Legal Commentary

Last week, the Court struck an unexpected blow for federalism, deciding in Wyeth v. Levine that Diana Levine's state-law claim for negligent failure to warn was not preempted by the FDA's regulation of Phenergan's labeling under the Food, Drug, and Cosmetic Act. Among the many interesting aspects of this unexpected (by me, at least) 6-3 victory for federalism was Justice Thomas' concurrence -- a concurrence that contradicts the popular notion that conservative justices are indifferent to federalism when preemption benefits industry. But while Justice Thomas' rejection of implied preemption seems to me to be a principled legal theory, it also strikes me as utterly misguided effort -- less a capitalist plot than a textualist blunder.

Justice Thomas' Wyeth concurrence argues that implied "frustration-of-purpose" preemption should be rejected, because the inquiry into statutory purpose is somehow untethered to text. Thomas' dissent in Altria Group v. Good argued that "no explanation [of statutory purpose] is necessary [to explain the scope of the cigarette Labeling Act's preemption clause, because] the text speaks for itself." In short, Thomas seems to draw a sharp line between statutes that have preemption clauses and statutes that do not, adhering to a presumption against preemption only for the latter.

This theory, however, pulls the hat lining out of a hat and calls it a rabbit: Preemption clauses do not magically resolve questions of preemption unless they are themselves clear -- and they rarely are. Generally, such clauses provide that a vaguely defined category of state laws (e.g., "requirements") may not bear a vaguely defined connection to ("related to," "with respect to," "based on", etc) a vaguely defined category of federally protected activities. These clauses do not specify in their text that the presumption against preemption should not be used to interpret the clause's own scope. How, then, is it especially 'textualist' to read them as rejecting a presumption about which their text is silent? Likewise, there is no textualist reason to reject implied "frustration-of-purpose" preemption: If a statute's text, fairly interpreted, bars anyone from interfering with some private activity, then a textualist ought not to think that an additional clause is needed to specify that "anyone" includes "states." Where a federal regulatory requirement is best read as a ceiling on further interference with the regulated activity, then that requirement is itself a sufficient preemption clause -- at least, if you take text seriously.

Justice Thomas' theory, in other words, treats preemption clauses as self-interpreting, even when they are as clear as malted milk. (Justice Thomas authored one of the more egregious ERISA preemption opinions in Egelhof v. Egelhof, 532 U.S. 141 (2001), in which he adopted the fiction that ERISA "plainly" preempted a state probate code having no special relationship to employment benefit plans). This sort of textualism treats the preemption clause as a ventriloquist's dummy: The judge puts words into the statute and blames some wooden prepositional phrase -- "related to," "with respect to," "based on," etc -- for the result, when any 1L can see that the clause cries out for as much explanation as any theory of implied preemption. Frankly, I'd prefer preemption based on a libertarian's love of free markets over preemption based on worship of empty text: At least the former takes responsibility for its policy choices.