The Roberts Court, admin law, & federalism

For aficionados of federalism, all of the important battles over state power are now being fought in the arena of administrative law and regulatory preemption. If agencies can easily preempt state law, then federalism is dead, because agency regs tend inexorably to expand over time, squeezing out most state policy-making space. If the court throws road blocks in front of agency preemption, then the states have a fighting chance of eking out a corner in which to resist national policies.

So how does the Roberts Court rate on federalism grounds? For fans of state power, the early returns are not promising. Last week, the Court handed Riegel v Medtronic, 128 S. Ct. 999 (2008), in which it held that the FDA’s premarket approval of a balloon catheter preempted state common-law claims against the manufacturer for negligence, strict liability, and implied warranty. Riegel follows on the heels of Watters v Wachovia, 127 S. Ct. 1559 (2007), in which the Court held that the National Banking Act preempted states’ laws banning “predatory lending” by state-chartered banks that were subsidiaries of banks with national charters.

On a brighter note (again, if you are inclined, like myself, to favor broad state power), the Court yesterday summarily affirmed , by a 4-4 vote, the Second Circuit’s decision in Desiano v. Warner-Lambert, 467 F.3d 85 (2nd Cir 2006), finding no preemption of Michigan law. The issue in Warner-Lambert was whether Michigan law could permit product liability claims against manufacturers of FDA-approved drugs for fraud, failure to warn, and defective design and manufacture if plaintiffs proved that the defendants had obtained FDA approval of their drugs by misrepresenting material facts to the FDA. Judge Guido Calabresi had found no preemption.

The bad news for the friends of federalism is that the 4-4 decision was likely the result of Chief Justice Roberts’ recusal: It is difficult to believe that Roberts who joined Justice Scalia’s dissenting opinion in Gonzales v Oregon, would tolerate state tort claims rooted in the allegation that a manufacturer committed fraud against the FDA. Robust state power, in short, is probably on the short side of a 5-4 split.

The case on which hopes of federalism hang is now Wyeth v Levine, a case set for argument in the coming Fall term. (In Wyeth, the Vermont Supreme Court held that the FDA’s approval of Phenergan, an anti-allergy drug, did not preempt failure-to-warn claims that injection of the drug posed risks to a patients’ arteries. Levine’s arm was amputated after her artery was damaged as a result of an injection of Phenergan).

With agency preemption at the front of the Court’s agenda, it is not surprising that recent legal scholarship has focused on the issue. One could divide the literature into three categories.

First, there are pro-agency pieces that stress the capacity of administrative agencies to take into account state interests and “the values of federalism.” As first-rate examples of this position, consider Mark Seidenfeld’s and Brian Galle’s article, “Admin Law’s Federalism.” (See article) As Seidenfeld and Galle see it, the administrative process tends to be more transparent than congressional processes, and agency officials tend to have less of an interest in catering to industry’s desire for preemption that members of Congress. To the extent that one thinks that states will be adequately protected through the congressional political process, one should a fortiori be confident about agency processes.

Second, there are anti-preemption pieces that urge that agencies cannot be trusted to take into account the broad institutional interests of states. My former colleague, Nina Mendelsohn offers the best statement of this position in Chevron and Preemption, 102 Mich. L. Rev. 737 (2004), where she argues that agencies simply are not equipped with either the political incentives or the jurisdictional mandate to consider the values of federalism. Focused on their narrow statutory mission – safe drugs or automobiles, cleaner air, healthy workplaces, etc – agency officials have neither the time or inclination to think about values like political legitimacy or democratic participation.

Third, there is the inevitable middle ground – essentially trusting administrative expertise to weigh federalism but insisting that agencies exercise such expertise in a particular context. My colleague, Catherine Sharkey, urges an “agency reference” model in “Products Liability Preemption: An Institutional Approach” (see article) under which the critical consideration is whether the agency has addressed the issue of preemption in its normal policy-making process, when assessing the risks and benefits of the underlying private conduct being regulated. Sharkey would give agencies Skidmore rather than Chevron deference on their preemption positions, and she would insist on a relatively specific fact-finding that state common-law liability increases some risk that the agency is entitled to control. Along similar lines, Gillian Metzger urges in “Administrative Law as the New Federalism” (See article) that federalism be protected through administrative law’s normal requirements of good process – substantial evidence in the record, responses to public comments in section 553 rule-making, etc.

My own instinct is that these contributions might sidestep a critical issue in preemption decisions – the fight between populism and expertise. Much of the literature asks whether agency processes protect state power “enough” without providing any normative theory for what federalism is for. But one cannot know whether federalism gets “enough” protection without such a normative theory. And such a theory will have to be more specific than the usual bland nods to “the values of federalism” – political; participation, experimentation, and so forth. These “values” tend to ignore the real basis for fights between the center and the periphery.

A traditional function of the American federal system has been to represent the values and beliefs of people without national influence against the claims of bureaucratic and educational elites at the metropolitan center. Andrew Jackson’s campaign against the Bank of the United States, the 1870s Granger movement’s campaign against the railroads, Debs’ 1890s campaign against the labor injunction -- all of these movements invoked federalism as a vehicle for giving political outsiders (usually but not inevitably located in the South and West) a platform on which to fight against the well-connected, well-heeled, or well-educated elites who lobby or staff the federal government. The essence of American federalism is, in short, Country Party suspicion of claims to authority based on education or the needs of national capital.

It seems improbable to me that ordinary agency rule-making is well-equipped to negotiate these sorts of demands for democratic legitimacy based on lay judgment. Agencies are relatively immune to popular opinion and have no accurate way to gauge popular values. Agencies also have no mandate to consider the distributive justice or injustice of having losses fall where they land, even when such distribution leads to efficient production of goods or services. Most of all, agencies simply ignore local leaders’ claims to legitimacy, which tend to be based more on skill at raising distributive claims before lay audiences – juries and voters – and not at making causal claims through sound clinical studies or accurate multivariate regressions.

Before one defends or attacks agencies as adequate or inadequate defenders of federalism, therefore, one needs to answer one fundamental question: Should agencies be influenced by such populist claims of legitimacy? If one answers this question negatively, then the strongest claims for broad agency preemption probably ought to be accepted. If one answers it affirmatively, then existing agency procedures probably do not “adequately” consider “state interests,” because those processes are ill-suited for thinking about the only interests that state juries and legislatures are well-equipped to represent – the populist claims to equal concern and respect for lay judgment.