Photo of Eric Alexander

Along with Shakespeare’s plays and painfully plodding Victorian novels, there is a good chance that your western high school (or perhaps college) education included at least a smattering of philosophy.  The line between political science and philosophy can be hard to draw—Kant, Hobbes, and Rousseau might be featured in classes under either heading, for instance—but more fundamental philosophical considerations seem to be resigned to the recesses of our brains by the time law school begins its takeover.  Once legal practice starts, forget about it.  Drafting legislation or regulation, or perhaps challenging them, may involve an amorphous thing called public policy, but that is not really philosophy.  While definitions vary, and the linguistics only sheds so much light, we can say that a question like “what is the difference between expanding a right and rejecting a proposed limit on a right” is a matter of philosophy.  We can also say that it is an important question to ask in the context of Erie restraint and one that more federal courts sitting in diversity should be asking.

Having followed the principles behind Erie restraint for some time and noted how often federal courts have extended existing state law past its reasonable limits when trying to get around preemption or some other hurdle to plaintiff recovery, especially in the MDL context, we think we know the “right” answer to the philosophical question posed above.

Some principles first.  Tort plaintiffs can recover only when they establish the elements for a cause of action recognized by applicable case law or statute.  Under the U.S. Constitution and basic federalism principles, states get to decide their own laws.  While there may be federal constitutional limits on state laws as written or applied, federal actors do not get to establish state law.  If a state court plaintiff brought suit based on a novel tort claim not established by state statute or controlling caselaw, then the state court, according state jurisprudence, state constitutional provisions, and perhaps other considerations, would get to decide if it could establish a new cause of action or just dismiss the suit for failure to state a claim upon which relief could be granted.  (Some might call the former route “judicial activism,” but an exposition of the line between legislative authority and common law is beyond the scope of this post.  The latter route, and the existence of Fed. R. Civ. P. 12(b)(6) and state equivalents, speak to the fact that some complaints include some wacky claims.  Outside of our area, pro se prisoner cases and post-election voter suppression cases may be good examples of such wackiness.)

The seminal decision authored by Judge Brandeis in Erie Railroad Co. v. Tompkins, 304 U.S. 64 (1938), recognizes that the role of federal court sitting in diversity is different.  The core, and oft-cited, holding is that, “[e]xcept in matters governed by the Federal Constitution or by Acts of Congress, the law to be applied in any case is the law of the State.”  Id. at 78.  The underlying reasoning, however, is rooted in federalism and has given risen to Erie restraint (or the Erie doctrine) over the last 82 years.  Many of our prior posts on this subject, go into detail (like here, here, and here), so we will be relatively brief in our recap and just focus on what the Supreme Court has said.  Erie said federal judges do not get to “brush[] aside the law of a state in conflict with their views” and “advance[e] the doctrine [the federal judge] thinks at the time should be the general law on a particular subject.”  Id.  Because states get to “define the nature and extent” of rights to recover relief, “[t]hat object would be thwarted if the federal courts were free to choose their own rules of decision whenever the highest court of the state has not spoken.”  West v. AT&T Co., 311 U.S. 223, 236 (1940).  “A federal court in a diversity case is not free to engraft onto those state rules exceptions or modifications which may commend themselves to the federal court, but which have not commended themselves to the State in which the federal court sits.”  Day & Zimmerman, Inc. v. Challoner, 423 U.S. 3, 5 (1975).  Importantly, there is nothing in 28 U.S.C. § 1407, which authorizes MDLs, or in an controlling interpretation of that statute, that gives an MDL judge more license to create new state law that any other federal judge presiding over an individual diversity case.

So, back to our philosophical question.  If there is a plaintiff who seeks recovery for an undeniably novel claim—let’s say, “malicious staring” for a stranger “looking funny” at the plaintiff while he walked down the street—then it should be easy for the presiding federal judge to conclude that it is the plaintiff asking for the creation of a new state right not the defendant asking the judge to impose a new limit on a state right.  By contrast, there could be a situation where the defendant is clearly asking for the creation of a new state law limit or defense—like the recognized state claim of “malicious staring” cannot stand if the plaintiff was, as a matter of law, “funny looking.”  (We vaguely recall a pro se prisoner case against a prison guard along these lines from our days as a clerk, for what it is worth.)

A perhaps harder question might come from an MDL where plaintiffs are suing over a PMA device and trying to come up with a parallel claim to squirm through the narrow gap between express preemption and implied preemption.  Or maybe a plaintiff who claims injury from her use of a generic drug has sued both the manufacturer of the drug she took and the manufacturer of the drug with the NDA; she needs to come up with a non-preempted claim against the former and some claim against the latter that gets her past the traditional product liability hurdles she would not clear.  Assume in each case that the applicable states each have a fairly well-developed negligence law with older cases adopting the basic elements and more recent cases setting out requirements for cases against the manufacturers of prescription medical products, but that they have neither product liability statutes abrogating all common law nor high court cases ruling on the claims these plaintiffs need to assert to get have some chance of recovery.  Negligent failure to report adverse events to FDA?  Negligent failure of the generic manufacturer to ask the branded manufacturer to update its label?  Innovator liability?  When the viability of these claims get decided on a defendant’s motion to dismiss or motion for summary judgment, the posture is almost always rightly viewed as an attempt by the plaintiff to create or expand state rights not an attempt by the defendant to impose a new limit on an existing right.  This means that federal courts deciding these issues should use their Erie restraint as a brake on creating new state law rights.

