Judge (and former Yale Law School Dean) Guido Calabresi gave the keynote speech last month at NYU Law School’s symposium on “Tort Law in the Shadow of Agency Preemption.”

We don’t want to steal his thunder. We assume (but don’t know) that Calabresi’s talk may be published in next year’s symposium issue of the Annual Survey of American Law.

But we thought the speech was pretty interesting, so we’re going to share a couple of ideas from it here.

The two of us typically think about preemption from deep in the trenches. We have a drug company client; the plaintiff claims a certain injury; we’re working with a specific package insert, administrative history of interactions with the FDA, and a governing body of case law. Is there an argument that the plaintiff’s claim is preempted?

End of thinking.

Judge Calabresi steps way back from the trenches. (Maybe that’s why Calabresi’s a judge and former dean, while your humble scribes are just . . . well, humble scribes.)

Anyway, Calabresi noted several choices that are implicit in any discussion of preemption as a policy issue. (Most of what follows are Calabresi’s ideas. But we were jotting down our own reactions as we listened to the talk, and now we can’t separate the two, so there may be a few of our ideas mixed in with Calabresi’s in the following discussion. Needless to say, whatever’s correct is probably what Calabresi said; we surely came up with the errors on our own.)

1. Who should bear the cost of error in the drug regulatory process? For example: Should the user harmed by a prematurely approved drug bear the cost, or should the non-user harmed by an improperly delayed drug? The user of a drug that contained too few warnings, or the non-user of a drug that contained too many?

Whatever choice society makes, there are distributional consequences, because different people will bear the costs.

2. The choice between “national” and “local” regulation is not necessarily a choice between the national Food and Drug Administration and juries applying varying state laws. Society could, if it cared to, have mini-FDA’s (at the state or local level) that regulated drugs or, on the other hand, a single national body of product liability law.

3. The choice between “expert” and “lay” regulation is also not necessarily a choice between the FDA and lay juries. Society could, if it cared to, create regulatory agencies run by laymen or, more likely, empanel blue ribbon juries consisting entirely of experts in a field.

4. What institution should set minimum and maximum regulatory standards? If government sets the maximum standard (“no warnings beyond these may be given;” “no treatment beyond X may be provided”), then the government is deciding who lives and who dies — what can be done to save a life. Perhaps it’s dangerous for federal agencies to assign values to human lives. Or perhaps it’s unseemly to require juries to assign values to human lives. Either way, this is a policy choice.

5. Should regulations be set before the fact (ex ante) or after the fact (ex post)? Regulatory agencies set standards in advance that companies must then follow. Tort law regulates from an after-the-fact perspective (no matter what the substantive tort rules are). The choice between ex ante and ex post regulation is a policy choice.

Calabresi had a bunch of other ideas, too. But, as we said, we really don’t want to steal his thunder. (First, that would be wrong. And, second, we’re not in the business of ticking off Second Circuit judges.)

So ponder the few thoughts that we’ve shared above, but be sure to look at the symposium issue of the Annual Survey of American Law next year to read Calabresi’s own words on this subject.