The first half of the article is where we thought we were being flattered. The author, a 2011 law graduate, got the bright idea of trying to analyze empirically a phenomenon that we’ve simply accepted – that we’ve referred to on more than one occasion as “strange things happen in tort preemption cases.”
Like the article, we include the Supreme Court in that statement. And we also agree that one of the strangest of the strange things that happens in tort preemption cases is what the article charitably calls Court’s “intermittent” resort to a presumption against preemption. 5 N.Y.U.J.L. & Lib. at 750. That presumption is one “doctrinal element which causes uncertainty and confusion in the lower courts and among academics.” Id. It confuses us, too, we might add - at least when we don't think of the presumption as something employed like a drunk uses a lamppost, for support rather than illumination.
Specifically, the article recognizes that the tort preemption defense, both because of the power of the defense and the increasing politicization of tort law, tends to draw highly ideological reactions from our purportedly “neutral” judiciary. “Preemption has invited strongly ideological reactions, and has been characterized as a fundamentally political issue.” 5 N.Y.U.J.L. & Lib. at 752.
But unlike us, this article purports to light a candle rather than just railing at the judicial darkness (there's got to be a connection to that lamppost somewhere) to measure the extent of the politicization of preemption using several parameters:
The empirical question I seek to answer is whether these factors have influenced judges in defining whether a claim is parallel. I look to certain variables, including the type of court deciding a case, whether the litigants live in the state where the judge sits, the political party of the judge or the person who appointed them, and factors which correlate with ideology like race and gender.Id. at 747. So the author's looking at whether it’s a state rather than a federal court, whether the court is favoring an in-state party (plaintiff or defendant), whether the judge was elected (or the person appointing an appointed judge) as a D or an R, and race or gender.
That seems OK to us. Preemption is a matter of federal law overriding state law, so it’s logical to think that state judges may have a different perspective than federal ones. All other things being equal (which, of course, they rarely are), we’d think a state court would be more protective of state law and a federal court more amenable to application of federal law.
“Home cooking” is another issue, although plaintiffs forum shop and defendants remove cases for strategic reasons, so “in-state” is subject to systemic manipulation. Lot’s of out of state plaintiffs flock to favorable state court forums – particularly in Philadephia and New Jersey – that also happen to be the home states of pharmaceutical defendants. This is sufficiently prevalent that our clients are starting to leave those states. With all the tactical considerations that go into where a plaintiff files a suit, and where it ends up, we wouldn't put too much weight on the stats. It's pretty clear, from both sides' behavior, which courts real lawyers think favor which side, and "home cooking" is only a minor factor.
Political party – that’s pretty obvious. Most plaintiffs’ lawyers tend to support (and more importantly, give money to) Democrats, and most Chamber of Commerce types do the same for Republicans. Most of our judicial races (except retentions) here in Pennsylvania are partisan elections, although cross-filing is allowed, so we’re well aware of the peculiar “iron triangle” of lawyer political contributions, judicial candidates, and political party endorsements that comes into play in such elections. Not that appointive systems are that much better (in the sense of being less political), except that at least the appointing person (as opposed to the electorate) has some idea who the would-be judges are. But, with plenty of exceptions for political log rolling, it is likely that Rs will appoint Rs and Ds, Ds. To paraphrase the late Philly mayor Frank Rizzo, who are politicians supposed to give jobs to, their enemies?
Race and gender? We’re not sure how much that adds to things like political parties. Both are the subjects of political skewing. This may end up as a double count of the same thing.
So what about the flattery?
We’re getting there. Before the article could conduct any sort of statistical analysis, the author had to find a universe of post-Riegel medical device preemption cases. He decided to use the selection of cases provided by our post-Riegel preemption scorecard, which at the time had seventy-five decisions in it.
Between February 20, 2008, the date the Riegel case was decided, and July 15, 2010, American courts have ruled on 75 cases involving devices the FDA preapproved for the market under § 360k of the Medical Device Amendments.755 N.Y.U.J.L. & Lib. at 760.
