American Journal of Law & Medicine

Cost-Benefit Federalism: Reconciling Collective Action Federalism and Libertarian Federalism in the Obamacare Litigation and Beyond


The lawsuits challenging Obamacare's (1) individual mandate (2) have exposed a rift in federalism theory. On one side of the divide is a view that the national government ought to intervene--and ought to be constitutionally permitted to intervene--whenever the states are "separately incompetent" (3) to regulate a particular subject. (4) According to this view, the primary purpose of the Constitution's enumeration of national powers is to authorize Congress to fix collective action problems among the states. (5) Borrowing from Robert Cooter and Neil Siegel's article of the same name, I refer to this view as "collective action federalism." (6) On the other side of the divide is a view that federalism exists for reasons other than efficiency of regulation and particularly that the Founders created the federal structure for the protection of individual liberty. (7) According to this view, there is inherent value to state power that ought to be preserved against national encroachments. (8) I refer to this view as "libertarian federalism." In the Obamacare litigation, believers in collective action federalism generally support the individual mandate while believers in libertarian federalism generally oppose it. (9)

This Article presents a standard cost-benefit theory to bridge the gap--to reconcile the two competing theories of federalism. The cost-benefit theory is premised on two basic views. First, federalism exists both to promote regulatory efficiency and to protect individual liberty. That is, collective action federalism and libertarian federalism both rest on sound foundations. Second, regulatory efficiency always counsels in favor of national authority while individual liberty always counsels in favor of state authority. Either collective action federalism or libertarian federalism, if followed to its natural conclusion, would do away with federalism altogether. Assuming, then, that federalism is worth preserving and that both views rest on good foundations, the Supreme Court ought not to adopt one view to the exclusion of the other.

Fortunately, it is quite possible to follow both views simultaneously by optimizing the balance between the two. Under cost-benefit federalism, the inquiry ought to weigh the efficiency losses of state action against the liberty losses of national action and ought to choose the approach that maximizes the value--the benefits minus the costs--in the distribution of governmental authority. That is, if the libertarian costs of federalization outweigh its efficiency benefits, then Congress ought to leave regulation to the states. But if the efficiency benefits outweigh the libertarian costs, then Congress ought to regulate. Unlike the collective action and libertarian theories, the cost-benefit approach allows for some generalizable distinctions in the kinds of regimes that should fall to Congress and the kinds that should fall to the states. Cost-benefit federalism does not argue monotonically for national or state control. Indeed, the cost-benefit theory suggests a federalism line that roughly tracks the current doctrinal distinction: the economic/noneconomic distinction in the Supreme Court's Commerce Clause cases. (10)

That said, the simple account of the cost-benefit theory is too simple, for two reasons. First, not all regulation is monopolistically state or national; many regulatory regimes involve both levels of government. (11) But while national action can preserve state advantages, state action has a much harder time capturing national advantages. (12) Congress can (and frequently does) write national legislation that preserves states' ability to protect individual liberty (through either "cooperative federalism" programs or narrowly preemptive national laws), but states rarely surmount their collective action problems to engage in more-efficient coordinated efforts. (13) It might be quite rational, therefore, for courts and commentators to take a less skeptical view of national claims to action than of state claims to action. Second, the notion that the best approach to constitutional federalism would optimize the balance between regulatory efficiency and individual liberty elides an institutional competence problem. The courts are not institutionally capable of calculating and enforcing the optimal federalist balance, but neither is Congress. Information about the costs and benefits of regulation simply is unavailable. Optimality is thus a theoretical first-best answer to federalism rather than a judicially or legislatively administrable test. Nevertheless, the legislature's greater democratic legitimacy gives it a constitutionally relevant claim to supremacy on the federalism question. Congress is structurally better suited than the courts to gather the information that is available and to make legitimate decisions based on that information. The courts therefore should and do defer to Congress's rational decisions. (14) As long as Congress has not chosen a federalist balance that skews to an extreme of efficiency at the cost of liberty (and the states have not asserted a power that skews to an extreme of liberty at the cost of efficiency), the courts should not intervene.

