Recent Terms have brought a sequence of cases challenging portions of the criminal code for unconstitutional vagueness.1 Criminal defen-dants have sought relief from long sentences on the grounds that the statutory definitions of their crimes gave insufficient notice of their actions’ consequences — notice the Due Process Clause requires — and therefore that these statutory definitions were void for vagueness. And the Court has shown itself willing to grant that relief.2 Last Term, in Sessions v. Dimaya,3 the Supreme Court took the next logical step, applying its vagueness analysis, set forth in Johnson v. United States,4 to a provision of the criminal code incorporated into the Immigration and Nationality Act5 (INA). Given the severe sanction the INA contemplates (deportation), the Court was right to apply Johnson. But the Court should limit its extension of Johnson to circumstances where the consequences are severe; to do so the Court ought to import the distinctions it has already drawn in the procedural due process realm. This will allow it to develop a determinate standard by which it can judge when applying Johnson is appropriate, and thus avoid needlessly encroaching on the civil code.
James Dimaya, a native of the Philippines, lawfully entered the United States in 1992 and lived in California as a legal permanent resident.6 He was twice convicted of first-degree residential burglary, first in 2007, then again in 2009, and sentenced to a two-year prison term for each.7 Under California law, burglary is defined as entering any of a list of dwellings “with intent to commit . . . any felony.”8 In 2011, the Department of Homeland Security (DHS) initiated removal proceedings against Dimaya.9 Under the INA, an alien is subject to removal if he is “convicted of an aggravated felony at any time after admission.”10 The statute defines “aggravated felony” by pointing to the criminal code, 18 U.S.C. § 16, and its definition of “crime of violence.”11 Under § 16(b), the “residual clause,” “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” constitutes a crime of violence.12 DHS sought Dimaya’s removal under § 16(b), arguing burglary inherently involved substantial risk of physical force and thus was grounds for removal.13
The immigration judge (IJ) agreed, finding that California’s burglary statute, which required entry into a residence, dealt with crimes that “by [their] very nature” likely create risk of violence.14 Because the crime satisfied § 16(b), the IJ ordered Dimaya deported.15 The Board of Immigration Appeals (BIA) affirmed.16
Dimaya appealed to the Ninth Circuit, arguing that the BIA had erred in classifying California burglary as a § 16(b) crime of violence.17 But while his appeal was pending, the Supreme Court decided Johnson.18 There, the Court addressed the Armed Career Criminal Act’s19 (ACCA) definition of “violent felony,” which — like the statute at issue in Dimaya — included a residual clause, encompassing any crime that “otherwise involves conduct that presents a serious potential risk of physical injury to another.”20 The Court held that this residual clause violated the Fifth Amendment’s due process requirement by “den[ying] fair notice to defendants and invit[ing] arbitrary enforcement by judges.”21 It was unconstitutionally vague and thus void.22 In light of the Court’s holding in Johnson, the Ninth Circuit ordered supplemental briefing and arguments on the question of whether § 16(b), too, was unconstitutionally vague.23
The Ninth Circuit ultimately reversed the BIA.24 Judge Reinhardt,25 writing for the panel, first explained that both Ninth Circuit and Supreme Court precedent supported applying vagueness doctrine to deportation proceedings because of the “harsh consequences attached.”26 Then, having affirmed that the INA was subject to Johnson’s vagueness inquiry, Judge Reinhardt compared § 16(b) to ACCA’s infirm language, concluding that the two clauses were similar in all relevant ways, making § 16(b) unconstitutional just as ACCA was.27
The Supreme Court affirmed.28 Writing for the Court, Justice Kagan29 held § 16(b) unconstitutionally vague.30 As Judge Reinhardt had done below, Justice Kagan compared § 16(b) of the INA to the text of ACCA struck down in Johnson.31 She found that the provisions were similar in two crucial respects: both of the provisions (1) “‘tied the judicial assessment of risk’ to a hypothesis about the crime’s ‘ordinary case’”;32 and (2) created “uncertainty about the level of risk that makes a crime ‘violent.’”33 On the first point, Justice Kagan noted part of the worry was that § 16(b)’s residual clause — like its ACCA counterpart — required courts to employ the categorical approach, which tasks courts with “‘imagin[ing]’ an ‘idealized ordinary case of the crime,’” rather than analyzing the specific facts of the case at hand.34 The problem, according to the majority, was that both ACCA and § 16(b) “‘offer[ed] no reliable way’ to discern what the ordinary version of any offense looked like.”35 On the second point, the Court held that, because it encompassed crimes creating a “significant risk” of violence, § 16(b) created too much uncertainty as to what level of risk makes a crime violent, offering merely an “imprecise ‘qualitative standard’”36 like the one rejected in Johnson. While Justice Kagan noted that the use of a “non-numeric standard”37 like § 16(b)’s was not necessarily impermissible, she held that when combined with the “judge-imagined abstraction”38 required by the categorical approach, it “cease[d] to work in a way consistent with due process.”39 After finding that the two statutes had the same two constitutional defects, the Court dispensed with the government’s counterarguments and struck down the offending provision.40