75 See Jim Beck & Mark Herrmann, (New) Medical Device Preemption Scorecard.
That’s pretty cool, since this is one of the areas where we’ve striven to be comprehensive, rather than just favoring our pro-defense worldview. To get cited as the key (indeed, only) authority for collecting these cases in a published law review article wasn’t something we expected when we began blogging, but hey, we’ll take it. The author could have used any source. He could even have done his own research. Instead he decided to use us. On this topic, anyway, somebody other than us thinks we’re the best around.
So what does the article discover?
There were 21 state cases decided by 47 judges. 81% of the state cases and 87% of the state judges (some were appellate cases decided by multi-judge panels) threw out the claims based on preemption. That compares to 54 cases decided by federal judges or magistrates. Most of those, at the time, were single-judge federal trial (district) courts. 44 of those found preemption, or about 81% of the federal judges and 75% of the cases. 5 N.Y.U.J.L. & Lib. at 760-61.
The article interprets these statistics as showing – counterintuitively – that “there is broad parity between state and federal courts on preemption.” Id. at 764. It ascribes this result to two factors: (1) “[p]arity may exist because state courts almost always mimic federal courts” on preemption, and/or (2) “the exception created by the 'parallel’ requirements language of Riegel is so narrow that lower courts have very little discretion. Id.
Having practiced in state and federal courts for decades, and having filed lots of preemption motions in both, we disagree. We think the result is due to a skewed sample. We're pretty sure about this because, after all, we compile the scorecard. First of all, a lot more stuff slips under the radar screen in state court, especially on the trial court level (where very little ends up reported), than in federal courts, where Westlaw and Lexis get a much greater percentage of the total results using PACER (a nationwide electronic docketing system).
With state courts, we’re much more dependent upon what our readers send us. Most of our readers are defense lawyers. All lawyers are more inclined to publicize their wins than their losses. So most of what we get in terms of state trial court decisions will favor preemption.
Also, to throw out a case on preemption almost always requires a judge to do some work and write an opinion. A preemption motion can be denied in a one-line order that may or may not even mention preemption. With state courts, we think there’s a relative overrepresentation of preemption wins in our chart, compared to federal courts, simply as a result of what we find out about.
That’s one reason the sample is skewed.
Another even more important skewing factor is procedural posture. Of the first 75 cases on the list, there are three state supreme court decisions (Despain, Troutman, and Blunt). Each of them found preemption unanimously, and in none of them were “parallel” claims asserted. That’s because these cases were appealing pre-Riegel preemption orders in which “parallel” claims weren’t as important, because the existential preemption question was at that time undecided. After Riegel, those plaintiffs were stuck with their pre-Riegel records not asserting “parallel” claims, and thus they got mowed down like sitting ducks. That’s at least 21 different state-court justices right there. On these small sample sizes that’s more than enough to wipe out an overall anti-preemption bias in state court.
We think a more accurate reading of the attitudes of state versus federal courts towards preemption would be on an open question, rather than the one substantially closed by Riegel. In that respect we offer the split on preemption as applicable to investigational devices that existed after Lohr and before Riegel. Lohr had nothing to do with investigational devices, which are the pre-approval incarnation of the PMA devices in Riegel. What happened? Well every federal court found preemption, and every state court did not. Compare Chambers v. Osteonics Corp., 109 F.3d 1243, 1248 (7th Cir. 1997); Martin v. Telectronics, 105 F.3d 1090 (6th Cir. 1997); Blinn v. Smith & Nephew Richards, Inc., 55 F. Supp.2d 1353 (M.D. Fla. 1999; Touchet v. Ace Medical Co., 1998 WL 531887 (E.D. La. Aug. 24, 1998); Chmielewski v. Stryker Sales Corp., 966 F. Supp. 839 (D. Minn. 1997); Berish v. Richards Medical Co., 937 F. Supp. 181 (N.D.N.Y. 1996) (all finding preemption), with, Baird v. American Medical Optics, 713 A.2d 1019 (N.J. 1998); Niehoff v. Surgidev, 950 S.W.2d 816, 821 (Ky. 1997); Connelly v. Iolab Corp., 927 S.W.2d 848 (Mo. 1996); Armstrong v. Optical Radiation Corp., 57 Cal. Rptr.2d 763 (Cal. App. 1996) (all finding no preemption).