What about Obamacare? Under the cost-benefit approach I outline here, the theoretical test for Obamacare's federalism is whether the libertarian costs of increasing the national government's control over individual health insurance purchases will outweigh the efficiency costs of the status quo ante's level of state control. Notably, that question is more complicated than the litigants and the courts have made it seem. The analysis ought to incorporate a fuller understanding of the balance that Congress actually struck in the statute as well as a fuller understanding of the balance that existed before Obamacare's passage. We ought not to pretend, as many of the litigants and commentators have, (15) that Obamacare represents a full national takeover of health insurance regulation or that the pre-Obamacare world was one of full state control. Of course, even with that more nuanced understanding, it will be impossible to give any precise estimate to Obamacare's true federalist value. Both the statute's efficiency gains and its liberty costs are impossible to calculate because so many of the costs and benefits are insusceptible to standardized measurement. In practice, then, the question of Obamacare's federalism ought to be whether the balance that Congress struck between efficiency and liberty is irrational. If not, then the law ought to be upheld.

This Article argues that most commentators have exaggerated all three of the relevant issues with Obamacare: its efficiency gains, its liberty costs, and its departure from the status quo ante's federalist balance. The collective action problem with state insurance regulation is not as bad as scholars of collective action federalism have argued; the liberty implications of the individual mandate are not as extreme as scholars of libertarian federalism have argued; and the shift from state to national power is not as significant as the litigants and courts have argued. Although I do not make the strong claim that Obamacare reaches the optimal balance between regulatory efficiency and individual liberty, I do make the weaker doctrinal claim that Obamacare strikes an eminently rational federalist balance, which deserves judicial deference.

This Article proceeds as follows. Part II makes the case that both collective action federalism and libertarian federalism rest on sound foundations but that either theory taken alone would argue against federalism--in favor of either full national or full state authority. That Part then argues that the first-best theory of federalism would seek the optimal balance between these competing visions. Part II fleshes out the two complicating factors in the simple version of cost-benefit federalism: the frequent hybrid state-national strategies that Congress pursues and the impossibility of discovering optimality with precision. Part III then makes the case for judicial deference to Congress's federalism. Part IV turns to Obamacare and argues that the literature and the litigation alike have exaggerated the statute's relevant federalism implications, including its efficiency gains, its liberty costs, and its departure from the status quo ante. Part IV also argues that Obamacare strikes a rational balance between state and national power that ought to be preserved.


Throughout the Obamacare litigation, two competing visions of federalism have vied for supremacy. According to collective action federalism, Congress should be empowered to address any problem that the states are separately incompetent to address. (16) If a regulatory regime suffers from interstate externalities (like adverse selection and cost-shifting in health insurance markets, according to the advocates of this theory), then Congress should be allowed to intervene. (17) Under the competing vision of federalism, libertarian federalism, the states should retain primary regulatory authority because state power better preserves individual liberty. (18) Especially if a regulatory regime imposes on individual autonomy in some new or troubling way (as does Obamacare's individual mandate, according to the advocates of this theory), each state should remain free to reject the imposition. Although both of these visions of federalism rest on legitimate theoretical foundations, each vision taken alone would argue for the end of federalism. Collective action federalism argues monotonically for national control while libertarian federalism argues monotonically for state control. In order to capture the foundational correctness of both views, federalism doctrine ought to seek a balance between the two visions--between the efficiency of national regulation and the liberty of state regulation.

This Part will demonstrate that the national government is better situated, from a purely structural perspective, to capture regulatory efficiency while the state governments are better situated, from the same purely structural perspective, to preserve individual liberty--and that these structural capacities apply to every regulatory regime. It will then make the case that efficiency and liberty should serve as mutual constraints for one another and that the goal of a federal (rather than either a purely central or a purely diffuse) system of government should be to optimize the balance between the two values.


According to collective action federalism, Congress should be empowered to address any regulatory problem that the states are "separately incompetent" to solve. (19) In elaborating this general theory of national power, Cooter and Siegel articulate an "internalization principle" (20) for determining which level of government should be authorized to address a problem. In their view, the Constitution should "assign power to the smallest unit of government that internalizes the effects of its exercise." (21) As we shall see, however, the scope of that principle depends on whether Cooter and Siegel mean "the smallest unit of government that internalizes most of the effects of its exercise" or "the smallest unit of government that internalizes all of the effects of its exercise." If they intend the former (as they seem to), then we need a theory for what constitutes most. How significant does an interstate externality need to be before Congress may constitutionally intervene? Cooter and Siegel provide no answer. If they intend the latter (which they seem not to), then the theory argues for nationalization of all social and economic policy. All such policy suffers from some interstate externalization because the citizenry is freely mobile.

