I certainly take Alex’s point that libertarians can disagree about the appropriateness of federal preemption of state tort law. Indeed, Justice Thomas’s concurring opinion lays out a strong textual case against implied preemption of state law in all cases. However, there are a few points I was not able to make in yesterday’s post that better elucidate why preemption is better policy given the narrow facts presented in Wyeth v. Levine. And, Justice Alito’s dissenting opinion -- after pointing out a few misleading statements of fact in Justice Stevens’s majority opinion -- shows why, if one accepts the current state of implied preemption caselaw, the Court’s majority decision is bad law. Stevens’s majority opinion asserts that, regardless of FDA’s regulation, “the manufacturer bears responsibility for the content of its label at all times.” It further argues that, FDA can’t keep track of all safety issues that arise after a drug is approved, that the agency never specifically made a determination regarding the safety of IV-push administration, and that Wyeth could have changed its label without FDA’s pre-approval after receiving information regarding the risk of arterial injection of Phenergan. However, as Justice Alito’s dissent makes clear, FDA repeatedly and intensively investigated this exact question after cases of severe tissue damage connected to the use of Phenergan began to emerge after the drug’s approval in 1955, and the agency approved several changes to the drug’s label that boosted the warnings accordingly. Indeed, Levine and her attorneys conceded that, in 1988, Wyeth proposed a label change that “if followed, would have prevented the inadvertent administration of Phenergan into an artery,” but the FDA rejected that language. Further, no new evidence of Phenergan’s risks has emerged in decades. The Court majority nevertheless concluded that there was some other conceivable way for Wyeth to have ramped up its warning short of Levine’s preferred route of ruling out IV-push injection altogether, which the FDA rejected. Thus, FDA regulation is a “floor” below which state law cannot fall, but that the agency’s drug labeling regulation should not preempt state tort laws that require a more strict approach. Unfortunately, that conflicts with reasonable and long-standing Court precedent regarding implied conflict pre-emption of state laws elucidated most recently in Geier v. American Honda. As Justice Alito’s dissent makes clear, “the ordinary principles of conflict pre-emption turn solely on whether a State has upset the regulatory balance struck by the federal agency.” That is exactly what has happened here. There is no such thing as a perfectly safe drug, but on balance most drugs offer more benefit than harm. Congress established the FDA and enacted the Food, Drug and Cosmetics Act and subsequent amendments giving the agency statutory authority over questions of safety and efficacy because it believed that a federal expert body could most effectively balance the benefits and risks of new medicines. FDA made a decision that permitting IV-push injection provides greater benefits than could be achieved with the alternative of deep tissue injection, and that those benefits outweighed the risks. Permitting state tort law to over-ride that determination upsets the regulatory balance struck by FDA. Thus, contrary to Alex’s suggestion, allowing Ms. Levine’s claim here would in effect act as a ban – not one that removes the drug from the market, but one that removes one “safe and effective” use of the drug from its label. And, while physicians may prescribe a drug for an “off-label” use, if a use is not identified on the label, the manufacturer can’t tell anyone about it. That restricts the ability of doctors to receive information about possible uses, and it means that some patients will not be able to benefit from some drugs. Finally, it is relevant, as Alex notes, that the existence of FDA regulation as a legal floor means that the tort system acts as a check on FDA decision-making, but only in one direction: toward greater regulation. Alex is right. “This IS one of these questions about what to do in the real world, where first-best solutions just aren’t politically possible” (emphasis added). So, I think it’s reasonable for we libertarians to support the less bad position that, if our society is going to create a regulatory gatekeeper for drugs and empower it to make risk-benefit balancing decisions on our behalf, then we should not permit lay juries to ratchet up that regulation in circumstances in which all reasonably available information about risks and benefits is internalized into the system.