[This post has been edited to reflect corrections received by email from Jon Adler, whom I sincerely thank for his thoughtful engagement.]
Federalism is in the air again—once again surrounding Obamacare’s financing structure—and some people are surprisingly surprised. Given Justice Kennedy’s questions at oral argument, it’s become quite clear that King v. Burwell is about federalism, and a couple of conservative commentators have written blog posts that amount to a cry of, “I am shocked—shocked—to find federalism issues in this case.”
Of course the case is about federalism. It has always been about federalism. The only thing that the JALSA brief (which I authored)—and Kennedy’s questions arising from that brief—have added to the case is the point that constitutional federalism cuts in the government’s favor, not the plaintiffs’. Perhaps political federalism cuts in the plaintiffs’ favor—if we join Jon Adler, Michael Cannon, and Michael Carvin in their extended daydream about Senator Ben Nelson’s intent—but constitutional federalism obviously supports the IRS’s interpretation (via the canon of constitutional avoidance as well as canons urging preservation of the traditional federal-state relationship and requiring clarity in the communication of conditions).
In short, the argument in the JALSA brief is that the plaintiffs’ interpretation embeds a severe regulatory threat into the statute. Under the plaintiffs’ reading, Obamacare not only deprives non-establishing states of subsidies (a traditional fiscal threat) but also imposes a perverse subset of Obamacare’s substantive regulations on non-establishing states (a novel regulatory threat), and the particular subset of regulations that applies under plaintiffs’ interpretation is perfectly designed to cause insurance market death spirals. That structure seems at least plausibly (perhaps clearly) unconstitutional under NFIB (and maybe also Shelby County, as I argue in the brief).
Given that this argument leverages two “conservative” cases, it seems strange that some conservative commentators have reacted so negatively to the argument. Rick Hills made this point beautifully, and he also identified some of the potential motivations for conservative hypocrisy in (particularly Ilya and Randy’s) recent King posts. I agree with the vast majority of his analysis, and in some respects, I would go even further. There’s a curious phenomenon occurring over this JALSA brief: Other than the Court, Rick, and me, folks don’t seem to like it much. In fact, several commentators (including Ilya Somin in his recent post) have insisted on citing and referencing the states’ brief instead, even while clearly confronting the arguments in the JALSA brief. (Are they trying to deflect attention from the version of the argument that’s harder for them to rebut?) Some Court watchers have asked me to stop writing so much about federalism now that the spotlight has gotten so intense. One reason, I think, must be the one Rick identifies: Liberals are worried that the brief takes seriously a constitutional doctrine that could threaten other liberal legislation, and conservatives dislike the brief because they want to win King. (Kennedy’s questions have even inspired Randy Barnett and Nick Bagley to write identically titled blog posts. Who’d have thought that was possible?)
Here’s the thing. Constitutional federalism doesn’t and shouldn’t have a partisan valence. There’s plenty to critique about the anti-coercion constraint, the anti-commandeering constraint, and the limits on the commerce power without turning your nose up at them for being “conservative” doctrines. Perhaps more importantly, they’re not “use it or lose it” doctrines; they won’t disappear if liberals ignore them, and they won’t become pure GOP tools if conservatives try to distort them when their proper application cuts against the conservative cause. Federalism constraints are law; they’re getting stronger as time goes on (consider Lopez to Morrison to NFIB); and at least some of them might be sensible if the Supreme Court would start explaining them more clearly and rationally. Yes, modern federalism doctrine marks a departure from the New Deal. But that doesn’t make it definitionally terrible, even from a liberal perspective. Our job as legal academics—even liberal ones—ought to be to confront the doctrine, help the Supreme Court to rationalize it, and (for the Dworkinians, at least) figure out best justifications for the doctrine as it evolves. Part of that project might include exploring the doctrine’s application to live cases, like King, in which we liberals might be able to use the doctrine to support our preferred outcomes.