I will describe and model the problem of interstate externalities (the collective action problem that defines collective action federalism) and will explain why the theory needs some kind of limiting principle in order to be a theory of federalism rather than a theory of nationalism. I will also provide two possible empirical limits for the theory, both of which I find legally unsatisfying.

1. Interstate Externalities

The core problem that collective action federalism seeks to address is the problem of regulatory spillovers between and among the states. To see this problem in action, I will use the canonical example of interstate externalities that justify national intervention: pollution. If Massachusetts, for example, undersupplies pollution abatement, some of the effects of that choice will travel to other states because some of the excess pollution will travel downstream to New Hampshire or Rhode Island (and beyond). Acting alone, therefore, Massachusetts's incentive to punish polluters is incomplete.

We can express the problem with a simple model. (22) Ideally, a government will regulate up to the point that the social benefits of the regulation (cleaner air and water) equal the social costs of the regulation (enactment and enforcement costs as well as lost economic productivity from pollution reductions). The constraint, then, is that a regulation should pass whenever:

[SB.sub.r] [greater than or equal to] [SC.sub.r]

If some of the regulation's benefits (some of the cleaner air and water) will travel to a different governmental system (a neighboring state), then voters and legislators will rationally spend less than the optimal amount of social cost (like time, effort, money, opportunity cost, and political capital) on regulating. The social benefit that they can expect to capture from the regulation will be less than the full social benefit that the regulation produces. The externalized benefits affect the rational regulatory constraint as follows, where [e.sub.1] is the externalized social benefit of regulating:

[SB.sub.r] - [e.sub.1] [greater than or equal to] [SC.sub.r]

When this kind of externality occurs, each state acting alone systematically underproduces socially valuable regulation.

Furthermore, the social costs of regulating might also increase due to interstate externalities. Imagine, for example, that Massachusetts imposes strict pollution controls on its industrial manufacturers while New Hampshire chooses a laxer regulatory environment. The regulated entities will move from Massachusetts to New Hampshire in order to avoid the higher cost of complying with Massachusetts's pollution abatement laws, causing Massachusetts to suffer an additional social cost of regulating: lost jobs and tax revenues. Massachusetts, then, has an incentive to regulate only until:

[SB.sub.r] - [e.sub.1] [greater than or equal to] [SC.sub.r] + [e.sub.2]


[SB.sub.r] - [e.sub.1] - [e.sub.2] [greater than or equal to] [SC.sub.r]

We can think of the [e.sub.2] as a standard first-mover problem or multilateral prisoners' dilemma, where each state wants to avoid being the first actor to impose pollution abatement regulations because the first-acting state will lose business to later-acting states.

To put the constraint in slightly different terms, a state's willingness to pay for a regulation should equal the total social benefit that the regulation will produce (WTP = [SB.sub.r]). Externalities cause the state to experience something less than the full social benefit, decreasing the state's willingness to pay accordingly (WTP = [SB.sub.r] - [e.sub.1] - [e.sub.2] - [e.sub.3] ... - en). Externalization therefore pushes states systematically towards underproduction of regulation. (23)

To make this picture more concrete, imagine that the total social benefit Massachusetts could produce through pollution abatement laws is 100 utils. (24) In other words, imagine that polluters in Massachusetts are producing 100 utils worth of harm that could be reduced through regulatory intervention. Ideally, the people of Massachusetts would be willing to spend up to 100 utils of their own regulatory inputs to address the problem and reduce the harm. That is, Massachusetts should be willing to pass any pollution abatement law that is likely to be effective and that has a total social cost of 100 utils or less to enact and enforce.

