Will Uncooperative Federalism Survive NFIB?By Abigail R. Moncrieff—Associate Professor Boston University School of Law—and Jonathan Dinerstein
76 Mont. L. Rev. 75
Abigail R. Moncrieff & Jonathan Dinerstein*
Surveying the political landscape, one might conclude that the dominant feature of modern federalism is the one that Jessica Bulman-Pozen and Heather Gerken described in their aptly named article of five years ago, “Uncooperative Federalism.” Our hyper-partisan era is one of matter-of-course state resistance to national programs, with Republican governors pushing back on President Obama’s directives.
At first blush, the doctrinal landscape seems to support the political tide. In National Federation of Independent Business v. Sebelius (“NFIB”), the Supreme Court openly facilitated states’ resistance to Obamacare’s Medicaid expansion, holding that the statute may not require the states to participate. According to the Supreme Court, the states must be free to refuse the national government’s instruction to expand their Medicaid programs. This holding looks like a tremendous victory for the “uncooperative federalism” model, ensuring that states can resist even the most significant of the national government’s priorities without threatening their participation in existing cooperative federalism schemes.
But on closer consideration, we think NFIB’s Medicaid holding is more likely to harm than help the era of uncooperative federalism—and might harm federalism generally. We predict that, in the long run, the holding is likely to cause more nationalization of policy decisions and policy administration. That might seem counterintuitive given that NFIB looks like an aggressively pro-federalism and pro-state holding. Let us explain.
There’s not much law left that is purely states’ jurisdiction. The national government has used its expansive spending power to touch every arena of modern law and policy, and its money comes with strings attached. Through the power of the purse and the purse’s many strings, the national government today influences how states exercise all of their traditional “police powers,” including the erstwhile local realms of education, health, safety, and welfare. States today exert their influence primarily through efforts to shape these “cooperative federalism” programs, whether they do so cooperatively or uncooperatively, and the end result is that a startling amount of policy administration happens in negotiation between national and state governments. That is why we live in this era of uncooperative federalism. When philosophical disagreements arise between the controlling faction in a state and the one in the national government, there are lots of opportunities for the state to assert itself by defying national directives.
But what happens when a state, whether intentionally in the course of active un-cooperation or innocently in the course of cooperative administration, violates a statutory condition for the receipt of national funds? Can individuals who are harmed by the violation force state compliance? Can the national government enforce its statute against the state? Is the state’s violation even illegal? As it turns out, the answers to these questions are complex and in flux, largely thanks to the Supreme Court’s ambivalence on the deeper questions of federalism. Given the current set of doctrinal answers from two sleeper cases, Gonzaga University v. Doeand Douglas v. Independent Living Center of Southern California, the answer seems to be that the state has done nothing wrong when it violates a statutory spending condition, but the national government has done something wrong if it acquiesces in the violation. The burden is thus on the national government to enforce its statutory spending conditions, and individuals can sue the national government (but not the state government) if the national government fails to enforce its statutes. But after the Court’s holding in NFIB, such enforcement will be harder than it used to be (and might be occasionally impossible) in any cooperative federalism program that uses money as the enforcement tool.
Because of this odd set of difficulties that the Supreme Court has created, the national government might need to switch to enforcement tools that do not raise the constitutional difficulty that NFIB identified. Within the realm of cooperative federalism, only one other tool seems to exist: conditional preemption. Other than money, the only enforcement tool that the national government sometimes uses in cooperative federalism programs is a crowd-out of state administration when states fail in their enactments of national directives—like the national fallback option in Obamacare’s exchange provision. After NFIB, it will be safer and easier for Congress to use conditional preemption than financial penalties when enacting or amending cooperative federalism programs. The result, however, will be to substitute nationalization for uncooperative federalism. If a state takes its resistance so far as to provoke a national crowd-out, it will lose the power of the servant and any benefits that come from the federal structure.
We do not necessarily mean to criticize NFIB with this prediction. Indeed, we do not necessarily think that a shift to conditional preemption would be bad for public policy in the United States. Our greater concern is with the sequence of holdings in Gonzaga, Douglas,New York v. United States, Printz v. United States, Dole, and NFIB and the uncomfortable position that the Supreme Court has created for both the state and national governments. All told, the Supreme Court’s federalism jurisprudence makes little sense and might ultimately prove self-defeating given the Supreme Court’s stated justifications for its holdings.
II. THE SUPREME COURT’S PRE-NFIB FEDERALISM JURISPRUDENCE
In 1992, the Supreme Court began a renewed interest in federalism limitations—a doctrinal adoption of President Reagan’s “New Federalism”—with the Justices starting to enforce constraints on national power that had lain doctrinally dormant since the New Deal. The Court began with an anti-commandeering doctrine in New York v. United States, progressed to a weak constraint on the commerce power in United States v. Lopez, reinforced the anti-commandeering principle in Printz v. United States,and reinforced the Commerce Clause constraint in United States v. Morrison.
Focusing only on these holdings, the New Federalism revolution seemed to issue a relatively clear set of rules for Congress. First, if the national government wants to regulate something directly, it has to enforce the law itself. It cannot require state actors to implement national policy. Functionally, this rule requires that Congress put its money where its mouth is, ponying up national funds to enforce national regulations, and it requires the national government to be clear with constituents about whose policy is whose, flashing FBI badges rather than local police badges when it arrests citizens for violating national policies. Second, these cases hold that there are some things that Congress may not regulate through this model of purely national, direct regulation. Non-economic and primarily local behaviors, like carrying guns near schools, abusing women, and refusing to buy health insurance, are beyond Congress’s direct regulatory jurisdiction. So far so good. The rules are not exactly simple, but they’re simple enough.
The pre-NFIB New Federalism holdings, however, left Congress’s Spending Power entirely intact. Most importantly, the New Deal holdings inUnited States v. Butlerand Helvering v. Davis survived the New Federalism’s onslaught without a scratch. In Butler, the Court invalidated a portion of the Agricultural Advancement Act, but in the process, it issued a critical holding for the future of American federalism: Congress’s Spending Power is a separately enumerated power, which the national government may use to pursue ends other than those otherwise granted to Congress in Article I, Section 8. In other words, Butler held that Congress could use its taxing and spending power to enhance the “general welfare” in ways other than those enumerated in the rest of the Section 8 list. But the opinion interpreted “general welfare” somewhat narrowly, keeping the spending power relatively contained. A year later, however, came Helvering, which held that the “discretion” to decide whether a given expenditure is for the general welfare “belongs to Congress, unless the choice is clearly wrong, a display of arbitrary power, not an exercise of judgment.” This holding gave Congress essentially unlimited authority to regulate through taxing and spending. After Helvering, if Congress could get the votes to pass a spending statute, it could regulate the subject of that statute, irrespective of the Constitution’s apparent attempt to reserve un-enumerated substantive powers to the states.
