« Want to Cut Costs? Release More Prisoners | Main | Glen on Kafka and Subjectivity in Law »

Friday, February 25, 2011

A Presumption Against Cost-Benefit Preemption?

The SCOTUS's decision in Williamson v. Mazda Motors is welcome to all friends of federalism not only because of its pleasingly unanimous holding but also because it is the perfect course-packet and casebook companion to Geier v. American Honda. The two decisions differ by a hair's breadth from each other in the basic facts: The question in both was whether the feds' giving a manufacturer an option regarding a safety precaution should preempt states' holding the manufacturer from exercising one of the options permitted by federal law. The Geier Court held that the manufacturer's choice between airbags and automatic seatbelts, permitted by FMVSS 208, was protected from state tort liability. The Williamson Court held that the manufacturer's choice not to install a lap-shoulder belt for the middle rear seat, also permitted by FMVSS 208, was not protected from state tort liability.

How can these holdings be squared? Here's one possible reading: The Court could be adopting a presumption against preemption whenever federal agencies exercise self-restraint because by not imposing a precaution on the ground that the precaution is not cost-justified. The agency's purpose of curtailing its own mission in the name of efficiency, in other words, is presumptively not preemptive of state law. If this turns out to be how the Court ultimately construes Williamson, then the decision is actually a big deal -- and welcome news to federalism fans who, like myself, believe that democracy works better when subnational politicians are freer to change the status quo.

Here's my basis for this "agency-efforts-to-promote-efficiency-don't-preempt" reading of Williamson. In Geier, the Court held that the DOT preserved the choice between airbags and automatic belts because the automakers' exercise of this choice would yield data about the relative merits of the two devices. Because the options preserved by the federal agency would contribute to auto safety, the federal agency's specific mission, the states could not second-guess the manufacturers' exercise of the agency-sanctioned option. In Williamson, by contrast, the federal agency refrained from requiring a lap-and-shoulder belt for rear inner seats only because the federal agency thought that this requirement would not be cost-effective. According to the Court, this sort of agency exercise in self-restraint for the sake of efficiency rather than for the sake of the agency's specific mission should generally be presumed not to preempt state law:

"Many, perhaps most, federal safety regulations embody some kind of cost-effectiveness judgment. To infer from the mere existence of such a cost-effectiveness judgment that the federal agency intends to bar States from imposing stricter standards would treat all such federal standards as if they were maximum standards, eliminating the possibility that the federal agency seeks only to set forth a minimum standard potentially supplemented through state tort law. We cannot reconcile this consequence with a statutory saving clause that foresees the likelihood of a continued meaningful role for state tort law."

In other words, because agencies routinely make cost-effectiveness judgments, those judgments cannot be treated as preemptive without radically undercutting state power.

Of course, the auto maker could respond that perhaps federal agencies want radically to undercut state power for the sake of promoting efficiency. Why presume that federal agencies' judgments about efficiency restrain the agency alone rather than the states as well?

Is there any justification for treating cost-effectiveness judgments as less preemptive than safety-effectiveness judgments? One possible reason might be that agencies specialize in particular regulatory missions -- health, auto safety, food safety, drug safety, clean air, clean water, etc. -- rather than in some general mission of making the law more efficient. It is the Department of Transportation, after all, not the Department of Regulatory Efficiency. There is no special reason, therefore, to defer to federal agencies' judgments about the costs imposed by state law when those costs do not involve interference with the federal agency's primary regulatory mission (e.g., auto safety, clean water, clean air, better medical technology, etc.) If a state wants to impose inefficient levels of liability because of (for instance) a hypochondriac risk aversion, or a desire for populist democracy implicit in the jury system, or a paranoid distrust of corporations, it is no business of the federal agency's.

I would even go so far as to say that, absent some fairly clear statutory mandate to promote generalized efficiency, no federal agency should be construed to have the legal power to preempt state laws that do not interfere with the agency's specific statutory mission. The National Highway Traffic Safety Administration (NHTSA) is entitled to impose safety-safety tradeoffs on the states, because NHTSA is an expert on safety. The FDA is entitled to impose health-health tradeoffs on the states for the same reason. And so forth, for the rest of the federal menagerie of agencies: Each has a specific area of expertise. But who appointed these agencies to be czars of efficiency in general? True, they might have authority to curtail their own missions in the name of of some general brand of social welfare. But it is hard for me to see why this statutory mandate for self-restraint -- if it exists -- entitles them to trump state judgments about social goals that simply fall outside the federal agencies' bailiwick.

I realize, however, that I, being temperamentally a federalism nut, take an extreme view. I am sufficiently happy that Williamson could be plausibly construed to have adopted a default rule in implied preemption cases that agencies' efforts at self-restraint in pursuit of their mission are presumptively non-preemptive.

Posted by Rick Hills on February 25, 2011 at 09:00 PM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference A Presumption Against Cost-Benefit Preemption?:


Well, the reply is that "safety-effectiveness" and "cost-effectiveness" is a false dichotomy. After all, it would be maximally safe to require both airbags and automatic seatbelts--and then we wouldn't need comparative data. The reason that the agency didn't choose that option was due to cost. I'm not sure that you can say that the agency left a choice between airbags and auto seatbelts only for safety reasons and not cost reasons.

Posted by: TJ | Feb 26, 2011 12:38:20 AM

TJ's comment indicates that I was unclear: I did not mean to suggest that the options left available by NHTSA in Geier were left available solely for safety reasons. Instead, I meant to say that the options left open in Geier (at least, as characterized by the Court) were at least in part left open for safety-based reasons -- and such safety reasons were necessary to justify implied "frustration-of-purpose" preemption. By contrast, there were no such safety-based reasons for the options left open by NHTSA in Williamson, and the absence of such reasons foreclosed "frustration-of-purpose" preemption in Williamson.

Thanks for the prompt to clear up the misunderstanding.

Posted by: Rick Hills | Feb 26, 2011 12:14:32 PM

Yes, I misread your test. So I can see the distinction, though I'm still not convinced that it is a wise test at the end of the day. Fundamentally, all regulation is ultimately about a cost-benefit trade-off, and the benefits are going to be what the agency is focused on doing (FDA does safety, EPA does environment, etc.) While the agency may explicitly say that it refrains from a regulation due to cost, implicit in that statement is there is not enough benefit, as assessed by the agency in its expertise. And when the agency says it imposes a regulation due to the benefit, implicit in the statement is that the expert-determined benefit outweighs the monetary cost. So the test turns less on the substantive judgment of expertise, and more on how explicit the agency relies on a particular side of an inherently two-sided coin.

Posted by: TJ | Feb 26, 2011 7:10:26 PM

Post a comment