American Enterprise Institute

February 21, 2008

[Edited transcript from audio tapes]

3:45 p.m.	Registration
4:00	Panelists:	Michael S. Greve, AEI
		Catherine Sharkey, New York University School of Law
		Daniel Troy, Sidley Austin
		Brian Wolfman, Public Citizen
	Moderator:	Theodore H. Frank, AEI
6:00	Adjournment

Proceedings:

Theodore H. Frank:  I would like to thank everyone for coming.  I’ll talk slow to give Michael a chance to get up to the podium.  Welcome to our panel on Federal Preemption and the Supreme Court.  We pulled some strings to get the Supreme Court to decide some cases yesterday.  You might have seen three preemption cases come down yesterday from the Supreme Court.  We will let Michael talk about this, but there are still others yet to come. 

On Monday, four days from today, on the 25th, Warner-Lambert v. Kent will be argued, a preemption case relating to Michigan state law and a case over Rezulin, a diabetes drug.  At the beginning of next term, Wyeth v. Levine, a case involving Phenergan in the Vermont Supreme Court, and more broadly on the preemption issue, Altria Group v. Good – I took the wrong set of notes up here – it will be argued in October.

We will start with Michael Greve, who is the John G. Searle scholar at AEI.  His research and writing cover American Federalism and its legal, political and economic dimensions.  He cofounded and directed the Center for Individual Rights.  He has written widely on these issues, including – he is supposed to hold up his book now – the AEI Press volume on Federal Preemption.  Michael.

Michael S. Greve:  Thank you, Ted.  I have to apologize in advance.  I have another event at George Mason Law School immediately following this arcane subject of federal preemption, which means I have to scoot a little early, and I truly apologize for that.  To make up for it, the usual pattern at AEI - the convention is to make three points.  Since I have to run off early, I’ll give you a bonus point.  I’ll make four points today, and they have all to do with the broader context of preemption jurisprudence, and then people who actually know something about FDA preemption.  We will talk about today’s actual issue.

My four contextual points are one, about statutory interpretation, another about a principle that Richard Epstein and I call anticircumvention principle, a third brief point about litigation economy in the preemption context, and a final point about the role of Congress.

Let me start on the statutory interpretation point.  One has to be very, very careful here, I think, in the preemption context about generalizations.  Still, with that usual caveat stated, there seems to be a trend - and it is evident in Riegel and some of the other cases this term - a trend towards the court starting and occasionally ending with a statutory language read in context, statutory language as opposed to what or to the exclusion of what.

Well, the first one is general presumptions and, especially, the so-called presumption against preemption.  That is nowhere to be seen in Riegel and it is nowhere in any of the other cases decided this term. That is, of course, different from Hillsborough; it is different from Lohr, and most important to my mind, it is very different from the Bates decision that, of course, was not about the FDA, but it was about FIFRA.  But you look at that decision, you put that next to Riegel, there is a real question not simply on this point but on others.  How much is really left of Bates, and to what extent the partial dissent in that case is really now the law?

The second thing that statutory interpretation turns against to push into the background is legislative history and the question of what Congress could have intended or should have intended or would have intended.  That, of course, is Justice Scalia’s point.

The third thing that tends to move into the background once the court focuses on statutory interpretation in particular is the agency’s litigation position or its position expressed elsewhere and its expertise.  That, again, is very prominent in Riegel, which was decided in a very different way from the way the Second Circuit looked at the case.  What this decision seems to say in so many words is, “Look.  Spare me the metaphysics of Chevron.  Spare me the metaphysics of Skidmore.  Give me the statute.”  It is not the first case in that vein. 

There is a very interesting 2006 case, Watters v. Wachovia, written, interestingly enough, by Justice Ginsburg, which dealt with the preemptive authority of the Office of the Comptroller of the Currency with respect to the National Bank Act where the government relied very heavily on OCC’s interpretation of the statute, and where Justice Ginsburg said, “Look.  I went to law school.  I know how to read a statute.  Do not bother me with the agency positions here.”  So it is not the first case in this vein, but again, it is a very different emphasis than in some other cases where the Supreme Court has tended to give great, great weight to the agency’s position.

This matters, of course, especially with respect to the FDA.  In my judgment, deference to the agency’s position, to its preambles and so forth cuts one way, my way so long as the FDA is run by Dan Troy and his friends.  It is very, very different under a McCain FDA where the pharmaceutical companies are “the bad boys.”  I cannot imagine that an Obama or a Clinton FDA would be much better.  So under those circumstances, the agency’s expertise may be to disclaim any preemptive effect and intent.

It is also important to note, I think, that the emphasis on the statute here is not invariably pro-defendant.  I can think of cases where the Supreme Court looked very, very meticulously at eight particular statutes and says, “There is no preemptive intent.”  Sprietsma v. Mercury Marine is, to my mind, the paradigm case there. 

Finally, it, of course, remains to be seen how this plays out in implied preemption cases where you do not have an expressed preemption provision as in Riegel, and I suppose we will find that out in Levine.

The second point, my famous anticircumvention principle of Epstein and Greve I mentioned -- actually, it is not our invention.  It is an ancient principle; it was a part of Roman law; it was a part of British law where it is called the anti-mischief rule.  And what it says is that statutes have to be read to prohibit not just direct violations but also attempts on the part of the regulated parties to achieve prohibited ends indirectly and through evasion.  That is a principle of statutory fidelity because, otherwise, the statute becomes empty.

Again, the Supreme Court has previously applied that in preemption cases, most notably in securities cases, now, I think firmly established.  You cannot avoid or evade preemption by slapping a different label on the prohibited state or private activity.  That is in Riegel, right?  A requirement remains a requirement even if you call it a general common law duty.  That is the rule. 

The point is directly addressed in another decision handed down yesterday, Rowe v. New Hampshire Transport Association, in an opinion by Justice Breyer.  So there you are dealing with a federal law that prohibits any state law related to a price, route or service of any motor carrier.  New Hampshire then regulated service for tobacco shipments by those shippers, and it said, “Hey.  That is not a regulation of the service of a motor carrier; that is a public health regulation because it deals with tobacco.”  Justice Breyer’s opinion basically said, “Look, a pig with a public health lip gloss remains a pig, and the statute says no pigs.”  Not verbatim.

The only exception that I can see to this becoming a central rule of preemption analysis is the environment where federal law currently prohibits any state law relating to a fuel efficiency standard.  And as you will know, I suspect, California and other states want to have CO2 standards for cars.  Those standards are not related to fuel efficiency standards; they are fuel efficiency standards, albeit under a different label. 

There are two federal lower courts now that have said, “There is no preemption here.”  The reason why they have done that is that they have read Massachusetts v. EPA as creating a global warming exception, which displaces any and all known principles of American administrative law.  We will find out at some point whether that interpretation of the case is actually correct.

Points three and four I’ll do more briefly.  There is a point about litigation economy.  There are, as you know, served petitions and tons of preemption cases any given year in which cases go up and in what order matters big time, as we say at AEI, and neither the plaintiff’s attorneys and AGs nor corporate parties have been very good at controlling the case flow.

Riegel is an example, to my mind, maybe in our cases where MDA preemption would appear shocking and harsh - an ambiguous label, somebody who obviously gets hurt, no medical error.  But this is not that case.  This is a patient for whom the label clearly contraindicated this particular product, and it was used in a way that was specifically warned against on the label and the guy had a rip-roaring med-mal case against the doctors and the hospital and recovered for that practice.

So, on all those counts, I think this is a case that should not have been brought from their perspective.  Rowe is just like that.  Maybe, there is a case for defending an end-run around the clearly stated statutory provision, but it is not this provision and not this Court with Justice Breyer on the Court because - guess what? – he was the guy who wrote that provision when he was a young aide to Senator Kennedy or the precursor to this provision.  He knows what that provision meant or what he meant when he wrote it.  I’ll spare you the details for now, but corporations have been equally undisciplined in cranking cases up there on the cert docket.

So there is a huge leaning now and for the near future on the Court’s sense of direction, its case selection, and there is a huge bonus for the side that figures out, at a minimum, to keep the dogs of cases off the docket.

Final point: the Congress.  Both sides to the preemption debate, it seems to me, say, “Look, it would be useful and it would be nice if Congress provided more clarity in this field.”  That is, I think, highly plausible.  For one thing, you are dealing with very ancient statutes here, such as the Medical Device Act, that were enacted on very different circumstances before the emergence of litigation in the street, before large-scale frequent consumer class actions and so forth.  You now have an inquiry that focuses on the Congressional purpose and intent, but what you really want to know is what Congress would have thought had it thought about a question that, acknowledgedly, it did not think about.

So it might be nice to update these ancient statutes.  And come to think of it, Congress overrules statutory interpretations all the time, so why not here?  Well, it does not.  Congress very rarely overrules preemption rulings.  The only example I can actually think of is securities litigation.  I have not done a systematic survey but, again, I cannot think of –- it is rare in any event.  In some civil -- yes, that is a good point –- some civil rights cases.  But it is a rare event, especially in these industry regulatory areas, and Congress does not go near these statutes even if it thinks that the court has gotten it wrong.

The Federal Arbitration Act that was one of yesterday’s cases -- you may not call that strictly preemption because the question is does that act apply in state court proceedings.  Justice Thomas has consistently held, “No, it does not.”  The Court has consistently held, “Yes, it does.”  My corporate masters will probably have a word with me now.  I think Justice Thomas was probably right about that, although it may be a little late in the day to insist on that.  The FAA is now in play in Congress but it has proven harder than I think anybody expected for Congress to deal with the adverse or pro-“preemptive” FAA rulings.

