The Yale Law Journal


Preemption as a Judicial End-Run Around the Administrative Process?

30 Apr 2012

The Medicaid providers (doctors, hospitals, and pharmacists) in Douglas sued to enjoin provisions of California law that reduced the amount of Medicaid reimbursements to providers, alleging that the law violated the federal Medicaid Act’s directive that reimbursements must be “sufficient to enlist enough providers.”2 Medicaid is a “cooperative federalism” scheme to assist the needy, whereby states act as coregulators with the federal agency, the Department of Health and Human Services (HHS).3 States must submit a comprehensive state Medicaid plan—and any amendments thereto—to the Centers for Medicare and Medicaid Services (CMS), a subagency of HHS, for approval. CMS must decide whether a state plan conflicts with the federal Medicaid scheme. CMS approval or rejection of a plan is considered final agency action,4 which can be challenged by private parties (as well as the state) under the APA.

In Douglas, California imposed payment-rate restrictions before seeking CMS approval. It subsequently incorporated the restrictions into a state-plan amendment that it submitted to CMS. At the time the Medicaid providers commenced legal action, CMS had not yet ruled on the plan. Because the providers would have had to wait (months or, most likely, years) until final agency action by CMS to bring an APA challenge, they instead framed their case as an affirmative preemption claim, invoking the Supremacy Clause to argue that the federal Medicaid Act trumped the conflicting provisions of the state Medicaid plan.

The question before the Court in Douglas was whether or not private parties, such as the providers, could bring a cause of action for preemption directly under the Supremacy Clause to enforce provisions of the federal Medicaid Act. In a five-to-four decision, however, the majority skirted the issue. Justice Breyer, writing for the majority, was not ready to close the door altogether to such private suits; instead, a final decision was put off (and the case remanded to the Ninth Circuit Court of Appeals) on account of “changed circumstances”—namely, CMS’s approval of the rate restrictions in the state plan amendment by the time of the Court’s decision.5 The dissent, by Chief Justice Roberts,6 addressed the question head-on. The dissent was prepared to shut the judicial door to such private causes of action under the Supremacy Clause, at least where Congress has not granted a private right of action under the relevant statute. The dissent construed such a cause of action as an “end-run” around the Court’s sharp constriction of federal statutory causes of action under either Section 1983 or implied rights of action.7

The specter of a different sort of end-run haunted the majority: that private parties would forgo the administrative-challenge route in favor of preemption challenges in court, thus “undermin[ing] traditional APA review.”8 According to the majority, private parties can expect “an authoritative judicial determination of the merits of their legal claim” in actions against CMS under the APA.9 Furthermore, the majority suggested that an APA suit against the agency is preferable to a suit against the state under the Supremacy Clause for two reasons. First, it ensures the primacy of the agency—which is “comparatively expert in the statute’s subject matter”10—in the interpretive task necessary to discerning federal-state conflicts. Second, it ensures uniformity in approach, with one centralized federal agency charged with conflict determinations, thereby guarding against the prospect of “subject[ing] the States to conflicting interpretations of federal law by several different courts (and the agency).”11

The foremost concern of the Douglas Medicaid providers was not which actor—courts or the federal agency—would ultimately determine the existence of any federal-state conflict regarding Medicaid reimbursement rates.12 Nor did the dissent think that the existence of the parallel administrative process was relevant to the question before the Court.13 But, for the majority, this question emerged as the key issue. Viewing the majority’s handling of the case in this light brings two unresolved questions to the fore: (1) Why permit a preemption cause of action at all in situations in which there is a viable administrative action? And (2) where parallel proceedings persist, should agency determinations receive varying degrees of deference in each?

I. Courts as Agency Prompters

One reason for courts to entertain direct preemption challenges under the Supremacy Clause (and the one most likely animating the majority’s reluctance to foreclose such suits altogether) is to enable courts to enjoin the states from acting until the administrative process concludes.14 On that view, APA review remains the preferred route, but private parties need an additional, limited court option to hold the states at bay until the agency makes a final determination. The issue was of particular concern to the Douglas Medicaid providers, given that nothing in the Medicaid Act prevented California from implementing its rate reductions before CMS approved them.

