Justice Alito on Criminal Law – Kate Stith

Posted by arispitzer on Apr 24, 2023 in Per Curiam

Justice Alito on Criminal Law

Kate Stith^*

I.               Introduction

Justice Samuel A. Alito is a natural judge—by temperament, character, disposition, and experience. What do I mean by a “natural judge”? It is difficult to conceive of Justice Alito accepting a legal position where he would have to perform as a pure advocate, which he knows may require mincing words, shading nuance, and hiding the ball. Indeed, Alito’s entire career as a lawyer—both within the U.S. Department of Justice and in the federal judiciary—has been defined, in part, by ethical norms and standards of straightforward and honest lawyering.[1]

This chapter concerns itself with the corner of Justice Alito’s jurisprudence dedicated to the criminal law. Justice Alito’s criminal-law jurisprudence reflects his aversion to reasoning that will leave the Supreme Court (or the police, citizens, and lower courts) out on a limb, in a place that threatens to undo social understandings and order.

To begin, I should clarify what I mean (and what I do not mean) by “criminal law.” When I say criminal law, what I really mean is substantive criminal law. Substantive criminal law refers to the set of laws within a jurisdiction that define and punish the acts and mental states that together constitute crimes. Criminal law is, of course, distinct from criminal procedure, which regulates the machinery by which the government can apprehend alleged violators of the criminal law and initiate a prosecution. Criminal procedure is largely a matter of constitutional interpretation, but the meat and potatoes of the criminal law is statutory interpretation.

As Justice Scalia once noted, “We live in an age of legislation, and most new law is statutory law.”[2] Every actor in a criminal case—whether the prosecutor, the defendant, or the judge—must engage in statutory interpretation. Prosecutors, first, must identify the statutory provision an individual allegedly violated and determine under that statute which facts must be proven beyond a reasonable doubt to the factfinder. Defendants, by contrast, will mine statutes to identify every burden the prosecutor must prove and what, if any, defenses the law affords. Judges must interpret criminal statutes to instruct the jury, assess the relevance of evidence, and impose a sentence within the lawfully authorized range.

I focus on two aspects of federal criminal law that have been of particular concern to Justice Alito—the categorical approach and mens rea. The former addresses primarily how Congress has instructed federal courts to sentence repeat offenders (or “career criminals” in the words of Congress),[3] while the latter addresses what mental state is required while committing the crime at issue.

Justice Alito’s opinions in these two areas epitomize his pragmatic approach to the criminal law. He is not interested in constructing or in deducing from a grand theory, or any theory at all. Pragmatism is less a unified theory than a collection of related ideas, including intellectual humility, resistance to abstraction, and concern with real-world consequences. Over the years, Justice Alito has expressed unease that the Court conceives of itself as a tribunal of theoreticians rather than a tribunal of judges who must grapple with the concrete realities of the criminal-justice system at large and the facts of a particular defendant’s case. If the Court nonchalantly opens the floodgates of litigation or delivers unclear instructions to the lower courts, Justice Alito is ready in the wings (often in solo concurrences or dissents) to remind the Court of its decisions’ problematic real-world consequences. In one criminal-law dissent emblematic of his pragmatism, Justice Alito noted that the “well-known medical maxim—‘first, do no harm’—is a good rule of thumb for courts as well.”[4] As we shall see, this is a precept Justice Alito follows too.

II.             The Categorical Approach

The categorical approach is, in Justice Alito’s words, the result of “pointless abstract questions” for “aficionados of pointless formalism.”[5] The Justice’s sharp words make the categorical approach an irresistible—though highly convoluted—window into his pragmatic jurisprudence.

In its primary application, the categorical approach is a method of statutory interpretation that the Supreme Court has said federal courts must use during the sentencing stage of some criminal prosecutions. Most notably, the categorical approach has been applied to provisions of the Armed Career Criminal Act (ACCA)—a federal statute first enacted as part of the Comprehensive Crime Control Act of 1984. Pursuant to ACCA, the sentencing judge must determine whether the defendant’s prior convictions are of the type that, under the 1984 statute (as further amended), trigger a higher penalty for the federal crime currently being sentenced.[6]

For instance, the bare crime of being a felon in possession of a firearm has a ten-year maximum penalty.[7] But if an individual convicted under the felon-in-possession statute has three or more previous felony convictions for a “violent felony” or a “serious drug offense,” the sentencing consequences are much more severe. Instead of a ten-year maximum sentence, the defendant is subject to a mandatory minimum sentence of fifteen years.[8] But how is the judge to determine whether a prior offense is “violent” or “serious”? In 1990, the Supreme Court in an opinion by Justice Blackmun,[9] set the course that, amazingly, it still follows today: ACCA requires sentencing judges to engage in an inquiry that is, to put it mildly, highly abstract. Moreover, the inquiry is truncated; it requires the sentencing court, in most contexts,[10] to ignore the facts of a defendant’s actual conduct and instead look only to the text of the statute under which the defendant was previously convicted.[11]

For both the lawyers and nonlawyers among us, the categorical approach may seem a soporific example to pick in cataloguing Justice Alito’s jurisprudence. But as the Justice’s opinions in this area reveal, the practical implications of the categorical approach are significant and alarming. The approach leads to wild variations in sentencing (and deportation) consequences depending on the precise wording of the (usually state-law) statutes under which defendants have previously been convicted. Moreover, the Court’s alteration and fine-tuning of the categorical approach has cascading effects across our entire criminal justice system. What may seem like an abstract, textual inquiry for the Court can undo the work of prosecutors, criminal-defense lawyers, and judges across the land. In this Part, I offer a brief historical primer on the categorical approach and highlight Justice Alito’s most important opinions on this topic, which put his pragmatism on full display.

* * *

As noted, the categorical approach’s journey began more than thirty years ago, with the Supreme Court’s first interpretation of ACCA.[12] In the 1980s, Congress turned its attention to “career” criminals and sought to “increase the participation of the Federal law enforcement system in efforts to curb armed, habitual (career) criminals” in the states.[13] Congress noted the proliferation of “crimes involving theft or violence . . . by a very small percentage of repeat offenders.”[14] In its contemporary and amended form, ACCA’s fifteen-year mandatory minimum applies to a person who commits a felony punishable by greater than one year and “has three previous convictions by any court . . . for a violent felony or a serious drug offense.”[15] A prior conviction may arise from any court (state or federal).[16] Because most felonies are prosecuted in state court, this means that in most cases, a federal court must determine whether a prior state offense counts as a “violent felony” or “a serious drug offense” under ACCA.

ACCA does not instruct judges how to make that determination. Neither Congress nor any state legislature has a list of “violent” offenses or of “serious” drug offenses; rather, each jurisdiction enacts prohibitions, which accumulate over the years.[17] Some crimes may on their face sound unquestionably violent, such as murder and rape. Yet, as they are wont to do, lawyers begin to pose hypotheticals that bend these intuitions. Is murder by starvation a violent felony? Is consensual sex with an underage teenager also a violent felony? What makes a particular drug offense “serious”—the particular type of drug, the amount of the drug, or the role of the offender (as a “kingpin,” for example)?

In the critical 1990 case that adopted the “categorical” approach to answering these questions, Taylor v. United States,[18] the Court confronted whether a defendant’s prior conviction for burglary in the state of Missouri constituted a violent felony within the meaning of ACCA. If Missouri burglary did count as a violent felony, then the defendant would be subject to the fifteen-year mandatory minimum.[19] But if Missouri burglary did not count as a violent felony, then the defendant would not be subject to the enhanced sentence under ACCA.

In resolving the question, the Court refused to look at the facts of Taylor’s prior burglary conviction, which might have revealed whether Taylor actually burglarized violently or with a dangerous weapon.[20] Instead, the Court adopted a “formal categorical approach, looking only to the statutory definitions of the prior offenses, and not to the particular facts.”[21] This approach requires a sentencing court to do a side-by-side comparison of two statutes: first, the federal statute defining a generic “violent felony” or a “serious drug offense,” and, second, the statute defining the crime of prior conviction (e.g., the Missouri burglary statute at issue in Taylor).[22]

After comparing the two statutes, the sentencing court must ask whether the statute of prior conviction punishes conduct that is not included in the generic definition of a “violent felony” or a “serious drug offense.” If the statute of prior conviction only reaches behavior within this definition of a “violent felony” or a “serious drug offense,” then the prior conviction counts toward the sentencing enhancement. If the statute of prior conviction sweeps more broadly and punishes conduct that is beyond the definition of a “violent felony” or a “serious drug offense” under federal law, then the prior conviction does not count toward the sentencing enhancement.

For example, in Taylor itself, the Court had to compare ACCA’s definition of “violent felony” to Missouri’s definition of second-degree burglary. ACCA includes “burglary” in a list of violent felonies, but it leaves that term undefined—a task the Supreme Court took on for itself. Whereas the federal definition of burglary only proscribes unauthorized entry into a “building or other structure,” the Missouri definition of burglary also includes unauthorized entry into a “boat or vessel.”[23] As a result of this mismatch, the Court held, Taylor’s prior conviction therefore did not count toward ACCA’s sentencing enhancement. The problem was that the Missouri statute punished certain types of conduct not prohibited under federal law (namely, burglary of a “boat or vessel”). It was irrelevant that Taylor may have, in fact, entered a building or structure and not a boat or vessel because the statute of prior conviction, in the abstract, swept more broadly than the generic federal definition.