Justice Kagan wrote for just a plurality of the Court in Part II.41 Here, she rejected the government’s contention that Johnson ought not apply because the INA is a civil statute; she responded, citing the government’s own brief, that “[t]he degree of vagueness that the Constitution [allows] depends in part on the nature of the enactment.”42 Noting deportation’s severity, Justice Kagan concluded that the proceeding called for an exacting application of Johnson’s vagueness analysis.43
Justice Gorsuch concurred in part and concurred in the judgment, writing with two points. First, Justice Gorsuch tackled Justice Thomas’s originalist concerns about vagueness doctrine, pointing to several historical texts to justify the doctrine’s application.44 He concluded that far from violating the separation of powers, vagueness doctrine was required by it, in that overbroad laws impermissibly delegate legislative power to courts to decide what the law should be in practice.45 Second, Justice Gorsuch challenged the plurality’s suggestion that removal proceedings occupy a unique place in the civil law. While the majority accepted vagueness doctrine’s traditional distinction between criminal and civil law (but simply qualified removal proceedings as a rare exception to it),46 Justice Gorsuch argued that the criminal-civil distinction should not be the alpha and omega of vagueness jurisprudence.47 He argued that if the distinguishing feature of removal proceedings, according to the plurality, was the “special gravity of its . . . penalty,” what difference was there between removal proceedings and, say, “subject[ing] a citizen to indefinite civil commitment, [or] strip[ping] him of a business license essential to his family’s living”?48 Justice Gorsuch could “think of no good answer.”49 Thus he suggested leaving the door open to applying Johnson in a wider variety of future cases.
Chief Justice Roberts50 dissented. He distinguished § 16(b) from ACCA, arguing that applying the categorical approach to § 16(b) provided consistent and fair results that were knowable ex ante.51 He identified several textual differences between the statutes — for instance, “potential risk” in ACCA versus simply “risk” in § 16(b), risk of injury in ACCA versus use of force in § 16(b) — that counseled against applying Johnson.52 He also ran through a number of crimes whose classification under § 16(b) was simpler than it would have been under ACCA53 and reminded the Court that its ACCA jurisprudence was muddled before Johnson, while the § 16(b) doctrine was clear before Dimaya.54
Lastly, Justice Thomas dissented to critique the majority’s application of the Fifth Amendment, to advocate for case-by-case application of vagueness doctrine, and to urge abandonment of the categorical approach. Justice Thomas disputed that vagueness doctrine had any home in the original understanding of removal proceedings and refuted Justice Gorsuch’s historical sources with his own.55 Far from agreeing that anti-vagueness was an originalist aim, he pointedly observed that vagueness doctrine’s emergence coincided with “a time when this Court was actively interpreting the Due Process Clause to strike down democratically enacted laws.”56 And in response to Justice Gorsuch’s argument that separation of powers mandated robust vagueness scrutiny, Justice Thomas countered that the separation of powers should be achieved through other constitutional provisions, not through the Due Process Clause.57 He went on to argue that vagueness doctrine required a case-specific inquiry and that § 16(b) was not vague as applied to Dimaya’s crimes,58 and he concluded with a plea to the Court that, if it was the categorical approach causing vagueness problems, it revisit the categorical approach rather than continue to invalidate the statutes themselves.59
On the whole, Justice Kagan’s diagnosis was accurate: “Johnson [was] a straightforward decision, with equally straightforward application” in Dimaya.60 Once the Court had done the analysis in Johnson, the only real question in Dimaya was whether or not the Court would extend this due process protection to deportation cases. Because the Court has previously acknowledged that deportation is just as severe as incarceration,61 justice demanded Johnson’s application here. But narrow though it is, the Court’s opinion could, if wielded by a deregulation-minded majority, do dramatic damage to the administrative state. While vagueness doctrine properly applies to civil deprivations that closely resemble criminal ones, the doctrine could easily be extended too far, doing violence to regulatory schemes that require some degree of vagueness to be effective, but that do not deprive individuals of their liberty in the same urgent way. Perhaps what is most important about Dimaya is what it did not do: it failed to give future courts binding precedent about how the vagueness doctrine should be extended (or not) going forward. Although the Court made the important move of applying Johnson to a civil law, it failed to agree on why it chose to do so and what that means for future cases. If, going forward, the Court hopes to protect vagueness doctrine from the type of foundation-shifting expansion that Justice Gorsuch advocated in Dimaya, it ought to develop a robust test for determining where and how it will apply the vagueness doctrine outside of the criminal and removal contexts. Ad hoc extension of vagueness doctrine could create arbitrary and unpredictable outcomes in individual cases — exactly what the doctrine is trying to prevent. The Court should look to its procedural due process precedent to develop a standard against which to judge the relative severity of various deprivations. In those cases, the Court has drawn precisely the lines Justice Gorsuch’s concurrence challenged the Court to draw and provided the Court with ample analogies against which to measure civil harms when determining their severity.