There is a similar example with the Consumer Product Safety Commission, which probably now two years ago, held that its mattress flammability standards were preemptive.  That is plainly something that Congress disagrees with, and the CPSC, after the toy scares, does not have a lot of credibility left. And yet, Congress cannot somehow figure out how to overrule those kinds of agency interpretations.

I think the reason for Congress not going back to preemption provisions is that preemption provisions are not just a part of the statute; they are the statute.  They are compromised because the pro-regulatory and anti-regulatory forces agree to fight another day; fight in the agencies and then fight in courts.  And without that compromise, without that agreement to fight another day, there would be no statute.

What that entails is that the preemption provisions will always be and almost always, much of the time, be ambiguous or be matched with a savings clause or articulated “in complete silence.”  And what it also means is that Congress will be very reluctant to go back to the statute and to revisit it.  So what you end up with is a preemption law that is a convoluted body of de facto federal common law.  And what it means is that at this particular front the Supreme Court will be, I think, for the foreseeable future, remain the biggest game in town.  Thank you.

Theodore H. Frank:  Thank you, Michael.  Our next speaker is, perhaps, the preeminent preemption person on the plaintiff’s bar side.  He has litigated several of these cases in the Supreme Court, and he is on the briefs in Warner-Lambert on Monday.

Brian Wolfman is the director of Public Citizen Litigation Group where his cases involve consumer health and safety regulation, freedom of information, expanding access to the courts and opposing federal preemption of state products liability law, consumer law and class actions.  For the past five years, he has taught a course on appellate courts at Harvard Law School and has previously taught similar courses at Stanford, American University and Georgetown Law.

Before joining the Litigation Group in 1990, he was a staff attorney at Legal Services of Arkansas where he did trial on appellate work in cases involving housing law, welfare law and consumer protection.  Before that, he clerked for Judge Anderson on the Eleventh Circuit.  Mr. Wolfman?

Brian Wolfman:  Thanks, Ted.  Actually, I really was not planning to talk about the specifics of the pending cases, and I think some of what I’m going to say overlaps or addresses some of the things that were just said.  What I want to do is I want to discuss first whether the theoretical basis for preempting state tort law works in the real world.  That is whether the basis for according preemptive effect, that the state law interference with federal objectives, is a reality.  Then I want to discuss why, in general, I think preemption of state tort law is a bad idea.  That is immunizing the makers of products that cause injury because, for instance, their products have been approved for marketing by an agency.  It is bad for the injured people and for society, generally.

But first, I do want to talk a little bit about doctrine and not so much about what the text means.  I think it was the implication of the prior speaker that the Court is moving away from looking at the legislative history, and I think it ought to do that as it takes a look at the text.  My disagreement, though, is the rather impoverished arid way that the Court goes about the textual reading.  You see that in yesterday’s decision, they focus on this word, “requirement”, and infuse it with so much more meaning than it could have possibly had. 

It is true that Justice Breyer probably knows what he meant when he wrote that but it is very unlikely that he meant what the Court says it meant because no one had this on their radar screen in 1976.  He may think he thought that but I do not think he really did.

In any event, what is the theoretical basis for infusing either the word “requirement” or a whole set of words in the statute with the kind of meaning that the Court has been doing it?  I think what it is is the people who want to find that meaning and it is that there is some equality between a positive law requirement and a tort law requirement.  That is where I differ from many members of the Court in some of the cases have been.  They have actually been schizophrenic on it, and I want to draw that out a little bit.

The first statement on this topic that I can find in the case laws, a labor law case from the late ‘50s, a Garmon case where the Court said that - and I’m quoting - “State regulation can be as effectively exerted through an award of damages through some form of preventive relief.”  Now that does make that equality but the Court really has been silent on that analysis for quite a number of years after that.  But it popped up again in the Cipollone case, the famous tort preemption case involving tobacco and the reason it popped up is because the defendant lawyers did a very good job in bringing the Garmon theory to the court’s attention.

Since then, tort defendants have repeated the Garmon Cipollone phraseology ad nauseam.  It is a smart thing for them to do, trying to show that tort law and positive law are inherently the same, inherently comparable.  The funny thing is in Cipollone itself, the majority opinion, the lead opinion in Cipollone itself, just a few a paragraphs after the endorsement of Garmon, the same justice - not even a different justice - said, and I quote, “There is no general inherent conflict between federal preemption of state positive law warning requirements and the continuing vitality of state common law damages actions.”  In that case, based on a failure-to-warn because that was the claim there.

The attention comes up again in the Lohr case where there is a plurality of opinion that says one thing and two other opinions that say the other thing.  There are a bunch of cases between Lohr and where we are today, but in Sprietsma - Michael referred to this - just a few years ago, the Court said, and again I’m quoting, “It is perfectly rational for Congress to preempt state positive law but not ‘common law claims’ which, unlike most administrative and legislative regulations necessarily perform an important remedial role in compensating accident victims.”  So the Court has just been schizophrenic about this.  Obviously, they were somewhere different from Sprietsma yesterday; I understand that. 

But I’m more interested in what is the basis for this.  Does it have a basis for reality?  That is what I want to turn to.  The reason I want to turn to it, whether it makes sense, is, I think, still even after yesterday, the doctrinal field is wide open because you can pick your case here.  There is no set law on that question.  To me, then, the question is an empirical one.  The question is a matter of regulatory impact.  To me, it is a huge leap from the proposition that tort is meant and, to some degree, does have a regulatory type effect - I can see that - to the proposition that its impact is equivalent to direct positive law regulation.

When the FDA wants to get a drug off the market, it can do so swiftly; I understand it does not always do so swiftly but it can.  It can seize products, for instance, like David Kessler did with misbranded orange juice when he first became the FDA commissioner.  I understand there are pressures for the FDA not to do that.  I understand it does not always use its full regulatory factor.  But it can deny regulatory approval; it can effectively impose advertising restrictions; it can demand data.

The FDCA gives industry various procedural roads in imposing agency action, but as a practical matter the agency can exert its positive law priorities with great force, assuming it has the funding and the structure to do it.  But it can do that.

Now, contrast that with the tort system.  Large industry players generally react quite slowly, actually, and sometimes not at all to liability pressures.  By that, I mean that in most instances of liability, those are absorbed without any change in manufacture or conduct of the kind that can effectively be attained swiftly by a regulator.  Again, not always, but it does not have to change to the extent that tort exerts regulatory effect against a drug manufacturer.  It occurs only after repeated suits, settlements, findings of liability.  And even then, the cause and effect relationship is rarely clear.

So what I am suggesting here -- and let me just back up and say I understand that there is a symbiosis between the two systems.  So the regulatory system exerts pressure on the tort system and vice versa, and both exert financial and political pressure and are capable of making public information that otherwise would stay locked away in corporate file cabinets.  So I understand that.  But I do not understand why you would build a body of legal literature and judicial doctrine, which is what has occurred, on the equivalency between tort and direct regulation when in my judgment it is not remotely accurate.  They are just not the same thing.

So this really brings me to the last point I wanted to get to, which is why preemption, generally speaking - not always - is a bad idea.  I think there are two reasons.  The first I have already touched on briefly, which is that the tort system has a non-regulatory component.  To me, really, its principal component, which is compensation -- what is interesting about that component -- it is virtually never a component of the American administrative law system.  That is, federal agencies that regulate virtually never compensate as well.  There are a couple of exceptions but virtually never do so.

It is sad in my view that present-day legal academics that are intrigued by the theoretical regulatory effective torts as a means of social control as well as the Supreme Court, although, as I say, schizophrenically, have only made passing reference to the compensatory component.  At the very least, the regulatory and compensatory components are major goals of tort law.  That said, I do not really see why compensation should give way when there is a perceived regulatory conflict between tort and the administrative state.

I say this particularly in light of the political reality where federal law guarantees neither health insurance nor accident insurance.  This is not to say that -- so I’m finding the basis in the political and reality of the situation.  I’m not saying -- and there are probably -- some of my compatriots in the plaintiff’s bar would shoot me for saying this.  I’m not saying that having a tort system is an essential component of a just society.  I’m saying it makes sense in this reality.  That is all I’m saying.

In this regard, it seems to me that one conception of strict products liability, which I understand the current conception of strict products liability is nothing like it was in the ‘70s.  But in one conception of it, we acknowledge that socially beneficial products can cause grave harm, so we compensate but we keep them on the market because they are also beneficial.  To me, that kind of system is perfectly consistent and in any event, not in conflict with a regulatory system that seeks but does not fully achieve optimal health and safety benefits.

I’ll put it another way.  What would be the conflict with the medical device scheme that we have, or the drug regulatory scheme that we have, if the states wanted to compensate people for injuries from drugs or devices without fault?  If there were liability without fault, all you had to show was causation of the liability without fault.  There is no conflict with the federal regulatory scheme there.  It is true you would be socking the manufacturers, but there is no conflict with the regulatory scheme.

That naturally brings me to my second reason why preemption, I think, is generally a bad idea.  The regulatory system is not intended to prevent all harm nor could it.  It was meant to balance risk and benefits.  I acknowledge, with the knowledge, though, that injuries will occur in a highly imperfect system where regulators are utterly dependent on profit-motivated players to submit all available relevant data, data that changes over time based on new information known only after a product was approved and marketed.  That being the case, even assuming that tort exerts some regulatory pressure - and as I say, I do assume that - why would we not want it to do so?

