by Michael Dorf
The opinion of Chief Justice Roberts in King v. Burwell proceeds concisely and persuasively via the following steps: (1) Read in isolation, the most natural reading of "an Exchange established by the State" as used in Section 36B of the Internal Revenue Code's provision governing the size of tax credits available for the purchase of health insurance would make such credits unavailable for individuals purchasing insurance through the federal exchange in those states that did not create their own exchanges; (2) however, statutory code sections should not be read in isolation, and given other provisions of the Affordable Care Act as well as its overall design, the language is ambiguous; and (3) accordingly, to implement rather than frustrate the manifest purpose of the law, in this provision (though not necessarily in others) "by the State" means "by the State or by the federal government acting in place of the State."
Before discussing the dissent, I'll note two very important features of the majority opinion. First, it does not in any way rely on constitutionally-driven presumptions of the sort I described in the aftermath of the oral argument (e.g., here). It does not say that reading Section 36B as the dissent does would amount to impermissible coercion under either the limits of the Spending Clause or the anti-commandeering doctrine.
Second, CJ Roberts expressly rejects the option of Chevron deference. He says, quite sensibly, that this particular ambiguity in the ACA cannot sensibly be read as a delegation to the IRS to fill in the gap. That seems plainly right and it's also important because it means that the IRS under a Republican president could not simply reverse course and decide that refundable tax credits for the purchase of health insurance are not available on federal exchanges (although, of course, a Republican president with a Republican Congress could amend or repeal the ACA).
As for Justice Scalia's dissent (joined by Justices Thomas and Alito), it seems to me either a rejection of his textualist philosophy or, if taken as exemplary of that philosophy, an indictment thereof. In his academic writings and in many of his opinions, Justice Scalia has been at pains to distinguish what he calls textualism from literalism or strict constructionism. For example, in the lead essay in A Matter of Interpretation, Justice Scalia writes (at p. 23) that "[t]extualism should not be confused with so-called strict constructionism, a degraded form of textualism that brings the whole philosophy into disrepute. I am not a strict constructionist, and no one ought to be . . . ." He says on the next page that "the good textualist is not a literalist . . . ."
Justice Scalia's King dissent pays lip service to the distinction between literalism and textualism. For example, he says that he "agree[s] with the Court that sound interpretation requires paying attention to the whole law, not homing in on isolated words or even isolated sections. Context always matters." And yet, throughout the dissent Justice Scalia gives no weight at all to the context that matters most: what Congress was trying to accomplish through the central provisions of the ACA and with the backup provision by which the federal government creates exchanges when states don't. To use an elegant turn of phrase by Justice Scalia from a case that CJ Roberts relies on in King, the King dissent finds that Congress has hidden an elephant in a mousehole.
The King dissent reads like literalism in one obvious sense. Justice Scalia repeatedly exclaims that "State" can only possibly mean "State but not federal" rather than "State or federal," festooning his increasingly adamant insistence with colorful adjectives for the majority's position ("absurd"; "impossible"; "shame[less]"; etc). But as in the four-Justice dissent in NFIB v. Sebelius, where the same lineup plus Justice Kennedy repeatedly insisted that a payment Congress labeled a "penalty" could not possibly be a tax for constitutional purposes even though CJ Roberts patiently explained why a provision that functions as a tax is a tax regardless of labels, so too in King, Justice Scalia's adamance that "State" cannot possibly mean "State or federal" leads anyone who does not share his commitment to the destruction of the ACA to ask "why not?".
What about context? Justice Scalia makes two moves to render his disregard of the broad context irrelevant. First, he thinks the meaning of "State" so darn obvious that there is no need to look at context--even though most of the judges to have examined the question think that it's at least unclear. Second, when he looks at context arguendo he looks mostly at linguistic context, describing how "established by the State" in other parts of the statute does appear to draw a distinction between states and the federal government. CJ Roberts responds that a single phrase can mean different things in different places in the same statute but Justice Scalia basically dismisses the point.
Only at one point in his dissent does Justice Scalia address the central question that has dogged the plaintiffs ever since they dreamed up this improbable case: Why would Congress have possibly meant to exclude subsidies on federal Exchanges? Justice Scalia's answer tracks the answer that the concocters of the lawsuit concocted: It's an incentive scheme to induce the states to create their own exchanges.
As CJ Roberts shows (and as numerous others have shown), that reading of this tiny elephant trapped in her mousehole makes a hash of the statute as a whole, but there is another problem with it, and one which even non-degraded textualism exacerbates: Before the ACA's enemies settled on Section 36B as the latest means of attempting to kill the law, nobody breathed a whisper of this supposed incentive mechanism. But that's not a problem for the dissenters, because the complete post-hoc-ness of the challenge can be dismissed as relying on legislative history--and even undegraded textualism abjures legislative history.
The core of the textualist argument against legislative history is that committee reports, floor statements, and the like are not reliable indicators of legislative intent. If one agrees, that's a fair reason to discount evidence in such materials, but it's not a reason to discount the absence of evidence for some position that has a very dramatic effect on the meaning of a law. Especially now that members of Congress and their staff go to great length to stuff support for whatever position they favor into the legislative history, if anybody in Congress thought that the ACA barred subsidies on federal exchanges and did so for the purpose of incentivizing the states to create their own, then surely one would expect to find something somewhere in the pre-enactment record to that effect. And yet there's nothing. With respect to legislative history of this sort, absence of evidence is evidence of absence.
To be sure, Justice Scalia believes he has another reason to disregard legislative history, even if reliable. Such history aims at uncovering legislative intent, but textualists say that unenacted intent is not law, and so courts should only look to construe the words that were enacted. Fair enough, but suppose that, in context, the meaning of the words is not pellucid. At that point, even Justice Scalia advocates construing texts "reasonably." And so the question arises: as between a construction of the statute that enables federal exchanges to function as backups in states that don't create their own (the majority approach) and a construction in which the ACA's other provisions will likely lead to the destruction of the individual health insurance market in states that don't create exchanges, which is more reasonable?
Justice Brennan once described originalism as "arrogance cloaked as humility." Textualism--at least as practiced by the King dissenters--is vulnerable to the same charge. It is supposed to constrain judges to implement the will of the legislature rather than their own. Textualism fails in that purpose, at least if one can regard King as an example of textualism. Justice Scalia accuses CJ Roberts of adopting special rules of interpretation in order to salvage the ACA. A psychologist might say that he is projecting in order to cover his adoption of special rules in an attempt to kill it.