« What about Calpurnia, Tom, and Scout? | Main | What is a heckler's veto? »

Tuesday, April 17, 2018

Is Gorsuch’s Dimaya concurrence the opening shot in a bid to revive the Non-Delegation Doctrine?

Eugene Volokh has noted that Justice Gorsuch’s concurrence in Sessions v. Dimaya looks like a “cross-over sensation,” because Gorsuch joined four liberal justices in ruling against the deportation of an immigrant. There is, however, a deeper signal buried in Gorsuch’s concurrence: By refusing to draw any distinction between civil and criminal cases insofar as VFV is concerned, Gorsuch seems to be reviving the non-delegation doctrine as a basis for striking down statutes. If the VFV doctrine applies, as Gorsuch says it does, “in civil cases affecting a person’s life, liberty, or property” and “criminal cases involving relatively modest penalties,” then it might also apply to any statute containing terms ambiguous enough to trigger Chevron.

Such an expansive version of VFV would fit with Gorsuch’s famously skeptical view of Chevron in his Gutierrez-Brizuela concurrence. So maybe the concurrence is best read as an opening bid to revise Chevron and/or the non-delegation doctrine, fulfilling his implied promise, in Eric Posner’s words, to “join Justice Thomas as one of only two justices to seriously oppose the administrative state in the last 50 years, at least.”

Just for the record, I do not think very much of Gorsuch’s attack on Chevron. It seems to rest on one of two equally odd views that (1) statutes should not have gaps to fill, or (2) only judges should fill them (i.e., “say what the law is”). (For an excellent analysis that spells out the weaknesses of such a position, see Asher Steinberg’s excellent post on the Yale JReg blog). As for the non-delegation doctrine, if constitutional doctrines could be void for being excessively vague, then the NDD would qualify — as, indeed, would the VFV doctrine itself. I do not see five votes for reviving a doctrine that paradoxically gives judges uncabined discretion to forbid agencies from exercising uncabined discretion.

Since the SCOTUS granted cert in Gundy, however, we will find out soon enough whether Gorsuch can persuade any of his colleagues to join him on his quixotic mission to expand dramatically the courts’ exclusive role in saying what the law is. If Dimaya is any indication, then Gorsuch might be trudging down a lonely path: Although Justice Thomas is supposed to be skeptical about the administrative state, Gorsuch could not get Thomas’s vote for the proposition that the Due Process clause requires civil statutes to be clear enough to give citizens notice of their contents.

Posted by Rick Hills on April 17, 2018 at 11:13 PM | Permalink


The short answer to the question above? Yes. Certainly, those skeptical of the administrative state were celebrating the opinion that way (see, for example, Prof. Josh Blackman's Twitter thread on the opinion yesterday evening).

Posted by: Robert Morse | Apr 18, 2018 10:14:38 AM

Josh beats everyone to the punch.

Posted by: Rick Hills | Apr 18, 2018 10:31:11 AM

Josh beats everyone to the punch.

Posted by: Rick Hills | Apr 18, 2018 10:31:13 AM


It seems to me that the separation-of-powers argument in his opinion at pages 8-9 applies equally to antitrust as it does to criminal statutes or civil statutes containing civil penalties. As far as I can tell, the argument is that Congress has all legislative powers, so the courts can't create new rules governing future conduct. That argument, which non-delegation devotees rarely make (though a few have), would seem to take out a lot of statutes that are understood to cede quasi-common-law lawmaking power to the courts, like the Sherman Act, or the Alien Torts Statute, or the Voting Rights Act, the key provisions of which are terribly thin. For that matter, it's hard to read the recent opinions about Title VII and sexual-orientation discrimination and not see Title VII as a massive delegation of quasi-common-lawmaking power.

As a matter of originalism, which his argument supposedly sounds in, I find it very hard to square with what very little I know of early federal legislation; when one looks at, say, the Crimes Act of 1790, one doesn't find many crimes defined in terms of elements. Rather, it picks up a number of common-law crimes that the new federal courts were left to elaborate on. Of course one might say that the Crimes Act of 1790 was meant to pick up precisely whatever the common law of murder of 1790 was, without change as the common law of murder developed, but is that what a federal court in 1850 would have done with the Crimes Act, were it still the law, as parts were -- examine 1790 common law, disregarding any recent developments? And would such an approach be constitutionally mandated? If not, why is it a separation of powers violation for federal courts to create a common law of substantial risk of physical force? And what of the Sedition Act, which left federal courts to expound on its vague terms, terms that were less known to the common law? Did it violate the non-delegation doctrine as well as the First Amendment (if it even did that)?

Of course, even if one were theoretically open to a judicial annex to the non-delegation doctrine, it would be all but impossible in practice to say which delegations to the judiciary go too far, and I doubt that the one in Dimaya is too close to the extreme end of delegation; there's a great deal of intelligible principle to work with there, even if decisions under those principles may be hard to make and harder to predict. I'm fine with saying the statute is void for vagueness as a matter of fair notice, although I suspect that there is a heartland of relatively obvious applications in this case and in the cases of most vague penal statutes and that the best approach is to split the difference between Gorsuch and Thomas and use lenity as an as-applied void-for-vagueness doctrine.

Posted by: Asher Steinberg | Apr 18, 2018 2:24:37 PM

Oh, and I should add that Gorsuch's view that courts are unconstitutional delegates under a robust nondelegation doctrine is in serious tension with his interest in overruling Chevron on nondelegation grounds, as making courts the interpreters of seriously vague provisions in agencies' organic statutes is only an improvement as far as nondelegation goes if you assume that it's constitutional to delegate to courts but not to agencies. It seems to me that a more coherent version of his position would be (a) an extremely stringent Step One that resolves most cases at the level of unambiguity, or a determination that, as Chevron actually described Step One, Congress spoke to the question at issue (however ambiguously) and didn't leave any gap to fill, and (b) invalidating what remains as a delegation to whomever one thinks Congress meant to delegate discretion to -- courts or agencies, it doesn't matter.

Posted by: Asher Steinberg | Apr 18, 2018 5:56:39 PM

Asher, I infer that originalists might subject both the President and Article III courts to a relaxed non-delegation standard, at least to the extent that (1) the President is exercising prerogative powers that the originalists believe have already been vested in him or her by Article II or (2) the courts are engaged in common-law “legislating” that was common practice in 1789 and implicitly incorporated into the “judicial power” by Article III. So perhaps the Sherman Act is safe?

Of course, as these doctrinal epicycles pile up, the question of what sorts of judicial policy-making are “analogous” to traditional 18th century common-law policy-making and what sorts of powers are truly prerogatives preserved by Article II’s vesting clause becomes more and more vexing. But if one is committed to this enterprise, I guess such a deep dive into historical metaphysics is not very troubling.

Posted by: Rick Hills | Apr 18, 2018 7:12:37 PM

So, RIck, how do you think Gundy (i) should and (ii) will be decided?

Posted by: Marty Lederman | Apr 19, 2018 4:37:26 AM

Post a comment