All of that said, the particular use to which I put NFIB in the JALSA brief might not be a mere application of existing doctrine. It might change and even expand that doctrine. (I don’t think it does, but there is certainly room for debate on this front.) And one of the questions that many colleagues and commentators have raised—including Jon Adler and Rick—is whether I am ready and willing to embrace the application of my King argument to various liberal statutes that seem to employ regulatory coercion, like the Clean Air Act and the Low-Level Radioactive Waste Policy Amendments Act of 1985 (the statute at issue in New York). The short answer is yes, but I think the question requires much more nuance than that. I am perfectly happy to analyze liberal statutes under the doctrine I advocate in the JALSA brief, but that might be partly because I don’t think the application of that doctrine would render those statutes unconstitutional.
An important note at the outset: I am not an expert on environmental statutes. These thoughts are largely impressionistic reactions to some issues that Jon and Rick raise, and they might contain errors or at least debatable ideas about the ways these statutes do and should work.
The Clean Air Act
As I argue in the brief, the problem with the King plaintiffs’ interpretation is that it causes the federal government to enforce different rules in different states based on the states’ passive tripping of a statutory trigger. That is, under plaintiffs’ interpretation, if a state fails to establish a compliant exchange (even if unintentionally), then the federal government will change federal rules within that state’s borders. That’s quite different from the Clean Air Act, which sets a federal default regime and then empowers the states to change the rules within their own borders if they choose to operate under a State Implementation Plan (SIP). Under the CAA, the states have a choice between regulating things themselves (under federal oversight) or letting the federal government regulate for them. The CAA does not give states a choice between two different federally-crafted and federally-enforced regulatory regimes (one of which is clearly punitive). That, to my mind, is a huge difference.
What about Texas’s claim, in the lawsuits Rick mentions, that EPA has threatened to impose a draconian regulatory rule—the suspension of all new emitting constructions—in states that refuse to incorporate greenhouse gases into their SIPs? Isn’t that exactly the same kind of regulatory threat that I complain about in King? I don’t think so. The threat here is not a new set of destructive federal regulations. It’s that EPA will revoke Texas’s SIP, which would prevent Texas from issuing any lawful permits for new construction. But the state could avoid the ban on new constructions, which emerges de facto from the SIP revocation, by authorizing EPA to issue permits instead (under a Federal Implementation Plan (FIP)). This structure is the standard federal fallback structure, common in cooperative federalism, that Congress actually intended for Obamacare’s exchanges; it is not the regulatory threat that plaintiffs read into Obamacare.
How about Jon’s point that the CAA might punish states for relying on the federal fallback regime by requiring a higher offset ratio for emissions under the NAAQS (a different provision than the one at issue in the Texas litigation)? Doesn’t that amount to an unconstitutional regulatory threat—a threat of tougher federal regulation in noncompliant states? Here, the question is closer and depends in part on the proper interpretation of the sanctions provisions in 42 U.S.C. s 7410(m) and s 7509(b). Again, I’m not an expert on this statute, and I don’t know how EPA or the courts have interpreted those provisions. On the face of the statute, section 7410(m) specifies that sanctions must be levied only “for the purpose of ensuring that the requirements of this chapter relating to [a SIP] . . . are met.” That, to me, implies that sanctions can be imposed only if a state is continuing to try to operate under a SIP, and such sanctions are unavailable to EPA if the state chooses to hand regulatory responsibility over to a FIP. That said, if I’m wrong about that and the statute does authorize the federal government to punish states merely for the choice to let the federal government regulate in their place, then that punishment should, in my opinion, be subject to the anti-coercion constraint I describe in the JALSA brief.
The question, in that case, would turn on whether the federal government is sincerely willing to impose the CAA’s tougher offset requirement when states choose the federal fallback regime. This answer is an application of Einer Elhauge’s argument, as I understand it: that the coercion constraint should depend on whether a threat is contrived, which depends on whether the federal government would want to impose the threatened regime if it couldn’t communicate the threat. I’m not an environmental law person, so I don’t know whether the more-stringent offset requirement looks as contrived as the subsidy-free insurance regime that plaintiffs are urging in King. (I doubt it given how completely insane the subsidy-free regime is.) But that, to my mind, is the relevant question here.
