This is part three of a three-part series on Judge Gorsuch.

Last summer when I was taking the New York bar, I ran into an acquaintance who was an incoming clerk for one of President Trump’s Supreme Court short-listers. At some point she asked me what a judge I knew was “like,” and I replied that he had a very studious and sincere but slightly naïve obsession with “getting every case right.” Confused by what could possibly be wrong with that, or perhaps unable to imagine that I’d hinted there was anything wrong with it, she said it sounded as if he were a very good conservative judge—to which I replied that he was, in the main, a very good conservative judge, but that that sort of thing could be a little trying if you were someone who happened to believe that some cases can’t be gotten “right,” at least as far as a single right answer is concerned. A little aghast, she dismissively waved her hand and said, “Oh, well if you’re one of those people — !”

I am one of those people, and it always disappoints me that so many of my conservative peers, even the most sophisticated ones, are not. There was once a time not so long ago when one didn’t have to be Mark Tushnet to believe in the existence of meaningful quantities of legal indeterminacy. Who wrote the following passages?

I am not so naïve (nor do I think our forebears were) as to be unaware that judges in a real sense ’make’ law. But they make it as judges make it, which is to say as though they were ‘finding’ it . . . .

In the vast majority of [ambiguous statutory] cases I expect that Congress neither (1) intended a single result, nor (2) meant to confer discretion upon the agency, but rather (3) didn’t think about the matter at all.

[Commenting on a hypothetical statute intended to illustrate a so-called pure question of law, providing that “the draft board shall conscript an army of 1,000,000 persons between the ages of 21 and 25”]: How does one know, to use the example you gave, that Congress did not intend to leave it to the agency to determine whether 21-year-olds should be drafted?

[Writing of the Benzene Case]: In giving content to a law which in fact says no more than that OSHA should ensure “safe places of employment” (whatever that means) and should maximize protection against toxic materials “to the extent feasible” (whatever that means), it was . . . the Court, rather than OSHA, that ended up doing legislator’s work.

[And responding to a call to narrow Chevron just four years ago]: The effect would be to transfer any number of interpretive decisions—archetypal Chevron questions, about how best to construe an ambiguous term in light of competing policy interests—from the agencies that administer the statutes to federal courts. We have cautioned that ‘judges ought to refrain from substituting their own interstitial lawmaking’ [emphasis mine] for that of an agency. That is precisely what Chevron prevents.

The answer, of course, is Judge Gorsuch’s would-be predecessor Justice Scalia.

Turning to the rather less realist work of Justice Scalia’s would-be successor, Judge Gorsuch’s concurring opinion in Gutierrez-Brizuela calling for an overturning of Chevron (or as we’re now told, merely “identifying an issue for my bosses . . . like any good employee”) has been hailed as one of the most prescient, comprehensive dissections of the doctrine” that some serious people have “ever read,” and at the very least an interesting, thoughtful piece of work. In truth, the opinion is the most half-baked critique of Chevron one is likely to encounter outside of a law school classroom.

Judge Gorsuch has three main arguments against Chevron. First, that it prevents courts from fulfilling their constitutional and statutory duty, under the APA, to say what the law is. Second, that it makes it considerably more difficult for “the people” to discern what the law is. Third, that the amount of interpretive authority Chevron vests in agencies violates the non-delegation doctrine. The first two arguments, it seems to me, rest on a denial of legal indeterminacy, while the third strangely assumes that, to the extent legal indeterminacy does exist, otherwise unconstitutional delegations of legislative authority are perfectly permissible so long as the courts are the delegates.

Beginning with the first argument, I was struck the other day by Judge Gorsuch’s comparison of Chevron to Lucy snatching the football away from Charlie Brown. Judge Gorsuch was comparing Mr. Gutierrez-Brizuela to Charlie Brown, the theory being that Gutierrez-Brizuela relied on a precedent (which the BIA had rejected before he relied) that Chevron, like Lucy, threatened to pull out from under his feet. But it strikes me that the real Charlie Brown in Judge Gorsuch’s mind is Judge Gorsuch. In a poignant or petulant passage, depending on your sensibilities, Judge Gorsuch complains that just when he thinks he’ll get to say what the law is, Chevron robs him of that prerogative:

[R]ather than completing the task expressly assigned to us, rather than “interpret[ing] … statutory provisions,” declaring what the law is, and overturning inconsistent agency action, Chevron step two tells us we must allow an executive agency to resolve the meaning of any ambiguous statutory provision. . . . Of course, some role remains for judges even under Chevron. At Chevron step one, judges decide whether the statute is “ambiguous,” and at step two they decide whether the agency’s view is “reasonable.” But where in all this does a court interpret the law and say what it is?

