Leslie Kendrick: "The Impossibility of Impossibility Preemption"

          When the Supreme Court hears oral argument in Mutual Pharmaceutical Co. v. Bartlett, it will embark on its third elaboration of “impossibility preemption” in the prescription-drugs context.  This line of cases is reshaping preemption doctrine, and it is doing so with little regard for a basic legal idea: the distinction between a property rule and a liability rule.

          The question at issue is whether it is “impossible” for prescription drug manufacturers to comply with both Food & Drug Administration requirements and state tort law.  State tort liability for design defect or failure-to-warn is predicated on a judgment that a drug or drug label was not designed safely.  But manufacturers generally cannot alter drugs or labels without FDA approval.  Thus, manufacturers argue, it is impossible for them to comply with both tort law and FDA requirements.

          The Supreme Court implicitly accepted this view in 2009’s Wyeth v. Levine, when it found no impossibility preemption of a failure-to-warn claim against Wyeth.  The Court rested this conclusion on the view that, under FDA regulations, Wyeth could have changed its drug’s warning label without prior FDA approval.  Wyeth thus could have avoided tort liability while complying with FDA rules. 

          This analysis was always dubious: labeling changes ultimately require FDA approval, and thus a manufacturer’s power to make unilateral changes is short-lived and mostly hypothetical.  But there soon arose a side effect which the majority in Wyeth, led by Justice Stevens, surely neither foresaw nor intended.  To the extent that the power to make unilateral changes exists at all, it is enjoyed only by brand-name manufacturers and not by makers of generic drugs.

          Thus when, in PLIVA, Inc. v. Mensing, the manufacturer of a generic drug made an impossibility preemption claim, the Court sprang the trap it had set in Wyeth.  If impossibility was avoided only to the extent that manufacturers could change their labels, then impossibility must exist with respect to generic drugs.

          In Mutual Pharmaceutical Co. v. Bartlett, the Court is being asked to extend this analysis to preempt some suits against generic manufacturers for design defects.  And the United States, for its part, appears to ask the Court to feel free to endorse this rule for all design-defect claims, for both generic and brand-name drugs, on the theory that no manufacturer can alter the substance of its drug without FDA approval.  What started as a ground for rejecting preemption in Wyeth is now poised to eliminate virtually all tort liability for prescription drug manufacturers, all in the name of “impossibility.” 

          But if we know one thing about tort law, we know that it is not impossible to comply with it and FDA requirements simultaneously.  We know this because drug manufacturers have been complying with both since the passage of the Federal Food, Drug, and Cosmetic Act in 1938.   

         Moreover, anyone briefed in the most basic attributes of tort liability and FDA regulation can explain why.  Tort liability requires manufacturers to pay damages, not to alter their products.  A manufacturer may market a drug as approved while paying tort damages.  The manufacturer has other options, such as seeking to modify a drug or label, or developing improved products.  But it is entirely possible to change nothing and comply with both sets of laws.

          These facts cannot have escaped the Court’s notice but may seem too mundane, or too literal, to matter.  The Court focuses exclusively on the “duty” imposed by FDA and tort rules.  FDA requires a drug or label to have a certain content, and the implication of tort liability is that it should have a different content.  Hence, impossibility. 

          But in focusing on duties to the exclusion of remedies, the Court is disregarding the fundamental distinction between a property rule and a liability rule.  FDA regulation subjects drug manufacturers to a property rule: drugs must conform to FDA requirements or be recalled from the market.  Tort liability, meanwhile, is the classic liability rule: the remedy for breach is the payment of damages. 

          It is no coincidence that FDA’s powers function like a property rule.  The point is for FDA to decide which drugs can be marketed and which cannot.  Nor is it a coincidence that the state tort claims at issue are liability rules.  Tort liability does not enjoin behavior.  It charges a price, without altering other entitlements.   It is not impossible for these two regimes to co-exist.  It’s impossible that they couldn’t.

          From this perspective, it also becomes clear that the so-called “stop-selling” doctrine at issue in Bartlett is not necessary to its resolution.  The First Circuit held that there was no impossibility preemption because the drug manufacturer could comply with both regimes by not selling the drug.  “Stop selling” may be a response when a product faces conflicting injunctions (e.g., “Do X or your product will be recalled,” and, “Do not-X or your product will be recalled.”).  But the whole point of a liability rule is that the manufacturers need not stop selling in order to comply with it.  They can pay and play. 

          This is not to say that no conflict may ever arise between tort law and FDA regulation.  But impossibility is not the right term, under any plausible definition.  It is not physically impossible to comply with both regimes.  There is no inherent logical contradiction in them.  Instead, the question is how much one gets in the way of the other. 

          The question thus involves the other form of implied preemption: obstacle preemption.  The Court rejected obstacle preemption in the context of Wyeth.  Available evidence suggests that Congress, in passing the FDCA, understood tort and FDA regulation not to conflict but to complement each other.  FDA, throughout most of its history, has held the same position.

          And there is good reason to adopt this view.  Given the nature of drug trials and the FDA approval process, there is always a risk of side effects whose gravity only becomes clear after a drug is on the market.  Drug manufacturers are generally in the best position to bear this risk, because they have more information about their drugs than either FDA or consumers and they have the ability to work with FDA to alter a drug or label.  To adopt another Calabresian term, drug manufacturers are the cheapest cost avoider.  Meanwhile, a system that immunizes drug manufacturers provides
suboptimal levels of deterrence.

          We may worry, however, that tort liability will be so costly that it will cause drug manufacturers to raise prices to unaffordable levels, stop developing new products, or withdraw from the market entirely.  Such possibilities could have serious public health consequences, and thus they are germane to the preemption question.   

          But this is an incidental effect of liability, not an indication of impossibility.  To conclude
otherwise would suggest, for example, that the special immunities given to vaccine manufacturers are unnecessary.  If it is impossible for all drug manufacturers to comply with regular FDA regulations and state tort law, then there is no need for special rules that treat the interaction of these two regimes as imposing particularly steep incidental costs on particularly necessary medications. 

          In short, whether tort law and FDA regulation conflict or complement each other is a question not of formalism but of facts on the ground.  Either way, what is happening is not a question of impossibility.  The Wyeth impossibility trap is one of the Court’s own devising, and the Court should extricate itself.

--Leslie Kendrick, University of Virginia School of Law