But why ignore the facts of Taylor’s case to determine if his prior offense constituted a violent felony under ACCA? The Supreme Court supplied a handful of justifications for a side-by-side statutory comparison instead of digging into the factual record, including the “daunting” reality of rifling through the record, “practical difficulties[,] and potential unfairness” of fact-bound inquiries.[24] The categorical approach’s prohibition on peering into the factual record relieves sentencing courts from the possibly cumbersome effort of retrieving state-court records, or being left without a paddle in some plea-bargaining cases. Avoiding judicial inquiry into the actual facts of the defendant’s prior conviction also avoids the question whether such inquiry comes within Apprendi’s demands that sentencing factors enhancing punishment must be admitted by the defendant or proved beyond a reasonable doubt to a jury.[25]

And so the Court embarked on a course that would employ “uniform, categorical definitions to capture all offenses of a certain level of seriousness . . . regardless of technical definitions and labels under state law.”[26] By carefully sticking with categorical definitions, sentencing courts would avoid the “unfairness” of enhancing a defendant’s sentence based on the mere “label employed by the State of conviction.”[27]

Fifteen years after Taylor, the Court (which Justice Alito would join the following Term) somewhat softened this aversion to reviewing the factual record of a prior conviction. In Shepard v. United States and its progeny, the Court has carved out an exception to the no-factual-record rule in what has come to be known as the “modified categorical approach.” Under the modified categorical approach, a sentencing court may review the terms of the charging document, plea agreement, colloquy transcript, or “some comparable judicial record” of the factual basis of the conviction.[28] Taylor had abjured any review of such documents (now known as Shepard documents). While Shepard opened the door to some consideration of the actual facts of the defendant’s prior offense, the Court has clarified in subsequent cases that the modified categorical approach is appropriate only when the statute of conviction describes multiple crimes and Shepard documents would be useful to determine which of those crimes the defendant was convicted.[29] Critically, this means that where a statute merely defines multiple means of committing a single crime, the traditional categorical approach of Taylor applies—and the factual basis of the conviction is entirely off the table.

When Justice Alito joined the Court in 2006, the contours of the (now modified) categorical approach had been set. In his first few Terms at the Court, Justice Alito showed himself to be a somewhat faithful adherent. In James v. United States, Justice Alito applied the categorical approach to find that the attempted burglary at issue counted as a “violent felony” under ACCA’s residual clause.[30] In United States v. Rodriquez, he similarly wrote for the Court that a Washington drug-trafficking felony counted as a “serious drug offense” under ACCA.[31]

Although he applied the categorical approach, even in these early cases he registered concerns about rigid application without attention to practical consequences. In James, he pointed out that this mode of statutory interpretation should not require “metaphysical certainty” as to the scope of the statute of prior conviction; rather, sentencing courts must look for a “realistic probability, not a theoretical possibility, that” the statute of prior conviction encompasses behavior not included in the federal definition.[32] Moreover, in Rodriquez, Justice Alito rebuffed the claim that inquiries into “novel questions of state law and complex factual determinations” are necessarily “difficult.”[33] Sentencing courts could easily look to Shepard documents, including formal charging documents and plea colloquies.[34] And a “mere possibility that some future cases might present difficulties cannot justify a reading of ACCA that disregards the clear meaning of the statutory language.”[35]

Justice Alito’s unease with the categorical approach became far more pronounced by 2010. In Johnson v. United States, the Supreme Court had to decide whether Florida felony battery by “actually and intentionally touching” the victim constituted a violent felony under ACCA.[36] Prosecutors sought an enhanced penalty under that Act, but Johnson objected to the categorization of his 2003 Florida conviction for simple battery as a “violent felony.”[37] Under Florida law, battery may occur in any of three ways: if the defendant “[i]ntentionally caus[ed] bodily harm,” “intentionally str[uck]” the victim, or “[a]ctually and intentionally touche[d]” the victim.”[38] The court records of Johnson’s prior simple-battery conviction were unavailable, so no Shepard documents could illuminate which of the three divisible crimes Johnson had committed. As a result, the majority, in an opinion by Justice Scalia, applied the pure categorical approach to look only at “the least of these acts,” namely “actually and intentionally touching.”[39] The Court read ACCA’s “physical force” provision to require force that is violent.[40] Following this definition, the majority reasoned that “touching” may lack sufficient violence to reach the level of physical force necessary to constitute a violent felony under federal law. Defendants in Florida, including Johnson, could therefore be convicted of felony battery without having committed a violent felony within the meaning of ACCA.[41]

In dissent, Justice Alito strongly objected on several grounds to the majority’s characterizations and reasoning. He first challenged the Court’s tortured understanding of “physical force” as requiring violence—which he persuasively demonstrated does not accord with the common-law definition of “force.”[42] From there, Justice Alito noted the inevitable “untoward consequences” of the majority’s interpretation of Florida’s battery offense for the purposes of ACCA.[43] Numerous states are like Florida: they have indivisible battery provisions “govern[ing] both the use of violent force and offensive touching,” and charging instruments and jury instructions that “simply track the language of the statute” without distinguishing the type of force used by the defendant.[44] The inevitable result would be a windfall to defendants who in fact have used violence in committing a battery, solely because of the statutory grouping-conventions and the record practices of the state of conviction. More generally, once a crime is labeled categorically “nonviolent” under the Court’s approach, it cannot qualify as an ACCA predicate even if committed in a violent manner.

The Court waved away Justice Alito’s concerns, noting that the government had on some occasions successfully used the modified categorical approach with Shepard documents. At the same time, the court did acknowledge that the “absence of records will often frustrate application of the modified categorical approach—not just to battery but to many other crimes as well.”[45]

Justice Alito also worried in Johnson that the majority would “hobble at least two federal statutes” that also contain the term “physical force.”[46] Under 18 U.S.C. § 922(g)(9), a person convicted of a “misdemeanor crime of domestic violence” may not lawfully possess a firearm, and “misdemeanor crime of violence” is defined to include crimes with “an element, the use or attempted use of physical force.”[47] And under 8 U.S.C. § 1227(a)(2)(E), an alien convicted of a “crime of domestic violence” is subject to removal, with “crime of domestic violence” defined to include an offense with “an element the use [or] attempted use . . . of physical force.”[48] If Johnson’s definition of “physical force” and its strict adherence to the categorical approach were to govern interpretation of these terms, many persons convicted of serious spousal or child abuse would be allowed to possess firearms or remain within the United States.

Justice Alito was prescient on this score. The interconnectedness of various criminal statutes has permitted defendants to apply categorical-approach arguments across different statutes, both state and federal. Consider, for example, the recent litigation over the constitutionality of “residual clauses.” Residual clauses generally encompass any “violent felony” (or the analogous “crime of violence”), as defined to include offenses that that pose a sufficient degree of “risk” of physical injury.[49] Since the dawn of the categorical era, judges had been required to apply that approach to determine whether the “ordinary case” of the prior crime at issue surpassed the risk threshold so to count as a violent felony under the applicable residual clause.[50] However, in 2015, the Supreme Court held (per Justice Scalia) in Johnson v. United States that ACCA’s residual clause was unconstitutionally vague because of the “unpredictability and arbitrariness” of judges applying the categorical approach to determine what conduct possessed sufficient risk.[51] And in 2018[52] and 2019[53] the Court applied Johnson to hold nearly identical residual clauses in two other federal statues unconstitutionally vague.

Moreover, even though the wording of these residual clauses is virtually identical, the consequences of the Court nullifying these clauses are not identical. In Johnson, the Court considered ACCA’s residual clause—which, had it not been struck down as unconstitutionally vague, would have had the effect of enhancing an already convicted defendant’s sentence on his current federal offense. In the 2019 case, United States v. Davis, the residual clause at issue also sought to define “crime of violence.”[54] But the operative effect of applying this residual clause would not be to enhance the sentence for the defendant’s current offense. Rather, the residual clause in Davis was part of the substantive offense that the defendant was convicted of in the case-at-hand—here, using a gun in furtherance of any federal “crime of violence.”[55] Simply put, if the residual clause in Davis was found unconstitutional, then the prohibition itself was unconstitutional and a defendant could not be prosecuted under it. In light of the Court’s holding, post-Davis defendants challenging their convictions under this residual clause will have their convictions thrown out entirely.

