants to raise an affirmative defense based on their real world conduct.  Specifically, the statute permits defendants to argue that a previous conviction does not qualify as a “serious violent felony” if the defendant can establish by clear and convincing evidence that “no firearm or other dangerous weapon was used in the offense and no threat of use of a firearm or other dangerous weapon was involved in the offense; and the offense did not result in death or serious bodily injury.”  18 U.S.C.A. § 3559(c)(3)(A).  Given that a defendant could challenge at least some prior convictions using her real world conduct, one might conclude that § 3559(c) is not void for vagueness.  The Johnson Court clearly permits the application of a qualitative standard to real world facts, 135 S. Ct. at 2561, and a defendant can put real world facts at issue by raising this affirmative defense.  In other words, one might argue that this affirmative defense renders the statute sufficiently definite to survive a vagueness challenge.

But there are reasons to think that this affirmative defense cannot save § 3559(c) from unconstitutional vagueness.  In its past vagueness cases, the Supreme Court has not looked at available affirmative defenses when assessing whether a criminal statute is sufficiently definite.  And providing a defendant with an affirmative defense is not the same as narrowing a criminal statute.  Defendants often bear the burden of proving an affirmative defense—as they do under § 3559(c)—and the Due Process Clause requires that the prosecution prove all elements of a crime beyond a reasonable doubt.  See Sandstrom v. Montana, 442 U.S. 510, 524 (1979); In re Winship, 397 U.S. 358 (1970).  Consequently, whether § 3559(c) is unconstitutionally vague most likely must be decided without considering the affirmative defense.

Johnson and the Federal Sentencing Guidelines

Lynch v. Dimaya is not the only case that the Supreme Court will hear this Term about the effect of Johnson on the vagueness doctrine.  Johnson also drew into question the constitutionality of several Federal Sentencing Guidelines, as well as sentences based on those Guidelines that have become final.  The Supreme Court will take up these questions in Beckles v. United States, No.15-8544.

The career offender sentencing guideline, U.S.S.G. § 4B1.2(a)(2), was amended earlier this year.  Before that amendment took effect, § 4B1.2(a)(2) used language that was identical to the ACCA residual clause.  If the previous version of § 4B1.2(a)(2) is void for vagueness, then several other guidelines that use the career offender guideline’s definition of “crime of violence” are also in doubt.  These other guidelines include:

  • U.S.S.G. § 2K1.3 & cmt. n.2 (explosive materials Guideline) 
  • U.S.S.G. § 2K2.1 & cmt. n.1 (firearms Guideline) 
  • U.S.S.G. § 2S1.1 & cmt. n.1 (money laundering Guideline) 
  • U.S.S.G. §§ 4A1.1(e), 4A1.2(p) (criminal history Guidelines) 
  • U.S.S.G. § 5K2.17 & cmt. n.1 (departure Guideline for semi-automatic firearms)
  • U.S.S.G. § 7B1.1(a)(1) & cmt. n.2 (probation and supervised release Guideline)

The government has conceded that previous version of § 4B1.2(a)(2) and all of these affected guidelines are impermissibly vague.  See Supplemental Brief for the United States at 6, United States v. Pagan-Soto, No. 13-2243 (1st Cir. Aug. 11, 2015).