Given these spending power holdings, the New Federalism limitations on Congress’s Commerce Clause authority are, functionally, requirements that the national government do some things through taxation rather than regulation. Because Congress can tax and spend for any purpose, the Lopez and Morrison limitation on national regulation of non-economic and local behaviors is a limitation of form rather than substance. Congress can still set incentives for individual non-economic behaviors, but it can do so only through its spending power.
Of course, there are real differences between taxation and regulation. After Lopez, for example, the national government can levy a tax against anyone who carries a gun near a school, which might dissuade many people from doing so, but the FBI cannot arrest and imprison those who willingly pay the tax for the privilege. Furthermore, because the Child Labor Tax Case survived the New Deal revolution, there are meaningful limitations on the heft of taxes and the means of administering them. A tax ceases to be a tax—and becomes a regulation—if the amount of the tax drastically exceeds the cost of complying with national policy, if the tax’s trigger includes a scienter requirement, or if the national government sets up an administrative structure separate from the Internal Revenue Service to determine whether an individual must pay the tax. Nevertheless, Helvering had a real and significant impact on national power; Article I, Section 8 no longer contains any substantive barriers to national action. Whatever barriers the Court creates under the Commerce Clause are rules of form rather than substance.
The next piece of the doctrinal federalism puzzle is Steward Machine Co. v. Davis and Dole, which survive NFIB, albeit with a significant injury. Steward Machine and Dole, although decided five decades apart, both held that Congress may use its expansive spending power to cajole the states into implementing national policy. The national government may offer money to the states that is conditioned on their implementation of national dictates. In Steward Machine, the Supreme Court upheld a statutory scheme under which states could relieve their citizens of national Social Security taxation only by implementing an alternative and equally good unemployment scheme. That statute was the inverse of today’s prototypical cooperative federalism program; it set a national regulatory default, which took money away from the states, but then gave the states the right to opt out of the national default by replacing the national program with equivalents of their own. In Dole, the question was whether Congress could threaten to take 5% of a state’s preexisting highway funds away from the state if it refused to raise the legal drinking age to 21. Again, that structure was not the prototypical “contract”-like cooperative federalism scheme in which the national government offers money to the states that is conditioned on their willingness to implement and enforce certain policy details. But the Supreme Court held that Congress could threaten states with a drop in their baseline funding as long as the drop was not so traumatic that the states would feel compelled to avoid it.
For our purposes, the punchline of these holdings is that Congress can use the Spending Power not only to circumvent the Lopez and Morrison restrictions on regulatory subjects but also to circumvent the New York and Printz restrictions on commandeering state governments. The federal structure of national policies that New York and Printz seemed to find distasteful is perfectly permissible as long as the national government pays the states for their acquiescence. Importantly, this holding limits the plausible justification for an anti-commandeering doctrine because it allows some obfuscation of policy responsibility. The national government can bribe state governments to become the face of a legal regime, convincing states to flash local police badges rather than FBI badges, as long as it pays the states to play that role. Notably, however, Dole contained a line of dictum that became the basis for the restriction in NFIB: “Our decisions have recognized that, in some circumstances, the financial inducement offered by Congress might be so coercive as to pass the point at which ‘pressure turns into compulsion.’” In other words, the states must retain a meaningful right to refuse this role—to refuse the acceptance of actual and apparent responsibility for national policy implementation. That requirement ensures that a state is, in fact, at least somewhat responsible for the policies it is enforcing because the state will have made a genuine choice to play the role of the national government’s policeman.
The final piece of the doctrinal federalism puzzle is the least well-known and the most puzzling. The New Federalism revolution has included two “sleeper” cases that limit the states’ responsibility for complying with national policies—even after they agree to participate in cooperative federalism schemes. In the first, Gonzaga, the Supreme Court held that 42 U.S.C. § 1983—the private right of action against state officials for deprivations of federal rights—does not provide a right of action for violations of national spending conditions unless the conditions, by their own terms, imply such a right of action. On its face, the Gonzaga opinion did not look like much of a departure from the Supreme Court’s earlier opinion in Pennhurst. Pennhurst had already noted that private rights of action were not an appropriate remedy under cooperative federalism: “In legislation enacted pursuant to the spending power, the typical remedy for state noncompliance with federally imposed conditions is not a private cause of action for noncompliance but rather action by the Federal Government to terminate funds to the State.” But Gonzaga had a significant impact on lower courts’ practices, limiting the number of spending conditions that the lower courts allowed individuals to enforce through § 1983 suits. Furthermore, Gonzaga reiterated and clarified the Supreme Court’s theory that spending conditions are not binding on states’ behavior; they are essentially contract terms between the national and state governments, which the national government has the power to invoke when deciding whether or not to give money to the states.
The other sleeper case is Douglas, which addressed the same issue of enforcing spending conditions against noncompliant states. That case has a strange and confusing procedural history, but the bottom line is that Justice Breyer, writing for the majority in what is probably a non-precedential portion of the opinion, ended up encouraging individuals to use Administrative Procedure Act (APA) challenges to compel national enforcement of spending conditions. His theory is that national agencies act arbitrarily and capriciously, in violation of their organic statutes and the APA, when they disburse money to states despite the states’ violations of statutory conditions on funding. In other words, Justice Breyer wrote that the national government is not only empowered but may be obliged, under the APA, to enforce the terms of its contracts with the states.
These last two cases create a strange situation for uncooperative federalism. Gonzaga and its predecessors were largely responsible for allowing states the flexibility they needed to resist national directives while still participating in federal programs—in other words, for allowing uncooperative federalism to exist. In many of these programs, the national government has been unwilling to use the blunt (and often perverse) instrument of withdrawing funds to bring states into line, and individual suits under § 1983 were thus, before Pennhurst and Gonzaga, the only functional mechanism for enforcement of spending conditions. Once that mechanism all but disappeared, the states became much freer to resist national directives within cooperative federalism regimes, without much risk to their funds.