What about the fact that a federal fallback regime would (or at least could) be substantively different from the choices a state would want to make under a SIP? Doesn’t that constitute impermissible regulatory differentiation under the JALSA brief’s analysis? On this one, I think the answer is clearly no. First, it’s important, perhaps decisive, that the choice under the CAA is between federal default rules and state crafted rules; the CAA does not require states to choose between two different sets of federally-crafted and federally-enforced rules, one of which is more draconian than the other. The federal law allows states some choice if they regulate themselves, but it does not reward states with a laxer federal regime if they agree to play ball in some other regulatory arena.
The reference to “some other regulatory arena” brings me to my second point: It’s important that the choice the CAA offers is self-contained; the threat of federal regulatory displacement is contingent on a state’s choice to regulate the same field for itself. If a state agrees to take on the regulatory task, then the federal government will reward the state with some regulatory flexibility, but if the state refuses the regulatory task, then the federal government, out of necessity rather than punition, will impose its own (preemptive) regulations, which the state might or might not like. But the state’s choice between the two regulatory worlds is not, like in King, contingent on its willingness to take on some other task that is only loosely related to the threatened federal regulation.
To make this point more concrete: The CAA and the real (government-interpreted) Obamacare are actually quite analogous. Obamacare makes this offer to the states: “If you set up an exchange (task 1), you can have some flexibility in how you manage the exchange itself and in how you regulate insurance plans sold through the exchange (some choices of how to do task 1), but if you rely on us to set up an exchange (task 1), we will set all of the rules for exchange management and insurance policies sold, whether you like our rules or not (our choice of how to do task 1).” That part of Obamacare is just like the CAA’s federal fallback rules. But that’s not what the King plaintiffs claim Obamacare does. Under the plaintiffs’ read, this is Obamacare’s threat: “If you set up an exchange (task 1), we will enforce a rational set of insurance market regulations in your state, some of which will apply both on and off the exchanges (our choice of how to do task 2), but if you rely on us to set up an exchange (back to task 1), we will enforce a destructive set of insurance market regulations in your state (a punitive version of task 2).” That structure would be brand new to cooperative federalism, and it would be a transparent attempt at coercion.
The Low-Level Radioactive Waste Policy Amendments Act (LLRWPAA)
Jon’s other question for me is whether my argument could undermine Part III.B of New York, which upheld one of the incentives for state regulation that was included in the LLRWPAA. The problem that Congress was trying to address in the LLRWPAA was that there were not enough disposal sites for low-level radioactive waste in the country, and the states were reluctant to build more. All states were free-riding on three states’ disposal sites, and those three sites were getting overburdened. The statute therefore tried to incentivize states to build disposal facilities.
In the part of New York that Jon and Ilya are citing against the JALSA brief, the provision at issue authorized states that built disposal sites (complying with the federal policy) to refuse to sell disposal services to producers of waste from states that had refused to build. In other words, compliant states could deny access to out-of-state producers from non-compliant states. (The authorization for that discrimination had to come from Congress because a state’s discrimination against out-of-state citizens would otherwise be unconstitutional under the Dormant Commerce Clause.) As Jon interpreted this provision in an email to me, the LLRWPAA said: “Comply or producers of low-level radioactive waste in your state (including hospitals and medical research centers) will have no place to dispose of their waste.” So stated, the LLRWPAA looks like it contains a severe, federally-imposed regulatory punishment—and certainly a contrived threat—for states that refuse to comply with the statutory directive.
In reality, though, the statute said something quite different. It said: “Comply or producers might have no place to dispose of their waste, depending on how you, the state, regulate the generation and disposal of low-level radioactive waste (because state regulations are not preempted) and depending on whether compliant states take up Congress’s invitation to exclude your producers.” That more-accurate statement of the statute’s structure contains two enormous and highly relevant differences.
First, as Justice O’Connor made clear two sentences after the quote that Jon blocks in his post, LLRWPAA was not preemptive of non-compliant states’ regulatory regimes. In O’Connor’s words: “Nor must the State abandon the field if it does not accede to federal direction; the State may continue to regulate the generation and disposal of radioactive waste in any manner its citizens see fit.” So the federal regulation was not really regulatory at all. It was a suggestion of one way to facilitate safe disposal of low-level radioactive waste, and it was an authorization of interstate compacts (which have to be congressionally approved under the Compacts Clause) to facilitate bargains between and among the states for the disposal of their wastes—to overcome a collective action problem. The states, however, remained free to address the problem of such waste however they liked. They could prohibit or limit the generation of waste; they could figure out disposal mechanisms other than the ones suggested in the federal act and amendments; or they could build disposal sites just like the ones recommended but without following the federal statute’s deadlines and guidelines. It’s hard to call that degree of flexibility “coercive,” no matter how you slice it.