Vagueness doctrine has primarily been used as a tool for defendants to challenge their convictions or arrests.62 Its core value is that laws ought to “give the person of ordinary intelligence a reasonable opportunity to know what is prohibited, so that he may act accordingly.”63 Two separate justifications undergird this doctrine: first, that the most severe deprivations of liberty can be rightfully effected only after proper notice (the “notice” component), and second, that the legislature abdicates its responsibility in passing loosely drafted statutes and grants outsized rulemaking power to courts (the “separation of powers” component).64 The plurality’s focus in Dimaya was on the notice component. Once one concedes that deportation is functionally as severe as imprisonment,65 a rigorous application of the vagueness test seems required.66
Justifying Dimaya on separation of powers grounds is more challenging. Where the right purportedly infringed upon is Fifth Amendment “liberty” (rather than a more clearly defined right, like free speech, for example), separation of powers might actually counsel against invalidating statutes. This is because expanding a demanding version of Johnson to statutes that threaten an intangible — and subjective — notion of liberty would invite judges to read their preferred values into the Fifth Amendment, turning vagueness doctrine into the modern-day equivalent of substantive due process.67 The Court should be vigilant against this creep of separation of powers–based vagueness doctrine, which risks delegitimizing the Court and exposing it to the charge of Lochnerism. In addition, such an expanded doctrine threatens to severely hamper the government’s efficacy. Modern-day administration is predicated on the understanding that some statutes are designed to be capacious, and it would be impossible to hold every statute to the same exacting test of linguistic precision that a rigorous vagueness doctrine might.68 Ambiguity, and thus some degree of vagueness, is tolerated — even presumed — to make the federal government run.69 A limiting principle would help to guard against severe deprivations of liberty and property without disturbing the core functions of the administrative state.
For that reason, the Court should reject any attempt to deploy a formalistic understanding of separation of powers against the administrative state more generally. An expansive approach to the “liberty” deprivation that triggers Johnson scrutiny, if adopted by a majority, could signal a radical departure from the premise of flexible federal government. In order to guard against a threatening extension of vagueness doctrine, the Court needs to develop a principled means for limiting the doctrine’s scope, or at the very least a standard against which to judge various deprivations of liberty.70 Although proceeding in an ad hoc manner may seem like the safest course, it also comes with risks — that bad facts will make bad law, or that particular coalitions will shape a bizarre path in vagueness doctrine.71 Instead, the Court should look to its own Fifth Amendment jurisprudence, where it has already drawn just such lines between various noncriminal deprivations of liberty or property. In Mathews v. Eldridge,72 the Court established a balancing test for determining what level of process the Fifth Amendment requires when the government seeks to terminate Social Security benefits.73 In doing so, the Court laid out a framework it later used for administrative proceedings more generally, holding that in determining due process’s requirements, courts should balance (1) the degree of potential deprivation, (2) the fairness and reliability of the existing pretermination procedures, and (3) the public interest.74 While prongs two and three are less applicable when the task at hand is statutory interpretation, rather than requiring a particular procedural mechanism, the first prong has considerable relevance to the question of whether — and how — to apply Dimaya to other civil statutes. This prong ought to be imported into the vagueness doctrine.75
Mathews’s core contribution to the vagueness doctrine is that the level of protection owed to citizens varies based on the severity of the harm they might suffer.76 Not every type of deprivation requires the same level of process; nor should every statute require the same level of scrutiny. Because Mathews offers an entire body of case law distinguishing deprivations that are by their very definition not criminal,77 the Court should lean on this precedent when weighing various civil injuries for purposes of vagueness analysis. Relying on Mathews cases that parse out varying levels of civil harms, the Court can then slot these cases into the maxim parroted by both the plurality and the concurrence: “The degree of vagueness that the Constitution tolerates . . . depends in part on the nature of the enactment.”78
Consider three kinds of deprivations that have given rise to Mathews challenges: (1) cases involving civil imprisonment; (2) cases involving less deprivation than imprisonment, but more than mere economic deprivation; and (3) pure monetary loss. Each category involves some type of deprivation worthy of consideration in its own right. However, in the Mathews context, the Court has refused to address them with a one-size-fits-all approach. Rather, it has weighed the importance of the deprivation at issue before deciding what standard it will use to evaluate the procedures that gave rise to the deprivation.