Douglas , however, creates pressure for the national government to start using its enforcement tools to push back on states’ un-cooperation. Justice Breyer’s opinion literally invites individuals to sue national agencies for allowing state resistance within cooperative federalism programs and predicts a rule under APA arbitrary and capricious review that the national government must force states to make a choice between their money and their principles. That rule, if it clearly emerged from APA litigation, could have put an abrupt end to uncooperative federalism in its most extreme form—the form that Bulman-Pozen and Gerken call “civil disobedience”—by making the threat of losing national dollars much more credible than it has ever been in the past. But then there’s NFIB.
III. NFIB’SDOCTRINAL ROLE
NFIB complicates the picture more than clarifying it. In the first part of the opinion, NFIB stands by the nonsensical division that has governed Congress since Lopez or, really, since Butler: the things that Congress cannot accomplish by direct regulation, it can nevertheless accomplish through taxation. Chief Justice Roberts’s compromise to save Obamacare is the same compromise that has facilitated national regulatory power since the New Deal, notwithstanding the limitations that the constitutional framers probably had in mind and that the Court has halfheartedly tried to resurrect in the New Federalism era. Congress holds an unlimited (legal) power to tax and spend.
What was new in the NFIB majority was the Medicaid holding. For the first time, the Supreme Court gave teeth to the Dole dictum that Congress’s financial inducements might become so tempting as to “pass the point at which ‘pressure turns into compulsion.’” By a vote of 7–2, the Court held that the potential tie between the Medicaid expansion and the states’ pre-existing Medicaid funding gave the states no choice but to accept the expansion, unduly coercing them to enact national policy. In other words, the Medicaid expansion needed to be meaningfully voluntary for the states, and by threatening states with loss of their pre-Obamacare Medicaid funding, the national government would make it too hard for states to refuse the expansion. The Court therefore held that the Obama Administration may not withdraw pre-existing Medicaid funds from states that refuse to expand.
This holding might seem sensible enough—if only it were a little less far reaching and a lot more robustly theorized. The problem with the Court’s analysis is that the move Congress made with Obamacare’s Medicaid expansion is one that it makes all the time: it changed the statutory conditions for an existing national grant. Furthermore, even assuming the Medicaid expansion was a new program rather than a mere amendment to an existing one, the threat of dropping states below their preexisting baseline of national funding if they refuse to enact a new national policy was not new; that was what the statute in Dole had done too. So what, if anything, was unique about the Medicaid expansion? If the answer is nothing, then a lot of current statutory spending conditions might be constitutionally unenforceable.
Chief Justice Roberts’s opinion, which spoke for three Justices but represented the governing plurality, gave the following reasons for finding the Medicaid expansion unconstitutional even though the financial punishment at issue in Dole was not. First, he reasoned that the expansion was so different from pre-Obamacare Medicaid that it could not be considered part of the preexisting program. So far, this point fails to distinguish Dole—Chief Justice Roberts would have to concede that, as a policy matter, expansion Medicaid was at least as closely related to pre-expansion Medicaid as the drinking age was to highway funding. Furthermore, Chief Justice Roberts gave very few meaningful criteria for distinguishing an amendment to an existing program from what he claims happened here: the creation of a new program under the same statute as an existing program. What he said was the “Medicaid expansion . . . accomplishes a shift in kind, not merely degree” by abandoning eligibility categories in favor of a blanket eligibility threshold of 133% of the federal poverty line. He compared that change to “[p]revious amendments to Medicaid eligibility[, which had] merely altered and expanded the boundaries of [the existing eligibility] categories.” What is not clear is why the abandonment of categorical eligibility is a change of kind rather than degree while additions of new eligibility categories, such as pregnant women (as Justice Ginsburg noted in dissent), is a change of degree rather than kind. Would Chief Justice Roberts’s analysis have gone out the window if Congress had instead amended the Medicaid Act to say that childless adults living in or near poverty constituted a new category of mandatory eligibility?
Chief Justice Roberts went on to say:
Under the Affordable Care Act, Medicaid is transformed into a program to meet the health care needs of the entire nonelderly population with income below 133 percent of the poverty level. It is no longer a program to care for the neediest among us, but rather an element of a comprehensive national plan to provide universal health insurance coverage.
But there are serious problems with this argument. Most importantly, many of Obamacare’s drafters probably believe that “the entire nonelderly population with income below 133 percent of the poverty level” is, today, a perfectly apt definition for “the neediest among us.” Also importantly, many members of the 1965 Congress (which created Medicaid) believed they were writing “an element of a comprehensive national plan to provide universal health insurance coverage.” They didn’t succeed, but it’s not at all clear that they weren’t trying. If these are the only grounds for treating the expansion as a new program, then that treatment seems wrong; the Medicaid expansion looks like an effort to amend the prior program so that it can better achieve its original goals. In the end, Chief Justice Roberts’s first point fails to distinguish Dole and provides tenuous (if not simply incorrect) arguments in support of the notion that NFIB presents a Dole case rather than a standard case of an amendment to an existing conditional grant.
The Chief Justice’s next attempt to demonstrate that expansion Medicaid is different from pre-expansion Medicaid is more successful. He noted that Congress structured coverage differently for the expansion population than for the preexisting population, offering a higher federal financial participation percentage and mandating lesser benefits for the expansion group. These differences are more concrete reasons to believe that Obamacare Medicaid is a different program from—and not a mere amendment to—pre-Obamacare Medicaid. Still, though, the point fails to distinguish Dole, which imposed a brand new requirement on the states, largely unrelated to highway funding, at threat of losing preexisting funds.
Chief Justice Roberts’s second major argument is the one that mattered—the one that successfully distinguished Dole: Medicaid puts too much money at stake. In Dole, the punishment for refusing to implement the national drinking age was 5% of highway funding, which amounted to .05% of South Dakota’s budget at the time. The punishment for refusing to expand Medicaid, by contrast, was (potentially) the entire federal portions of the state’s Medicaid budget, which, as Chief Justice Roberts noted, “accounts for over 10 percent of a State’s overall budget . . . .” To Chief Justice Roberts, this level of financial inducement “is a gun to the head.” He concluded that such a threat “is economic dragooning that leaves States with no real option but to acquiesce in the Medicaid expansion.”
The problem with this argument is that it puts every single Medicaid condition—or at least every single condition that was not in the statute the day the state joined the program—at risk of unenforceability. Withdrawal of the national portion of a state’s Medicaid budget is the only enforcement tool that the Medicaid Administrator has to ensure compliance with Medicaid Act conditions. If the amount of money is the problem, then all Medicaid enforcement is unconstitutional.