As far as party, there was a distinct split, as “Democratic” judges let plaintiffs skate past preemption, even after Riegel, 30% of the time, while “Republican judges did so only 13.5% of the time. 5 N.Y.U.J.L. & Lib. at 762. We don’t doubt that this is an accurate finding.
In-state residents escaped preemption 28% of the time, while out-of-state plaintiffs escaped preemption 15% of the time. Turn that around and look at defendants, and the article indicates that in-state defendants lost preemption motions 21% of the time, and out of state defendants lost their motions 24% of the time. So there was “home cooking” on both sides, but more pronounced with in-state plaintiffs, whose success rate was almost twice as high. Id. at 762-63. There are further breakdowns by federal and state courts, but some of the sample sizes are so small, that we don’t think the conclusions are meaningful.
As for race/gender, women and minorities were actually more pro-preemption in the sample. The white men refused to find preemption almost 22% of the time, while the women/minorities did so only 14% of the time. Id. at 763. The article notes a small sample size, and we agree. The difference is not enough to be significant.
The article should have stopped there, instead it goes on to lambaste a preemption proposal made by one of our founding fathers, Mark Herrmann (as well as Dave Alden, who’s still at Jones Day), in a law article, “The Meaning Of The Parallel Requirements Exception Under Lohr And Riegel,” 65 N.Y.U. Ann. Survey of Am. L. 545 (2010). We discussed Mark’s article, here.
The author of the article obviously thinks he has a great career ahead of him as a plaintiff’s lawyer, and maybe he does, but we have to take issue with his criticism of Mark’s article as “nontextual and controversial” and contrary to Supreme Court “dicta.” 5 N.Y.U.J.L. & Lib. at 774.
The argument that express and implied preemption work in tandem to limit “parallel” claims is firmly supported by controlling Supreme Court precedent. The first Supreme Court case to discuss “parallel” claims was Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996). Lohr was a pre-TwIqbal case decided on the pleadings – not a procedural posture we would have recommended at the time. Because the only thing in the record was the complaint, Lohr involved only extremely vague allegations of supposed FDCA violations, and the Court knew it. “[T]he precise contours of their theory of recovery have not yet been defined” but “it is clear that the [plaintiffs’] allegations may include claims that [defendant] has, to the extent that they exist, violated FDA regulations.” 518 U.S. at 495.
Lohr addressed only that issue. Declaring that “[n]othing in §360k denies [a state] the right to provide a traditional damages remedy for violations of common-law duties when those duties parallel federal requirements,” it found those claims unpreempted. Id. The mere addition of a damages remedy did not render such duplicative state-law claims “different from” federal requirements. Id. Lohr didn’t consider anything other than express preemption under the Medical Device Amendments (where §360k came from).
Implied preemption was addressed in Buckman Co. v. Plaintiffs’ Legal Committee, 531 U.S. 341 (2001), where the Court looked into whether there was something in the FDCA – other than the §360k(a) preemption clause – that preempted the fraud on the FDA claims the plaintiffs claimed fell within Lohr’s “parallel” violation rationale. Those fraud on the FDA claims undoubtedly came within the scope of Lohr’s “parallel claims.
That didn’t get them past preemption. The Court unanimously held that not every “violation of the FDCA will support a state-law claim.” Id. at 353. The unpreempted violation claims in Lohr “arose from the manufacturer’s alleged failure to use reasonable care in the production of the product, not solely from the violation of FDCA requirements.” Id. In other words, there was some pre-existing state-law liability theory for the claim in Lohr to be “parallel” to.
Not so in Buckman. Violation claims that “exist solely by virtue of the FDCA disclosure requirements,” id., were impliedly preempted for a different reason – because the FDCA “leaves no doubt that it is the Federal Government rather than private litigants who are authorized to file suit for noncompliance with the medical device provisions.” Id. at 349 n.4 (citing 21 U.S.C. §337(a)). Absent any independent state-law basis, private FDA fraud claims conflicted with and were preempted by the FDCA’s exclusive grant of federal prosecutorial authority. Id. at 353-54.
