The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
At some point between now and the end of June, the Supreme Court will decide King v. Burwell and, in the process, determine whether the phrase "established by the State" actually means "established by the State." This phrase, which appears twice in the Patient Protection and Affordable Care Act (PPACA) provisions authorizing tax credits for the purchase of health insurance in exchanges, has bedeviled defenders of the IRS rule purporting to authorize credits in federally established exchanges. Some claim this phrase is "convenient shorthand" for "exchange" (even though it's neither more convenient nor shorter), while others argue the phrase is effectively meaningless, or actually means something else (such as established in the State). The plaintiffs, on the other hand, maintain that the language means precisely what it says.
If Congress did not mean to refer to state-established exchanges, why did it use the phrase "established by the State"? According to story by Robert Pear in the New York Times "the words were a product of shifting politics and a sloppy merging of different versions. Some described the words as 'inadvertent,' 'inartful' or 'a drafting error.'" In other words, by this account it was a mistake—and a mistake no one noticed until well after the bill's passage.
This may be how congressional staffers and legislators characterize the drafting process now, but that's not what the federal government and its supporting amici told the Supreme Court. The solicitor general, for instance, proclaimed in its brief that the phrase "established by the State" was a "statutory term of art." At oral argument, the SG also denied that there were any last-minute revisions (as I discussed here). Throughout, the federal government has insisted that the statute actually authorizes the contested IRS rule.
Why hasn't the SG embraced the "drafting error" argument? Because that would be a sure way to lose. As I explained years ago, when first discussing this issue on the VC, the courts are not in the business of fixing purported "drafting errors," and the burden on those making such arguments is quite high. Justice Elena Kagan made the same point in her opinion last year in Michigan v. Bay Mills Indian Community, in which she explained that the "Court has no roving license, in even ordinary cases of statutory interpretation, to disregard clear language simply on the view that . . . Congress 'must have intended' something broader." It's also hard to maintain that the inclusion of a phrase is a "drafting error" when it was added in multiple places and multiple separate times during the drafting process, as "established by the State" was in Section 1401.
Later on in the Times story, in discussing how the PPACA emerged from the Senate Finance and HELP Committee bills, Pear writes:
there were substantial differences between the two committees' versions of the bill, and the task of merging them in October and November 2009 fell largely to Mr. Reid. Both bills called for insurance exchanges and provided subsidies to lower-income people, but the health committee measure clearly allowed subsidies in all states. The Finance Committee version was not so explicit.
As Pear recounts, there was language in the HELP bill clearly authorizing subsidies in federal exchanges, but no such language in the Finance Committee bill. What Pear fails to mention is that the HELP bill also expressly withheld subsidies in states that failed to adopt specified reforms. The staff members Pear interviewed may insist there was never any discussion of withholding subsidies, it is beyond dispute that this is precisely what the HELP bill did.
Not only did the HELP bill hold off subsidies for up to four years in states that refused to create their own exchanges, it also barred subsidies in states that failed to enact other desired reforms, as Michael Cannon and I explained in our amicus brief (see p. 26). Professor Tim Jost, who contests our interpretation of the PPACA on other points, described the HELP provisions in the same way in 2009 (see p. 7). Further, this point was conceded in the pro-government legislators' amicus brief filed in the D.C. Circuit in Halbig and (less explicitly) in FN18 of the government's King brief. So anyone who claims that the Senate never considered withholding subsidies in recalcitrant states is either a) dishonest, or b) doesn't know what they're talking about.
In stitching together the final legislation, Mr. Reid took the language on tax credits from the Finance Committee, and he generally followed the health committee in allowing the secretary of health and human services to operate a federal exchange.
But in borrowing from the health committee bill, Mr. Reid did not adopt an important provision that could perhaps have avoided the current fight. That provision said that a state with a federal exchange "shall be deemed to be a participating state," and that its residents could receive federal subsidies to help pay premiums.
In other words, according to Pear's account, the Senate drafters failed to include language expressly authorizing subsidies in federal exchanges, even though that language was in one of the precursor bills. This sort of equivalence language exists in other parts of the PPACA, however. Section 1322(c)(6), for instance, conditions recognition of an organization as a "qualified nonprofit health insurance issuer," in part, on whether a state has adopted insurance market reforms or "the Secretary has implemented [the reforms] for the State." So the bill's drafters knew such language was necessary where equivalence was desired and knew full well how to authorize the federal government to fill the state's shoes where that was intended. The bill's drafters just did not include such language here.
Pear's sources uniformly claim that it was never anyone's intent to deny subsidies in states that failed to create their own exchanges, but such claims really don't have much legal relevance. Post-hoc accounts of are not reliable indicia of legislative intent particularly where, as here, they are not supported by objective contemporaneous evidence. Yet to this day, no such evidence has emerged—no memos from committee staff discussing the drafting revisions, no e-mails or contemporary analyses of the actual bill's language, no documents indicating how House staffers interpreted the Senate bill they were forced to accept, nothing. And insofar as the bill was not everything its supporters wanted, and there was no way to fix it, they made the choice to pass it anyway.
The most reliable indicator of legislative intent is the text of the statute Congress enacted and the president signed into law. That statute only authorizes tax credits for insurance plans purchased in an "Exchange established by the State." As the Pear article makes clear, the relevant statutory text is on the side of the plaintiffs. Were it not, there would no need to claim anything was an "error" or proclaim a contrary intent.
A powerful line of judicial thinking holds that courts do not have a license to disregard or revise the clear language of a law.
What matters, Justice Antonin Scalia has said, is "not what Congress would have wanted, but what Congress enacted."
And that is precisely why the Court should side with the plaintiffs.
* * *
My co-author Michael Cannon has additional thoughts, and identifies some problems with some of the claims made by Pear's sources. He concludes:
While the Times' sources may have intended to undercut the King challengers' case, they, like others before them, inadvertently bolstered it. They acknowledged, perhaps inadvertently, that they did not pay as much attention to the bill as maybe they should have, that they had no choice but to enact a bill they knew was flawed, that the statute's meaning is plain, and that the government's "term of art" argument is unbelievable. They demonstrated why courts should rely on the plain meaning of statutes and ignore post hoc statements by members of Congress about what they really intended.
Charles Cooke reaches a similar conclusion.