The most charitable (and probably correct) understanding of the Chief Justice’s opinion is that both features need to be present for enforcement to be unconstitutional: the national government must impose (or threaten to impose) (1) an enormous financial penalty for (2) refusal to implement a brand new policy (as distinct from an amendment to an existing policy). If that is the right understanding of the NFIB plurality’s holding, then perhaps the other provisions of the Medicaid Act—even the conditions that Congress added to the statute after the states opted in—are safe from constitutional attack despite their newness to the statute and despite the potential financial gravity of their enforcement. Unfortunately, however, the Chief Justice’s rules for what constitutes a brand new program are so unclear that it will be hard for future courts to figure out which post-1965 statutory amendments are constitutionally enforceable.
The dissenters’ opinion—which added two necessary and two additional votes to the holding that the Medicaid expansion was unconstitutional—engaged in an importantly different analysis. First, the dissenters thought it mattered that a given state’s citizens would have to pay for the Medicaid expansion whether or not they agreed to join. Because the national portion of a state’s Medicaid program comes from the nation as a whole, not just the citizens of the participating state, national grants to states create a collective action problem that (eventually) coerces all states to get with the program. As each state joins, all states’ taxes rise, until the marginal cost of joining feels negligible.
If this argument had been decisive, then Medicaid should have been unconstitutional in 1965, and all of Medicaid should be unconstitutional today. When Arizona became the last state to join the program in 1982, its citizens were already paying for Medicaid programs throughout the rest of the country. According to the dissent’s logic, Arizonans had not meaningfully avoided paying for the national policy; they had only avoided administering that policy within their borders.
Fortunately, the dissenters do not seem to have taken this argument particularly seriously. They immediately went on to say: “Whether federal spending legislation crosses the line from enticement to coercion is often difficult to determine, and courts should not conclude that legislation is unconstitutional on this ground unless the coercive nature of an offer is unmistakably clear.” Thereafter, the dissent’s analysis tracked Chief Justice Roberts’s argument that the heft of the potential penalty was what mattered in the case. The feature of the Medicaid expansion that made the difference for the dissenters was the size and consequence of the financial penalty for the failure to expand. Medicaid is too significant a source of states’ revenues and expenditures for the national government to threaten to take it away.
The dissenters’ analysis, however, did not seem to include Chief Justice Roberts’s point that the expansion was a different program from pre-Obamacare Medicaid. They seemed to believe that any threat of withdrawing national Medicaid funding would be an unconstitutionally coercive means of enforcing compliance with national directives, whether those directives were part of the original Medicaid program or not.
As noted above, that view, if it ever carried a majority on the Court, would make the vast majority of Medicaid Act conditions entirely impossible to enforce. There are few Medicaid conditions that individuals can enforce through § 1983 actions after Gonzaga, and the joint dissent’s constitutional analysis would forbid the Medicaid Administrator from using her one and only enforcement tool—withdrawal of funds—to enforce conditions, no matter how many APA suits were filed against the agency after Douglas.
IV. NFIB’S PRACTICAL CONSEQUENCE
The final vote in NFIB’s Medicaid holding is 4–3–2, with seven justices finding the mandatory Medicaid expansion unconstitutional. Unfortunately, however, the divided opinion leaves tremendous uncertainty as to what made the expansion unconstitutional. The size of Medicaid and its importance to state budgets were undoubtedly key. Does that mean that the holding is unique to Medicaid, which is the biggest or second-biggest line item in every state’s budget? Or will other programs look equally big and important when states bring challenges in other cooperative federalism contexts? For future challenges, will it matter whether a condition that is being enforced is a part of the preexisting program or is, instead, a different kind of program that has been added to the same statute? How will courts know which new conditions are new programs and which are amendments to existing programs? In short, the NFIB opinion is troublingly confused and confusing.
One thing, however, is clear: for the first time in American history, the Supreme Court has enforced a ceiling on Congress’s power to financially induce state administration of national policy. And the certainty that this Court—including two of its liberal appointees—is sometimes willing to enforce such a ceiling, combined with the uncertainty as to the ceiling’s breadth and contours, will provide the states with an unusually large bargaining chip in uncooperative federalism negotiations.
This point is key. Even if the Supreme Court never again enforced its new constraint on spending conditions, the NFIB holding, particularly combined with Justice Breyer’s Douglas suggestion to enforce spending conditions through APA litigation, could dramatically affect cooperative federalism programs. Most federal policy administration does not happen through statutes or courts; it happens in negotiation between national and state agencies, contained entirely within the Fourth Branch. When a state wants to change its Medicaid program, it approaches the national Centers for Medicare and Medicaid Services (CMS) with a State Plan Amendment and seeks CMS’s blessing to move forward, with the state’s matching funds secure. Before NFIB, the state agent in that context knew that outright noncompliance with Medicaid Act conditions could result (even if it never actually had resulted) in a severe financial penalty. CMS had the threat of its 42 U.S.C. § 1396c power to withdraw national money from all or part of the state’s Medicaid program, and that possibility kept negotiations relatively civil. After NFIB, however, a recalcitrant (or cash-strapped and desperate) state agent could dare her national counterparts to try withdrawal of funding, threatening constitutional litigation if national bureaucrats called the bluff. Imagine, for example, a state Medicaid administrator who wants to cause trouble for the sitting President, or a state administrator that simply does not have enough money to pay for rising Medicaid costs and wants to slash the program in contravention of Medicaid Act directives. Such an administrator might be willing to push a lot harder against CMS in the shadow of NFIB than she ever would have dared to push before the shadow was cast. Meanwhile, CMS knows that if it does not do something to enforce Medicaid Act requirements, it is likely to face an APA suit for arbitrarily and capriciously approving a State Plan Amendment that violates the national statute. But what can it do? Its only enforcement tool might be unconstitutional.
In short, given the uncertain basis for NFIB’s holding, national bureaucrats will have lost a significant amount of their bargaining power in federal administrative negotiations––at least in those programs that rely on money for inducement and enforcement.
