Last Christmas, I spent a somewhat panicky inter-semester break writing an amicus brief for King v. Burwell. I was worried that five Supreme Court justices were going to be too tempted by the plaintiffs’ legalistic interpretation of Obamacare’s text, despite ample evidence beyond the text that Congress never intended to deprive citizens in 34 states of health insurance subsidies.
In a seminar I taught at Boston University, one of my students had proposed a legalistic version of the common sense point that Congress could not possibly have intended the plaintiffs’ result—a legalistic argument that could be fatal to the plaintiffs’ case but that the government could not make—and I decided to spend my break writing and submitting it.
The Canon of Constitutional Avoidance
Together with my student and the Jewish Alliance for Law and Social Action (JALSA), I wrote a brief arguing that the plaintiffs’ interpretation of § 36B, which would have treated health insurance subsidies as an incentive for states to establish exchanges, would raise serious constitutional problems. The plaintiffs’ interpretation, we argued, plausibly violated both the anti-coercion constraint that the Court had announced in NFIB v. Sebelius (the first Obamacare case) and the fundamental principle of equal sovereignty that the Court had announced in Shelby County v. Holder (the Voting Rights Act case).
The constitutional problem we highlighted was that, if the plaintiffs’ understanding prevailed, the federal government would end up enforcing a senselessly destructive regulatory regime in non-establishing states, while enforcing a sensible regime in compliant states, and it would do so with the intent of forcing all states to manage their own exchanges.
Under Shelby County, the federal government is constitutionally obligated to treat all states the same, and under NFIB, the federal government is constitutionally prohibited from threatening states with destruction if they refuse to implement federal programs. We weren’t arguing that § 36B was actually unconstitutional.
We were raising a rule of statutory construction, called the canon of constitutional avoidance that urges judges to avoid any statutory interpretation that would raise constitutional questions. Even the potential constitutional defects we highlighted would have justified the Court in choosing the government’s interpretation over the plaintiffs’.
At the King oral arguments, Justices Sotomayor, Ginsburg, and Kennedy all drew from the JALSA brief in questioning the plaintiffs’ lawyer. Indeed, Justice Kennedy made explicit reference to one of the hypothetical questions we posed in the brief; Kennedy noted that the federal government would not be allowed to lower the speed limit to 35 MPH in states that refused to raise their drinking age—a callback to the Supreme Court’s first anti-coercion case, South Dakota v. Dole, that we had included to highlight the potential problems with regulatory coercion.
In light of the justices’ questions, I was confident that Kennedy would vote for the government, but I was also completely sure that he would write an opinion, whether a majority or concurrence, that rested on our mutual constitutional concerns with the plaintiffs’ interpretation.
I was right on the first count, but deeply wrong on the second. Not only did Kennedy join Chief Justice Roberts’s majority opinion without writing separately, but also Roberts’s opinion completely ignored the constitutional problems that his colleagues and I had highlighted.
For me, this result was a little bit ego blow, a little bit regret for the lost Christmas break, and a large bit relief. The justices declined an opportunity I had given them to strengthen a couple of constitutional doctrines that are wildly unpopular among progressive academics (i.e., my friends and colleagues). But the result is also deeply puzzling—not just for me—and I’d like to take a moment to explain why.
Destruction v. Coercion
The statutory provision at issue in King, § 36B, contained three words—“by the State”—that Roberts admitted he was rendering superfluous. Typically, courts try to give effect to all words in a statute if possible, so Roberts needed to give some justification for his embrace of an interpretation that mooted three of § 36B’s words.
The reason I spent my Christmas break on the brief was to supply the Court with such a justification. If giving effect to those three words would render the statute plausibly unconstitutional, then the Court might end up with a choice between mooting the words or invalidating the statute. When given that choice, courts basically always choose to moot the words, either through avoidance in a statutory case or through severance in a constitutional case.
And Roberts is generally a fan of “savings constructions”—interpretations of statutory provisions that avoid constitutional problems. Indeed, he used such a construction in NFIB, deeming the individual mandate a tax rather than a penalty in order to save it from unconstitutionality.
But Roberts didn’t use that justification in King. Instead, he made a much simpler argument. The Congress that enacted Obamacare intended to fix health insurance markets, Roberts said, not to destroy them. The plaintiffs’ interpretation would threaten to destroy the markets, so that interpretation can’t possibly be what Congress intended. Roberts’s simple argument is essentially the same one we made in the JALSA brief—a concern over potential destruction of states’ insurance markets—but without all the constitutional jiggery-pokery.
Here’s the puzzle. Justice Scalia, in dissent, articulated the plaintiffs’ rebuttal to that argument, and Scalia’s rebuttal clearly raised the constitutional specter we had identified in the JALSA brief. Scalia said that the plaintiffs’ interpretation might cause destruction in health insurance markets, but if it did, then the states could simply establish exchanges in order to avoid the destruction. And if the plaintiffs’ interpretation turned out to be as destructive as Roberts predicted, Scalia argued, then all states would surely establish exchanges.
