The Volokh Conspiracy
Mostly law professors | Sometimes contrarian | Often libertarian | Always independent
Defenders of the administration's position in King v. Burwell have recently emphasized two federalism-based arguments in support of their claim that the Affordable Care Act must be interpreted to allow federal subsidies for health insurance purchased in federal government exchanges, as well as in those established by states. The first, emphasized by Marty Lederman, holds that any ambiguities in interpretation of the law should resolved in favor of the government in order to adhere to the Supreme Court's rule that federal statutes must be construed to avoid interpretations that upset "the usual balance" between federal and state power. The second—advanced in an amicus brief by state governments that support the ACA—claims that the plaintiffs' interpretation might raise constitutional problems because it would result in possibly unconstitutional "coercion" of state governments, who would be pressured into setting up their own exchanges in order to avoid the loss of federal subsidies. The two arguments suffer from the same common flaw: Both try to use doctrines developed to protect state governments against potentially excessive federal intrusion to justify an interpretation of the ACA that expands federal power rather than limits it.
Like Randy Barnett, I don't have any strong opinion on the statutory interpretation issues that are at the heart of this case. As a federalism scholar, my role in this debate is very limited. I can only skulk around the edges of the battlefield and pick off vulnerable straggling arguments that happen to overlap with my areas of expertise. But these particular stragglers have gotten a lot of attention recently, and deserve a response.
I. The Federalism Canon Against Statutory Interpretations that Upset the "Usual Balance" between State and Federal Power.
In several cases, the Supreme Court has ruled that interpretations of federal statutes that upset the "usual constitutional balance of federal and state powers" are disfavored, and can only be accepted if this result is unmistakably clear from the text of the law. Defenders of the federal government's position in King v. Burwell argue that the plaintiffs' interpretation of the law upsets the balance because it would pressure the states into setting up exchanges, lest they risk losing federal subsidies and unbalancing their health insurance markets.
The problem with this argument is that the federalism canon does not protect states against all painful choices created by federal statutes, but rather, as the Court put it in the recent case of Bond v. United States (2014), against the expansion of federal power into "areas of traditional state responsibility." For example, Bond rejected an interpretation of a federal law that would "convert an astonishing amount of traditionally local criminal conduct into a matter for federal enforcement, and involve a substantial extension of federal police resources" (internal quotations omitted).
The federalism canon therefore can't be used to justify interpretations of federal law that actually expand the scope of federal power rather than restrict it. And that is precisely what defenders of the federal government's position are doing in King v. Burwell. If the plaintiffs prevail, states have the option of setting up their own Obamacare insurance exchanges and thereby getting federal subsidies for their residents, or refusing to do so and thereby foregoing the subsidies while also freeing many of their residents from the ACA's employer and individual insurance mandates. For states that choose the latter option, the net result is a lower level of federal intrusion into their health care markets than those that choose the former. Several conservative states have filed an amicus brief indicating that that is precisely the alternative they prefer.
By contrast, under the federal government's interpretation of the ACA, state governments are left with little meaningful choice. They must either establish ACA-compliant exchanges for their residents themselves, or have the federal government establish nearly identical institutions (supposedly acting on their behalf). In either case, they end up with both federal subsidies and full application of the individual and employer mandates within their territory. This is a considerably more expansive federal role in health care policy than would exist under the plaintiffs' approach to the ACA.
Of course, it's also possible that the federalism canon just doesn't apply to this case at all. As Obamacare defenders emphasized throughout the previous round of ACA litigation that culminated in NFIB v. Sebelius, health insurance has been heavily regulated by the federal government for decades, including by means of statutes that impose restrictions and obligations on state governments. If this view is correct, then neither side's position in King upsets the "usual" balance between federal and state power, because health care policy is no longer an area generally left to the states. But if the canon does apply, it favors the side supporting a lower level of federal involvement and greater freedom of choice for states. And, not surprisingly, that side isn't the federal government.
II. "Coercion" of State Governments.
The coercion theory has many of the same weaknesses as the federalism canon argument. It's hard to credibly argue that the plaintiffs' interpretation of the ACA is coercive when your own preferred alternative imposes tighter constraints on state governments and includes a greater degree of federal control of health care regulation. The coercion argument also has some additional weaknesses.
The Supreme Court's current state coercion precedents apply only to conditions attached to federal spending grants to state governments. And even then, they have only been invalidated when the condition is so severe that it amounts to a "gun to the head," as Chief Justice Roberts put it in NFIB v. Sebelius (the only Supreme Court ruling that has ever invalidated a spending condition as coercive). As Jonathan Adler points out, current Supreme Court precedent holds that there is no constitutional problem with federal statutes that use the threat of increased federal regulation to incentivize states to regulate on their own.
Justice Sandra Day O'Connor's majority opinion emphasizes in New York v. United States (1992), the key distinction between regulatory statutes and conditional spending statutes is that, in the former case, "[t]he affected States are not compelled by Congress to regulate, because any burden caused by a State's refusal to regulate will fall on [private citizens], rather than on the State as a sovereign."
This distinction might seem artificial. After all, if a state's residents are harmed, the state government often suffers also, even if indirectly. But if the Supreme Court were to expand the coercion doctrine to cover federal laws that burden states indirectly by imposing economic costs on private citizens and firms, that would likely have radical consequences. Jonathan Adler points out that there are many other federal programs that offer states a choice of adopting state regulations or accepting greater federal regulation of their citizens.
Moreover, many federal laws that do not explicitly incentivize states to regulate still implicitly impose painful choices on them by inflicting costs on state residents. Federal minimum wage laws increase unemployment and harm businesses that could otherwise prosper by hiring low-wage workers. Federal environmental laws depress economic activity in industries they regulate. Federal immigration restrictions weaken the economies of states that could otherwise attain greater prosperity by attracting foreign workers. In each of these cases and many others like them, the harm inflicted on the private sector often ends up reducing tax revenue available to state governments, indirectly forcing them to adopt new regulatory and spending policies in order to adjust. If the plaintiffs' interpretation of the ACA in King is coercive, the same is likely true of other federal laws that impose large economic costs on private sector actors, and thereby have a major effect on state governments.
Advocates of the coercion argument in King do not go quite so far as to claim that the plaintiffs' interpretation is coercive enough to be unconstitutional. They merely claim that its possible coerciveness is great enough to create a potential constitutional problem that the Court should interpret the ACA to avoid. Even so, their argument suggests that numerous other federal laws also raise similar constitutional problems, and should also be construed in ways that reduce the burdens they impose on state governments.
The federal government's position in King v. Burwell could well be right for a variety of reasons. But federalism arguments are not among them. You cannot defend a net expansion of federal power by appealing to legal doctrines intended to protect states against federal intrusion.
UPDATE: I should clarify one point that is perhaps only implicit in Part II of my post. Since the relevant part of the ACA is not a statute that offers conditional funds to state governments, but only private citizens, none of the rules laid out in conditional spending cases such as South Dakota v. Dole (1987) apply to it under current Supreme Court precedent, except (of course) the one stating that conditions can't violate any other part of the Constitution. Thus, Dole's clear statement requirement, which Michael Dorf argues supports the government's position in King, does not apply. As the Court stated in Dole, the rules it lays out in that case apply to conditions on "the States' receipt of federal funds." Here, the federal funds in question would go to individuals, not states.
In the case of Dole's "coercion" standard, its lack of applicability to cases not involving federal grants to state governments was made clear in New York v. United States, as discussed above.
UPDATE #2: NYU law professor Rick Hills responds to this post here. I originally posted my reply to his insightful critique as an update to this post. But I instead decided that it deserves a separate post of its own, which can be found here.