V. NON-FINANCIAL ENFORCEMENT?
So far, our analysis seems to indicate that NFIB has increased rather than decreased state power in cooperative federalism programs, perhaps allowing the states even greater leverage to behave uncooperatively. They now have a powerful threat of their own that they can use to counterbalance the national agency’s threat of withdrawing funding: a constitutional challenge to agency action. But remember that the Supreme Court, in the same term as NFIB, told private litigants that they could and should force national agencies to enforce their statutes. In Douglas, Justice Breyer invited APA litigation against CMS for its approval of a state plan that violated a provision of the Medicaid Act. If private stakeholders accept Justice Breyer’s invitation, then national agencies will not be able to surrender under the state bureaucrats’ threat of constitutional litigation. They will have to show some serious attempt to force state compliance with statutory conditions, and they will not be allowed to permit the kind of uncooperative federalism that arises from political disagreement.
This situation could devolve quickly. A serious threat of financial enforcement from national agencies will provoke constitutional litigation under a standard that defies simple line-drawing, but the national agencies’ failure to enforce will provoke administrative law litigation under a standard that invites judicial intermeddling. The question, then, is whether Congress could give national agencies some other enforcement tool that would avoid NFIB’s constitutional morass. Is there another tool that Congress could give CMS that would enable it to avoid both constitutional litigation from states and APA litigation from private stakeholders?
Indeed, Congress seems to have exactly two mechanisms for encouraging state implementation of national policy: conditioning national money and conditionally pre-empting state action with national regulation (crowding-out state administration). Some cooperative programs use a mix of these two strategies, but these two mechanisms appear to be the only two in existence today. Both financial incentives and conditional preemption allow state legislators and residents to make a choice regarding their involvement in national policy, thereby complying with the anti-commandeering constraint, but the programs that rely on financial penalties facilitate greater policy discretion for the states than those that allow conditional preemption.
Typically, if a state chooses to forgo national funding in a cooperative federalism program, it can entirely avoid the implementation of the national policy within its borders. Arizona’s refusal to implement Medicaid from 1965–1982, for example, meant that there was no Medicaid program operating in the state during that time, which might have been Arizonans’ goal. Similarly, Wyoming’s willingness to suffer the 5% reduction in highway funding allowed it to maintain a drinking age of 18 until 1988, when it became the last state to join the National Drinking Age Act of 1984. Puerto Rico still enforces a drinking age of 18, choosing to forgo 10% of its highway funding so that it can keep its preferred policy.
By contrast, a conditional preemption mechanism does not allow the state or its residents to avoid the existence of a national program in the state. When the state refuses to implement the national program, the national government steps into the state’s borders to administer the program itself. Conditional preemption thus allows the state to avoid only the responsibility for implementing a program. The state government does not get drafted into the national government’s service, but the state’s citizens do not avoid the imposition of national policy.
For example, consider the health insurance exchanges under Obamacare, which use a conditional preemption mechanism. The statute requires states to establish compliant exchanges by a set date, but it provides that, if the state refuses or fails by the deadline, then the national government will establish an exchange on the state’s behalf. The thirty-four states that refused to establish their own exchanges, thus, have not kept exchanges out of their borders. They have escaped the responsibility for running the exchanges and have perhaps maintained a clearer line of responsibility than the states that are running exchanges of their own, but they have not shielded their citizens from the national policy.
For another example, consider the Resource Conservation and Recovery Act (RCRA). RCRA features a conditional preemption enforcement mechanism rather than conditional funding. Congress sought to regulate solid and hazardous waste through three methods: national agency enforcement through the Environmental Protection Agency (EPA), regulation through approved state programs, and citizen suit enforcement. States can regulate solid and hazardous waste under RCRA only if EPA approves their plans. States that want a greater degree of control over waste management have an incentive to create an EPA-approved plan so that they can implement Congress’s policy themselves rather than triggering the conditional preemption provision by which EPA takes over full implementation responsibility. But, under this scheme, no state can avoid the implementation of some nationally-regulated waste management plan within its borders. Its choices are between a cooperative program and a national program. It cannot avoid the policy in its entirety the way that a state could under a standard conditional grants program.
As noted above, the NFIB ruling strains Congress’s ability to condition national funding on states’ implementation of regulatory policies. Without clear rules for when conditional funding “passes the point at which ‘pressure turns into compulsion,’” the national government might be hesitant to use money as the primary enforcement mechanism in future cooperative programs, and Congress might become tempted to turn to conditional preemption even in existing cooperative programs. In short, the holding might have a profound effect on the national government’s choices about how to manage cooperative federalism.
That said, national agencies have occasionally been hesitant to use their conditional preemption authority just as they have usually been hesitant to use financial penalties. Currently, for example, EPA has authority to take direct regulatory control under the Clean Water Act (CWA) if states fail to meet national standards. The EPA, however, has a lot of practical constraints that prevent it from using this authority, including the lack of necessary personnel to take over CWA enforcement in a state. Indeed, in general, state agencies are better-equipped than national agencies for the day-to-day administration of these complex policies. Historically, national agencies have capitulated to state demands rather than developing the infrastructure required to take over administrative responsibilities from the states.
If, however, our prediction about Douglas’s effect comes to fruition, requiring the national government to enforce spending conditions, and if nationalization becomes a more common enforcement mechanism in light of NFIB, then the national government will overcome its practical obstacles out of necessity. Congress and the national agencies will predict the need for conditional preemption in new cooperative federalism schemes, and they will find the resources they need to implement national programs (or will decrease national regulation to conserve resources). Consider, for example, Medicare. The entirely national Medicare program demonstrates that practical barriers to nationalization in Medicaid are far from insurmountable and could, in fact, disappear if duly anticipated. Given time, the same may be true across the catalogue of cooperative federalism programs. Furthermore, although some agencies may be loath to take on additional administrative burdens, they are certainly no more hesitant to use conditional preemption than they have been to use withdrawal of funding. Withdrawal of funding is ultimately a perverse enforcement tool, making perfect the enemy of the good. A state that is doing some positive work along national policy lines—while violating many national policy directives—is better than a state that cannot afford to do anything at all.
In the end, we predict that, over a long timescale and assuming no dramatic amendments to prevailing Supreme Court doctrine, Douglas and NFIB will push the national government to greater use of conditional preemption for enforcing spending conditions. This result might be good for national policy, but it will leave the states with less freedom to resist, avoid, or influence national policy. It will, in short, end uncooperative federalism by eliminating state involvement in states that disagree with national programs.