Writing in dissent,[56] Justice Kavanaugh—joined by Justices Thomas and Alito and in relevant part by Chief Justice Roberts—despaired the practical implications that Justice Alito had predicted in Johnson: namely, all sorts of offenders convicted under the residual clause could now seek to vacate their convictions. To illustrate the absurdity of the Court’s decision, Justice Kavanaugh offered up several examples of defendants now off the hook due to the nullification of the firearm-in-furtherance residual clause: a defendant convicted of assault with intent to murder after shooting his wife multiple times, a defendant convicted of arson for throwing a Molotov cocktail to firebomb a shop, a defendant who kidnapped a man and severely beat him with threats to kill him, and so on.[57] By constraining the Court to consider the “imagined conduct of a hypothetical defendant rather than [] the actual conduct of the actual defendant,” the categorical approach has yielded “serious consequences.”[58]

Justice Alito reserved his strongest criticisms of the categorical approach for his dissent in Mathis v. United States, issued in 2016.[59] As others have noted, Justice Alito’s Mathis dissent is “crucial [to an] understanding of his jurisprudence.”[60] In Mathis, the Supreme Court confronted a categorical-approach question nearly identical to Taylor’s: whether Iowa burglary, which reaches unauthorized entry into any “building, structure, [or] land, water, or air vehicle,” counts as a burglary under ACCA, which only reaches unauthorized entry into a “building or other structure.”[61] The more precise (if mind-numbing) question before the Court was whether the modified categorical approach (where the sentencing court may review Shepard documents to narrow its inquiry) applied to a statute listing “multiple, alternative means of satisfying one (or more) of its elements,” as opposed to alternative elements.[62]

The Iowa burglary statute was undoubtedly broader than the federal definition of burglary under the approach of Taylor. But in Mathis, the Solicitor General invited the Court to loosen its formalism. If the sentencing court could review Shepard documents, it might conclude that Mathis had in fact burglarized a “building or other structure” within the meaning of ACCA and his prior conviction for Iowa burglary would “count” toward ACCA’s sentencing enhancement.[63] Six members of the Court declined the Solicitor General’s invitation,[64] reasoning that the Iowa burglary statute’s elements are broader than the federal definition of battery no matter “[h]ow a given defendant actually perpetrated the crime . . .[or] the ‘underlying brute facts or means’ of commission.”[65] The Court refused to stray course from Taylor and Shepard based on the text of ACCA, Sixth Amendment Apprendi concerns, and avoiding unfairness arising from possible “errors” in the trial record related to statutory means of committing an offense.[66]

In his memorable and withering dissent, Justice Alito compared the Court’s refusal to deviate from the categorical approach to the story of Sabine Moreau, a Belgian woman whose refusal to deviate from her GPS led to her driving 900 miles in the wrong direction toward Zagreb instead of Brussels.[67] With the categorical approach first programmed into the Court’s GPS in Taylor in 1990, “the Court set out on a course that has increasingly led to results that Congress could not have intended.”[68]

Here we may review a few examples from cases in which Justice Alito had previously opined. As the Justice noted in Mathis, the result of that decision would be that burglary convictions in many states could be disqualified from counting as violent felonies under ACCA,[69] just as under Descamps v. United States, no California burglary conviction could count under ACCA.[70] Moncrieffe v. Holder had rendered convictions in nearly half the states for large-scale drug trafficking to not count as “illicit trafficking in a controlled substance” under the immigration laws.[71] We may add that the year before Mathis, Justice Alito joined Justice Thomas’s dissent in Mellouli v. Lynch,[72] rejecting the majority’s holding that if a state’s drug schedule includes substances not included on the federal drug schedule, a state drug offense may not constitute a “violation of . . . any law . . . relating to a controlled substance,” which is a ground for removal under the Immigration and Nationality Act.[73] This Term, in United States v. Taylor, Justice Alito dissented from the Court’s “veer[ing] off into fantasy land” when it held that an attempted Hobbs Act robbery did not constitute a “crime of violence” in a case where a defendant’s accomplice shot and killed the attempted-robbery victim.[74] As Justice Alito proclaimed in Mathis, the Court had ignored the “warning bell” of such anomalous results and “ke[pt] its foot down and drive[n] on” with the categorical approach.[75] Justice Alito despaired that such anomalies are the inevitable result of the Court’s unceasing formalism.

Adding insult to injury, the categorical approach’s premium for abstract inquiry often leaves sentencing courts up a creek without a paddle. The threshold element/means distinction at issue in Mathis is hardly an insignificant undertaking for a sentencing court, which must typically identify a state-court precedent addressing whether a provision of a criminal statute is a means or element. Where no precedent exists, a sentencing court has to make this distinction itself. The means/element determination in Mathis only seemed “easy,” Justice Alito explained—in a not-atypical insight borne of his penchant for legal realism—because Mathis had a “fortified legal team that took over [his] representation after this Court granted review [and] found an Iowa case on point.”[76] This belated discovery evinces, in the real world of state statutes being consulted by federal sentencing judges, the inordinate difficulty of determining whether a statutory provision constitutes a means or an element.

Drawing on his decade on the Supreme Court bench at the time of Mathis, Justice Alito offered an alternative, an approach for the “real world,”[77] that would avoid the mess of the categorical approach:

Allow a sentencing court to take a look at the record in the earlier case to see if the place that was burglarized was a building or something else. If the record is lost or inconclusive, the court could refuse to count the conviction. But where it is perfectly clear that a building was burglarized, count the conviction.[78]

As Justice Alito had suggested before in Descamps,^[79] the Court should drop its formalistic inquiry into whether a statute is divided into elements or means and instead delve into the factual record to settle whether the prior conviction can trigger a sentencing enhancement under federal law. If the factual record is insufficient to determine that a prior conviction falls within the definition of a “violent felony” or “serious drug offense,” the prior conviction won’t count. In Justice Alito’s view, the Court should discontinue its practice of concocting hypothetical crimes and fact patterns and shed the conceit that “[r]eal-world facts are irrelevant.”^[80] Like Ms. Moreau, the Court has driven past numerous signs that it is “off course,” but it has rebuffed “opportunities to alter its course. . . , traveling even further away from the intended destination.”[81]

* * *

In the categorical-approach cases, Justice Alito has shown an abiding disdain for abstract inquiries that turn a blind eye to real-world consequences. In his more than fifteen years on the Court, several of his prognostications have been proven correct, and his consistent critique may have won over some of his colleagues.[82] For example, in the United States v. Taylor case decided in 2022,[83] Justice Thomas expressed openness to abandoning the categorical approach and adopting a conduct-based approach akin to Justice Alito’s proposal in Mathis. At oral argument, Thomas asked both the government and respondent to game out what would happen “if we could abandon the categorical approach.”[84] Naturally, the government noted that it had not briefed the issue but would welcome such a change in light of “the judicial . . . chorus of complaints about the categorical approach that has been growing ever louder.”[85] Although the case was decided 7-2 with Justices Alito and Thomas in dissent, Thomas took the opportunity to recommend overruling the Court’s categorical-approach precedents, which have “led the Federal Judiciary on a ‘journey Through the Looking Glass.’”[86] Like Justice Alito, Thomas would extinguish the categorical approach’s reliance on hypothetical defendants committing hypothetical crimes and instead adopt a “conduct-based approach” into the defendant’s actual conduct to determine whether a prior offense constitutes a violent felony or crime of violence.[87]

Justice Alito’s pragmatic concerns with the categorical approach have indeed generated a judicial chorus of complaints outside One First Street. In a recent Second Circuit opinion, Judge Michael H. Park (a two-time law clerk of Justice Alito) noted the “absurdity of the exercise” of the categorical approach, which requires judges to “ignore the actual facts before them and instead to theorize about whether certain crimes could be committed without violent force.”[88] The categorical approach “perverts the will of Congress, leads to inconsistent results, wastes judicial resources, and undermines confidence in the administration of justice.”[89] Judge Park went on to cite sixteen federal-court opinions concurring with this sentiment, further measuring the reach of Justice Alito’s concerns.[90]

Judge Reena Raggi, also of the Second Circuit, had recent occasion to opine on the practical consequences of the categorical approach in a case vacating a conviction for Hobbs Act robbery, resembling the recent Term’s Taylor decision concerning attempted Hobbs Act robbery.[91] Judge Raggi noted the irony that in a case where there is “no question” that the crime of conviction “was violent, even murderous,” the conviction must be vacated in part because it cannot be deemed a crime of violence through the “commands [of] the categorical approach.”[92]

Judge William Pryor of the Eleventh Circuit put his disdain for the categorical approach more simply: “It’s nuts.”[93] He asked, “How did we ever reach the point where” we “must debate whether a carjacking in which an assailant struck a 13-year-old girl in the mouth with a baseball bat and a cohort fired an AK-47 at her family is a crime of violence? . . . Congress needs to act to end this ongoing judicial charade.”[94] If Justice Alito could respond to Judge Pryor’s charge that the criminal-justice system must navigate out of the categorical-approach quagmire, he might warn Judge Pryor, “Don’t trust your GPS.”

III.           Mens Rea

In his opinions construing the mens rea requirement for a variety of federal crimes, Justice Alito has exhibited his characteristic pragmatism and decried the far-reaching ramifications of the Court’s decisions. Unlike the categorical-approach context, where criminal liability is not typically at issue,[95] mens rea is often a defendant’s best and last line of defense. Mens rea is derived from the classic maxim, actus non facit reum, nisi mens sit rea,[96] or as William Blackstone translated it, “an unwarrantable act without a vicious will is no crime at all.”[97] Mens rea is a foundational concept in our criminal law and requires that an individual must have a culpable mental state corresponding to a particular element of a crime (whether it’s an act, result, or the circumstances surrounding the crime).

Like most legal precepts, the necessity of mens rea is not without its exceptions,[98] but as a general matter, courts interpreting criminal statutes must identify the mens rea associated with the other various elements that together comprise a crime. In the contemporary era, the Model Penal Code provides the generally accepted standards of mens rea, which come in four increasingly culpable levels: negligence, recklessness, knowledge, and purpose.[99] In most situations, the higher the level of mens rea, the steeper the government’s evidentiary burden in proving criminal liability. Where there is no direct evidence of the defendant’s mental state, but the defendant clearly engaged in the charged conduct, the defendant’s primary jury argument may be that the government has failed to prove the requisite mens rea by proof beyond a reasonable doubt. Defendants may also argue that the government has put forth insufficient evidence to demonstrate mental culpability or that a criminal prohibition requires a higher tier of mens rea than the government has proven or than has been charged to the jury.