Thus Herrmann’s article was fully supported by binding Supreme Court precedent when it observed that implied preemption operates to knock out those “parallel” claims that, in fact aren’t “parallel” to anything in state law. Those claims are simply improper attempts to enforce a claimed FDCA violation through private litigation – which Congress prohibited in §337(a) – applicable to both drugs and devices, but unfortunately not to food.
But there are indeed violation claims, like manufacturing defect claims, that are, in fact “parallel” to well-established state common law claims. We know that. That brings us to our third Supreme Court case, Bates v. Dow Agrosciences LLC, 544 U.S. 431 (2005), a non-FDCA case involving violation claims under a different statute (involving pesticides/herbicides). So what must a court do with an otherwise unpreempted “parallel” violation claim? Bates tells us. Courts have to make sure that such claims actually are “parallel” to relevant federal requirements, and not some lawyer twisting a malleable regulation is some way the relevant agency would never dream of doing. Id. at 454. To “survive preemption” the state law claim, while it “need not be phrased in the identical language,” must be “genuinely equivalent.” Id. “[A] manufacturer should not be held liable under a state [parallel] requirement . . . unless the manufacturer is also liable” under federal law “as defined by [the agency].” Id.
And what’s the best way to make sure that something’s genuinely equivalent? That would be for the FDA to have already decided the issue. Two justices concurred on precisely that ground in Buckman:
This would be a different case if, prior to the instant litigation, the FDA had determined that petitioner had [done what the plaintiffs claimed it did] and had then taken the necessary steps to remove the harm-causing product from the market. Under those circumstances, respondent’s state-law fraud claim would not depend upon speculation as to the FDA's behavior in a counterfactual situation but would be grounded in the agency’s explicit actions. In such a case, a plaintiff would be able to establish causation without second-guessing the FDA’s decisionmaking or overburdening its personnel, thereby alleviating the Government's central concerns regarding fraud-on-the-agency claims.Buckman, 531 U.S. 354 (Stevens & Thomas concurring) (emphasis added).
That’s what Herrmann’s article has behind it – the holdings of the only three Supreme Court cases that have actually considered “parallel” violation claims. The admitted “dicta” that the article wants to apply instead is from Riegel, where the Court “decline[d] to address” such claims “in the first instance” because they had been waived. 552 U.S. at 330.
To sum up our reaction to the second half of the article, it’s pretty clear that the author doesn’t want to work for our side after graduation. One doesn’t go disparaging the only relevant Supreme Court precedent (as opposed to dictum) as “nontextual” unless one labors on the other side of the “v.”. There’s lots of Supreme Court precedent we don’t like – see Wyeth v. Levine, 129 S. Ct. 1187 (2009), as Exhibit A – but we don’t disregard it in favor of dictum.
That’s why we don’t deny Herrmann’s argument is “controversial.” So is ours, which is essentially the same as Herrmann’s. Everything about preemption is controversial.
It’s also appropriate to note that, in attacking Herrmann, the article’s very first argument is the least valid of the propositions it sought to draw from its statistical survey of post-Riegel preemption cases – that there’s no difference, in terms of attitudes towards preemption, between state and federal courts. The article states:
First, the data in this article suggests that state courts are generally trustworthy in carefully applying preemption, and dismiss claims at similar rates as federal courts. That consistency rebuts the concern that state courts will create disparity and uncertainty throughout the United States. It also demonstrates that state courts take preemption quite seriously, and will not simply disregard Supreme Court precedent to protect local interest groups. That undermines the assertion that further federal preemption is needed to successfully constrain state courts.5 N.Y.U.J.L. & Lib. at 774. That’s a great deal of weight to put on such a badly skewed data set. With apologies to Lloyd Bentsen, we know that data set – that data set was a friend of ours – heck, we created that data set. We know that, as to state courts, it’s doubly biased by both selection criteria and the procedural posture of the state court decisions that we were able to lay our hands on.
But then again, there are lies, damn lies, and statistics. It may well have been that the entire purpose of the statistical analysis (and thus the first half of the article) was to create bogus data to support a bogus legal conclusion, in order to attack Herrmann’s (and our) preemption position with respect to “parallel” claims. That’s certainly how the statistical analysis was used.
We do believe that the author will be right at home working for the other side – he already understands their expert witnesses.