City of Arlington v. Federal Communications Commission, a big and hugely interesting AdLaw case, was decided last week. I still haven’t fully digested it. (I’ve been busy grading exams and assuring my disconsolate econ grad daughter, of whom I am immensely fond and proud, that there is life after Boston College.) Still: while a conservative-libertarian meme-in-the-making says that Arlington marks yet another triumph for a headless and heedless Fourth Branch, my preliminary noodling has led me to suspect that this may actually be wrong. There’s far more good news here than bad.

The case, competently summarized by others so I don’t have to do it (again), asked whether administrative agencies get so-called Chevron deference when they construe the scope of their own jurisdiction. The majority opinion, written by Justice Scalia, says “yes” (as does a lone concurrence by Justice Breyer), so long as the statute is ambiguous. All we’re asking is whether the agency had the authority to do what it did, as we do in countless AdLaw case; there is no separate “jurisdiction” category. Chief Justice Roberts, joined by Justices Kennedy and Alito, says “no.” For matters within an agency’s jurisdiction, Chevron grants judicial deference when a statute is ambiguous and withholds it when it’s not; but the question of whether or not the applicable statute granted the agency jurisdiction over the question in the first place must be for the courts, de novo. Otherwise, you have a serious “fox in the henhouse” problem, and you may have a serious Marbury problem. (It’s the judiciary’s job, not someone else’s, to “say what the law is.”)

Both opinions are very good, and very forcefully argued. Note that both were written by conservative justices (and note that Justice Thomas was in the majority). This is, in a manner of speaking, amongst ourselves; the liberals reverted to their default position (“anything goes”) and were just along for the ride.  I’m not at all sure, though, which conservative side I’m on.

The Chief’s dissent reads over long stretches like something from the Cato Institute: The administrative state is out of control. Congress is creating agencies left and right. No one is minding the store. This has to stop, and it’s our business to stop it. The Chief not being one to simply vent, his opinion is intended as a marker: enough is enough. This is unambiguously good news, and it marks a sea change in conservative AdLaw sensibilities. Go read conservative opinions from the late 1970s or the 1980s (starting perhaps with Vermont Yankee (1978)): they’re all about judicial “deference,” the better to let expert agencies and the unitary executive do their job. This opinion says the opposite. And while I don’t mean to over-read it, it has the ceterum censeo flavor of the dissents written by the Chief’s one-time boss, then-Justice William Rehnquist, in early federalism cases: one of these days, we’ll return to constitutional norms. (A bit more on this below.)

So what’s there not to like, and whence the intra-conservative disagreement? Several things.

To start where Justice Scalia starts, “jurisdiction” is a messy category even with respect to courts, let alone administrative agencies. For example (mine, not the Court’s), when the Corps of Engineers claims “jurisdiction” over some wetland, is that a purely legal question (as the term “jurisdiction” suggests)? A question of fact? Of policy? Maybe you don’t want appellate courts to sort through hundreds of statutory provisions, flipping an on-off switch called “jurisdiction.”

On a more substantive note, the intuition that drives the “no deference on jurisdiction” train of thought is a quasi-delegation theory: it’s okay to give agencies discretion on smallish “fill-in-the-gaps” questions but not on big “what’s-our-reach” questions, where judicial deference spells bureaucratic empire-building. I sympathize, but I don’t think “jurisdiction” is a good way to get at the problem. The real-world issue in City of Arlington was pretty trivial, and the agency’s action—not even a rulemaking, for what that’s worth—was no power grab. (It had to do with the local approval process for cell phone towers.) Contrast that with the unquestioned authority of OSHA or EPA to set exposure standards or technology requirements: the agencies can make or break entire industries, and all you get is arbitrary and capricious review. So far as I know, there’s no competent evidence to suggest that agencies build empires exclusively or even primarily on “jurisdictional” margins. In short, “jurisdiction” doesn’t capture the real problem.

To be sure, an agency-constraining doctrine may well be worth having even if it doesn’t do all you might want it to do. It’s not at all clear, though, that a “no deference on jurisdiction” doctrine will have any constraining effect. Suppose the question is whether greenhouse gases are “pollutants” for purposes of the Clean Air Act, and suppose EPA says, “not on our reading of the statute.” Ordinarily, the agency would get deference. But on a “no deference on jurisdiction” theory, a federal court could come along and say, “On our de novo review, GHGs are pollutants. Go regulate: you don’t have discretion not to.”

Pow! Right in the kisser: that’s Justice Stevens’ majority opinion in Massachusetts v. EPA (2007) (about which more in coming posts). Nor was Mass v. EPA Justice Stevens’ first such stunt: in Babbitt v. Sweet Home Chapter (1995), he decided that the Endangered Species Act’s prohibition against the “taking” of an endangered species covers “habitat modification,” as when your home addition impedes a toad’s amorous desires. (The opinion burbles about Chevron deference but conspicuously fails to apply it.) Of course, John Paul Stevens is retired. But David Tatel on the D.C. Circuit is not. He is equally smart and clever, and he decides more AdLaw cases in a month than the Supremes will see in a decade. You don’t want to arm him.

There are merits dissents in both Mass v. EPA and in Sweet Home Chapter, and both are written by (drumroll!) Antonin Scalia. It’s simply wrong to paint him as a bureaucrats-have-at-it judge, and I don’t think his City of Arlington opinion reflects any such disposition. The way he thinks about Chevron goes like this: mere linguistic ambiguity can be found just about anywhere; so that alone can’t give you deference. Reasonable ambiguity is what remains after we’ve read the statute and applied ordinary canons of construction. Those canons tell me that a construction of the Clean Air Act that turns Frisbees and flatulence into “air pollutants” is crazy, and they tell me that an ESA construction of “take” that conscripts vast tracts of private land to national zoological use is likewise insane. (The formulations, sans “crazy” and “insane,” are too good to be mine; they appear in the opinions.) The more you follow this line of thinking, the more constraining it becomes for agencies. But it’s the statutory interpretation that does the work, not AdLaw doctrine.

And, speaking of “doctrine”: for all of Chevron’s flaws and discontents, we know how it shakes out. We don’t know how the Chief’s approach would shake out over the general run over appellate cases. That is not a trivial point, and it brings us back to the “Rehnquist federalism” parallel—actually, two parallels. First: in the federalism arena as in administrative law, everything that matters is in the weeds. All the grandiose, constitutional federalism theories (Congress can’t regulate local commerce! Congress can’t grant money for disaster relief in Oklahoma unless the tornado came from Texas!) failed; all the subterranean stuff on sovereign immunity, private rights of action, clear statement rules, and “commandeering” has stuck and, to my mind, made a salutary difference. AdLaw is also a game of inches, and perhaps more so. And, second: if the so-called “federalism revolution” fizzled, that’s because the Court never had a coherent theory. Individual justices had very different notions of the construct, and the result wasn’t pretty—an “Upside-Down Constitution,” to coin a phrase.

We don’t have a coherent theory of “the administrative state,” either. Until we do, small doctrinal steps, taken in the right spirit, may well be the best course of action.