In the categorical-approach context, Justice Alito voiced his concern with formalism obfuscating the facts of a defendant’s case and generating adverse consequences at odds with congressional purpose. These pragmatic concerns are also on display in the mens rea context, where Justice Alito has taken exception to the Court apparently hiding the ball or contorting statutory language. Throughout his service on the Court, Justice Alito has adopted a decidedly non-abstract approach to interpreting mens rea. Even where he is willing to be guided by a default “general presumption”—such as that a statutorily “specified mens rea applies to all the elements of an offense”—he pragmatically insists on leaving room for “instances in which context may well rebut that presumption.”[100] As the following cases reflect, “context” to Justice Alito typically entails the possibility of “odd results,” the risk of opening the floodgates of litigation, and the need for clear and stable precedent.[101]

A revealing example is Justice Alito’s dissent in Elonis v. United States, decided in 2015.[102] In Elonis, the defendant was convicted of violating 18 U.S.C. § 875(c), which makes it a crime to transmit in interstate commerce “any communication containing any threat . . . to injure the person of another.”[103] After his wife left him, Elonis posted rap songs on Facebook containing violent imagery.[104] Although Elonis included disclaimers about his innocent intentions, his wife sought a state-court order of protection.[105] Elonis remained undeterred.[106] At issue was the proper mens rea corresponding to Elonis’s communication of the threat. The statute itself was silent on what mens rea (if any) was required regarding the threat itself, but the Third Circuit inferred that the appropriate level of mens rea was negligence—the lowest level of mens rea.[107] In other words, the government had to show that Elonis was negligent with respect to the threatening nature of his communications.

In an opinion by Chief Justice Roberts, the Supreme Court reversed, finding that a mental state higher than negligence should have been inferred. Invoking its strict-liability precedents,[108] the Court observed “the conventional requirement for criminal conduct [is] awareness of some wrongdoing.”[109] This conventional requirement instructs reluctance to infer a negligence standard.[110] The silence on mens rea in the prohibition Elonis was convicted of violating did “not mean that none exists” and “mere omission from a criminal enactment of any mention of criminal intent” should not be read “as dispensing with it.”[111] After correctly noting that the negligence standard adopted by the Third Circuit did not require the government to prove that Elonis was in fact aware of the threatening nature of his behavior, merely that he was negligent toward its threatening nature, the majority opinion then concluded only that negligence is insufficient for liability under § 875(c).[112]

Quite deliberately, the Court did not answer whether recklessness, knowledge, or purpose would suffice for liability under § 875(c).[113] In light of the brief lip service to this question during oral argument and there being “no circuit conflict over the question” in the majority’s view, a merits decision on the precise mens rea required under § 875(c) was inappropriate.[114]

Without missing a beat, Justice Alito picked up on the Court’s omission. Noting that Marbury v. Madison’s had “famously proclaimed” that the judicial department must “say what the law is,” Justice Alito said the majority opinion had failed in that regard and instead had announced, “It is emphatically the prerogative of this Court to say only what the law is not.”[115] The Court’s decision not to clarify the required mens rea under § 875(c) “is certain to cause confusion” and “regrettable consequences” among the lower courts.[116] Unlike the Supreme Court, which “has the luxury of choosing its docket,” lower courts and juries “must actually decide cases,” which means “applying a standard.”[117] Elonis and the government had in fact both briefed the issue of mens rea, and if the Court thought it lacked sufficient information to reach a merits decision, it could have ordered further briefing and argument.[118]

Concurring in the judgment, Justice Alito would have found that recklessness is enough. He largely agreed with the majority’s default presumption that § 875(c)’s silence as to mental state should require a mens rea more than mere negligence. Following the Model Penal Code, Justice Alito would infer recklessness “when Congress does not specify a mens rea in a criminal statute,” but go no further toward knowledge or purpose.[119] In his view, “[t]here can be no real dispute that recklessness regarding a risk of serious harm is wrongful conduct.”[120] Justice Alito might have also cited the colloquy that he had at oral argument with Deputy Solicitor General Michael Dreeben concerning what Justice Alito referred to as the Model Penal Code’s “razor-thin distinction[s]” between purpose and knowledge and the “considerable difference between” knowledge and recklessness.[121] Depending on the proper level of mens rea, the government’s burden could therefore vary significantly without further instruction from the Court.

Failure to reach an answer on the proper level of mens rea would also have plain adverse consequences, the Justice explained. If purpose or knowledge is required under § 875(c) and a district court instructs the jury that recklessness is sufficient, a defendant may be wrongfully convicted. Yet, if recklessness is enough under § 875(c) and a district court instructs the jury that proof of knowledge or purpose is required, a guilty defendant may be acquitted. With “[a]ttorneys and judges . . . left to guess,”[122] all parties—defendants included—are left in the lurch because the majority decided that hiding the ball (or stopping it short of the goal) was more prudent than reaching the mens rea merits question.

Four years later, in Rehaif v. United States, Justice Alito expressed similar, though distinct, concerns that the Court’s novel reading of the commonly charged firearm-in-possession prohibition[123] would both make it extremely difficult to prove mens rea in many cases, as well as “open[] the gates to a flood of litigation.”[124] Hamid Rehaif had entered the United States on an immigrant student visa, but after receiving poor grades he was kicked out of his university and forfeited his immigration status.[125] Thereafter, Rehaif visited a firing range, and he shot two firearms.[126] The government then charged Rehaif under 18 U.S.C. § 922(g), which provides “[i]t shall be unlawful for any person . . . who [inter alia], being an alien is illegally or unlawfully in the United States[,] . . . [to] possess in or affecting commerce, any firearm or ammunition.”[127] Under the relevant sentencing provision, § 924(a)(2), “[w]hoever knowingly violates” § 922(g) “shall be fined . . . [or] imprisoned not more than 10 years, or both.”[128] The question presented to the Supreme Court was whether the government only had to prove that Rehaif “knowingly” possessed a firearm, or whether the government additionally had to prove Rehaif had a mens rea of knowledge as to his status as “an alien . . . illegally or unlawfully in the United States.”[129]

The majority, in an opinion by Justice Breyer, answered in the affirmative: under § 922(g) the government had to prove both that Rehaif knew he was in possession of a firearm and of his status as an unlawful alien. The Court adopted this interpretation for several reasons, including its “ordinary presumption in favor of scienter”[130] and the grammatical construction of the statute.[131] Because the government failed to show Rehaif knew of his immigration status, the Court reversed Rehaif’s conviction and remanded to the lower court.

One immediate consequence of the Court’s decision in Rehaif was the decision’s retroactive application. Because Rehaif placed knowing possession of a firearm without knowledge of one’s immigration status beyond the reaches of the extant federal criminal law, the decision would apply retroactively under the rule of Teague v. Lane, permitting individuals currently imprisoned under § 922(g) to challenge the validity of their convictions within one year on federal collateral review.[132] Defendants on direct review of § 922(g) convictions could also seek new trials on this basis.

Justice Alito dissented, joined by Justice Thomas. He began by critiquing the majority so “casually” overturning an interpretation of § 922(g) “adopted by every single Court of Appeals” and “used in thousands of cases for more than 30 years.”[133] The Court’s decision was “no minor matter” and disabled one of the nation’s chief tools “to combat gun violence.”[134] Moreover, the decision would create a “mountain of problems” and “swamp the lower courts” with thousands of prisoners seeking collateral relief on the claim that their § 922(g) convictions were defective for failure to charge or prove knowledge with respect to their status.[135] Justice Alito, of course, recognized that the Court must enforce the laws of Congress “even if we think that doing so will bring about unfortunate results,” but usually the Court requires “clear indication of congressional intent” before wreaking such havoc.[136] Yet, in Rehaif, the Court was intrigued by a “superficially appealing but ultimately fallacious argument” and diverged from its usual practice of resolving conflicts among the lower courts, and preserving a long-established interpretation absent evidence that it had “worked any serious injustice.”[137]

Justice Alito tried to set the record straight after the majority presented a “bowdlerized version of the facts.”[138] The Court, in his view, sought to paint an “entirely imaginary case, a heartless prosecution,” that would evoke sympathy for the Court’s ultimately baseless statutory construction.[139] In yet another clear nod to legal realism, the Justice explained that in fact, Rehaif was not a down-on-his-luck immigrant student. Rather, after his expulsion and visa revocation, Rehaif moved into a hotel facing the airport, paid more than $11,000 in cash for his lodging, and frequented a firing range over the course of fifty-three days.[140] Justice Alito appeared perturbed that the Court was pulling the wool over readers’ eyes, stretching and molding the story of a relatively unsympathetic defendant to produce a defendant-friendly decision at odds with thirty years of precedent and with untoward consequences. These sentiments undoubtedly remind us of his categorical-approach jurisprudence, which critiques the Court for proscribing review of the full factual record and only permitting “bowdlerized” Shepard documents to reveal the facts underlying a conviction.