This interim period, moreover, could be significant, as there could be years of delay between when the state acts and when the federal agency completes its review. Douglas itself is instructive on this point. Final agency action took place more than three years after California first enacted the payment cuts.15

To push the point further, what if the agency process is not merely delayed, but altogether stymied owing to agency inaction? According to former HHS officials, the agency lacks sufficient resources to comprehensively oversee states’ compliance with federal Medicaid provisions,16 and limited agency resources can lead to inaction. Neither private parties nor courts can directly force agencies to act, but a court-ordered injunction that preserves the status quo should galvanize states (and other interested parties) to pressure the agency to act.17

More expansive agency-forcing measures on the part of courts might be contemplated in contexts outside of Medicaid (and other Spending Clause legislation). Consider, for example, the arena of health and safety, where direct APA challenges to preemptive rules promulgated by the Food and Drug Administration or the National Highway Traffic and Safety Administration are few and far between. In these contexts, standing and ripeness barriers at the time of agency rulemaking loom large (in marked contrast to the Medicaid context, where the approval process contemplates final agency action as a matter of course with immediate repercussions for affected parties). Perhaps where direct APA challenges are not viable, preemption challenges could provide an alternate, albeit more indirect, route to challenge an agency’s conflict determination embedded in an adjudication or rulemaking process.18 Courts can play an agency-forcing role here, encouraging agencies to act in adjudication and rulemaking proceedings with the “force of law” necessary to receive heightened deference in both APA and preemption challenges.19

It is one thing (as the Supreme Court has done) to leave the door partially open to some form of agency-prompting court actions. It is quite another to define precisely which private parties should be able to initiate such actions. When preemption arises as a federal defense, it is necessarily limited in scope. Where it is brought as an affirmative action under the Supremacy Clause, however, there is concern about launching a “roving commission” to find all situations where state law violates federal law.20 The majority offers no guidance here, but I have two preliminary thoughts.

First, consistent with numerous Supreme Court precedents, a private party should be able to bring an action for injunctive relief as an anticipatory preemption defense.21 Second, and more controversially, with respect to implied rights of action under the Medicaid Act, I would alter the analytic framework to make room for the agency’s view of whether implied rights of action would be consistent with the federal regulatory scheme. In Alexander v. Sandoval,22 the Supreme Court rejected the claim that there was a private right of action to enforce disparate-impact regulations promulgated under Title VI of the Civil Rights Act of 1964.23 The majority did not take heed of the federal agencies’ contrary view that “private enforcement is consistent with, and in many instances necessary to ensuring that individuals have effective protection against discriminatory practices by recipients of federal funds.”24 Should the agency itself believe that federal regulatory purposes would be best served by the enhanced enforcement provided by private actions, why not give some deference to this position?25

II. Courts as Arbiters of Federal-State Conflicts

If a court does review an agency’s determination of a conflict between federal and state law in a direct action against the state under the Supremacy Clause, should the court’s level of deference to that agency’s determination be the same as the level of deference a court would give to the agency in a suit against the agency under the APA?

APA review of final agency determinations is highly deferential.26 In practice, court review of CMS approval or rejection of state Medicaid plans and amendments is particularly deferential, given that the federal Medicaid statute is the prototypical “complex and highly technical regulatory program.”27

In Douglas, Justice Breyer suggests that CMS’s approval of the plan should be accorded the same heightened deference in a preemption challenge.28 Counsel for the Medicaid providers agreed.29 He elaborated: “[U]ltimately, you have to demonstrate . . . by clear and convincing evidence, a conflict between Federal and State law. And the agency that . . . evaluates the standards of Federal law will have said in a very authoritative way that there is not a violation under those circumstances.”30

But the Supreme Court has sent mixed signals about the level of deference owed to an agency’s determination of the existence of a conflict between federal and state law in preemption suits.31 To the extent that courts accord lesser deference to agency determinations in preemption challenges than would be the case in an APA challenge, the Douglas majority’s fear that preemption could too easily serve as an end-run around the administrative process seems well-founded.