Judge Gorsuch’s anxiety to perform his duties is commendable, but completely misunderstands Chevron doctrine. Chevron says that if Congress has “spoken to the precise question at issue,” courts “must give effect” to the law Congress made on that question. But, if “the court determines Congress has not directly addressed the precise question at issue”—if the law is indeterminate on that question—courts cannot say anything further about what the law is, because within the range of interpretations the statute can bear (which the court must delimit), there is no definite law. Having, as the APA commands, “decide[d] all relevant questions of law” and determined that what the agency is doing is legal, a court is left to review the agency’s policy choice as between the options the statute permits for arbitrariness and caprice, as the APA also commands. That’s the doctrine, anyway, and whinging about how it doesn’t let courts say what the law really is misses its point completely (though a more thoughtful critique of Chevron’s conflation of ambiguity with indeterminacy wouldn’t).

Judge Gorsuch seems to both understand and not understand this. After complaining that Chevron prevents him from saying what the law is, he acknowledges that “some,” including Chevron itself, say that Chevron is “about letting agencies fill legislative voids.” Quite so, which should solve his problem; courts can’t say what the law is in a void. But Judge Gorsuch refuses to take ¯_(ツ)_/¯ for an answer. Voids and delegation, he says, don’t “rescue[] us from our riddle,” because “whatever the agency may be doing under Chevron, the problem remains that courts are not fulfilling their duty to interpret the law and declare invalid agency actions inconsistent with those interpretations.” This seems obtuse. How can courts interpret a void? Or find that agency actions are inconsistent with one?

The trouble, apparently, is that Judge Gorsuch just doesn’t believe there are any legislative voids. Judge Gorsuch has told us this elsewhere, in a lecture on Scalia that he pointedly cites in Gutierrez-Brizuela. He begins his discussion of indeterminacy there with the disclaimer that he did “not mean to suggest that traditional legal tools will yield a single definitive right answer in every case,” which is the sort of thing people say when they’re about to deny the existence of legal indeterminacy. However, he says, “the amount of indeterminacy in the law is often (wildly) exaggerated.” In an unusually strained example of the argument from judicial consensus, he notes that 95% of cases in the courts of appeals are decided unanimously, and only 0.014% of cases filed in the federal courts end with a split Supreme Court decision. And “even accepting some hard cases remain—maybe something like that 0.014%,” he says “there’s no reason why we cannot make our best judgment depending on (and only on) conventional legal materials,” and “[n]o reason, too, why we cannot conclude . . . that one side has the better of it.”

Now, attempts to quantify indeterminacy in this way are wildly over-simplified. In suggesting that overwhelming legal determinacy explains the data, they ignore, inter alia, panel effects, herding, path dependence, the severe limits on the Supreme Court’s certiorari jurisdiction, and of course, the uniformity that Chevron (like other ambiguity-triggered canons wrongly criticized by Gorsuch’s peers) has created by requiring lower courts to defer to a single agency interpretation where there is indeterminacy. But the quantity of hard cases is somewhat besides the point. What matters for purposes of Chevron, Chevron being a doctrine of hard cases, is that Judge Gorsuch’s solution for whatever hard cases do exist is to act as if they’re cases of ultimately determinate law too. When courts encounter one, he says they should keep pottering away at their conventional legal materials until they arrive at an answer. This assumes a peculiar definition of a “hard case”: one where it’s hard to find what the law is, but there is some definite law to be found. (Compare it to the one offered by H.L.A. Hart, a legal philosopher heavily drawn on by originalists and textualists: a case where “on some point no decision either way is dictated by the law and the law is accordingly partly indeterminate or incomplete.”) The possibility that there may just be no answer—outside of the interpretations Congress instructed agencies to render on these questions—to whether, e.g., a “stationary source” is a plant or a single smokestack, or whether the statutory rape of a 17-year-old by a 21-year-old is “sexual abuse of a minor,” seems utterly beyond Judge Gorsuch’s comprehension.

Because Judge Gorsuch thinks that conventional legal materials have all the answers, he also believes that ordinary people will have a much easier time guessing what a court is likely to make of an ambiguous statute than gauging their chances under an agency interpretation already on the books. Without Chevron, he says, the people would merely have “the duty to conform their conduct to the fairest reading of the law that a detached magistrate can muster”; with it, they’re “required to guess whether [a] statute will be declared ‘ambiguous’ . . . and required to guess (again) whether an agency’s interpretation will be deemed ‘reasonable.’” Doesn’t that sound much harder?

It’s to Judge Gorsuch’s credit as a writer that he makes it sound harder, but to his distinct discredit as a legal thinker that he thinks it is. Consider the following two real scenarios. In the system Gorsuch wants, an undocumented immigrant who’s illegally reentered the country has to decide which of two complex and, according to Judge Gorsuch, conflicting statutes controls his entitlement to adjust his status. If he guesses right, he may get to become a lawful permanent resident; if he guesses wrong, he’s ineligible, has admitted his illegal reentry to the authorities by applying for adjustment, and will be deported. In the other legal system, the one we’ve actually got, the immigrant reads a guidance document that says which statute controls and knows it will control his fate unless the guidance document is unambiguously wrong—which it almost certainly isn’t given the inherent ambiguity posed by two conflicting statutes. Those are the stylized facts of Gutierrez-Brizuela itself.

