The Illegal IRS Rule on Health Insurance Exchanges – A Reply to Bagenstos

One component of the PPACA (aka Obamacare) provides for the creation of health insurance exchanges in each state in which consumers may purchase health insurance. The PPACA’s supporters anticipated that every state would create its own exchange, and the law provides for tax credits and subsidies for the purchase of qualifying health insurance plans in state-run exchanges. Yet as Michael Cannon and I pointed out the PPACA does not authorize tax credits and subsidies in exchanges run by the federal government. This is a potential problem because at least twenty states are refusing to create their own exchanges and are defaulting to the federal option. In response, the IRS issued a rule to authorize tax credits and subsidies in federal exchanges. The only problem, as Cannon and I explain at length in a forthcoming article in Health Matrix, the IRS rule is illegal. Now the rule is being challenged in court by Oklahoma in a suit legal analyst Stuart Taylor calls “by far the broadest and potentially most damaging of the legal challenges” to PPACA implementation, and more suits are likely.

The IRS defends its rule, but has had difficulty providing much by way of justification beyond vague references to congressional intent. Now comes my friend Samuel Bagenstos, at his Disability Law blog and Balkinization, arguing that Cannon and my arguments are “nonsensical” and “deeply legally flawed.” Bagenstos is a serious scholar, and his arguments are clever, but they do not sustain the case for the IRS rule.

Bagenstos’ central argument is similar one advanced by Tim Jost on the Health Affairs blog (and to which Cannon and I responded here). Essentially he argues that the PPACA makes federal fallback exchanges the legal equivalent of state-run exchanges and therefore any tax credits or subsidies authorized in the latter must be available in the former as well. Here is the argument as Bagenstos lays it out:

Although the tax-credit provision twice uses the phrase “Exchange established by the State under section 1311,” see 26 U.S.C. § 36B(b)(2)(A), (c)(2)(A)(i), that phrase does not have the exclusionary meaning Cannon attributes to it. That is because Section 1321 (codified at 42 U.S.C. § 18041) makes clear that, when a state fails to set up an exchange, the federally-operated exchange will stand in the shoes of the state exchange for purposes of Section 1311. Thus, Section 1311 provides that “[e]ach State shall” set up an exchange by January 1, 2014. 42 U.S.C. § 18031(b)(1). Section 1321 provides that if a state “will not have any required Exchange operational” by then — that is, an exchange required by Section 1311 — then the federal government “shall (directly or through agreement with a not-for-profit entity) establish and operate such Exchange within the State.” 42 U.S.C. § 18041(c)(1) (emphasis added). “[S]uch Exchange” in Section 1321 clearly refers to the “required exchange” — that is, the Section 1311 exchange. When the federal government operates an exchange pursuant to Section 1321, then, it is not operating some wholly foreign entity; it is operating the state exchange that Section 1311 required the state to set up but that the state failed to create. Because Section 1321 provides that a federally-operated exchange will stand in the shoes of a state-operated exchange created by Section 1311, there is no basis for denying participants in federally-operated exchanges the same tax credits obtained by participants in state-operated exchanges.

So, according to Bagenstos, when Section 1321 directs the federal government to create “such Exchange,” it is authorizing the federal Section 1321 exchange to operate as a Section 1311 exchange. As I said, it’s a clever argument, but it’s incomplete. Just because a federal exchange created under Section 1321 is subject to all the same requirements as a state exchange created under Section 1311 does not mean that tax credits available in a state exchange much be available in a federal exchange as well, particularly when the plain text of the statute provides otherwise.

First, Section 1311 expressly requires that an authorized Exchange must be “established by a State.” Section 1304(d) also expressly defines “state” as “each of the 50 States and the District of Columbia.” Later amendments to the PPACA also provide that Exchanges created by territories are to be treated as the equivalent of state-run Exchanges, but there is no such language concerning federally run Exchanges. But let’s set this aside and concede, for the sake of argument, that a Section 1321 Exchange is the equivalent (or “stands in the shoes”) of a Section 1311 Exchange. This is still not enough to sustain Bagenstos’s claim.

The eligibility requirements for the tax credits are laid out in Section 1401. This section repeatedly defines qualifying health insurance plans eligible for tax credits as those purchased “through an Exchange established by the State under section 1311.” We can read Section 1311 to incorporate Section 1321 as Bagenstos urges, but a federal Exchange is still not an Exchange “established by the State.” Bagenstos claims he has provided “the most plausible reading of the statutory text,” but to accept Bagenstos’s argument is to render this repeated language both redundant and surplusage. The “most plausible” reading of any statute is not one that requires the reader to ignore relevant statutory text or deprive it of any meaning, yet that is precisely what Bagenstos does.