The recent preemption controversy, the most recent one that has garnered the most attention, is the plea for preemption from the FDA - Dan has been involved in this - which claims they just have discovered that, in 1938, when Congress enacted the FDCA and gave the FDA authority to issue labeling rules, it intended that such rules would preempt state tort claims premised on a failure-to-warn.  That failure-to-warn is the principal type of claim raised by people that are injured by drugs.  I just think that claim is ill-timed.  Recently, there have been two independent government reports describing the dangerous shortcomings in FDA oversight in drug safety.

A couple of years ago, the GAO concluded that FDA lacks clear and effective processes for making decisions about providing management oversight of post-market safety issues; that the Institute of Medicine at FDA’s request found that the nation’s -- the report anyway was that the FDA’s request found that the nation’s drug safety system is impaired by “serious resource constraints that weaken the quality and quantity of the science that is brought to bear on drug safety.”  They said that, “the organizational culture at FDA is not optimally functional.”

Former chief counsel Peter Barton Hutt says in a report issued recently that “the FDA is in a precarious position in terms of both personnel and money to support them; the agency is barely hanging on by its fingertips.  It is a paradigmatic example of the hollow government syndrome - an agency with expanded responsibilities, stagnant resources, and the consequent inability to implement or enforce its statutory mandates.”

That is the situation wherein the final report I’ll refer to is the Union of Concerned Scientists survey released the summer before last where 60 percent of the survey respondents -- 60 percent of the FDA employees knew of cases “where commercial interests have inappropriately induced or attempted to induce the reversal, withdrawal, or modification of FDA determinations or actions.”

Now, I’m going to read you a list of drug names:  Rezulin, which is involved in the case, it is going to be argued on Monday on the Supreme Court, Lotronex, Celebrex, Vioxx, Zoloft, Prozac, Accutane.  You have heard of all or most of these.  They are among the many drugs in just the past few years that have required post-approval labeling changes to add or strengthen the warnings on those drugs.  In some of those cases, the need for labeling changes, at least in my judgment, was demonstrated, at least, in part with information and pressure derived from the tort system.

Now, according to the Physicians’ Desk Reference, which I assume most of you know, the doctors’ bible of drug labeling, many serious adverse drug reactions are discovered only after a drug has been on the market for years.  In fact, in nearly half of them, the drug has been on the market for over seven years or more.  So with all that said, why are you putting some of the preemptive stock and the preemptive effect off this original approval?  It is hard to see why in light of that reality.

So I want to close with this.  I was emailing about something entirely different - not on the topic of preemption - with someone I have known for many years -- the name I’m not going to disclose.  But for the food and drug lawyers in the audience, there is not a food and drug lawyer in America who would not know this person, a defense lawyer who does food and drug law.  He just writes to me, and I’m not going to disclose the name because his clients would not like this.  But he said to me -- he raised the issue of preemption and I’m quoting his email:  “I do not think the opponents of preemption have pushed what I consider the most telling arguments.  The proponents are arguing that preemption occurs because of the legal scheme,” - this is the point I’m making, not just the legal scheme - “without regard to the ability of the FDA to function in that scheme.  If they succeed in the Supreme Court, then they have every incentive to starve and hobble FDA’s resources and operations.  That results in complete freedom from regulation.  As the conservative movement has shown over recent decades, they cannot succeed by frontal assaults on government programs.  They are too politically popular.  Instead, the strategy has been” - and he is putting this in quotes - “not to dismantle the government but to shrink it so that it could be drowned in a bathtub.”  Thank you.

Theodore H. Frank:  I think that was Grover Norquist who said that.  There are those who say that the biggest free agent signing in New York in the last year was not Johan Santana and the New York Mets, but Catherine Sharkey moving from Columbia to NYU.  She is a prominent scholar in the area of torts, products liability, punitive damages and class actions.  She has published articles in leading law reviews.  As a student at Yale Law School, she was executive editor of the Yale Law Journal and served as a clerk for Guido Calabresi on the Second Circuit and Supreme Court Justice David Souter.  Cathy?

Catherine Sharkey:  Thank you, Ted.  Unfortunately, those who said that, I think, were all family members, but I appreciate the introduction, nonetheless.  I also appreciate the opportunity to come here and speak.  It is interesting:  I’m placed in the center of those actively involved on one side or the other of this debate, but it has been an issue that has actively involved me as an academic or intellectual issue.

I want to actually start where Michael ended with respect to the focus on Congress because I think a lot of what Michael said about Congress, with which I completely agree, raises a real paradox in the following sense.  When courts begin - and they have for time immemorial, basically - and say that the touchstone of preemption analysis is the intent of Congress -- that is where we begin.  And as Michael was saying, if that is where we are to end, then there is a real paradox if we also see that in practice, Congress again and again, as Michael suggested, enters into compromises where they decide that they will fight another day, and they will fight another day in either courts or agencies.

So if that is the case, and I believe it is the case in this realm, particularly the realm of preemption dealing with products liability, which is what I have specifically studied, that leads to this paradox with respect to a legal analysis of preemption that is going to begin and end with statutory text; in other words, that is not going to resolve all of the issues of preemption that are out there.  And they have been left to fight in another day.

And to me, the really interesting question is, given this acknowledgment - and we will get to in a minute - in my mind, actually, I might take a more radical position and say, actually, Congress not deciding these issues might be a good thing.  But even if we do not go that far, if we just say as an empirical matter, Congress, by repeatedly passing statutes where they say requirements, where they very much with a stroke of the pen, could say exactly what they mean with respect to common law tort liability, and they do not again and again, and the argument that they did not back in the 1960s and ‘70s before the explosion of tort liability in products rings a little hollow to me, at least, in present day.  It is very clear that there is a rife litigation landscape at times where Congress is either acting or amending statutes.

If we couple that - Michael said something else that I agree with; it is interesting - with the idea that Congress not only does not decide the issue when it is enacting these statutes, but does not get back in the game after the Supreme Court rules, there is actually a very good node, a student node, in the Harvard Law Review that came out recently that does this systematic study, that looks at preemption cases.  It takes on -- actually Michael and John clicked - did a very nice empirical study of preemption cases in the First and Second Rehnquist Courts.  And this node looks at all of the cases from 1983 to 2003, 127 cases.  Unfortunately, the punch line is not going to be as good because I cannot remember.  It is something - it is less than a handful.  It is something like two or three instances in those 127 cases, and he is looking particularly at all decisions where Congress overrules preemption determinations by the Court, and it is really negligible.

So the idea that we can rely on Congress either ab initio, or after the fact, to be resolving our preemption conundrum to me rings very hollow.  So to me, the interesting question is -- I read this as Congress is affirmatively punting this issue, and to whom?  Is it the courts or to agencies?  I think it is worthwhile for us to think hard about which of those institutions where the balance between the two we would like to resolve these issues.

So we start -- at least I start from the premise that text is not going to be able to resolve our questions.  I agree again - there is a lot of -- I agree with much of what each of the previous speakers has said, at least, as a descriptive matter.  I agree that the Court, particularly Scalia’s opinion, which I find very curious -- I always find, though, the preemption decision that comes out eight-to-one is going to have a lot of curious reasoning because they have cobbled together a majority, and justices that have very different views on things like once you get beyond the text, the presumption against preemption, deference to agency views, et cetera. 

But here, I agree that the Court seems, at least in this opinion, to be saying we can rely just on the text, but I do not think that is actually the case.  In fact, in Watters v. Wachovia, again, a case Michael alluded to, which was decided by Justice Ginsburg, focusing on the text of the National Banking Act, it is interesting that there is a vociferous dissent written by Justice Stevens, joined by Justice Scalia, that says whatever the Court says this is a case about an administrative agency’s power to preempt state laws.

So there, what this all leads to me is that I feel as though in preemption at least - this is not an argument against textualism writ large - in preemption because of the empirical reality of how Congress is writing these statutes, that this focus on text is inevitably going to cloak something else.  Again, to my mind, it is worth thinking about what that is.  So we had today -- we have had where the text does not resolve these issues.  We have competing canons of interpretation. 

One has been the presumption against preemption.  It looks like this one is getting interred, and I think that is good thing, actually.  In the area of products liability preemption, it is completely haphazard.  Which of the cases this presumption rears its head, which one it is entirely quiet, does not come to the floor. 

It has been absurd in my mind that in the expressed preemption cases where there actually is some actual language that Congress has written in, saying that there should be some preemptive effect.  That is where the Court has reared the head of the presumption against preemption, but in the implied ones where you would think you need some statutory interpretation canon, it has been relatively silent.  So it has been deployed at this point too strategically, I think, as an empirical matter.  It has not been doing actual work, so I think it is a good thing that that presumption seems to be getting buried at the moment.

The second one would be to bring up a metaphysical point, I suppose.  I would call it administrative law point, not metaphysical, necessarily -- would be Chevron deference or mandatory deference to agency interpretations of statutes.  Here again, I find it very interesting that in a study that I have done at least in the products liability realm -- and Riegel, I would now add as a new data point for this study -- in every single one of those cases, save Bates - so put Bates to one side - in every other one, the Court’s decision aligns with the position argued before it by the relevant agency.  There are not that many.  Interestingly right, the U.S. Supreme Court this term, and now in the carryover to the next term, is hearing three products liability preemption cases, and there are about half a dozen from 1992 with Cipollone up until Bates.