Moreover, court determinations of federal-state conflicts unmoored from the agency’s position undercut the majority’s desire for uniformity in approach. Indeed, CMS has complained that “court decisions [in preemption suits against the states] have not offered consistent approaches to compliance with the [Medicaid provider] access requirement.”32 These preemption decisions have wreaked havoc on the states, leaving them “without clear and consistent guidelines and . . . subject[ing] them to considerable uncertainty as they move forward in designing service delivery systems and payment methodologies.”33

In my view, harmonizing the standards of judicial deference to agencies across preemption and administrative challenges need not result in courts giving agencies carte blanche in either realm. Particularly in “cooperative federalism” regimes such as Medicaid, in which the federal agency’s incentives may be closely aligned with those of the states, courts may need to step up in their role as guardians and enforcers of the supremacy of federal law.34 Judges should therefore apply a veritable “hard look” in APA challenges and require substantial evidence in the record to back the agency’s conflict determination.35 For preemption challenges, in an approach I have developed elsewhere called the “agency reference” model,36 courts should apply a “State Farm with teeth”37 standard in reviewing the evidence in the agency’s regulatory record supporting the conflict between state and federal law. Courts should accord Skidmore “power to persuade” deference38 (rather than mandatory Chevron deference39) to an agency’s interpretative views on preemption based upon the consistency, care, formality, and relative expertise of the agency.40

This framework should have the salutary effect of nudging agencies toward more thorough determinations of federal-state conflicts based on substantial evidence in the administrative record. An agency, anticipating more searching judicial review, will have an incentive to make its conflict determinations based upon record evidence from either a more robust adjudicative process or notice-and-comment rulemaking. Consider, in this regard, CMS’s recent proposed rule—prompted by APA challenges and preemption lawsuits—that aims to “create a standardized, transparent process for States to follow” to ensure compliance with the Medicaid Act’s equal access provision.41 CMS concedes that, to date, it has “generally relied upon State assurances” that state-plan amendments were consistent with federal law.42 The proposed rule provides “a more consistent and transparent way to gather and analyze the necessary information to support such reviews.”43 CMS conflict determinations, embedded in decisions to approve or to reject state plans, would then be based upon empirical data of the effect of payment-rate changes on access to care for Medicaid beneficiaries, not simply “State assurances” of compatibility.

Conclusion: Preemption as Agency-Forcing

The Douglas majority sides with expert agencies over generalist courts to resolve a potential federal-state conflict. Because there is a concern that courts adjudicating preemption disputes will fail to accord appropriate deference to agency conflict determinations, the majority strongly hints in its remand to the Ninth Circuit that it would close the door to such preemption actions once the agency has approved the state statutes. Nevertheless, the majority leaves the door open to preemption challenges in the future—either where the agency has yet to take, or has decided against taking, any action.

Instead of thinking of preemption challenges as an end-run around APA review, one could conceptualize them as agency-forcing measures.44 The agency reference theory of preemption harnesses the federal administrative process—including State Farm’s “hard look” review of an agency’s factual determinations, and Skidmore’s “power of persuasion” deference—into judicial review. Douglas only skims the surface of the potentially rich interface between the administrative process and preemption challenges.

Catherine M. Sharkey is the Crystal Eastman Professor of Law at New York University School of Law. She is also a John Simon Guggenheim Fellow for 2011-2012. She gives thanks to Rick Hills and Matthew Shahabian for provocative discussion and comments and to Andrew Fine for excellent research assistance (on a short deadline).

Preferred citation: Catherine M. Sharkey, Preemption as a Judicial End-Run Around the Administrative Process?, 122 Yale L.J. Online 1 (2012),