From questioning today in King v. Burwell, there is quite a buzz that Justice Kennedy appears concerned about whether interpreting the ACA to deny subsidies to citizens of states with federal exchanges would unconstitutionally coerce states to set up their own exchanges. The alleged coercion would result from the damage caused to the insurance markets of these states by the other mandates in the ACA — for example, by “community rating” that restricts the ability of insurance companies to set their rates according to actuarial risk, and “guaranteed issue,” that is, preventing carriers from refusing insurance based on pre-existing conditions. The concern is that, because these provisions absent a subsidy would cause a “death spiral” in those states, states would be unconstitutionally coerced to setting up their own exchanges lest their insurance markets be destroyed. Therefore, it is contended, under the doctrine of “constitutional avoidance,” the ACA should be interpreted to avoid this unconstitutional result and allow the IRS regulation to stand so subsidies will flow to the states.

Readers may note that I have not previously opined here on the merits of King v. Burwell. I have remained silent because there were no constitutional issues involved upon which I considered myself expert, and I was not prepared to offer an expert opinion on the statutory construction question presented by the case. I freely admit that I hoped and believed that the challengers of the IRS regulation had the better argument, but I remained open to arguments to the contrary, and was unprepared to take a public position on the question presented. Nor did I file or join any amicus brief in the case.

But the invocation of the constitutional avoidance doctrine starts to trench into the constitutional domain in a particularly annoying way.  Here are just some of the concerns raised by employing this doctrine now to save the IRS regulation.

1) As a threshold matter, this constitutional concern seems misplaced in the case that is before the Court. First, 8 States filed amicus briefs in support of petitioners, saying they don’t want exchanges OR subsidies — so obviously those States aren’t being “coerced.” Second, neither party in this case has ever raised the constitutional concern, so we lack adversarial briefing on this issue. Third, if the relevant wording of the statute is unambiguous and this wording exposes the statute to constitutional attack in some later case, then so be it. This is similar to later Origination Clause challenges to the “individual insurance mandate” cum “option to buy insurance or pay a modest tax” that could only be brought once it was established that what looked like a Commerce Clause “penalty” was really a noncoercive tax. We must take up these matters one step at a time.

2) But suppose that a State later raises a constitutional coercion claim: How does that support the IRS regulation making the subsidies available on federal exchanges at issue in this case?  Assuming there is a “constitutional concern,” what’s the proper remedy? Do you rewrite the statute to make subsidies available in states that don’t establish exchanges, or do you strike down the federal insurance regulations that allegedly create the “death spiral” and threaten to “destroy” state insurance markets unless states set up exchanges? It is the latter expressed and unambiguous regulations that will cause the “coercive” death spiral. To remedy the adverse affect of these regulations, should the Supreme Court judicially authorize billions of dollars in subsidies that Congress refused to authorize?

3) The supporters of the government have urged the importance of “state flexibility.” Yet states will have more flexibility if the Court later invalidates the community rating and guarantee issue requirements as unconstitutionally coercive (unless Congress provides subsidies). If the Court does that, states would be able freely to choose between (a) setting up their own exchanges + Obamacare regulations + federal subsidies to pay for them OR (b) having a federal insurance exchange without subsidies, but no costly guaranteed issue and community ratings regulations, which is what what makes the federal subsidies necessary.

4) This is essentially the same remedial option that Justice Kennedy and the four other justices embraced in their NFIB dissent: with respect to Medicaid, the remedy should be to invalidate the Medicaid expansion, not to invalidate the condition between the new and old Medicaid funds. Moreover, Justice Kennedy and the other dissenters further argued that the invalidity of the Medicaid expansion (together with the unconstitutional individual mandate) required invalidating the whole Act as non-severable.

5) Thus, if any justice thinks the coercion of states is a serious constitutional problem, the best way to “avoid” it is to limit subsidies to exchanges “established by the state,” and then invite states who feel coerced to choose to bring a coercion challenge in the future. Should such a challenge be brought, the Court could at that time find the community-rating and guaranteed-issue regulations (which are not now before the Court) to be unconstitutionally coercive in states that decide not to establish exchanges (and assess the severability consequences of that invalidation).

  • What the Court should not do  is decide the case on an issue without the benefit of full briefing and argument.
  • What the Court should not do is rewrite one part of a statute to avoid the “coercion” that is allegedly caused by unambiguous parts of the law that are not presently before the Court.
  • What the Court should not do is refuse to enforce the ACA as written to uphold an IRS regulation that is contrary to the meaning of the ACA in context to redress the onerous consequences of other clear and unambiguous provisions of the ACA.

But this then returns us to where the focus of this case has been all along: does the meaning of the ACA restrict subsidies to “exchanges established by a State” or not? And that is the statutory construction question presented by King v. Burwell on which I still do not wish to opine.