Justice Alito’s penchant for pragmatism is perhaps matched by his knack for metaphor—in Mathis analogizing the categorical approach to a discombobulated GPS, and in Rehaif, analogizing the majority’s “purportedly textualist argument” to “a magic trick.”[141] Because the firearm-in-possession statute’s “knowing” mens rea requirement is housed in § 924(a)(2)—an entirely separate provision from the firearm-possession prohibition itself, which is in § 922(g)—“any attempt to combine the relevant language” of the two statutory provisions “necessarily entails significant choices that are not dictated by the text of those provisions.”[142] Rehaif naturally preferred applying the knowledge requirement broadly to include the status elements of § 922(g), because this would increase the government’s burden. The Court fell for the defendant’s move, which Justice Alito referred to as the trick “presto chango.”[143] But Justice Alito asserted that “[t]he truth behind the illusion,” is that under ordinary usage, four different readings of the statute are possible.[144] The majority’s sleight of hand was to suggest that among the four plausible interpretations, Congress intended for the option with “a very high mens rea requirement,” requiring knowledge for the status element.[145]

While this might not make much difference in many cases where the § 922(g) status at issue is unlawful presence in the country, it would enormously increase the government’s burden in prosecuting the most common firearm-possession crime, where the possessing defendant is a convicted felon. The government would now need to prove that the defendant knew he had been convicted not just of a crime, but of a crime within the category of “felony.”[146] Typically, to avoid the introduction of evidence concerning a prior offense, a defendant will stipulate to his felon status, but after Rehaif, the prosecution may need to offer evidence about the nature of the prior felony to allow the jury to infer knowledge—quite a defendant-unfriendly consequence of the majority’s holding.[147] Moreover, if the knowledge requirement of a gun-possessor’s status also applies to the prohibition on sale of firearms to persons falling within a § 922(g) category,[148] it seems highly unlikely that most sellers will know whether the purchaser falls into one of the § 922(g) status categories.[149] Finally, the Court’s decision contravenes the “practical unanimity” of the courts of appeals on these questions; instead the Court invented hypothetical, conflicting interpretations on its way to approving an interpretation that no circuit had thought to adopt.[150]

Echoing his concerns in the residual-clause context discussed previously,[151] Justice Alito’s most prominent concern with Rehaif was its inevitable opening of litigation floodgates. Because the Court’s decision applied retroactively,[152] the “[t]ens of thousands of prisoners . . . currently serving sentences for violating 18 U.S.C. § 922(g)” may be eligible for relief, such as a new trial if the case is still on direct review or even release through collateral review under 28 U.S.C. § 2255.[153] Those currently imprisoned for § 922(g) convictions may have their convictions vacated if they can demonstrate they are innocent of violating § 922(g), which, after Rehaif, only requires showing they did not know they fell into a § 922(g) status category. The requirement that district courts “hold a hearing . . . and make a credibility determination as to the prisoner’s subjective mental state at the time of the crime” many years before “will create a substantial burden on lower courts, who are once again left to clean up the mess the Court leaves in its wake as it moves on to the next statute in need of ‘fixing.’”[154] This too will not “necessarily be limited to § 922(g)” and may spread to other statutes, Justice Alito worried.[155]

Lower courts, prosecutors, and defendants would all pay the price of the Court’s purportedly textualist decision, and Justice Alito quantified the number of § 922(g) offenders that could raise Rehaif claims. According to the U.S. Sentencing Commission, in fiscal year 2020 alone, that number was 6,782 individuals.[156] Justice Alito’s concern with the real-world consequences of Rehaif evince his hesitance to destabilize or set unclear precedents, especially in a case like Rehaif where there was no lower-court conflict warranting the Court’s review.

As Justice Alito prophesied in 2019, Rehaif’s effects have now reached well beyond the ambit of § 922(g) status offenses. This past Term, in Ruan v. United States, the Court applied Rehaif’s “mens rea canon,” as Justice Alito dubbed it, whereby “the Court interprets criminal statutes to require a mens rea for each element of an offense ‘even where the most grammatical reading of the statute does not support’ that interpretation.”[157] In Ruan, the relevant provision of the Controlled Substances Act (CSA) makes it a federal crime “[e]xcept as authorized[,] . . . for any person knowingly or intentionally . . . to manufacture, distribute, or dispense . . . a controlled substance.”[158] The majority held that the mens rea requirement (“knowingly or intentionally”) applied to the “except as authorized” provision, requiring the government to prove beyond a reasonable doubt that a defendant “knew that he or she was acting in an unauthorized manner, or intended to do so.”[159] The majority offered four justifications for this interpretation,[160] although as Justice Alito points out, “[i]t bases this conclusion not on anything in language of the CSA” but rather the mens rea canon established in Rehaif.[161]

To Justice Alito, the Court’s effort to apply the mens rea canon to the CSA “rests on an obvious conceptual mistake.”[162] The “[e]xcept as authorized” clause represents an affirmative defense, not an element, and the mens rea canon is inapt in this context.[163] Alito bases this interpretation on several factors. First, “[a]s a matter of elementary syntax,” the “knowingly and intentionally” clause modifies the verbs that follow and do not operate backwards to the “introductory phrase ‘except as authorized.’”[164] As Justice Alito quipped at oral argument, “[W]e are interpreting statutes and regulations, and maybe we ought to start with what they actually say.”[165] Second, the authorization clause “lacks the most basic features of an element of an offense,” such as mandatory inclusion in every § 841 indictment.[166] Yet, the CSA specifically provides that it is not “necessary for the United States to negative any exception or exception set forth in [the relevant subchapter],” implying the authorization clause lacks one of the key indicia of statutory elements.[167] Third, the authorization clause operates as a proviso giving “justification or excuse” for conduct that otherwise satisfies the elements of an offense.[168] Under the Court’s precedents, “an exception made by a proviso” designates an “affirmative defense that the Government has no duty to ‘negative.’”[169] Fourth, the majority, without reference, reverses the common-law rule that defendants bear the burden of production and persuasion of any affirmative defense by instead holding the government must prove unauthorized use beyond a reasonable doubt after a defendant has made a showing that their activity was authorized.[170]

In this most recent example of Justice Alito’s mens rea jurisprudence we observe two of his trademarks. First, the Justice recoils at the Court’s expansion of the judicial role at the expense of Congress. By reading mens rea into every provision of a criminal statute to craft criminal offenses in a “sound” and “just” manner, the Court has effectively usurped Congress’s role in defining the elements of a criminal offense.[171] Not to mention the majority’s trampling of ordinary usage and grammar. The Ruan Court also stumbled over the subtle distinction between the mens rea canon illuminating what Congress intended to include as an element and what the Justices want to include as an element as a matter of lenity. A Court capable of rewriting criminal statutes proves hard to square with the Constitution’s command of separation of powers.

Second, and relatedly, Justice Alito warns that when the Court reaches for a “sound” or “just” result in the criminal law, it ignores the cascading consequences. The Ruan Court’s elision of the element / affirmative defense distinction in the name of lenity makes it unclear “[h]ow many other affirmative defenses might warrant similar treatment.”[172] Such a blasé attitude toward fundamental criminal-law concepts “leaves prosecutors, defense attorneys, and the lower courts in the dark.”[173]

* * *

So, what are we to make of Justice Alito’s mens rea jurisprudence? A common thread (reminiscent of his categorical-approach jurisprudence) is an abiding concern with the Supreme Court engaging in theoretical expeditions at great cost to the administration of criminal law. Whether it is befuddling lower courts as in Elonis, inviting dubious collateral attacks as in Rehaif, or confusing the status of statutory affirmative defenses as in Ruan, Justice Alito is often one of the few voices on the Court calling out real-world consequences.

Moreover, the Justice is clearly concerned not just with the lower courts, but also the victims of crimes. As Justice Alito has asked advocates in oral argument, “[W]hat do you say to the amici who say that if your position is adopted, this is going to have a very grave effect . . . . [Are] they[] just wrong, they don’t understand the situation?”[174] Dashing any air of pretension, he directs advocates to quantify the concrete effects of their preferred position, asking in Rehaif, for example, “[h]ow many people are now serving time in federal prison under the felon-in-possession statute?”[175] Or in Ruan, asking whether the petitioner’s interpretation of the CSA would require dismissal of “all the other indictments” in every case the Department of Justice brought under the relevant provision.[176] With such high practical stakes on the line, Justice Alito prefers to prioritize context over abstraction, and reality over theory.

IV.           Conclusion

In opening this chapter, I noted that Justice Alito is a natural judge. By this I meant that Justice Alito conceives of his job as getting a case right and not “winning.” As a former Department of Justice employee, U.S. Attorney for the District of New Jersey, and Third Circuit Judge, Justice Alito is well-versed in how a Supreme Court opinion coming down from on high can wreak havoc on people working in or confronting the nation’s criminal-justice system. In his categorical-approach and mens rea jurisprudence, Justice Alito has demonstrated a discerning sense of how a particular decision can unleash a chain reaction of negative consequences borne by the lower courts, prosecutors, defendants, and victims. He abjures pure textual formalism in statutory interpretation, if, for instance, giving shrift to a curious and perhaps errant comma would produce real-world results that are arbitrary, inconsistent, and contrary to Congress’s evident purpose in enacting the statute.

It’s hard to neatly define the Justice’s pragmatism, but his criminal-law jurisprudence reveals an unyielding commitment to avoid prejudging a case without a full accounting of the facts, the statutory landscape, and the practical consequences. Justice Alito’s refusal to give in to the “cavalier treatment of . . . important question[s]”[177] has secured his position as the Justice who steadfastly acknowledges real-world consequences and Congress’s purpose in the criminal law. Often through his concurrences and dissents, Alito has served as the Court’s criminal-law oracle, time and again accurately predicting how the Court’s decisions prioritizing abstraction over text and practical consequences will yield adverse consequences. But there is nothing supernatural about his prophecies. Justice Alito is simply a natural when it comes to judging, thinking two steps ahead of the curve.