VI. NORMATIVE THOUGHTS
In the end, the Supreme Court’s pronouncements on federalism doctrine have created a strange set of incentives for the national government, and the Court’s rules might push Congress and the Fourth Branch away from financial penalties and towards their only alternative for enforcing spending conditions: conditional preemption. Such a swing would mark a significant blow for uncooperative federalism. When the national government withdraws funds to enforce spending conditions, the state retains power over the policy regime. The state might become cash-strapped or desperate, but it is still in charge of shaping and administering policy. When the national government uses conditional preemption, though, the state loses all of its powers within that regime, whether those powers were fully sovereign or not before the national takeover. The various “powers of the servant” that Bulman-Pozen and Gerken identified disappear, and the state reverts to a regular “outsider” in the shaping of policy. It still has the power of a lobbyist, trying to convince the national government to run its policy in particular ways, but it lacks the power of either sovereign or servant.
All of that said, there might be many advantages to this trend, should it emerge. Uncooperative federalism has significant drawbacks, and those drawbacks might gradually disappear if the national government starts crowding out defiant states and mandating true cooperation from the states that want to stay involved in federal programs. Consider the theoretical arguments of functional federalism. Functional federalism argues that the national government should be in charge when uniformity, redistribution among states, or economies of scale are important or when the national government needs to prevent spillovers or a race to the bottom among the states. By contrast, the theory argues that states should be in charge when experimentation, voice, exit, or diversity is important or when we want regulation to occur only during times of economic growth (when the states can afford to regulate without violating their balanced budget requirements). Cooperative federalism allows some combining of these virtues of national and state regulation, respectively. Under cooperative federalism programs, the national government engages in financial redistribution among the states and sets a regulatory floor to ensure basic uniformity, but it allows state experimentation, diversity, voice, and exit in regulating above the floor.
The problem with uncooperative federalism—for all that it might enhance deliberation—is that it undermines the virtues of national involvement. Imagine that spending conditions in a particular program are carefully designed to optimize the balance between national and state regulation. Imagine, that is, that the conditions are all necessary to create needed uniformity, to avoid interstate spillovers, or to prevent races to the bottom. In this hypothetical cooperative federalism regime, there are no superfluous or gratuitous spending conditions. By hypothesis, then, each and every spending condition is needed to counteract an incentive that the states would face in the absence of the national statute: incentives to create inefficient dis-uniformity, to engage in a race to the bottom, or to externalize regulatory costs onto neighboring states. That’s the idea of efficient regulation. In the era of uncooperative federalism—in which spending condition enforcement is relatively difficult and rare—the states have had tremendous license to take national money while continuing to regulate in ways that are, by this hypothesis, inefficient. If the national government could engage in efficient enforcement of its efficient regulations, then cooperative federalism would strike an impressive federalism balance.
If, however, the national government shifts to a conditional preemption enforcement tool, then the virtues of state involvement might disappear. Imagine again a cooperative federalism program that Congress has designed carefully to optimize efficiency, but this time, consider the absences of national regulation. A well-designed scheme of cooperative federalism leaves states with flexibility in those areas of the federalism program that benefit most from experimentation, diversity, voice, and exit—the states’ strengths. If the national government shifts from non-enforcement or financial enforcement to conditional preemption, its takeover from the states could undermine or obliterate those virtues of state involvement. Although the national government might be able to run slightly different programs in different states after a crowd-out technique of nationalization, thereby maintaining some experimentation and diversity values, the voice value will all but disappear. The voice advantage of state power hinges entirely on the smallness of the governmental entity that’s in charge. If the national government is in charge in a particular state, it will be very difficult for the state’s citizens to influence the shaping of policy.
To evaluate normatively the future that we predict will emerge from the Supreme Court’s interventions, we would want to know whether the cost of losing the states’ strength with respect to citizen participation and voice will outweigh the costs that the system has been incurring from weak enforcement of spending conditions. We, the authors of this article, suspect that the cost of uncooperative federalism is higher than the cost of nationalization, so we are relatively happy with the Supreme Court’s odd and seemingly internally contradictory set of decisions. But the normative question is a difficult one to test empirically because it depends to some extent on citizens’ feelings about their state and national governments. Some citizens might be quite unhappy with the loss of voice that accompanies nationalization, whether that loss is real or merely perceived.
In the end, the Supreme Court’s federalism jurisprudence seems to run contrary to its stated goals. The New Federalism era, up to and including NFIB, creates an incentive for the national government to flex its own muscles more, not less. Maybe that result will be good for voters’ clarity and for uniformity of national policy, but it is not good for uncooperative federalism or for states’ autonomy—the values that the Supreme Court seems to be trying to protect.
* Abigail R. Moncrieff, Peter Paul Career Development Professor and Associate Professor of Law, Boston University School of Law. Jonathan Dinerstein, Boston University School of Law, J.D. expected 2016.
 Jessica Bulman-Pozen & Heather Gerken, Uncooperative Federalism, 118 Yale L.J. 1256 (2009).
 132 S. Ct. 2566, 2608 (2012).
 Id. at 2607–2608.
 U.S. Const. art. I, § 8, cl. 1; S.D. v. Dole, 483 U.S. 203 (1987).
 See No Child Left Behind Act of 2001, Pub. L. No. 107-110, 115 Stat. 1425 (2002).
 See Patient Protection and Affordable Care Act, Pub. L. No. 111-148, 124 Stat. 119 (2010).
 See Occupational Safety and Health Act of 1970, Pub. L. No. 91-596, 84 Stat. 1590 (1970).
 See Personal Responsibility and Work Opportunity Reconciliation Act of 1996, Pub. L. No. 104-193, 110 Stat. 2105 (1996).
 See Erin Ryan, Negotiating Federalism, 52 B.C. L. Rev. 1 (2011).
 536 U.S. 273 (2002) (holding that individuals may not sue heads of state agencies under 42 U.S.C. § 1983 to correct violations of federal spending conditions).
 132 S. Ct. 1204 (2012) (holding that individuals should sue federal agencies under the Administrative Procedure Act’s arbitrary and capricious standard when those agencies agree to fund state programs that violate statutory spending conditions).
 See id. at 1210 (urging individuals to sue the national government under the Administrative Procedure Act).
 132 S. Ct. at 2608.
 42 U.S.C. § 1321 (2012) (Patient Protection and Affordable Care Act).
 505 U.S. 144 (1992) (holding that the federal government may not simply require state legislatures to implement federal policy).
 521 U.S. 898 (1997) (holding that the federal government may not simply require state law enforcement officers to implement federal policy).
 483 U.S. 203 (holding that the federal government may withhold federal funding from states that refuse to comply with federal directives as long as the funding is reasonably related to the directives and the financial inducement is not coercive).