Bagenstos goes on to argue that “nor is there any reason” why Congress would have conditioned the availability of tax credits on state cooperation. But of course there is. The authors of the Senate bill wanted states to create exchanges. The statute even purports to require states to do it. But Congress cannot tell states what to do. Thus it needed to provide them with an incentive to play along, and committing to create a federal exchange as a fallback is not much of a threat. So the Senate bill provides modest financial support for state creation of exchanges (but no money for oerating expenses) and threatens to withhold benefits – the tax credits and subsidies — in states that don’t comply. Where did the Senate get this idea? One possibility is our friend Tim Jost, who wrote in 2009 that Congress could try to induce states to create exchanges by, among other things, “by offering tax subsidies for insurance only in states that complied with federal requirements.” While less common than threatening to withhold funds (as was done with Medicaid) this approach is not unprecedented, and several of the pre-enactment Senate health care reform bills contained similar provisions explicitly designed to encourage state cooperation.

Bagenstos thinks the threat of withholding tax credits was unnecessary because the PPACA provides for a federal fallback exchange for noncompliant states, and this was enough to give the states a meaningful choice. But the PPACA’s authors did not want states to have such a choice, as indicated by the language of the bill, repeated statements that all states would (or would be required to) create their own exchanges, and the lack of any money in the bill to pay for creating federal fallback exchanges. The requirement that HHS create federal fallback exchanges in noncompliant states is hardly much inducement for state cooperation – and everyone knows it. Even when Congress puts a big pot of money on the table, as it did with the Medicaid expansion, states are reluctant to play along at their own expense. Conditional preemption is common in cooperative federalism statutes, but it is rarely sufficient to get states to toe the line. That’s why Congress sought to condition other funds on state participation on Medicaid’s expansion and why it would have been hopelessly naïve to believe that every state would create an Exchange – as PPACA advocates repeatedly claimed – without something else on the table. That something else is the availability of tax credits. The PPACA’s authors’ mistake was not in how they drafted the statute, but in believing they had done enough to get every state to create an exchange.

Bagenstos also takes issue with our reference to a colloquy in the Senate Finance Committee as further evidence that the Senate bill’s authors knew what they were doing. At a hearing, Senate Finance Chairman Max Baucus was asked to explain how the committee had jurisdiction to impose requirements on state-run health care exchanges. The answer, according to Sen. Baucus, was that the Committee could impose “conditions” on the receipt of tax credits. Precisely. Bagenstos argues “there is nothing the Finance Committee’s jurisdiction that required it to limit tax subsidies to participants in state-operated exchanges.” That’s true, but beside the point. If the Senate Finance Committee wanted to tell states how their exchanges were to be run, it needed a hook. That’s the role the conditional tax credits play. The Committee could write requirements for state-run exchanges because these requirements were “conditions” for receipt of tax credits.

In sum, the text of the PPACA is clear, as even some of our opponents have acknowledged. The alternative interpretation Bagenstos offers requires disregarding portions of the text and ignoring the statute’s history. Neither Bagenstos nor anyone else has been able to identify any statutory language or contemporaneous evidence from the legislative history that is inconsistent with Cannon and my account. There are lots of statements that tax credits would be available in every state, but even more comments that every state would create its own exchange, and the latter explains the former. The CBO scored the bill as if tax credits would be available in every state, but has admitted that it conducted no legal analysis of the Senate bill. Congress expected states would willingly create exchanges if given mild inducement, and the Senate bill was written accordingly. This was in error, but it does not justify rewriting the statute.

A final point: Bagenstos characterizes our paper as a “rearguard challenge to Obamacare,” implying the arguments were manufactured after the fact to “unravel” the Supreme Court’s NFIB decision. The suggestion is that these are arguments of convenience. Not so. Cannon and I raised these issues well before NFIB was argued, let alone decided. I first noted how the text of the PPACA limits tax credits to state-run exchanges in a talk at a health care reform conference in February 2011 (subsequently published that spring in the Kansas Journal of Law & Public Policy), without realizing how this language could affect PPACA implementation. Cannon and I also first collaborated on this issue in November 2011, before the IRS rule was finalized and before many thought judicial invalidation of the individual mandate (let alone the PPACA as a whole) was even a remote possibility. There are also larger issues at stake. Federal agencies only have that authority they have been delegated by Congress and are not authorized to rewrite statutes in response to changing political conditions. Therefore, just as I criticized the Bush EPA for taking liberties with the Clean Air Act, I am now criticizing the Obama IRS and HHS Administration for taking liberties with the PPACA.

For those who want more, Michael Cannon also responded to Bagenstos at the Cato@Liberty blog, prompting a rejoinder from Bagenstos here. For those who want more, Bagenstos and I will have the chance to debate this issue tomorrow during a Federalist Society teleforum. (Podcast to follow.) Cannon and I will also be posting a revised and updated draft of our paper on SSRN shortly.