Or take another scenario from a case the Court will decide this term, Esquivel-Quintana. In Gorsuch’s system, a 21-year-old non-citizen in college has to guess whether a court is likely to think her relationship with her high-school-senior boyfriend is “sexual abuse of a minor,” an undefined term. If she guesses wrong and is caught, she gets deported. In the system we’ve got, she reads a guidance document that says in the clearest possible terms that her relationship is sexual abuse of a minor and has to decide whether she’s so confident the agency’s decision is unambiguously wrong or unreasonable that she’ll run the risk of deportation. I know which system I would rather have if I were an undocumented immigrant, and I’m a lawyer with the ability (and Westlaw subscription) to conduct multi-jurisdictional surveys, study the doctrine of immigration lenity, and engage with all the rest of the complicated interpretive apparatus that would bear on a court’s de novo interpretation of the question.

Judge Gorsuch, however, thinks it would be easier for the undocumented immigrants in my examples to play at being statutory interpreters than to predict their chances at beating the government under Chevron. His thinking so rests on a wildly exaggerated notion of how easy it is for non-lawyers to conform their conduct “to the fairest reading of the law that a detached magistrate can muster,” which in turn rests on a denial of statutory opacity. Even if Judge Gorsuch were right about the unbounded determinacy of the law, one would think he would acknowledge a truism that every lawyer knows: that it’s much easier to predict the outcome of a case with a deferential standard of review than a case without one. After all, even Judge Gorsuch admits that hard cases, if ultimately determinate, are hard. Chevron makes them relatively easy.

Finally, there’s delegation. Switching tack to an argument that at least acknowledges Chevron’s premise—that statutory ambiguities contain no hidden determinate law, but rather delegate interstitial lawmaking authority to the executive—Judge Gorsuch suggests the delegations Chevron presupposes are unconstitutional under even the current instantiation of the nondelegation doctrine, to which no statute has fallen victim in eighty years. He cutely reasons that if the Schechter Court could, in 1935, invalidate a statute that gave an agency the power to write a code of competition for the poultry industry without any guidance on how to do it, then Chevron, “a rule that invests agencies with pretty unfettered power to regulate a lot more than chicken,” is likely unconstitutional too.

I think Judge Gorsuch is owed the respect of assuming that he doesn’t take this argument seriously and really wants to expand the nondelegation doctrine beyond its current bounds, because this is not a remotely serious argument. As Justice Scalia recounted in Whitman, since Schechter the Court has upheld unvarnished authorizations to regulate in the public interest, or to fix prices at a “generally fair and equitable” level that would “effectuate the purposes” of a statute. The delegations Chevron presupposes (a) are within the much smaller confines of a bounded ambiguity—that is to say, some agency interpretations are off the table at Step One, (b) are subject to arbitrary and capricious review, and (c) are informed by and channeled through the intelligible principles and purposes that Congress has spelled out or can be discerned from a statutory scheme, as well as trans-substantive norms, like cost. Arguing, as Judge Gorsuch does, that it violates the current version of the nondelegation doctrine for Congress to permit the EPA to choose between a limited range of permissible interpretations of “stationary source” by weighing Congress’s concern for “allow[ing] . . . reasonable economic growth” over a certain theory of pollution control, would be very-near sanctionable if a litigant tried it in federal court.

But even if Judge Gorsuch were right about nondelegation, what follows? The remedy for an unconstitutional delegation is to invalidate the delegation and suggest that Congress write a clearer law. Judge Gorsuch seems to think an equally good remedy is to pretend the delegation isn’t there. Holding that courts get to decide the meaning of an indeterminate term like “stationary source” or “sexual abuse of a minor” wouldn’t make the delegations embedded in those indeterminacies go away; it would just re-route the delegations to courts, a result that (besides not being what Congress wanted or making any practical sense) isn’t any more constitutional than letting agencies keep them. Courts, to be sure, would claim they were finding what “sexual abuse of a minor” and the like “really” meant, but who besides Judge Gorsuch would believe them?

I am reminded here of a passage I already quoted, from Justice Scalia’s commentary on the Benzene Case. In the Benzene Case, the Court avoided a nondelegation problem by giving definite content to an extremely vague statute. Justice Scalia, perhaps surprisingly, disapproved, writing that the Court should have struck the statute down or let the agency have its way:

In giving content to a law which in fact says no more than that OSHA should ensure ‘safe places of employment’ (whatever that means) and should maximize protection against toxic materials “to the extent feasible” (whatever that means), it was the plurality of the Court, rather than OSHA, that ended up doing legislator’s work.

Justice Scalia was realist enough to know and honest enough to admit that absent the enforcement of a nondelegation doctrine that he came to recognize was unenforceable, denying deference to agencies would result in courts doing legislator’s work. I wish his likely successor were too.

 Asher Steinberg writes The Narrowest Grounds, a law blog. He can be reached at asher.steinberg@outlook.com.