Unfortunately, there are two externalization problems that will lead Massachusetts, quite rationally, to set its willingness to pay at less than 100 utils. First, some of the harm from pollution produced in Massachusetts occurs in other states, such that Massachusetts cannot expect to capture the full 100-util benefit of eliminating the pollution produced within its borders. Some of that benefit will accrue in the states to which Massachusetts's air and water travel. Imagine, then, that the out-of-state portion of the pollution abatement law's benefits--the [e.sub.1] of this scenario--is twenty-five utils. (In other words, imagine that twenty-five percent of the pollution generated in Massachusetts travels to and harms other states, such that twenty-five percent of the benefits from eliminating that pollution will accrue in other states.) That externality reduces Massachusetts's willingness to pay for regulation from 100 to seventy-five. The state therefore has a less-than-optimal incentive to produce pollution abatement laws.

Second, individuals' and corporations' freedom of travel among the states might heighten Massachusetts's social cost of enactment, particularly if its close neighbors refuse to regulate polluters. Imagine, then, that the cost to Massachusetts in industrial business lost to competing states--the [e.sub.2] in this scenario--is ten utils. That externality alone would reduce Massachusetts's willingness to pay for a 100util reduction in pollution to ninety utils, or if combined with the imagined [e.sub.1], it would reduce the willingness to pay to sixty-five utils. The [e.sub.2] therefore also decreases Massachusetts's incentive to produce pollution abatement laws, causing it to fall below the social optimum.

In short, the externalization of regulatory costs and benefits from one state to the next distorts the states' incentives to regulate. In the case of pollution abatement and in any other regime in which [e.sub.1] and [e.sub.2] both have positive value, each state is underincentivized to regulate relative to the social optimum. (25)

What about Congress's incentives? Assuming that the costs associated with both [e.sub.1] and [e.sub.2] are entirely internal to the United States, Congress's willingness to pay for the same pollution abatement in Massachusetts will be the full 100 utils. Congress captures all of the benefits to every state in the union of reducing pollution that originates in Massachusetts, and congressional policy-making can avoid the first-mover problem by setting a policy for all states at one time. In other words, assuming that all benefits and costs of regulating pollution in Massachusetts are internal to the United States, national decision-making will not be distorted; el and [e.sub.2] will be zero, and the regulatory incentive will be optimal.

Of course, the assumption is false, at least with respect to pollution. In this regime, both [e.sub.1] and [e.sub.2] have international as well as interstate implications, meaning that Congress will suffer non-zero externalities just as Massachusetts would. With respect to el, some of the pollution produced in Massachusetts will travel to Canada, and some of it will contribute to global environmental degradations like climate change. With respect to [e.sub.2], national pollution restrictions might cause some manufacturers to set up shop overseas in less-restrictive developing nations like China, increasing the social cost to the United States of regulating polluters today relative to the cost that later-moving nations will experience from the same regulations.

Nevertheless, both of the externalities will be less impactful on the central government's willingness to pay than on the state governments' willingness to pay. As for [e.sub.1], the magnitude of the externality relative to the total social benefit will be smaller. This point is quite easy to see in the hypothetical case: a higher percentage of the pollution produced in Massachusetts will stay in the United States than will stay in Massachusetts. If the [e.sub.1] for Massachusetts is twenty-five utils, then the [e.sub.1] for Congress must be less than twenty-five utils because not all of the externalized harm from Massachusetts's pollution will reach Canada. Some of it will impact New Hampshire and Maine on its way to Canada, and some of it will go south instead, impacting Rhode Island and the states below. In the abstract, this point is mathematically true for all regulatory regimes imaginable. Simply by virtue of its size, the United States government must internalize a higher percentage of the harms it produces than any state government. As a ratio of its total geographic area, Massachusetts has more borderland than the United States, providing more opportunities for harms to sneak out of the territory. (26) That is, a harm produced in the dead center of Massachusetts has to travel less distance to reach another state than a harm produced in the dead center of the United States has to travel to reach another nation, and a harm produced at any random point in Massachusetts has to travel a smaller average distance to reach another state than a harm produced at any random point in the United States has to travel, on average, to reach another nation. It is therefore mathematically certain that a larger government will internalize a higher percentage of the physical harms it produces than a smaller government, regardless of how many governments exist globally.

With respect to [e.sub.2], the point is equally easy to see within the constraints of the hypothetical, and the same mathematical certainty applies in the abstract, at least within the constraint of a finite global system (i. …

Log in to your account to read this article – and millions more.