But in those cases, and just to run through them very quickly, the FDA’s position, which was a pro-preemption position in Riegel and in Buckman, the Court goes that direction.  It is an anti-preemption position in Medtronic.  The Court goes in that direction.  NHSTA, The National Highway Traffic Safety Administration, argues in favor in Geier, and the Court goes that way against in Freightliner, and then the Coast Guard argues against preemption in Sprietsma.

Interestingly, just to not hide the ball here, there is definitely an issue with respect to potential political flip-flop going on here aside from Sprietsma.  This is why this is the case that enrages many - I’ll put it this way - many conservatives who might otherwise be okay with the idea of Chevron deference to an agency.  What do I make of this?  The Court is not always up front about its reliance on agency views.  It certainly never has said that it is giving Chevron deference, and it is very cagey.  In Medtronic, the Court was substantially informed by the FDA’s regulation 808.1D1. 

In this case, it is interesting.  Scalia loved this line; Scalia goes through trying to evaluate what to make of this regulation.  This is the FDA’s own regulation interpreting the preemptive scope of the expressed preemption clause in the Medical Devices Amendment, which is at issue in Riegel.  So he says, after saying how it can go either way, “All in all, we think that 808.1D1 can add nothing to our analysis but confusion.”  So that is not a strong signal that we are going to be relying on what the agency is saying.

Moreover, I also find this interesting that even though in the majority opinion, he says he has found it unnecessary to rely upon that agency view because the statute speaks so clearly, he then goes on to say - and this is interesting language - “If, however, we had found the statute ambiguous and we had gone there, we would have given Skidmore deference to the agency’s view here.”

To me, that is important because it, again, accords with a model that I have been trying to develop, which I have called an agency reference model.  In other words, I actually think the Court should look to agencies for certain information in deciding preemption questions.  I think the Court should distinguish between the agency acting in its regulatory review process, and I think the Court has done that.

So for example, the distinction between the Lohr case, the Medtronics v. Lohr and Riegel -- I think it is a very strong distinction; it has to do with the type of review the FDA was giving to the particular medical device.  In Lohr, it was substantial equivalence.  The FDA was not involved in doing safety and effectiveness review and, therefore, it makes a lot of sense to me as a matter of regulatory policy, not statutory language per se, that there would not be preemption in that case. 

Whereas, if it is true that the FDA goes through a rigorous process and does the risk-risk analysis and does the same type of analysis that we would be asking a state law jury to do, then it makes sense that there would be preemption.  Here, interestingly enough, I agree entirely with Brian that if our system were different, in other words, if we were basing liability in the medical devices or the drug context for design defect or failure-to-warn as opposed to manufacturing defect -- and remember, manufacturing defects claims are not preempted -- I actually think that is the right conclusion as well. 

But for design defect and failure-to-warn, if we had a system that was a no-fault system, that was a system that says, “You should get compensated, like, say, under the Vaccine Act that Congress has passed or something else,” if that were the system, then there would be no inherent conflict between having federal requirements as to what constitutes a breach of this risk-risk analysis and getting compensation to injured parties.  But given the system that we have and the theory of basically negligent-based liability for design defect and failure-to-warn, I do think there is a conflict there.

So in other words, I actually think that is the right way to see the preemption analysis there.  The reason why this agency reference model is what I’ll describe as a middle course approach is because it would distinguish between looking at the regulatory action taken by the relevant agency and just what I’ll call their interpretive sphere.  And in the interpretive sphere, I think we need to be a little bit wary. 

We need to be a little bit wary about agencies that will do things -- for example, will propose a preemption preamble that does not go through notice and comment review; that switches position from when the notice of the proposed rule goes out and the final rule in the following sense.  Because my model really wants to locate as a normative basis that the agency might be the right place; not Congress, but the agency might be the right place to have the rife law-making type debates.  If that is so, then we have to make sure that all the relevant interests come in.  For example, those who are interested in state’s rights and state’s interests, et cetera, those might be better represented in proceedings before agencies.

I tend to think, notwithstanding the greatness of Herbert Wexler -- and it is hard to say this, given my betrayal from Columbia to NYU of late -- but the idea that the political safeguards of federalism, the idea that state’s interests are going to be so well-protected in Congress rings a little hollow to me as well.  It seems to be that a lot of state interest groups, including the National Association of Attorneys General and the like have actually caught on to the idea that they might fight some of these battles before the relevant agency.  So if that is the case, then it is particularly important to make sure that these debates are open to all of those interests coming in.

I also think that we need to be worried -- and here again, I actually agree with the point that Brian made with respect to -- I do not think, for example, in the drug context that manufacturers should have blanket immunity based on initial approval of a drug.  If it turns out that the relevant risk that have injured the particular plaintiff has not been considered by the FDA, it is a new risk that comes to light after the fact, not one that was either looked at by the FDA at the time of initial approval, or manufacturers can go back to the FDA.  In fact, in the Wyeth v. Levine case that is going to argued next term, that is in fact what happened.  The FDA had another juncture at which the manufacturer went back to it to weigh in on that risk.  You could come up with a position that is not all or nothing on preemption based on initial approval.

This gets me back to one of my points about -- if that is so, if the right position is the middle course position that I believe actually is right, then Congress might not actually be able to decide preemption as an all-or-nothing matter.  It might as a normative matter make sense that this is a more context-specific type determination.

I think the last thing that I will just say is it is interesting to note Riegel has been decided.  We have these two pharmaceutical cases pending.  The one that is being argued on Monday is about kind of narrow doctrinal issue with respect to claims of fraud on the agency, the Wyeth v. Levine.  It is a much broader case about what to do with respect to when a drug label has been approved by the FDA, are state tort claims preempted. 

It is worth at least emphasizing that there are some significant differences between the Riegel case and medical devices and drugs.  I guess we will get back to the first point where we started.  You might fear an obsession on statutory text and language as soon as you realize that the Medical Devices Amendment Act has an expressed preemption clause that is entirely absent from the provisions dealing with drugs.  So for that reason, there is a significant distinction there and the Court plays that out in the Riegel case.

The other thing that I just find interesting is in Riegel I do not think too many people were surprised.  Some people might have been surprised, for example, that Justice Stevens did not dissent, and instead, concurred in the opinion.  But I do not think too many people were surprised that the Court went in the direction that it went because the lower courts were basically all aligned in terms of lining up that the expressed preemption clause of the MDA should preempt common law tort claims.

In the drug context, I, for one, was actually surprised that the Court took the case because, in fact, in that context, the Third Circuit is the only circuit that heard arguments on the case.  As people like to say, it is not an issue that has percolated among the circuits, et cetera, and it seems to be a little bit more of an aggressive move on the part of the Court. 

I found at least some of the tea leaves about Justice Ginsburg’s dissent, for example, talking a lot about the drug context in this case, and Justice Scalia making clear that the Court had not yet decided that issue and yet at the same time emphasizing that there was a difference with respect to the expressed preemption clause dealing with devices and not drugs quite interesting.  Thank you.

Theodore H. Frank:  Thank you.  When these preemption cases get argued, on one side you will find Brian Wolfman and in the other you will often find Dan Troy.  He is a partner at the Washington D.C. office of Sidley Austin LLP where he is a member of the Life Sciences Practice in the Appellate Litigation Group.  He is the former chief counsel of the FDA, the Food and Drug Administration, in addition to providing strategic counseling on FDA-related matters. 

He practices administrative and constitutional law and litigation with particular focus on industries regulated by the FDA.  He served in the Office of Legal Counsel of the Department of Justice, and formerly clerked for the Honorable Robert Bork on the D.C. Circuit.  He is the author of Retroactive Legislation for the AEI Press.  He has been published in many journals and newspapers.  Mr. Troy?

Daniel Troy:  Thank you very much.  To start out with a full disclosure, my conflicts are just too deep and too broad to declare, so should you just assume that I’m deeply, deeply conflicted, but the views I’m going to state are my own.

Before I opened a bottle of champagne last night with my wife to celebrate the Riegel decision, I did try and go through Catherine Sharkey’s incredibly prolific production:  Preemption by Preamble: Federal Agencies and the Federalization of Tort Law - this just the last few years.  Federalism in Action: FDA Regulatory Preemption in Pharmaceutical Cases in State vs. Federal Courts, The Fraud Caveat to Agency Preemption, the 80-page Products Liability Preemption: An Institutional Approach.

And I have two thoughts.  One is, Catherine, if I had gotten the start, what would you be writing about?  The second is after I heard about your free agent signing --

Catherine Sharkey:  Punitive damages.

Daniel Troy:  And your betrayal of Columbia, I figured you owe me big.  But I did want to chat a little bit about how some of this got started because I do think that it addresses the real-world point that Brian has tried to bring forward.  But the real-world perspective that I want to bring forward is the real-world perspective from the perspective of the agency.

The first case I’m going to talk about -- and I know it is not technically a tort case, but it is a case called Dowhal.  This is a case where the product at issue was a nicotine replacement therapy, a product that would help people stop smoking.  Under Proposition 65, a plaintiff’s lawyer actually tried to get this California Attorney General to take up this cause and I think he at the time refused.  So they went to the agency and said, “We want a warning that requires that the nicotine replacement therapy say ‘Nicotine can harm your baby.’”

FDA, in a citizen petition response and then later in letters to the company, said, “We do not want that warning.  That warning might cause pregnant women to think that smoking is just as bad as this product, or worse, that this product is just as bad as smoking.  We want a much more nuanced warning.  We want a warning that says, ‘Try to stop smoking without this product.  See a health case practitioner.  This product can be useful.’”  And then it did say something about the potentially harmful effects of nicotine.  FDA tried to make this clear as possible that this was the warning that it wanted.