Indeed, the absence of preemption under the LLRWPAA clearly distinguishes New York from King. The problem with plaintiffs’ interpretation of Obamacare (according to the JALSA brief) is that it threatens federal imposition of preemptive regulations that are designed to be destructive. It threatens a direct exercise of federal power within the non-compliant state, which the non-compliant state would be powerless to avoid through any route other than caving to the federal order and establishing an exchange. A holding that this structure is constitutionally problematic would not undermine New York at all. The relevant provision of LLRWPAA didn’t threaten any preemptive regulations, much less preemptive regulations that were designed to be destructive.
Second, under the actual LLRWPAA (not Jon’s statement of it), any harm that occurred to private citizens in non-compliant states would arise, proximally at least, from other states’ regulatory choices, not from federal regulation. The federal government was not threatening to punish non-compliant states for their non-compliance; it was empowering compliant states to refuse to expend their own resources in facilitating non-compliant states’ non-compliance. The statute empowered all states to protect themselves against other states’ free-riding. To make Obamacare analogous to that structure, the statute would have to say: “States that establish exchanges are authorized to refuse to sell insurance policies to citizens of non-establishing states, but non-establishing states remain free to regulate insurance in any way they like.” That’s very different from the plaintiffs’ statute, which leverages federal regulatory power to destroy non-compliant states’ citizens.
If LLRWPAA had really threatened to ban the safe disposal of low-level radioactive waste produced in non-compliant states, then it would have been clearly unconstitutional under the principle I articulate in the JALSA brief. But the LLRWPAA’s actual structure simply does not involve federal imposition of destructive regulations within the states. Indeed, the relevant provision didn’t involve federal imposition of any regulations at all, and the provision that did involve federal regulation of the states (the “take title” provision considered in Part III.C of the opinion) O’Connor deemed unconstitutional.
A concession: Jon and Ilya pick up on this line from O’Connor’s opinion: “The affected States are not compelled by Congress to regulate, because any burden caused by a State’s refusal to regulate will fall on those who generate waste and find no outlet for its disposal, rather than on the State as sovereign.” If that line were serious, then plaintiffs’ interpretation in King would be constitutionally unproblematic because the threatened regulatory regime would, indeed, apply directly against citizens rather than against the states. Rick responds extremely well to this point in a follow-up to his original post, and I confront it head-on in Part I.B.3 of the JALSA brief. I will say here only that I find it decisive that all of anti-coercion law grows from Steward Machine, and Steward Machine is a case of direct regulation, not punishment of states qua sovereigns.
A Final Note
So far, it probably seems like I’m feinting when I say that I will happily apply my federalism principles to liberal legislation. I’ve applied them here, but I’ve also rejected both of the challenges presented. The reason, I think, is that there aren’t many (if any) federal statutes that violate the principle I articulated in the JALSA brief. All I can say to reassure conservative readers of my sincerity is this: I wish I had published a formal prediction that the Supreme Court would invalidate the Medicaid expansion. I certainly made that prediction on several occasions, and I think the Supreme Court reached the right result, as much as I wish, politically, that all states would expand Medicaid. I also think that if the plaintiffs win King then Maine should win Mayhew v. Burwell; the Court should invalidate the Medicaid maintenance of effort provision if that provision becomes an additional regulatory punishment for non-establishing states (which it would after a pro-plaintiff holding in King). And, as a general matter, I’m not a big fan of cooperative federalism. It causes inefficiencies, and the Supreme Court has lately been making it worse. And strengthening the anti-coercion constraint might mean more purely federal regulation, which would be a great result from my perspective. So, I’m not sure there’s anything we can do to overcome the prisoners’ dilemma Rick identifies, but I am sure that I wish we could overcome the pervasive tendency towards opportunistic federalism.