In the cases involving imprisonment, the Court — recognizing the paramount liberty interests at stake — has required more robust procedures before deprivation is considered to comport with the demands of due process.79 In cases involving less than imprisonment but more than mere economic deprivation (job loss, for example) the Court has required less rigorous procedures before sanctioning the deprivation.80 And in cases involving simple monetary harm, the Court has proven itself even less inclined to invalidate governmental deprivation.81 These examples demonstrate that the Court is able to condition its treatment of a particular legal concept (in the Mathews context, the procedures used to deprive a person of property) to the severity of deprivation associated with that concept. The Court could easily extend this reasoning to the vagueness doctrine. Just as not every type of deprivation will require the same amount of process, not every type of statute will require the same amount of scrutiny.
It is possible to make reasoned comparisons between various liberty deprivations, as the Court has shown. Thus, with Mathews precedent in hand, the Court is better equipped to take a more systematic look at the vagueness challenges that may soon arise. For example, using Mathews will help disentangle the three scenarios Justice Gorsuch posited in his concurrence: “indefinite civil commitment,” “business licens[ing],” and government seizure of property.82 These three scenarios map rather cleanly onto the three Mathews scenarios described above — indefinite civil commitment akin to civil imprisonment, business licensing akin to job loss, and confiscating a home akin to simple property loss.83 Though Justice Gorsuch professed to find no way to distinguish between these three scenarios, the Court has repeatedly done precisely that when deciding how much process to provide in an administrative proceeding. It can do so equally adroitly when deciding how closely to review a statute for vagueness.
Comparing the deprivations cannot itself completely answer the question of what level of scrutiny is needed. The Court will still need to determine how closely to parse a statute, if not quite at the level of Johnson scrutiny; it will also need to determine whether other factors (analogous to the second and third Mathews prongs) will counsel in favor of more or less scrutiny. But the sliding scale of process at the core of Mathews provides a guide for a sliding scale of interpretive deference in the vagueness context. Justice Gorsuch thus brings up a good point: civil commitment might pose a similar challenge for the Court. But when the consequences are merely monetary — as with environmental or consumer protection statutes — the Mathews line of cases would urge reading them permissively, to allow for functionality, rather than scrutinizing each word as the Court did in Dimaya. This would preserve administrative flexibility to administer critical laws, while still providing protection for unwary citizens risking extreme liberty deprivations.
On one level, Dimaya is a straightforward case about applying precedent. Johnson all but preordained the outcome, if the Court chose to apply it to the INA. The Court was right to recognize that the deprivation of liberty at issue in deportation proceedings is every bit as severe as the deprivation at issue in criminal ones. But in the coming Terms, the Court will likely be asked to apply the Johnson analysis beyond these settings. And Justice Gorsuch’s concurrence indicates that at least one Justice is happy to do so expansively. The Court should resist the invitation to conflate various civil harms and should avoid proceeding, as in Dimaya, with case-by-case inquiries. Instead, the Court should look to its cases in the Mathews line to draw from the methodical estimations it has already made about respective deprivations and use these to inform the level of scrutiny with which it will assess statutes. In this way, the Court can set principled limitations and prevent the vagueness doctrine from swallowing the entire civil code.