Well, the California Court of Appeals said, “This is Prop 65; California gets to speak.  More warnings are always better.  Put that warning on the label.”  This, as you might imagine, caused no end of consternation to the federal officials, the career officials at the Center for Drugs who, after all, think that they are the FDA, and they get to decide what the warning should be on products that they regulate.

So I actually had the situation where the lawyer for the civil division called me up and said, “Would you mind if we asked the California Supreme Court for time to argue in the California Supreme Court?”  “Yeah, sure.  Fine.”  So they went on their own motion, sua sponte, so to speak, to the California Supreme Court and, fortunately, the California Supreme Court, 9 -0, reversed and said, “More warnings are not always better.  There is such a phenomenon as over-warning.”  FDA gets to decide these questions where it has, in the language of Chevron, directly spoken to the precise question at issue.  So there we are - off to the races.

The next case that FDA got involved in, and I will not go through all of them, was a case much more controversial, I think, involving antidepressants.  People who are depressed tend to, tragically, commit suicide in somewhat higher percentage than the rest of the population.  And so it has always been very difficult to tease out the relationship between treating people who are depressed and suicidality.  And there is also -- it has long been known that there is some thought that when people first start to be treated, they are able to get out of their torpor but, potentially, they could use that time to take harmful action against themselves. 

Anyway, the advisory committee had considered whether or not to put a suicide warning on an antidepressant - in this case, Zoloft - and they looked at the data and they said, “No, we do not want to warn people away from taking these products.”

Both Public Citizen and the scientologists do not like these products and time and time again, went to the agency and said, “Put a suicidality warning on the product.”  And the FDA looked at it time and time again, at least two, three or four times and said, “No.  We do not think that the data is there.  You cannot, do not put such a warning on the label.”

So the case arose when someone started to take Zoloft and within six days tragically committed suicide.  And the essence of the case was that Pfizer was negligent for not putting a warning that this product can cause suicide.  This was a case in federal court applying state law.  The federal court said, “More warnings are always better.  They should have done it.  No preemption.”

The Ninth Circuit ended up punting and decided the case on causation grounds.  And so the Ninth Circuit never addressed the preemption question but the FDA had submitted a brief saying, “Look, we are FDA.  We get to decide.  We looked at this time and time again, and we do not think such a warning should be on the label.  And, in fact, we think it would be misbranded if you did this.” 

And so that brief was relied upon by others and that really set off the entire, let’s just say, industry of people actually coming to the agency and asking the agency to get involved in these cases to which we most of the time - 98 percent of the time - said no.  We, FDA, have to have something at issue.  Our decisions have to be undercut, and that did lead to FDA deciding to try and instead of communicate these issues on a one-off basis, an amicus brief here, an amicus brief here to write what I call the one-size-fits-all amicus brief.  And that is by addressing the issue in the physician labeling rule. 

But the reason why it fit into the physician labeling rule so nicely was because what happened in the physician labeling rule was FDA proposed something that is very rational.  That is, they wanted to rewrite, you know, the physician labels that package insert and the tiny little type that is almost incomprehensible written for doctors.  But it really tells a scientific story and starts out with the way the drug works and the pharmacology, and it really is not formulated for easy access.

Also, over time, the label had gotten from the perspective of people at the FDA junked up.  There were all sorts of stuff that the lawyers would advise people to put in and so the labeling went from being a communication from FDA physicians to physicians to basically a litigation avoidance device.  What we now know about risk communication is that less can be more, and that having these prolix labels does not actually help the public health.  It is not good risk communication.

So FDA had proposed these highlights, meaning it had to be no more than a half-page that would really capture all of the highlights of the labeling.  The industry came in and said, “It is a great idea but we are going to get killed in product liability litigation.  We are going to get killed because every case is going to be -- it should have been in the highlights and it was not.”

So the issue was very much teed up in the -- it is true that the NPRM had said, “We do not think that there is going to preemption,” but it is not uncommon based on the comments for an agency to change its mind.  And as long as you are within the scope of what there had been adequate notice of, and there was sufficiently adequate notice that more than one comment addressed this issue, it seemed appropriate to address the preemptive effect of the physician labeling rule, which, again, was an attempt to recapture the labeling from the lawyers for doctors.

The way we got into the device context is, perhaps, an even more interesting story.  Many of you know that there are three different classes of devices.  There is Class One; we do not need to talk about that; those are everything from Q-Tips, scalpels; things that are very pretty basic, relatively low-risk devices. 

Then there is Class Three, the PMA products, pre-market approval products, which are as extensively regulated as drugs and you have to submit a PMA, which is very much like a new drug application; many years of clinical trials, many documents, et cetera, et cetera.  And then there is this area of 510(k)’s, which means you are substantially equivalent to something that had been on the market before 1976 which is -- so it is called a 510(k).

Well, FDA’s primary mission is to say your PMA device, your this, your that, and in this case what happened was the company had filed a PMA; FDA approved the PMA, and the plaintiffs came in.  In trying to avoid preemption, they said, “This is not a PMA device.”  They had an FDA expert.  He said, “This is a partial PMA, partial 510(k) device.”  Well, there is no such thing, and the state court decided to submit the question of whether this was a PMA device or a 510(k)/PMA device to the jury.  And we at FDA said, “Well that cannot be.”  So we filed a statement of interest. 

Eventually the court kind of backed off on that, went to the state Court of Appeals.  And that led to the government’s views being requested in the Horn case and that led to the FDA declaring its view that PMA regulation was preemptive.  The reason why FDA is taking the position that both PMA and NDA regulation is preemptive is because this is the fundamental debate in this area:  Is FDA’s regulation just a floor or is it a floor and a ceiling? 

And FDA -– Rich Cooper, who is the Chief Counsel under President Carter and other chief counsels long before I ever came on the scene, had written that FDA regulation is both a floor and a ceiling.  Why?  Because FDA is balancing the benefits and the risks, and it is trying to come to a happy medium about what the appropriate level of warnings are.  And it makes, as I think Bryan said correctly, a risk-benefit decision that the jury is simply incapable of. 

The jury is presented with the one person who is injured; that is why they were there.  And they see a person who has been injured and often an out-of state big company, and they do not and they are not really institutionally capable of making the kind of balanced risk-benefit judgment that FDA makes.

Now FDA knows when they put a product on the market that there is some number of people who are going to be harmed by it.  All drugs and all devices have risks and so that is why if one had a no-fault compensation scheme that might make sense, and we could have that public policy debate.  But as Catherine said, we -- our current tort law is based on wrongs.  And so long as the situation is that a company must be proven to have done something wrong before someone can recover then I think there is, as Catherine said accurately, a conflict because a company has not done anything wrong if they disclose everything about their product to FDA and the FDA approves that product on, again, understanding that it is going to have some adverse events.

Now I do want to say -– the other thing is this notion that, somehow, because of whatever is submitted at the outset then there is preemption forever.  I do not think that is accurate.  I do not think that is what anybody is claiming.  FDA does, especially after the recently enacted FD3A, have incredibly extensive, post-marketing authorities.  Companies have very extensive post-marketing responsibilities; they have to submit adverse event reports within certain periods of time.

And I do not think anyone is talking about wholesale immunity based on the initial approval.  I do not think that FDA has argued that and I do not think that the companies are arguing that.  Certainly, again, manufacturing defect and other kinds of claims I think will continue to survive whatever the outcome is in Riegel and in Wyeth.

And so I do think that Catherine in her extensive writings is on to something when she talks about looking at this agency reference model.  But the agency, in this case, FDA, took a look at the effect of product liability and decided that it has some pretty harmful effects.  The most serious concern is, again, this over-warning, defensive labeling.  And, again, the Physician Labeling Rule was in part meant to address that.  But I think it is fairly unassailable that the current product liability environment does impact innovation.  As the AMA observed, “innovative new products are not being developed or being withheld from the market because of liability concerns or inability to obtain adequate insurance.” 

The current liability environment does clearly impact price; it is one of many reasons why prices are higher in the U.S.  And then for the some products, it is almost entirely the reason why prices are higher in the U.S. and it does impact the availability.  A product called Bendectin, the only treatment that FDA ever approved for morning sickness is not on the market here in the U.S., is in the market in Europe only because of product liability.  Another product, a contraceptive product, Norplant, is off the market only because of product liability concerns.  Now some of the social conservatives they might not think that is so bad, but we will leave that aside. 

And the idea that product liability does not impact decision making, I will just say again, from the real world perspective, is totally contrary to my experience.  I primarily do FDA regulatory law and counseling but I am not a hardcore product liability lawyer.  I’m really -- do obviously know this area.  But I cannot tell you the number of times that I have on the phone with me one of my product liability lawyers, and often the advice they are giving is different than the advice that we FDA regulatory lawyers are giving because every company lives in fear and terror of product liability.  So the idea that it somehow does not impact decision-making and only lays the way down the road after a company has been hammered the way Merck was on Vioxx that that is the only time that it impacts decision-making I think is just wrong.

A couple of last points on Riegel, actually -- this is a terrible thing to do but I actually hate to disagree with Michael Greve about something.  But the Medical Device Amendments, you know, Michael, are not so ancient.  They were in 1976 and they were actually after the litigation explosion involving the Dalkon Shield.  It is fair to say that the Food, Drug, and Cosmetic Act is an ancient statute back in 1938 and so you are not wrong in your fundamental point. 