In other words, Scalia’s argument was: If the coercive threat turns out to be coercive, then all states will be coerced, but no states will be destroyed. Given that Roberts had articulated a concern about destruction but no concern about coercion, Scalia could assert coercion as a safeguard against destruction. Coercion became an uncomplicated rebuttal to Roberts’s only articulated justification for mooting three words of the statute.
This rebuttal might seem strangely casual for the justice who took coercion so seriously in the NFIB oral arguments and dissent, but for Scalia, it actually makes sense. Scalia doesn’t believe in the canon of constitutional avoidance, and he would rather invalidate an entire statute—allowing Congress to try again—than moot three of its words—transforming Obamacare into SCOTUScare. To Scalia, whether mooting words occurs through severance or avoidance matters not at all; he believes the Court’s obligation is to enforce whole statutes or to invalidate whole statutes.
Still, though, why didn’t Roberts respond? As Scalia pointed out, the plaintiffs’ understanding of congressional intent would have avoided both the mooting of statutory language and the destruction of health insurance markets. The only problem with the plaintiffs’ interpretation, from a purely legalistic perspective, was the plausibly unconstitutional coercion it would cause. Why didn’t Roberts make some mention of that flaw?
Roberts openly admitted that he was rendering three words of a statute superfluous; he gave a justification for ignoring those words that the plaintiffs, through Scalia, effectively rebutted; and he gave no rebuttal to the rebuttal, despite the availability of a fully-formed counter-argument in an amicus brief that his colleagues had discussed at oral arguments. Not only that, but I think Roberts must have convinced Kennedy not to write a concurrence highlighting the constitutional issue. Roberts seems to have gone pretty far out of his way to avoid avoidance.
I’m not terribly good at reading justices’ minds, so I won’t pretend to know why Roberts wanted to avoid the constitutional argument. Given the overall tenor of his opinion, though, and given what I know about the public debates leading up to King, I have one guess. I think Roberts—like most Americans who were familiar with Obamacare—thought it painfully obvious that Congress intended subsidies to be available nationwide.
There was a mountain of evidence that the three words in § 36B were there by mistake, and there was no evidence, beyond the text, that Congress wanted to use subsidies as an incentive for exchange-creation.
Unfortunately, because the statute was hastily and inartfully drafted and because the Solicitor General’s office could not make convincing sense of the words “by the State” in the relevant provision, Roberts could not prove the obvious conclusion through the usual tools of statutory construction. Because Roberts lives in the age of textualism, he was unwilling to reach beyond the text of the statute for proof of congressional intent, and the plaintiffs and their allies had come up with answers to all of the purely textual and structural arguments that the government and its allies had made.
Roberts seems to have run head on into an age-old conflict between legalism and realism. The legalistic case for the government’s interpretation was a close call, but to anyone who was willing to look beyond the text of the statute, the reality of the government’s rightness was a slam dunk. The justices don’t live in a legalistic vacuum, but for reasons that are not entirely clear, some of them like to pretend they do. Roberts is sometimes among them. But in King, Roberts seems to have been unwilling to walk the legalistic garden path, which would have ended in constitutional avoidance.
Roberts did not write a broadly purposivist opinion. He cited none of the mountain of extrinsic evidence of congressional intent that supported the government’s interpretation. He carefully referred to the legislative “plan” everywhere that a latter-day purposivist would have referred to the legislative “purpose.”
He sidelined Chevron deference rather than blessing his purposive approach as a legitimate argument for Chevron Step One. But he refused to close the loop on the legalistic case for the government’s construction. He refused to walk his legalism all the way to avoidance.
Avoiding Avoidance
In his decision to avoid avoidance, there may have been a lot of thoughts more prominent in Roberts’s consciousness than a resistance to legalism. He might have thought that a constitutional argument would invite further litigation that he had no interest in inviting. He might have thought that repetition of his NFIB argument—this time to “save” rather than weaken the statute—would look confusing or strange to a public that has been unusually attentive to the Obamacare cases.
He might have wanted to avoid a public misperception that he had deemed the statute unconstitutional but upheld it anyway. He might have wanted to avoid quasi-constitutional discourse—which is what constitutional avoidance holdings create—on two seriously under-developed constitutional doctrines.
But I suspect and hope that some little voice in Roberts’s head was objecting to the willfully irrational and unrealistic legalism of the plaintiffs’ case. I suspect and hope that some part of Roberts’s thinking was a rebellion to the textualists’ willful blinders to an empirically obvious congressional intent. Ultimately, I think we should understand Roberts’s avoidance of avoidance as a quiet but profound statement against the creeping unreality of textualist interpretation.
Roberts shouldn’t have needed avoidance to win the case, so he simply refused to need it. It’s a puzzling little opinion. But I have never been so pleased to be rendered superfluous.
No comments:
Post a Comment