^* Lafayette S. Foster Professor of Law, Yale Law School. I thank several Yale Law School students who aided my efforts to understand the complex layers of the “categorical” approach and Justice Alito’s concerns about this doctrine—Sarah Jeon ’23, Caroline Lefever ’24, and Valerie Silva Parra ’23. I especially want to acknowledge and thank Joshua Altman ’22 for his prodigious research and our many conversations about Justice Alito’s jurisprudence.

[2] Antonin Scalia, A Matter of Interpretation 13 (1997).

[3] As discussed in Part I, infra, the categorical approach also applies in the context of defining certain substantive criminal offenses. See, e.g., United States v. Davis, 139 S. Ct. 2319 (2019); United States v. Taylor, 142 S. Ct. 2015 (2022).

[4] Manuel v. City of Joliet, 580 U.S. 357, 385 (2017) (Alito, J., dissenting).

[5] Mathis v. United States, 579 U.S. 500, 538, 543 (2016) (Alito, J., dissenting).

[6] Armed Career Criminal Act of 1984, Pub. L. No. 98-473, 98 Stat. 2185, amended by Career Criminals Amendment Act of 1986, Pub. L. 99-308, 100 Stat. 459 (codified as amended at 18 U.S.C. § 924(e) (2018)). The Armed Career Criminal Act (ACCA) has proven fertile ground for successful criminal appeals over the decades, and in recent Terms, the Supreme Court has continued to grant certiorari in ACCA cases. See, e.g., Johnson v. United States, 576 U.S. 591 (2015) (voiding ACCA’s residual clause on Due Process grounds); Welch v. United States, 578 U.S. 120 (2016) (holding that Johnson applies retroactively on federal collateral review under 26 U.S.C. § 2255 (2018)); Borden v. United States, 141 S. Ct. 1817 (2021) (holding that under ACCA’s elements clause, a criminal offense with a mens rea of recklessness does not qualify as a “violent felony”); Wooden v. United States, 142 S. Ct. 1063 (2022) (holding that a criminal defendant’s ten prior burglary offenses—all of which arose from a single criminal episode—did not constitute distinct “occasions” and thus counted as a single prior conviction for ACCA purposes).

Although the categorical approach is most closely associated with ACCA, it also operates with respect to some substantive provisions of the criminal code that define the very federal crime of which the defendant has been convicted. See, e.g., United States v. Davis, 139 S. Ct. 2319 (2019) (applying the categorical approach to 18 U.S.C. § 924(c)(3) (2018)); United States v. Taylor, 142 S. Ct. 2015, 2020–21 (2022) (applying the categorical approach to hold that attempted Hobbs Act robbery does not constitute a “crime of violence” under § 924(c)(3)(A)’s elements clause, which prohibits use of a firearm in connection with a “crime of violence”); see also infra notes 51–55, 80–84, discussing the categorical approach’s application to substantive criminal offenses outside of the ACCA ambit in Davis and Taylor).

The categorical approach is also central to immigration law, where courts determine whether an immigrant’s prior convictions may trigger removal. See, e.g., 8 U.S.C. § 1227(a)(2)(B)(i) (2018) (authorizing deportation of an alien “convicted of a violation of . . . any law or regulation of a State, the United States, or a foreign country relating to a controlled substance”); id. § 1227(a)(2)(A) (authorizing deportation of an alien convicted of crimes of moral turpitude, multiple criminal convictions, and aggravated felony, among other crimes). As such, the categorical approach’s pedigree may stretch back as far as the early twentieth century in the immigration context despite the Supreme Court’s more active use of this tool over the last thirty years. See Mellouli v. Lynch, 575 U.S. 798, 805–06 (2015) (“The categorical approach ‘has a long pedigree in our Nation’s immigration law.’ As early as 1913, courts examining the federal immigration statute” assessed past criminal convictions based on analysis of the statutory offense, not the underlying facts of the case (quoting Moncrieffe v. Holder, 569 U.S. 184, 191 (2013).).

With immigration statutes incorporating provisions of the criminal code, the Court’s use of the categorical approach in the criminal context may generate collateral consequences in immigration law. For example, if the Court has struck down a criminal provision using the categorical approach, immigration consequences will follow. See, e.g., Sessions v. Dimaya, 138 S. Ct. 1204 (2018) (striking the residual clause of 18 U.S.C. § 16(b) as unconstitutionally vague under the categorical approach in Johnson v. United States, 576 U.S. 591 (2015)); see also infra notes 46–55 and accompanying text (discussing the Johnson line of cases concerning the constitutionality of various residual clauses).

[7] 18 U.S.C. § 924(a)(2) (2018).

[8] Id. at §§ 922(g)(1), 924(e)(1).

[9] Taylor v. United States, 495 U.S. 575 (1990).

[10] As discussed below, see infra notes 24–27 and accompanying text, the “modified” categorical approach permits a federal court to look both to the statute of conviction and a “limited list of judicial sources,” referred to as Shepard documents. U.S. Sent’g Comm’n, Primer on Categorical Approach 2 (2021), https://perma.cc/BL47-X3EE; see Shepard v. United States, 544 U.S. 13 (2005).

[11] U.S. Sent’g Comm’n, supra note 10, at 1–2.

[12] See supra notes 8–10 and accompanying text.

[13] H.R. Rep. No. 98-1073, at 1 (1984); see also S. Rep. No. 98-225 (1983).

[14] H.R. Rep. No. 98-1073, at 1 (1984).

[15] 18 U.S.C. § 924(e) (2018).

[16] Id. § 922(g).

[17] In § 924(e), Congress has provided broad categories of what constitutes a “violent felony,” but it has provided few other details. A prior offense may constitute a violent felony if it falls within the elements clause of § 924(e)(2)(B)(i) (defining a violent felony to include an offense that “has as an element the use, attempted use, or threatened use of physical force against the person of another”) or within the enumerated-offenses clause of § 924(e)(2)(B)(ii) (defining a violent felony to include an offense that “is burglary, arson, or extortion, involves use of explosives”). In 2015, the Court held that the so-called “residual clause” of ACCA is unconstitutionally vague. See Johnson v. United States, 576 U.S. 591 (2015) (Scalia, J.) (holding 18 U.S.C. § 924(e)(2)(B)(ii) unconstitutional on the grounds of vagueness. That provision defined a “violent” felony to include an offense that “involves conduct that presents a serious potential risk of physical injury to another”).

[18] 495 U.S. 575 (1990).

[19] Id. at 578–79.

[20] Id. at 600.

[21] Id.

[22] Id. at 602.

[23] Id. at 599–600.

[24] Id. at 601.

[25] The rule of Apprendi v. New Jersey, 530 U.S. 466 (2000) (Scalia, J.), would not surface until ten years after the Supreme Court’s adoption of the categorical approach in 1990. The interplay between the two lines of doctrine is complex. For the most part, the Court has resisted the suggestion that sentencing judges peering into the factual record would violate Apprendi’s requirements of jury fact-finding and proof beyond a reasonable doubt. See In re Winship, 397 U.S. 358 (1970). Apprendi held that “[o]ther than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt.” Apprendi, 530 U.S. at 490. As Justice Alito and others have noted, the categorical approach—as well as the modified categorical approach discussed below, see infra notes 24–27 and accompanying text—is more akin to statutory interpretation than to the judicial fact-finding addressed in Apprendi. To be sure, however, there is disagreement on this question. Compare James v. United States, 550 U.S. 192, 213–14 (2007) (Alito, J.) (noting that the categorical approach is statutory interpretation and thus not subject to Apprendi), and Moncrieffe v. Holder, 569 U.S. 184, 198 (2013) (Sotomayor, J.) (“But those [Apprendi] concerns do not apply in this context. Here we consider a ‘generic’ federal offense in the abstract, not an actual federal offense being prosecuted before a jury. Our concern is only which facts the CSA relies upon to distinguish between felonies and misdemeanors, not which facts must be found by a jury as opposed to a judge, nor who has the burden of proving which facts in a federal prosecution.”), with James, 550 U.S. at 231 (Thomas, J., dissenting) (dissenting on the ground that ACCA runs afoul of Apprendi because its sentencing enhancements require judges to “make a finding that raises [a defendant’s] sentence beyond the sentence that could have lawfully been imposed by reference to facts bound by the jury or admitted by the defendant.” (citing United States v. Booker, 543 U.S. 220, 313 (2005) (Thomas, J., dissenting)).

Justice Alito suggested in a recent case that the Apprendi-motivated push to adopt the categorical approach is inconsistent with the original meaning of the Sixth Amendment, which more likely envisioned sentencing as largely discretionary. See United States v. Taylor, 142 S. Ct. 2015, 2033 n.1 (2022) (Alito, J., dissenting) (first citing Michael McConnell, The Booker Mess, 83 Denver U. L. Rev. 665, 679 (2006)); then citing Jonathan Mitchell, Apprendi’s Domain, 2006 Sup. Ct. Rev. 297, 298–99; then citing Rory K. Little & Teresa Chen, The Lost History of Apprendi and the Blakely Petition for Rehearing, 17 Fed. Sent’g Rep. 69, 69–70 (2004); and then citing Stephanos Bibas, Judicial Fact-Finding and Sentence Enhancements in a World of Guilty Pleas, 110 Yale L.J. 1097, 1123–32 (2001)).