 See generally Herbert Wechsler, The Political Safeguards of Federalism: The Role of the States in the Composition and Selection of the National Government, 54 Colum. L. Rev. 543 (1954) (arguing that the Supreme Court did not need to enforce federalism constraints because the states could protect themselves); but see Lynn A. Baker & Ernest A. Young, Federalism and the Double Standard of Judicial Review, 51 Duke L.J. 75 (2001) (arguing that the Supreme Court ought to enforce federalism doctrines more strictly).
 N.Y. v. United States , 505 U.S. 144.
 514 U.S. 549 (1995).
 521 U.S. 898.
 529 U.S. 598 (2002).
 Lopez , 514 U.S. 549.
 Morrison , 529 U.S. 598.
 NFIB , 132 S. Ct. 2566.
 297 U.S. 1 (1936).
 301 U.S. 619 (1937).
 297 U.S. at 65–66.
 301 U.S. at 640.
 There are today, and were beforeCongress passed the national Gun Free School Zones Act and Lopez was decided, many state laws against carrying guns near schools. It is only the national government that cannot regulate this issue through direct regulation.
 See Bailey v. Drexel Furniture Co ., 259 U.S. 20 (1922).
 The substantive individual rights enumerated in the bill of rights and implied in the Fifth Amendment’s Due Process Clause, of course, still constrain national power, but thanks to Fourteenth Amendment incorporation, they also constrain state power. Those substantive barriers are not about federalism (though federalism may be a means of protecting the same individual liberties that the substantive rights seek to protect).
 301 U.S. 548 (1937).
 483 U.S. 203.
 301 U.S. 548.
 Id. at 574 (“If the taxpayer has made contributions to an unemployment fund under a state law, he may credit such contributions against the federal tax, provided, however, that . . . the state law shall have been certified to the Secretary of the Treasury by the Social Security Board as satisfying certain minimum criteria.”).
 483 U.S. at 211.
 Pennhurst St. Sch. & Hosp. v. Halderman , 451 U.S. 1, 17 (1981).
 Dole , 483 U.S. at 211 (“When we consider, for a moment, that all South Dakota would lose if she adheres to her chosen course as to a suitable minimum drinking age is 5% of the funds otherwise obtainable under specified highway grant programs, the argument as to coercion is shown to be more rhetoric than fact.”).
 Id. at 211 (quoting Steward Mach. Co.,301 U.S. at 590).
 Erwin Chemerinsky & Martin A. Schwarz, Section 1983 Litigation: Supreme Court Review, 19 Touro L. Rev. 625, 663 (2003) (quoting then-Chief Justice William Rehnquist, who referred to Gonzaga University v. Doe as his “sleeper” case of the 2002 term).
 See Gonzaga U. , 536 U.S. at 282–286 (citing 42 U.S.C. § 1983).
 451 U.S. at 28 (quoted in Gonzaga U., 563 U.S. at 280).
 See generally Nicole Huberfeld, Bizarre Love Triangle: The Spending Clause, Section 1983, and Medicaid Entitlements, 42 U.C. Davis L. Rev. 413, 413, 427, 451–453 (2008) (noting that lower courts “have inconsistently and confusingly applied the Gonzaga framework.”).
 132 S. Ct. 1204.
 Id. at 1210.
 See Bulman-Pozen, supra n.1, at 1278.
 NFIB , 132 S. Ct. at 2584–2601 (holding the individual mandate is not a valid exercise of Congress’s power to regulate interstate commerce or of its power to enact laws that are necessary and proper to carry out commercial regulation but going on to conclude that the mandate is a valid tax that Congress may constitutionally impose).
 Its political power is a different story.
 Dole , 483 U.S. at 211 (quoting Steward Mach. Co., 301 U.S. at 590).
 NFIB , 132 S. Ct. at 2605–2608.
 Id. at 2604–2606.
 Id. at 2605–2606.
 Id. at 2605.
 Id. at 2606.
 Id. at 2606.
 NFIB , 132 S. Ct. at 2631 (Ginsburg, Sotomayor, Breyer, & Kagan, JJ., concurring in part and dissenting in part).
 NFIB , 132 S. Ct. at 2606 (majority).
 Id. at 2606.
 Id. at 2604.
 NFIB ,132 S. Ct. at 2605.
 Chief Justice Robert explicitly declined to “‘fix the outermost line’ where persuasion gives way to coercion,” leaving coercion theory quite vague. Id. at 2606 (quoting Steward Mach. Co., 301 U.S. at 591).
 See Id. at 2642–2643 (Scalia, Kennedy, Thomas & Alito, JJ., dissenting).
 Id. at 2661–2662.
 This theory must rest on diminishing marginal returns. The actual marginal cost should be about the same for each state to join, but by the time you are spending $7 billion per year on other states’ Medicaid programs, adding another $150 million for a program of your own might not seem so bad. See Id. at 2604 (majority) (noting that the national government will spend $3.3 trillion from 2010 to 2019 on Medicaid.). Our numbers are $3.3 trillion/50 for each state’s nine-year share, divided by 9 for each state’s annual share ($7 billion). We then divided that share by 50 to figure out how much each state pays in national taxes for a marginal state to join. This math is a rough approximation of a state average; different states’ programs cost different amounts, both because the programs’ eligibility and benefits differ and because the national government’s percentage share differs.
 Kaiser Family Foundation, Medicaid: A Timeline of Key Developments, http://perma.cc/ZQF2-A6MC (http://kaiserfamilyfoundation.files.wordpress.com/2008/04/5-02-13-medicaid-timeline.pdf) (Apr. 28, 2008).
 Perhaps. The math is complicated here because the national government pays for only 50–83% of each state’s program, and the states have a fair degree of flexibility in designing their programs. 42 U.S.C. § 1396(d)(b). Medicaid therefore costs different amounts in different states, and when a state joins, that state still takes on a large additional expenditure of its own—17–50% of an expensive program comes out of state taxes. Nevertheless, the dissent’s logic seems to imply that the cross-subsidization of states’ cooperative federalism programs through national taxation is a reason to suspect that any “offer” of national grants to the states is fundamentally impossible for the state to avoid accepting.
 NFIB , 132 S. Ct. at 2663 (Scalia, Kennedy, Thomas & Alito, JJ., dissenting).
 Id. at 2662.
 Id. at 2662–2668.
 Id. at 2575 (majority).