But the Medical Device Amendments, I think, one of the reasons why it did have an expressed preemption provision is because there was that experience with respect to litigation.  And one of my favorite lines in the Riegel decision is when Justice Scalia says, “If we were to speculate about congressional motives, which of course we should never do, the only indication available, the text to the statute suggests that the solicitude for those injured by FDA-approved devices, which the dissent finds controlling, was overcoming Congress’s estimation by solicitude for those who would suffer without new medical devices if juries were allowed to apply the tort law of 50 states to all innovations.”  I have to say the fact that at least seven Justices signed on to that statement gives me some hope for the Wyeth case.

So to close, there are some points about which I clearly agree with Catherine.  I am not at all sure and I’m not sure I have ever asserted that FDA’s pronouncements on this issue, whether I am there or not, should warrant Chevron deference, at least the kinds of things that went into the Preemption Preamble.  I also think as the Preemption Preamble said - and as I think we are going to be arguing on Monday - that if there is fraud on the FDA and FDA has actually found fraud, that you should not necessarily get off scott-free in that context.  That is consistent with another one of Catherine’s very impressive articles.

And so what happens on a going-forward basis?  I do think that we are now going to have an experiment, at least in the device context, with a sort of high-regulation - I hope and think - low-litigation model.  Does that impose certain obligations on industry?  I think it does.  I do think that one of the reasons why there are at least a fair number of people even in the -- people who are very pleased with the FDA’s position is because it really drives companies to come to FDA as quickly as possible to insulate themselves from liability, to be able to show, “Yes, we told FDA everything.”  And that argument gets a lot of traction at the FDA and so industry does have an obligation, I think, when they find out new information to communicate with the FDA.

I also think that the industry will continue as it has to try and get as much money for FDA as possible.  The idea that the industry is trying to hollow out the FDA is just not consistent, again, with the real world.  There are two -- not one but two coalitions to get the FDA more money, which span everywhere from the biotechnology industry to -- I think the Center for Science and the Public Interest is part of the coalition.  I mean, we have got people from Secretary Thompson and former Secretary Shalala all trying to urge Congress to give the FDA more money.  I think that industry has come to recognize that when there is a strong regulator that helps them, it does not necessarily hurt them. 

Note that a lot of these attacks on the FDA, which are often cited by the plaintiff’s bar, only seemed to happen after FDA got involved in the preemption debate.  Before that, particularly during the Clinton administration when the FDA was working hand-in-glove with the plaintiff’s bar, the plaintiffs’ bar seemed perfectly happy with the FDA.  So in closing, I do think that we are going to have an interesting time ahead.  There is no doubt in my mind that the tort bar will try and overturn these Riegel, and whatever happens in Wyeth on the Hill.  And so the one thing that we can be assured of is that Catherine will continue to have lots of fodder to write about.

Theodore H. Frank:  I want to give the earlier panelists a chance to respond to what they have heard after them.  We will start with Michael and work our way down the line.

Michael S. Greve:  So I have been shown up now by Dan Troy and really embarrassed by Cathy Sharkey.  I did not know that Harvard Law Review article.  Which goes to show not only do I not read the Harvard Law Review; I do not even read their student notes that are about or based on my own articles.  Other than that, I am a real scholar.  I want to start by agreeing with Brian at a conceptual level when he says they are cramming an awful lot of meaning into a single word - “standard requirement.”  It then turns out that the word “requirement” prior to this decision at any rate does not mean the same when it appears in different statutes.

I want to give you a different example where I think that conceptualism of that kind of stuff - this clause-bound statutory or regionalism or whatever you want to call it - got away from the Court.  And my example is Sprietsma.  If you look at that case, okay, what it is about is about the Federal Boat Safety Act, okay?  And the question is if the agency fails or deliberately decides not to establish a federal boat safety standard, is that preemptive of state tort suits that are predicated on the design defect, which says you fail to have something that the agency explicitly declined to require. 

And in that case, the Supreme Court said it is entirely consistent to have these state tort lawsuits running next to the statute.  Why?  Well, among other things, the difference to the Geyer [phonetic] case and some of the other cases is this:  In those cases, the statutes speak of preempting state law.  Here, the statute, the Federal Boat Safety Act, speaks of preempting a state law or requirement.  And by a state law, they must have meant something like regulatory requirements and statutes as opposed to common law duties.

And Steve Shapiro, who argued the case, knew this in advance; I mean that the word “a” would be a big, big problem for him.  Now you look at the substance of the case, what this is actually about.  What did Mr. Shapiro actually want from the Court?  What he argued was something like this:  Look, dear Justices.  Suppose there is no preemption here.  In that case, if we do not put this required item - a propeller guard - if we do not put that on the boats or on the motors in the wake of this ruling, we are going to be sued out the wazoo.  On the other hand, if we do put the propeller guards on, then we will also be sued out the wazoo because there is this Coast Guard decision from the ancient past which said, “You know what?  We should not require these propeller guards because they might do more harm than good.” 

So we are now actually putting something on these motors that says it will likely kill more people than it will save lives.  That is punitive-damages land.  So we are damned if we do and damned if we do not, and you are telling me that is the real purpose and intent and the real meaning of the Federal Boat Safety Act?  The only people who are rendered safe by a ruling to that effect are the trial lawyers of America.  That cannot be the meaning of the statute.  And the Supreme Court - none, zip - was totally unimpressed by that argument, that is to say, by the structure of this statute and what has to have been its actual intent.

There are then cases where the text, if you parse it, runs against the architecture of the statute as a whole and it matters big time whether the Supreme Court looks to one or the other, and there are lots of cases that go to that.  I also want to agree with Brian but disagree - just like Dan - with the defense lawyer that he cited.  I would not fear a starving of the FDA.  The reason why I’m a little bit nervous about sort of ruthless across-the-board preemption, weird as it sounds, is that I think that will increase the political pressure on the FDA to approve nothing at all unless and until it is absolutely, totally 100 percent certain that nothing can happen because there is not this “compensation” at the end. 

For that reason, I have considerable sympathy with Cathy’s well thought-out position that would leave some room for tort lawsuits in these kinds of cases.  Probably, unlike Cathy, I’m not particularly scrupulous about the accuracy of these verdicts.  I do not really care whether the guilty parties get punished and the harmed parties get compensated.  It is enough that somebody gets punished and somebody gets compensated at a level that is high enough to tell the American public everything is in good order; the bad guys are getting punished.  And there is compensation at the end.  That is all I need.

The reason why I’m a little bit nervous about that is, as Brian conceded, the compensation itself has -– it is inextricably intertwined with the deterrence or regulatory function.  And there you have to worry about distorting the calculus and about a mechanism, namely, compensation for injured parties that tilts the balance one way.  There are also victims on the other side.  There are dead people because a drug does not get out fast enough, does not get approval for enough uses.  We just do not know where they are, who they are, who compensates those people and their descendants.  So one way or the other, you bias the process and I’m always worried about an error, correction, or compensation mechanism that can cut only one way.

Theodore H. Frank:  Brian?

Brian Wolfman:  Just a few things.  You know it is funny because that is the -– I think it is important if you are going to address what I think is -– I will just put it a little stronger than Michael that really impoverishment of this textual analysis.  Again, I mostly agree with Justice Scalia about the use of legislative history.  Again, I want to make clear that is not what I am talking about.  I’m talking about the ridiculous focus on this word or that word and imparting meaning that could not possibly be there.  Actually, I think it is important to use examples that are both pro-preemption and anti-preemption, if you will.  To make the point in the case, I was going to point to Sprietsma. 

Let me just talk about the case for a second because I actually think the result was correct and I think it was an easy case, and that is why it was nine-nothing.  And I think it was drawing on a different part of the jurisprudence, which is that inaction is not sufficient.  Regulatory inaction, generally speaking, unless the agencies occupy the field - no one claimed that, in that case is insufficient to preempt. 

There was a case that I was somewhat involved in; it was an earlier case.  Both preemption cases that I was involved in had gone up to the Court, but the case was argued and then settled after argument, I think because the defendant saw the handwriting on the wall and saw it the next year with a nine-nothing decision.  But the plaintiff’s lawyer in the Sprietsma case asked me to go with her to talk to the Solicitor General about the case.  And, actually, just a slight correction to Catherine:  Actually, it is true that they agreed with the government.  They did not agree with the agency; the Coast Guard would not sign the brief.  It was filed just by the Solicitor General because actually the Coast Guard was in favor of preemption in that case. 

But what happened was in that case I think what occurred was that the -– we talked to the Solicitor General about this and I think one of the things we convinced the Solicitor General about is that the level of inaction there, this Coast Guard letter, the only thing that can be pointed to for inac -- for action, because there was no formal regulatory statement, was just not sufficient as an administrative law matter; it did not have the requisite formality to speak with the force of law.  And I think that is what was really going on in that case. 

But the notion -- and I did not know Steve Shapiro had it on his mind because it did not come up at argument; it was not in a single brief; no one had focused on the word “a.”  I mean it was almost -- to me it was like they pulled it out of their hat.  I mean, it was ludicrous.  I think that even though my side of the case won the case, just the notion that that is -- you know that the Congress sat there, and said, “Well okay, yes, we are going to use the singular and that means there is no preemption of common law claims, but there is preemption of positive law.”  That is not a serious argument in my mind. 

So I think this works both ways and I think if -– I’m not an academic but I would like to see the academic literature move in that direction and talk about that the mode of analysis strikes me as just not sufficient to deal with what I think is a very serious problem.