Moreover, Apprendi concerns do not underlie all categorical-approach cases. For example, in the context of § 924(c) substantive offenses, the jury, not a sentencing judge, will determine whether a defendant’s conduct amounted to a “crime of violence” in breach of § 924(c)’s prohibition against use of a firearm in connection with a “crime of violence.” See Taylor, 142 S. Ct. at 2026–33 (2022) (Thomas, J., dissenting); id. at 2033 n.1 (Alito, J. dissenting) (“[N]o Sixth Amendment concern is implicated under § 924(c).”).

[26] Taylor, 495 U.S. at 590.

[27] Id. at 589.

[28] Shepard, 544 U.S. at 26.

[29] See Mathis v. United States, 579 U.S. 500 (2016); Descamps v. United States, 570 U.S. 254, 258 (2013). The Court permits review of such documents where a statute is “divisible,” meaning it contains several different offenses or alternative elements under which a conviction may be sustained. See U.S. Sent’g Comm’n, supra note 10, at 9–10. Yet, where a statute describes a single crime and enumerates alternative means of committing the crime, the statute is considered indivisible, and the court may not use Shepard documents as permitted under the modified categorical approach. Id. at 3. However, if a statute of prior conviction is divisible (meaning Shepard documents are fair game), the Court has indicated that the sentencing court should look to the “least of [the divisible] acts” under the statute as the point of comparison to the federal statute. Johnson v. United States, 559 U.S. 133, 137 (2010).

[30] James, 550 U.S. 192.

[31] 553 U.S. 377 (2008).

[32] James, 550 U.S. at 207–08.

[33] Rodriquez, 553 U.S. at 388.

[34] Id. at 389.

[35] Id.

[36] 559 U.S. 133, 136–38 (2010).

[37] Id. at 136.

[38] Id. at 137.

[39] Id. at 137.

[40] Id. at 140.

[41] Id. at 138–40.

[42] Id. at 147–48 (Alito, J., joined by Thomas, J., dissenting).

[43] Id. at 151.

[44] Id. at 152.

[45] Id. at 145 (majority opinion).

[46] Id. at 152 (Alito, J., joined by Thomas, J., dissenting).

[47] 18 U.S.C. §§ 921(a)(33)(A)(ii), 922(g)(9) (2018).

[48] 8 U.S.C. § 1227(a)(2)(E) (2018); 18 U.S.C. § 16(a) (2018).

[49] See, e.g., 18 U.S.C. §§ 16(b), 924(c)(3)(B), 924(e)(2)(B)(ii) (2018); see also supra note 16 and accompanying text.

[50] United States v. Davis, 139 S. Ct. 2319, 2326 (2019) (Gorsuch, J.) (invalidating the residual clause of the federal felon-in-possession statute).

[51] 576 U.S. 591, 592 (2015).

[52] Sessions v. Dimaya, 138 S. Ct. 1204 (2018) (Kagan, J.) (concerning California first-degree burglary).

[53] Davis, 139 S. Ct. 2319 (concerning the use of firearms in connection with a federal crime of violence).

[54] 18 U.S.C. § 924(c)(1)(A) (2018).

[55] Id.

[56] Davis, 139 S. Ct. at 2354 (Kavanaugh, J., joined in part by Roberts, C.J., Thomas & Alito, JJ., dissenting).

[57] Id. at 2353–54 (citing cases).

[58] Id. at 2355.

[59] 136 S. Ct. 2243, 2266 (Alito, J., dissenting).

[60] Steven G. Calabresi & Todd W. Shaw, The Jurisprudence of Justice Samuel Alito, 87 Geo. Wash. L. Rev. 507, 513 (2019).

[61] Compare 18 U.S.C. § 924(e) (2018), with Iowa Code § 702.12 (2013). See generally supra notes 24–27 and accompanying text.

[62] Mathis v. United States, 579 U.S. 500, 503 (2016) (Kagan, J.) (emphasis added); see also supra note 27 and accompanying text. What’s more, Justice Alito has also held the Court to account when it fails to properly adhere to its modified-categorical-approach precedents. In United States v. Taylor, for example, the majority overlooked the fact that § 924(c)(3)(A)’s definition of “crime of violence” contains disjunctive elements, which would typically trigger the modified categorical approach. 142 S. Ct. 2015, 2033–37 (2022) (Alito, J., dissenting). In Taylor, applying the approach of Shepard, Justice Alito looked to Taylor’s plea agreement, which admitted that Taylor “and his accomplice intended to lure [the victim] into an alleyway, hold him at gunpoint, and take his money ‘by force’ in the event that he resisted.” Id. at 2036. In the Justice’s view, this should have been more than enough “to show that Taylor’s actual crime ‘ha[d] as an element the . . . use of physical force against the person . . . of another” under § 924(c)(3)(A). Id.

Justice Alito’s willingness to remind the Court of its own precedents recalls recent administrative-law cases where Justice Alito has noted the Court’s failure to even mention Chevron deference where such deference is likely owed to a federal agency’s interpretation of a statute. See Pereira v. Sessions, 138 S. Ct. 2105, 2121 (2018) (Alito, J., dissenting) (“[T]he Court’s decision implicates the status of an important, frequently invoked, once celebrated, and now increasingly maligned precedent, namely, Chevron . . . [but] the Court, for whatever reason, is simply ignoring Chevron.” (citing Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, 467 U.S. 837 (1984))). This Term, the Court arguably continued its sub silentio overruling of Chevron in West Virginia v. EPA (this time with Justice Alito joining in Justice Gorsuch’s concurrence). 142 S. Ct. 2587 (2022). In this case, the Court refused to defer to the EPA’s interpretation of the Clean Air Act as permitting the promulgation of carbon-emission caps based on a generation-shifting approach. Id. Instead, the majority held that through the Clean Air Act Congress did not intend to house such authority in the EPA because carbon-emissions regulation represents a “major question” of “economic and political significance” best left for congressional resolution absent an unambiguous delegation to an agency. Id. at 2605 (quoting Util. Air Regul. Grp. v. EPA, 573 U.S. 302, 324 (2014)). Justice Alito joined Justice Gorsuch’s concurrence, which made no mention of Chevron but supplied further support for the majority’s use of the major-questions doctrine. In her dissent, Justice Kagan critiqued the majority for failing to follow the Court’s Chevron precedent, specifically “step one” of the Chevron framework where courts are meant to apply the “normal principles of statutory interpretation” before reaching substantive canons of interpretation, such as the major-questions doctrine. See id. at 2635 (Kagan, J., dissenting).

[63] Mathis, 579 U.S. at 503–05.

[64] In addition to the dissenting Justice Alito, Justice Breyer, joined by Justice Ginsburg, also rejected the “means/elements distinction” in a separate dissent. See id. at 523 (Breyer, J., joined by Ginsburg, J., dissenting).

[65] Id. at 501 (majority opinion) (quoting Richardson v. United States, 526 U.S. 813, 817 (1999)).

[66] Id. at 510–513.

[67] Id. at 536–537 (Alito, J., dissenting).

[68] Id. at 538.

[69] Id.

[70] Id. n.2 (citing Descamps v. United States, 570 U.S. 254 (2013)).

[71] Id. (citing Moncrieffe v. Holder, 569 U.S. 184 (2013)).

[72] 575 U.S. 798, 813–14 (2015) (Thomas, J., joined by Alito, J., dissenting).

[73] Id. at 808–13 (majority opinion).

[74] United States v. Taylor, 142 S. Ct. 2015, 2033 (2022) (Alito, J., dissenting).

[75] Mathis, 579 U.S. at 538 (Alito, J., dissenting).

[76] Id. at 540.

[77] Id. at 539; see also Taylor, 142 S. Ct. at 2035 (Alito, J., dissenting) (“The whole point of the categorical approach that the Court dutifully follows is that the real world must be scrupulously disregarded.”).

[78] Id. at 541.

[79] Descamps v. United States, 570 U.S. 254, 281 (2013) (Alito, J., dissenting) (“I would give ACCA a more practical reading. When it is clear that a defendant necessarily admitted or the jury necessarily found that the defendant committed the elements of generic burglary, the conviction should qualify.”).

[80] Mathis, 579 U.S. at 543 (Alito, J., dissenting).

[81] Id. at 543–44.

[82] See, e.g., Borden v. United States, 141 S. Ct. 1817, 1856 (2021) (Kavanaugh, J., joined by Roberts, C.J., Alito & Barrett, JJ., dissenting) (“Because courts use the categorical approach when applying ACCA’s violent felony definition, the Court’s decision today will thus exclude many intentional and knowing felony assaults from those States.”); see also United States v. Davis, 139 S. Ct. 2319, 2337 (2019) (Kavanaugh, J., joined in part by Roberts, C.J., Thomas & Alito, JJ., dissenting).

[83] United States v. Taylor, 142 S. Ct. 2015, 2033 (2022) (Alito, J., dissenting).

[84] Transcript of Oral Argument at 5, 77, Taylor, 142 S. Ct. 2015 (No. 20-1459).

[85] Id. at 5.

[86] Taylor, 142 S. Ct. at 2026 (Thomas, J., dissenting) (quoting Lewis Carroll, Alice in Wonderland and Through the Looking Glass 227 (Julian Messner ed., 1982)).

[87] Id. at 2028; cf. supra notes 31–33 and accompanying text (outlining Justice Alito’s endorsement of the conduct-based approach). Justice Thomas also recommends overruling the Court’s residual-clause decisions, particularly United States v. Davis, 139 S. Ct. 2319 (2019), and adopting a conduct-based approach to § 924(c)(3)(B)’s residual clause that mitigates vagueness worries associated with the categorical approach. Taylor, 142 S. Ct. at 2031–32 (Thomas, J., dissenting).