 See N.A. of St. Budget Officers, State Expenditure Report: Examining Fiscal 2012-2014 State Spending 45–52 (2014) (available at http://perma.cc/WP5K-XYA6 (http://www.nasbo.org/sites/default/files/ State%20Expenditure%20Report%20%28Fiscal%202012-2014%29S.pdf)).
 See generally Erin Ryan, The Spending Power and Environmental Law after Sebelius, 85 U. Colo. L. Rev. 1003 (2014); Georgina Jones Suzuki, Clearing the Air Following National Federation of Independent Business v. Sebelius: The Clean Air Act and the Constitutionality of Highway Sanctions , 93 B.U. L. Rev. 2131, 2145 (2013); David Baake, Student Author, Federalism in the Air: Is the Clean Air Act’s “My Way or No Highway” Provision Constitutional after NFIB v. Sebelius? 37 Harv. Envtl. L. Rev. Online 1 (2012).
 See Bulman-Pozen & Gerken, supra n. 1, at 1285 (dissing administrative safeguards of federalism and state power within the Fourth Branch).
 Medicaid.gov, Medicaid State Plan Amendments, http://perma.cc/MRH2-68VR (http://medicaid.gov/state-resource-center/medicaid-state-plan-amendments/medicaid-state-plan-amendments.html) (accessed Jan. 12, 2015).
 See Motor Vehs. Mfrs. Ass’n v. St. Farm Automobile Ins. Co. ,463 U.S. 29, 43 (1983) (“Normally, an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.”).
 See Mass. v. EPA , 127 S. Ct. 1438, 1441–1444 (2007) (holding that political justifications for denying a petition for rulemaking were not valid justifications under arbitrary and capricious review); Jody Freeman & Adrian Vermeule, Mass. v. EPA: From Politics to Expertise, 2007 Sup. Ct. Rev. 51 (interpreting the Court’s holding in Massachusetts v. EPA as a requirement that agencies exercise expertise rather than make policy choices based on political preference).
 See 42 U.S.C. §§ 1451, 1456b(g) (the Coastal Zone Management Act (CZMA) suspends any funding to states under the Act if the states fail to fulfill the actions required of them under “coastal zone enhancement” federal grants); 42 U.S.C. § 1397aa, 1397ff(c)–(d) (the Children’s Health Insurance Program (CHIP) withholds federal funds from states that substantially fail to follow federal standards); 42 U.S.C. § 609 (the Temporary Assistance for Needy Family (TANF) program provides specified penalty calculations for different violations of TANF terms); 45 C.F.R. pts. 260–265 (2014) (TANF penalty calculations).
 See 30 U.S.C. §§ 1211, 1251, 1253, 1254 (2012) (the Surface Mining Control and Reclamation Act (SMCRA) allows states to submit their own programs to the federal government to regulate coal mines, but the federal government can substitute a federal program in its place if the state fails to submit or maintain an adequate regulatory program); 42 U.S.C. § 300g-1 to 300g-3 (the Safe Drinking Water Act (SDWA) permits states to develop their own drinking water enforcement programs but also strips states of their regulatory power if they fail to meet federal standards); Robin Kundis Craig, Federalism Challenges to CERCLA: An Overview, 41 Sw. L. Rev. 617, 622 (2012) (the Comprehensive Environmental Response, Compensation, & Liability Act (CERCLA) allows states to perform and regulate environmental cleanup operations themselves, with certain conditions); 42 U.S.C. § 9604(c)(1)–(2), (d)(2) (states are required to maintain federal standards under CERCLA or risk losing their regulatory discretion).
 Robert L. Fischman, Cooperative Federalism and Natural Resources Law, 14 N.Y.U. Envtl. L.J. 179, 192 (2005) (the Clean Water Act (CWA) grants a great deal of discretion to states to implement their own clean water regulations, provided they meet federal minimum standards); 33 U.S.C. §§ 1256(e), 1342(c)(3) (2012) (states can lose federal funding and regulatory discretion if they fail to meet the minimum federal standards).
 N.Y. , 505 U.S. at 168.
 Wisconsin Briefs, The Minimum Drinking Age in Wisconsin 6 (Jan. 1995) (available at http://perma.cc/W4FQ-WETG (http://legis.wisconsin.gov/lrb/pubs/wb/95wb3.pdf)).
 42 U.S.C. § 18041.
 King v. Burwell , 759 F.3d 358, 364 (2014).
 Pub. L. No. 94-580, 90 Stat. 2795 (1976) (codified at 42 U.S.C. §§ 6901– 6992(k)).
 42 U.S.C. § 6926(e).
 Jason M. Levy, Conflicting Enforcement Mechanisms under RCRA: The Abstention Battleground between State Agencies and Citizen Suits, 39 Ecol. L.Q. 373, 375 (2012).
 Id. ; 42 U.S.C. § 6926.
 Dole , 483 U.S. at 211 (quoting Steward Mach. Co., 301 U.S. at 590).
 33 U.S.C. § 1342(c)(3).
 Fischman , supra n. 85, at 192; Roderick M. Hills, Jr., The Political Economy of Cooperative Federalism: Why State Autonomy Makes Sense and “Dual Sovereignty” Doesn’t, 96 Mich. L. Rev. 813, 869 (1998) (explaining that while the federal government has “greater capital resources . . . [and] a comparative advantage in delivering capital-intensive services,” state governments have an advantage “in delivery of labor-intensive services” because they serve a smaller number of people); see also Martha Derthick, Agency under Stress: The Social Security Administration in American Government 37–46 (1990) (recounting that the Social Security Administration lacked the trained personnel to review eligibility for applicants when California refused); James E. Krier & Edmund Ursin, Pollution and Policy: A Case Essay on California and Federal Experience with Motor Vehicle Air Pollution, 1940–1975, 233 (1977) (summarizing the EPA’s inability to regulate drivers in California when the state refused to implement a plan under the Clean Air Act).
 Fischman , supra n. 85, at 192 n. 37.
 See Bulman-Pozen & Gerken, supra n. 1, at 1288 (generally discussing the usefulness of cooperative federalism to make states “insiders” who can, from that position, more fully represent “outsider” interests).
 See generally Abigail R. Moncrieff & Eric Lee, The Positive Case for Centralization in Healthcare Regulation: The Federalism Failures of the ACA, 20 Kan. J.L. & Pub. Policy 266 (2011).
 See Abigail R. Moncrieff, Cost-Benefit Federalism: Reconciling Collective Action Federalism and Libertarian Federalism in the Obamacare Litigation and Beyond, 38 Am. J.L. & Med. 288 (2012).