You know there are obviously a lot of things that people said with which I do not agree here but most of them are small things.  But I will just deal with one more of them, which is that I want to make it sort of a partial concession, anyway.  You know, the examples that Dan talked about at length seem to me to be the most difficult cases, right?  I have briefed these cases and I will not bore you.  I have arguments, of course, about why there ought not to be preemption in the Dowhal case.  And the next case he talked about is a case called Motus. 

These are situations where you could argue about the level of formality at which the agency expressed its view on whether that left all the formality sufficient to warrant preemption.  But the agency did express a view about a topic and did express a view about a warning and said there ought not to be a warning.  And the plaintiff, it is true, claims that a warning at least approximating that warning would have prevented his or her injuries.

So that is a more difficult case.  But what is going on in, for instance, the Solicitor General’s brief that it was asked to file prior to the grant of certiorari in Wyeth v. Levine is quite a different approach.  That approach essentially says -- we do not have time to get into the nuance, but essentially says that you do not have to have that level of overlap between what the agency did and what the plaintiff is arguing to warrant preemption.  But that the regulatory approval itself is sufficient in most cases to warrant preemption, really, regardless of what the warning is that the plaintiff says would have prevented the harm. 

And that has the potential -- I do not think the Court will buy this, but that has the potential to go far, far beyond the Dowhal and Motus cases.  So that is really, I think, what was -- again, I do not think the court will go that far.  I think there is a hint of that in the Riegel decision, but I think -- I just want to be clear; it is quite different from those kinds of cases. 

After all, you know, preemption - let me just put it this way - preemption -- there is a place for preemption.  If the federal government says, “In a regulation you need to come up on the accelerator,” and there is a state regulation that says, “No, you have to go down the accelerator,” well, you know, that is a conflict and the supremacy clause is a choice of law principle that says when there is a conflict you choose federal law.

Catherine Sharkey:  So, I actually like to be challenged and stand to be corrected.  I tell my students all the time that they should try; it sort of keeps me on my toes.  But I have to say, Brian, I’m curious.  One of the things -- as Dan mentioned, I write as I have to as part of a works project in the academy, long law review articles.  And as a result or somewhere along -- as a result, they stay with you a long time.  So this Products Liability Preemption piece that I wrote, some of you might have heard me speaking about it probably about a year ago, it is still actually not out in print.  So that gives me the ability of rereading again and again my own 80 pages and the footnotes. 

And on this, I’m just going to quote because it is sort of -– as you said I got this wrong; I thought, “Wait, I have read this millions of times.”  I’m just going to quote from the Sprietsma opinion:  “In a case before us today, the Solicitor General, joined by Counsel for the Coast Guard, has informed us that the agency does not review its refusal to regulate as having any preemptive effect.”  So I think I’m right on that point, but I’m sure I’m wrong on many others.

The couple of points that I wanted to raise, though - and I will try to be brief so we have some time for questions - are, one, I actually think that this issue of accountability and, in fact, just a preview -- this is an issue that has become -- some of my colleagues have kind of an obsession.  I’m writing an article right now called “Agency Accountability Preemption in 2020” because I was asked to envision preemption down the road.  When I echo -– and I think accountability works in two different directions.  Dan mentioned something that I actually found curious when I was doing research on this topic that during the Clinton administration when the FDA was taking a very strong anti-preemption position there were very few people -– there was one scholar, Lars Noah, who was running some interesting things, saying that there is a real accountability loophole here. 

And I think it is worth remembering the idea that if we have an agency that we have decided we want, you know, the U.S. does not have a huge amount of ex ante regulation but the FDA, in doing ex ante regulations of drugs, we do a lot of.  So once we have decided that then the idea that we are going to kind of let a regulator off the hook by saying we put out these regulations, there are floors. They are not anything other than that. 

And so if we put out things and we approve them and they are harming people -- that is not our problem.  The real problems with that strong form approach on accountability grounds.  To me, the flipside is if we are going to envision a lot of this action taking place as the agency -- and I mentioned this in my opening comments.  And I disagree slightly, at least, with the public record information about what went on with respect to the notice of proposed rule-making in the preamble context.  It appears, at least, there was a lot of objection that certain parties representing various states’ interests, et cetera, did not have that opportunity to come in because they were not giving the sufficient notice and the original notification. 

And I think it is important to me that courts maybe are conditioning deference to agency action, looking very closely at the procedures that the agency has used, either to enact regulations or to allow for this kind of debate.  Part of this -- I mean, that part -- some of you might not like to hear.  Part of it you might like to hear is, to me, it seems like this kind of view of agency accountability will be much more robust with respect to consideration and protection of vital states’ interests in some of these cases, and will at least have to hear those and respond to them.

The final thing that I just want to say -- maybe it is returning a favor to Dan since we all –- he already went clean up, so he got to respond to all of us and maybe give him something to have to answer about and it actually echoes -– Brian talked about the Solicitor General’s brief in the Wyeth case.  The Solicitor General’s brief in the case that is getting argued on Monday is quite extraordinary in the following respect. 

This is a case we did not talk a whole lot about it - the Warner-Lambert v. Kent - where the issue is, as Dan was suggesting, and what the party -- what Warner-Lambert has argued is if there has been fraud perpetrated on the FDA, and the FDA has made such a finding, then state law tort claims based upon that antecedent finding a fraud should not be preempted.  But where there has been no finding of fraud, you should not be able to have a state cause of action that is actually litigating whether there was fraud perpetrated.  The Solicitor General’s brief, if you read it, actually takes a much stronger, more aggressive position. 

So this is a kind of caveat towards courts just deferring wholesale to whatever position the SG might be arguing.  Because they actually argue even where there has been such a finding of fraud by the FDA that any state law tort claim should be preempted.  And to my mind that is not the right position.  I find it a little bit extraordinary the effects -- I'm sure the parties in the case -- Warner-Lambert probably does not like this particular contribution from the Solicitor General because without going into too much detail that kind of an argument would really put at issue the viability of the statute that is a Michigan regulatory compliance defense with a fraud caveat.

And presumably, one of the reasons why it makes sense that you would keep the regulatory compliance statute with this fraud caveat is that the fraud caveat has some kind of meaning and that meaning would be that in cases where the FDA had found fraud, then those kinds of claims are not preempted.  So in my mind it is a little bit extraordinary and I would be curious to hear your views about that.

Daniel Troy:  Anybody wants to hear our views on that can hear it on Monday at the Supreme Court.  All I want to say is Michael made an interesting point about case selection and I think case selection does matter.  But I'm going to ask this rhetorically since Brian is on the briefs and I presumably know what he thinks.  But I will note that in the Wyeth v. Levine case, which will be argued next term, there was a situation where Wyeth had gone back and forth with the FDA on the mode of administration. 

What happened in this case was there was a product that could be administered in two ways through a push IV or a drip.  Actually, Wyeth had proposed that the first sentence of the warning -- well, what happened was there was a push-IV and it leaked and so the woman’s arm got gangrene and needed to be amputated, quite tragically.  And so actually Wyeth had proposed a label that said there were reports of necrosis leading to gangrene requiring amputation.  Following injection of Phenergan usually in conjunction with other drugs, the IV was intended in these cases but partial arterial placement of the needle is now a suspect. 

The Chief Justice of the Vermont Supreme Court who is in dissent says, “While the proposed change in the warning language may not reflect what plaintiff would require in a warning, it cannot be disputed that Wyeth’s proposed alternative warning, one, placed greater emphasis on the risk of necrosis and amputation, and, two, gave the FDA the opportunity to consider this specific alternative warning advanced by Wyeth, as well as the adequacy of the warning in general.  Despite this opportunity, the FDA mandated that Wyeth retain the language of the existing warning.”  So, this seems to me now -- I know that some can argue, “Well, that was not with sufficient formality.” 

But the way the FDA operates is very much in kind of a back-and-forth letters, sometimes even emails and a notion that FDA has to somehow make some kind of super-formal finding on what the labeling should say before there is preemptive effect would very much hamper the agency. 

And so I will conclude with a sort of curtain raiser on Wyeth.  Another quote from the Chief Justice’s very long and, I would say, ardent dissent: “In 2000, FDA confirmed its assessment that healthcare professionals should be permitted to choose IV administration in its various forms as a means of delivering the drug where appropriate.  Wyeth could not both list all forms of IVP administration as approved as required by the FDA, and exclude all or some forms of IV administration as unsafe as required by the by the jury’s verdict in this case. It would be impossible to comply with both requirements.”  That is often the net of these cases that what is being asked for in the state court is something that is entirely incompatible with what is in the content of the FDA labeling. 

Again, nobody is here defending a situation where some company is not disclosing information that they clearly should disclose.  And I should tell you there are criminal penalties attached to not giving that information.  Whether you need to add on to a tort liability to me is somewhat dubious.  Pharmaceutical executives are plenty afraid of the criminal penalties these days, but that is not what is at issue in the Riegel case; it is not at issue in the Wyeth case, and it is not at issue in the Kent case.

Theodore H. Frank:  Did you want to respond?

Brian Wolfman:  I do want to respond to that last thing, because I think that this is important.  Just to have a sense of the Wyeth case: the reason I say that the Wyeth case is quite different from, say, the Dowhal is because, truly, those are accurate quotes from the dissent. 

But there were trial court findings of fact and findings of fact that in the majority opinion that are just directly contrary to that on all those counts that there was never any discussion of the IV push before the FDA at all and the method of administration that was at issue here.  And that the FDA was not presented with nor rejected a warning on that topic.  Those are findings of fact.  And what is interesting about it is that one could, of course, go back and disagree with that; I do not think there is any basis -- I read the entire trial transcript; I do not think there is any basis to disagree with it in this case. 