Although the government conceded that the previous version of § 4B1.2(a)(2) was void for vagueness, not all federal courts accepted this concession.  The Eleventh Circuit refused to accept the concession, deciding that previous version of § 4B1.2(a)(2) was not unconstitutional because the vagueness doctrine does not apply to the now-advisory Federal Sentencing Guidelines.  United States v. Matchett, 802 F.3d 1185 (11th Cir. 2015), reh’g en banc denied, -- F.3d. --, 2016 WL 475211 (11th Cir. Sept. 13, 2016).  A number of circuits have disagreed and held that the Federal Sentencing Guidelines are subject to a vagueness attack.  See United States v. Hurlbert, -- F.3d --, 2016 WL 4506717 (7th Cir. Aug. 29, 2016) (en banc); United States v. Sheffield, 832 F.3d 296 (D.C. Cir. 2016); United States v. Pawlak, 822 F.3d 902 (6th Cir. 2016); United States v. Madrid, 805 F.3d 1204 (10th Cir. 2015); United States v. Townsend, 638 F. App’x 172 (3d Cir. 2015).  A number of courts did not weigh in on this issue, but simply assumed that the guideline is vague based on the concession from the government. See United States v. Fields, 823 F.3d 20, 33 (1st Cir. 2016) (noting the government’s concession that this guideline is unconstitutionally vague); United States v. Martinez, 821 F.3d 984, 988 (8th Cir. 2016) (same); United States v. Maldonado, 636 F. App’x 807, 810 (2d Cir. 2016) (same).  Because the government has conceded that the vagueness doctrine applies to the Federal Sentencing Guidelines, the Court has appointed an amicus to argue in support of the judgement below on this issue.

If the Court in Beckels decides that the Federal Sentencing Guidelines are subject to vagueness attack, that decision may have other consequences.  For example, if 18 U.S.C. § 16(b) is deemed to be unconstitutionally vague in Lynch v. Dimaya, then U.S.S.G. § 2L1.2(b)(1)(C) may also be void for vagueness.  That is because § 2L1.2 incorporates the relevant language from § 16(b).  See, e.g., United States v. Hernandez-Lara, 817 F.3d 651 (9th Cir. 2016) (holding that, because 18 U.S.C. § 16(b) is unconstitutionally vague, so too is § 2L1.2(b)(1)(C)).  

Whether the vagueness doctrine applies to the Federal Sentencing Guidelines is not the only question presented in Beckles.  The case also raises the question whether a determination that a Federal Sentencing Guideline is unconstitutionally vague applies retroactively.  The Supreme Court decided that the vagueness ruling in Johnson was retroactive in Welch v. United States, 136 S. Ct. 1257 (2016).  The reasoning of that opinion suggests that anytime the Court finds a statute to be unconstitutionally vague then such a ruling would be retroactive.  

The government has not conceded that Johnson should apply retroactively to defendants sentenced under the career offender guideline.  The government argues that retroactivity is improper for the Federal Sentencing Guidelines because the now-advisory Guidelines are more procedural rules than substantive rules.  Whether a rule applies retroactively is determined largely by assessing whether the rule is substantive or procedural.  See Teague v. Lane, 489 U.S. 288 (1989).

It is hard to predict how the Court will rule in Beckles.  Both whether the Guidelines should be subject to vagueness attack and whether rulings about the Guidelines should be retroactive are questions that involve the legal status of the Guidelines.  That status has been a matter of some dispute since the Supreme Court deemed the Guidelines “advisory” in United States v. Booker, 543 U.S. 220 (2005).  Before the Court decided Booker in 2005, the Guidelines were mandatory—that is to say, they were largely indistinguishable from criminal statutes setting sentences.  So any defendant who was sentenced prior to the Booker decision should not have to confront the difficult issues raised in Booker.  Those defendants should be treated no differently than defendants who were sentenced under the residual clause of the ACCA.  The pre-Booker career offender Guideline was a void law that set a punishment range, and Johnson should apply to sentences imposed under that Guideline even though the sentences have become final.

* **

Johnson v. United States is of the most cited U.S. Supreme Court cases from recent Terms.  Johnson obviously affected the large number of defendants who were sentenced under the residual clause of the Armed Career Criminal Act.  It may, however, have a lasting impact on the vagueness doctrine itself.  By questioning the viability of the categorical approach and by clarifying that the doctrine applies also to laws that fix sentences, Johnson has called into doubt the constitutionality of other federal criminal laws and various Federal Sentencing Guidelines.  We will have to await the decisions in Lynch v. Dimaya and Beckles v. United States in order to fully assess the legacy of Johnson.  If the government loses those cases, then we are likely to see a further challenges to laws that fall within the long shadow of Johnson.