[88] United States v. Scott, 990 F.3d 94, 125–26 (2d Cir. 2021) (Park, J., concurring).

[89] Id. at 126.

[90] Id. at 126–27 (citing cases).

[91] This case was a follow-on to the Supreme Court’s decision in United States v. Davis, 139 S. Ct. 2319 (2019), which held the residual clause of 18 U.S.C. § 924(c) (2018) to be unconstitutionally vague. The defendant in this case had his conviction vacated because of its reliance on the unconstitutional residual clause.

[92] United States v. Barrett, 937 F.3d 126, 128 (2d Cir. 2019) (on remand from the Supreme Court). Although the categorical approach would apply in this case, the Second Circuit vacated the case primarily based on Davis’s holding that the residual clause of 18 U.S.C. § 924(c)(3)(B) is unconstitutionally vague.

[93] Ovalles v. United States, 905 F.3d 1231, 1253 (11th Cir. 2018) (Pryor, J., concurring).

[94] Id.

[95] But see Davis, 139 S. Ct. at 2354–55 (Kavanaugh, J., concurring) (“[D]efendants who successfully challenge their § 924(c) convictions will not merely be resentenced. Rather, their § 924(c) convictions will be thrown out altogether.”). Davis as well as its follow-on, Taylor are discussed supra at notes 49–55 and accompanying text.

[96] Sanford H. Kadish, Stephen J. Schulhofer & Rachel E. Barkow, Criminal Law and Its Processes: Cases and Materials 258 (2017).

[97] 4 William Blackstone, Commentaries *20–21.

[98] Of course, various jurisdictions recognize strict-liability crimes, where criminal liability is assigned without the government needing to show that the defendant had a culpable mental state with respect to one or more elements. See, e.g., Model Penal Code §§ 1.04(1), (5), 2.05(1) (Am. L. Inst.1986); United States v. Dotterweich, 320 U.S. 277, 281 (1943) (permitting strict liability for public-welfare offenses, “dispens[ing] with the conventional requirement for criminal conduct—awareness of some wrongdoing . . . [i]n the interest of the larger good [by] put[ting] the burden of acting at hazard upon a person otherwise innocent but standing in responsible relation to a public danger”); Staples v. United States, 511 U.S. 600 (1994) (holding that either of two conditions may be sufficient to permit strict liability with respect to at least one element: either the legislature’s clear intention to dispense with mens rea, or the non-felonious activity in which the defendant engaged was sufficiently dangerous to put the defendant on notice such that those engaging in that activity are not wholly innocent).

[99] Model Penal Code § 2.02(1) (Am. L. Inst. 1986).

[100] Flores-Figueroa v. United States, 556 U.S. 646, 660 (2009) (Alito, J., concurring); see also id. at 659 (“I write separately because I am concerned that the Court’s opinion may be read by some as adopting an overly rigid rule of statutory construction.”).

[101] Id. at 661.

[102] 575 U.S. 723 (2015).

[103] 18 U.S.C. § 875(c) (2018).

[104] Elonis, 575 U.S. at 726–27.

[105] Id. at 728–29.

[106] Id. (noting that after the court’s grant of a “three-year protection-from-abuse order against Elonis,” Elonis subsequently posted, making threatening reference to the order of protection and how he had “enough explosives to take care of the State Police and the Sheriff’s Department”).

[107] Elonis, 575 U.S. at 732 (summarizing the court of appeals’ holding that defendant can be found guilty if “a reasonable person would view [his words] as a threat”).

[108] See supra note 95 and accompanying text.

[109] Elonis, 575 U.S. at 738 (quoting Staples v. United States, 511 U.S. 600, 606–07 (1994)).

[110] Id. (citing Rogers v. United States, 422 U.S. 35, 47 (1975) (Marshall, J., concurring)).

[111] Id. at 734 (quoting Morissette v. United States, 342 U.S. 246, 250 (1952).

[112] Id. at 740.

[113] Id. at 741. (noting that § 875’s mental state requirement would be “satisfied if the defendant transmits a communication for the purpose of issuing a threat, or with knowledge that the communication will be viewed as a threat” and declining to decide whether “recklessness would [ ] be sufficient” because that issue had not been briefed (emphasis added)).

[114] Id. at 742. By avoiding a holding as to the mens rea required by § 875(c), the Court also avoided the question of whether the First Amendment implications of the statute require a high mens rea level.

[115] Id.at 742 (Alito, J., concurring in part) (emphasis added).

[116] Id.

[117] Id.

[118] Id. at 743.

[119] Id. at 745 (citing Model Penal Code § 2.02(2)(d) (Am. L. Inst. 1986)).

[120] Id. at 745.

[121] Transcript of Oral Argument at 43, Elonis, 575 U.S. 723 (2015) (No. 13-983).

[122] Elonis, 575 U.S. at 742.

[123] 18 U.S.C. § 922(g) (2018). This statute makes it a crime for people with a specified status to possess a firearm. Although the status categories are quite expansive, relevant here are the categories for persons convicted of any felony or being unlawfully present in the United States. See id. § 922(g)(1), (5)(A).

[124] 139 S. Ct. 2191, 2213 (2019) (Alito, J., dissenting)

[125] Id. at 2194 (majority opinion).

[126] Id.

[127] 18 U.S.C. § 922(g) (2018).

[128] Id. § 924(a)(2).

[129] Id. § 922(g).

[130] Rehaif, 139 S. Ct. at 2195 (noting that courts should presume that Congress intends to require mens rea regarding “each of the statutory elements that criminalize otherwise innocent conduct” absent contrary indication (citing United States v. X-Citement Video, Inc. 513 U.S. 64, 72 (1994))).

[131] Id. at 2196.

[132] See 26 U.S.C. § 2255(f) (2018) (providing a “1-year period of limitation” that runs from “the date on which the right asserted was initially recognized by the Supreme Court, if that right has been newly recognized by the Supreme Court and made retroactively applicable to cases on collateral review”); Teague v. Lane, 489 U.S. 288, 307 (1989) (providing that a new rule “should be applied retroactively if it places ‘certain kinds of primary, private individual conduct beyond the power of the criminal law-making authority to proscribe’” (quoting Mackey v. United States, 401 U.S. 667, 692 (1971)).

[133] Rehaif, 139 S. Ct. at 2201 (Alito, J., dissenting).

[134] Id.

[135] Id.

[136] Id.

[137] Id.

[138] Id.

[139] Id.

[140] Id. at 2202.

[141] Id. at 2204.

[142] Id.

[143] Id.

[144] Id.

[145] Id. at 2206 (emphasis omitted).

[146] Id. at 2209.

[147] Id. As Justice Alito noted, the requirement that the government prove other § 922(g) statuses, such as felon status, threatens to undo longstanding precedent in the realm of evidence law, that defendants may offer to stipulate a prior conviction to prevent the prosecution from introducing more prejudicial evidence concerning the nature of their conviction. See, e.g., Old Chief v. United States, 519 U.S. 172 (1997).

[148] Id.; see also 18 U.S.C. § 922(d)(5)(A) (2018).

[149] Id.

[150] Id. at 2210.

[151] See supra notes 47–54 and accompanying text.

[152] See supra note 127 and accompanying text.

[153] Rehaif, 139 S. Ct. at 2212–13 (Alito, J., dissenting).

[154] Id. at 2213 (citing Mathis v. United States, 136 S. Ct. 2243, 2269–70 (2016) (Alito, J., dissenting)).

[155] Id.

[156] Quick Facts: Felon in Possession of a Firearm, U.S. Sent’g Comm’n, https://perma.cc/29F3-E6B3.

[157] 142 S. Ct. 2370, 2384 (Alito, J., concurring) (quoting Rehaif, 139 S. Ct. at 2197).

[158] 21 U.S.C. § 841(a) (2018).

[159] Ruan, 142 S. Ct. at 2375.

[160] The majority’s four reasons included (1) the explicit inclusion of a mens rea term in § 841, (2) the importance of the “[e]xcept as authorized” element in “distinguishing morally blameworthy conduct from socially necessary conduct,” (3) the “serious nature of the crime and its penalties,” and (4) the “vague, highly general language of the regulation.” Id. at 2386 (Alito, J., concurring).

[161] Id. at 2383.

[162] Id.

[163] Id.

[164] Id. at 2384.

[165] Transcript of Oral Argument at 22, Ruan, 142 S. Ct 2370 (No. 20-1410).

[166] Ruan, 142 S. Ct at 2385 (Alito, J., concurring).

[167] Id. (citing 21 U.S.C. § 885(a)(1) (2018)).

[168] Id.

[169] Dixon v. United States, 548 U.S. 1, 13 (2006) (quoting McKelvey v. United States, 260 U.S. 353, 357 (1922)).

[170] Id. at 9 (citing Smith v. United States, 568 U.S. 106, 112 (2013)).

[171] Ruan, 142 S. Ct. at 2384 n.* (Alito, J., concurring).

[172] Id. at 2383.

[173] Id.

[174] Transcript of Oral Argument at 60, Elonis v. United States, 575 U.S. 723 (2015) (No. 13-983).

[175] Transcript of Oral Argument at 59, Rehaif v. United States, 139 S. Ct. 2191 (No. 17-9560).

[176] Transcript of Oral Argument at 78, Ruan, 142 S. Ct. 2370 (No. 20-1410).

[177] Ruan, 142 S. Ct at 2383 (Alito, J., concurring).