But I think what was interesting at the cert stage in this case is that there is no question that Wyeth, thinking that the Dowhal - Motus mode of analysis was the one that they would want to use.  In other words, the more direct conflict, at least from their perspective.  They essentially wrote their petition - I think it would be useful to read it if you are interested in this - trying to in essence get those findings overturned.  And the challenge in these findings is that they were open about it.  I mean, they were not trying to mask that at all. 

But the Solicitor General in its brief does not come in and say the same things.  In other words, the Solicitor General’s brief does not say, “Oh, we considered this warning and rejected it.”  It does not say that at all.  In fact, the only thing it says is that Wyeth could if it chose try to contest those findings of fact.  But the SG did not go that route.  The SG says that, basically -- and you can see this is on page eight of its brief, and at the cert stage it has a heading that says, “Generally speaking, the approval of a label preempts all claims for failure to warn design defect.”  And so far it is a proposition that I do not think will be accepted. 

But if it is accepted, there is no appellate court in the history of American jurisprudence that has accepted that to date -- that notion to date.  And I think Dan does not dispute this.  He said even these other cases have not gotten up to the Court.  You see the heading on page eight, I believe it is.  So anyway, that is just to give you a backdrop on that case.  And so I think my only point here is really not to dispute the facts of that case but to say that at least from some perspectives that are going to be before the court in Wyeth, there is quite a broad submission being made and I do not think the Court will adopt it.

Theodore H. Frank:  Let’s open up to questions.  I will be having somebody come around with a microphone and if you can wait for the microphone and identify yourself.  And we have lost Michael, but you can ask questions to the other three panelists.

Larry Ebner [phonetic]:  My name is Larry Ebner.  I have a question for the panel, actually, a specific question that I’ll combine with a general question.  The specific question relates to the notification that Justice Scalia in Riegel gave to Congress that the word “requirement” is a pregnant word and that Congress should understand that the Court will be interpreting that word as encompassing common law duties unless there is a statutory indication to the contrary.  I'm wondering if any of you have a reaction to that. 

And my general question is there seems to be quite a few preemption cases that have been or are being considered by the Court at this time.  By my count, I think there are seven of them.  Maybe four or five of them -- product liability cases.  What do you think, if anything, is going on?  Is the Court trying to straighten out the muddle of the past 15 years, or what is happening?

Daniel Troy:  I’ll start.  I think that Congress should have been on notice that the word “requirement” encompasses common law duties.  In fact in Cipollone, there are five votes for that position in the lower.  And so, I think even Bates -- although it goes in the anti-preemption position -- even Bates says that.  So I do not see that as new news.  That is point one.

As to what is going on, I think Michael has written about this so I’ll channel Michael in.  Just pretend I'm, you know, I'm much more articulate with a funny accent.  I think that this Court is very concerned about lawyer-driven litigation.  And so much of product liability litigation these days is lawyer-driven litigation, and we have seen it in the securities context.  I think we are seeing it here.  I think that very few of the justices are persuaded that the way to ideally seek justice and to have a better world is through state courts that are hampered in some cases by these nightmare jurisdictions.

Brian Wolfman:  Sorry, a brief response to both.  I agree with what Dan says; I think that some sense that Congress could have been on notice as early as 1992.  But I think what Scalia has done here -- I do not agree with the decision, obviously.  I mean, people in this room now know I would not agree with that decision.  But I do think what he did is fair enough.  I mean he was more explicit now, he said it: “You are now on notice.”  That is fair enough.  But that does not undo the problem of having sort of infused this word with sort of a bizarre meaning and too much meaning before. 

And the trouble with Dan’s point is that none of these statutes were enacted after 1992.  So the fact that notice was given in 1992 does not really have anything to do with the outcome of any of these cases.  He could be just saying be careful in the future.

What is going on?  You know, I think Dan may be right that looking at all of these issues that have come out the last term - punitive damages, pleading requirements, preemption, so forth - it is possible to say but I think maybe too early to say.  But it is possible to say that the thing -- bring them all together, tying them together is sort of hostility to litigation.  But I think you have to be very careful.  First of all, the preemption cases have been much more of a mixed bag, okay?  And the one thing you can say about them -- I know Catherine from an academic perspective is trying to bring some theory to this.

But I think the thing you can say is that they are not driven by some overriding theory; that the devil is in the details of these cases, the details of the regulatory structure.  And that is why Bates is different from Sprietsma is different from Lohr, which is different from Riegel and so forth.  I think it is very hard to get a handle on the preemptionary of what is going on.

Catherine Sharkey:  I think it’s included in your packet. I wrote this little piece on Riegel and I made the point that Dan made as well that, to me, that question about whether state common law tort claims can be considered requirements.  There were five votes for that in Lohr.  There was some ambiguity on the question and I thought it was resolved as well by Bates.  So my reading of why there is this emphatic statement is because there are potentially some justices and the one that comes to my mind would be Stevens, who wrote Cipollone and then also wrote Bates that there might be some waffling or wavering.  So once you have a majority, you just make a clear proposition. 

But as I recall at oral argument, there was not a whole lot of argument on that question that they just moved on.  So I do not think that was a real big issue in this case.  I guess I should say I hope -- I'm not going to prognosticate as to what is going on.  I guess I hope the Court is interested in fashioning some kind of framework for preemption analysis because I think it is -- I would be distressed if what they come up with is we are just going to look to the statutory text for the reasons that I explained before.

And the final thing I’ll say is, again, I’ll reiterate I was surprised that they took the Wyeth v. Levine case because usually the Court waits for issues to percolate amongst the lower courts.  And, clearly, as it is deciding medical devices and drugs, to my mind it suggests that the Court is not going to just take a haphazard approach but is at least thinking if they are able to pull it off.  That is a different question about fashioning framework.  Otherwise, why take that case what seems to me a little bit prematurely, except that they are thinking overarchingly about some kind of framework?

Eric Lasker [phonetic]:  Hi, I'm Eric Lasker.  Let me -- I have a question and a preface to the question.  As I look back at the preemption case along the past 20 years or so, I do think, at least in my mind, there is a common theme that I think goes through all the opinions, which is regardless of expressed preemption or implied preemption or whatever the theory is, that the Court is looking at conflict.  And whether there is a conflict and then there is a discrepancy as to how you define what a conflict is. 

But with that backdrop and with the FDA’s current positioning in mind, I want to talk about deference to the agency because during the presentation we talked about deference to the agency just in one context, the deference in the interpretation of the statute.  But there are different positions that the agency has taken that have an impact, I think, on how the Court uses -- and you saw this in Justice Scalia’s opinion because he had two different views, one statute and one regulation. 

And that, of course, is going to be important in Levine because of these changes that affected regulation, which is essential regulation in the preemption context.  But there is also a third level which is, the agency’s definition or understanding of what is a conflict.  What causes them to have a difficulty in airing out their regulatory process?  And it goes back to what Dan was talking about, about labels being both – warning being both — a  floor and a ceiling.  Got to go up.  And so, I was wondering if you could talk about those three different levels of agency determination and whether or not there is regulations there is a doctrinal difference.  I’m not sure that there is one with [inaudible].

Catherine Sharkey:  [Audio glitch] just as Breyer’s probably the one who has thought the most about that last rubric.  Namely, he has talked in early [audio glitch] opinions without stating explicitly what level of deference.  But in terms of articulating why there should be any kind of looking to the agency is because the agency is well-poised to be the institution that is able to see whether or not there is any kind of conflict with the federal regulatory program. 

In my article, I actually argue that on that point I'm looking at whether or not we should be regulating something at a federal level or state level, and the kind of conflict the agency would be a good body to get some information on that question.  But so far, doctrinally, I do not think the Court has put that all together.

Daniel Troy:  You thought about tying that issue to the issue of primary jurisdiction?

Catherine Sharkey:  Yes, a little bit.  I mean particularly, I guess to my mind the primary jurisdiction point in the fraud on the agency context seems the strongest.  That is where I have begun to think a little bit about that. It is harder for me at least at the moment to -- given the FDA’s role in the federal interest in policing fraud on itself, that doctrine seems to fit more naturally there.  But one could think more broadly.

Ilya Shapiro:  Ilya Shapiro from the Cato Institute.  For those of us who are not products liability or FDA lawyers, I want to ask a slightly broader question about preemption and that is yesterday being preemption day at the Supreme Court and more cases coming along the way, seemingly every time the Court decides for preemption.  And I'm wondering whether there is -- what is to that?  Is that just a happenstance or is there a growing trend, or is that the way that it has always been and with this Court at least, based on yesterday’s cases, overwhelmingly going to be for the near future?

Catherine Sharkey:  We’re not the ones bringing the cases.

Male Voice:  The observer of the Court?

Daniel Troy:  I mean I think we have sort of addressed that.  I mean I do not know that the Court has a very strong pro-preemption bias.  I think I do not, you know, fall-off-your-chairs agree with Brian that a lot of this is really at the level of the details in two ways.  I think the regulatory scheme matters an enormous amount.  I think the argument -- my food clients would not like to hear this but, you know, the argument for preemption in the food context when federal government regulates food to a much lesser extent is much weaker than it is in the drug context where there are very few, if any, products that are more comprehensively regulated by the federal government than drugs have been; perhaps, nuclear power plants.  I mean very, very extensive federal regime.  So that is one thing that matters.

And the other thing -- a