Sentencing, Vagueness, and Facial Invalidation in Johnson v. United States

In a rare decision that will earn plaudits from both the defense bar and many government attorneys, Johnson v. United States held that the “residual clause” of the Armed Career Criminal Act is unconstitutionally vague. Enjoying a kind of personal vindication, Justice Scalia wrote for the six-justice majority and so got to turn several of his prior dissenting opinions into the holding of the Court. By contrast, Justices Kennedy and Thomas would have found simply that the statute didn’t apply to this defendant. Finally, Justice Alito dissented on grounds partly endorsed by Kennedy and Thomas.

I’ve covered Johnson before. In short, I’ve basically argued that the Court’s repeated interactions with the residual clause are what rendered it vague. That explains why, for nearly 30 years, the allegedly vague residual clause has been able to function on such a massive scale, including during numerous trips to the Court. Only recently has there been any serious interest in finding the residual clause to be vague, for only after the clause had generated repeated judicial opinions did that vagueness come to exist.

So that means that I tend to agree with important features of both the majority and the dissent in Johnson: the majority is right that the residual clause is vague today, but the dissent is right that the vagueness is the judiciary’s own doing. This raises the question: what to do about it?

The majority argues that the entire residual clause must be declared unconstitutional. By contrast, Alito proposes two ways out of the bind. Though six justices disagreed with this aspect of Alito’s opinion, both of Alito’s proposals are thoughtful and reward serious consideration.

Alito’s first proposal is to revisit the Court’s whole approach to the residual clause. Instead of asking whether a predicate offense poses a substantial risk of violence in the typical case, Alito proposes asking whether each offender’s actual conduct in fact posed such a risk. But that approach would require overruling or narrowing a decision from 1990 that the Johnson majority says was correct as a matter of statutory interpretation. Perhaps more significantly, it would be a bold move for the judiciary to reboot this area of law after so plainly failing to manage it in recent years. While doing so might arguably be legally correct in some sense, Alito may not give his colleagues quite enough credit in criticizing them for being “tired” of handling these cases.

Alito’s alternative proposal is even more interesting. Quoting strongly supportive case law, Alito argues that facial invalidation is inappropriate in due-process vagueness cases unless a statute is vague in all its applications. And, Alito continues, the residual clause isn’t vague in all of its applications, since some crimes—like attempted rape—are surely violent in the normal case. The majority responds that the precedent Alito quotes, while clearly supportive of him, is dicta. Further, even paradigmatically vague laws that have been facially invalidated—like proscriptions on “annoying” public behavior—aren’t vague in every single one of their applications. After all, everyone knows that spitting in someone’s face is annoying. The Court accordingly narrows its own precedent on facial invalidity by reading the relevant statement as a mere “tautology.”

This reasoning about facial invalidation may be the most doctrinally far-reaching aspect of Johnson’s holding. Not only does it arguably change the baseline rule for remedying an important category of due process claims, but the Court’s reasoning also supports making facial invalidation more broadly available in general. As Alito points out, his preferred vagueness rule “is simply an application of the broader [Salerno] rule that, except in First Amendment cases, we will hold that a statute is facially unconstitutional only if ‘no set of circumstances exists under which the Act would be invalid.’” Alito suggests (or hopes) that the Court’s refusal to follow Salerno’s logic is confineable to vagueness, but it may not be.

In this respect, Johnson is a companion to this term’s Fourth Amendment case Los Angeles v. Patel, which also allowed a facial claim. In his Patel dissent, Scalia noted almost in passing that the claim there “must fail because the law is constitutional in most, if not all, of its applications. See United States v. Salerno, 481 U. S. 739, 751 (1987).” In retrospect, Scalia’s choice of language in Patel may be significant. As Alito rightly points out in Johnson, Salerno said that facial invalidation is generally appropriate only if a law is valid in “no set of circumstances,” whereas Scalia's Patel dissent seems to allow for facial invalidation when a law is invalid in “most, if not all” situations. So perhaps Johnson was meant to be a blow against Salerno after all.

Notably, Scalia’s opinion for the Court presents a certain irony. On the same day that Johnson came down, Scalia dissented in the same-sex marriage case Obergefell v. Hodges on the ground that the Court lacked legitimate authority to override the political system based on due process and about fifteen years' worth of experience with same-sex marriage. Yet in Johnson, Scalia led the charge in invalidating a federal statute based on due process, while also overcoming stare decisis based on a scant “[n]ine years’ experience.” Obviously, these cases are different in many ways. But still, a little ironic.

Courts now have to grapple with the question of remedy: who will get out of prison because of this decision? For thoughtful discussion of the practical fallout, see this short piece by Leah Litman, as well as posts over at SentencingBlog. And, of course, Johnson also raises important questions for legislators, since Congress might try to fashion a less vague replacement statute and thereby restore the punitive policy that the Court has now undone.