Even skipping our many posts on federal courts deciding purported parallel claims or innovator liability, the philosophical underpinning should be pretty clear.  In these situations, the plaintiff needs the law to expand or he loses, whether because of preemption, lack of duty, lack of causation, or something else.  The defendant wins absent an expansion, so it is not seeking a new limit.  We think this is what was going on in In re Testosterone Replacement Therapy Prods. Liab. Litig., MDL No. 2545, Case No. 17 C 3775, 2020 WL 6487327, *1 (N.D. Ill. Nov. 4, 2020), where an MDL judge evaluated whether Florida law imposed a duty on a prescription drug manufacturer “to provide warnings in some manner apart from the product’s FDA-approved package insert.”  The basic facts are that the particular plaintiff (hereafter, Davis) was prescribed and used the drug entirely, but shortly, after the label had been changed to include a paragraph on the relevant risk specifically required by FDA.  Although the opinion never said this, this clearly meant that Davis could not assert the same failure to warn claim that other plaintiffs had asserted because it would probably be preempted under Albrecht if it did not fail for lack of proximate cause.  Also absent in the opinion was any mention of Erie.  The concept of restraint—but not the word—is introduced only in the penultimate paragraph in a way that gives away the result but also frames the philosophical issue about which we have blathered thus far:

For the foregoing reasons, the Court disagrees with Actavis that under Florida law, a prescription drug manufacturer has no duty to provide warnings in a manner apart from the FDA-approved package insert.  In reaching this result, the Court recognizes that it must take care not to expand Florida law “beyond the boundaries established” in the state’s jurisprudence.  King v. Damiron Corp., 113 F.3d 93, 97 (7th Cir. 1997) (internal quotation marks omitted); AbbVie Reply at 7 (quoting same).  Florida courts have not addressed the question Actavis presents, and there is ample support for the conclusion that Florida courts would reject Actavis’s narrow interpretation of the duty to warn. Therefore, the Court denies Actavis’s motion for summary judgment.

Id. at *10.

Yes, the court engaged in a lengthy exercise of attempting to ascertain existing Florida law on the duty of prescription drug manufacturers in relation to warnings.  Yes, upon determining that the particular issue had not been decided, the court looked at cases from other jurisdictions to try to predict what the Florida Supreme Court would do if it decided this issue.  And, yes, those are the basic steps a federal court is supposed to take in a situation like this, assuming it does not avail itself of the opportunity to certify a question to the Florida Supreme Court as happened in Lehman Brothers v. Schein, 416 U.S. 386 (1974), and many other cases.  However, we think the entire analysis was misguided because it was premised on the idea that the defendant wanted to impose a new limit on its duties under Florida law instead of the reality that plaintiff wanted to create a new duty for drug manufacturers under Florida law because he had no claim otherwise.

We could belabor this by walking through the court’s actual analysis, but it is not really necessary.  Like in the recently published Zitney decision from a Pennsylvania state court, which did not have Erie to restrain it, the proposition that a drug manufacturer with an adequate FDA-approved package insert has a dutyto provide warnings in some manner apart from the product’s FDA-approved package insert” is a minority position.  The majority position, intertwined with the learned intermediary doctrine, is that a plaintiff cannot win a failure to warn claim when the package insert is adequate.  Most jurisdictions to weigh in also recognize that a prescription within two months of FDA approval of the package insert will lead to preemption of any warnings claim—especially where, like Davis—the FDA had specified the language to be included about the risk at issue.  It should not be a matter for distinction that Zitney or Florida law cases like Metz and Guarino (discussed in Davis) involved generic drugs.  Id. at **5-6.  Like in those cases, the plaintiff in Davis needed to expand existing state law to create a new duty to allow him to offer a non-preempted warning claim against the defendant he chose to sue.  (It really seems that his issue about the dissemination of the approved label for the drug he took was with his pharmacy, but that would not have fit neatly within the MDL framework.)

In that context, the question is not whether the defendant had proven why Florida would not follow a few minority cases (often older and rejected by later cases).  Showing appropriate restraint, the prediction of a proposed expansion in Florida state law should draw on actual indications in Florida state court decisions to see if clear support for expansion exists.  When it does not, then the federal court should not “go there.”  We submit that Davis, like far too many MDL courts in the past, should not have predicted a clear expansion of state law duty by claiming it was rejecting an imposition of a new state law limit.  You do not have to take our word for it.  The Seventh Circuit decision on Erie restraint cited in Davis makes just this point:

We have counseled before that “federal courts sitting in diversity ought to be circumspect in expanding the law of a state beyond the boundaries established in the jurisprudence of the state.”  Dausch v. Rykse, 52 F.3d 1425, 1438 (7th Cir. 1994) (Ripple, J. concurring, joined by Coffey, J. concurring).  See also Twin Disc, Inc. v. Big Bud Tractor, Inc., 772 F.2d 1329, 1333 (7th Cir. 1985) (“Because we are unable to predict what the Wisconsin courts would do in such a case, and because of the substantial policy considerations involved, we hesitate to expand Wisconsin law to allow these tort claims in the absence of a more direct indication of intent by the state courts or Legislature.”); Torres v. Goodyear Tire Rubber Co., Inc., 857 F.2d 1293, 1296 (9th Cir. 1988) (“In this case, however, we hesitate prematurely to extend the law of products liability in the absence of an indication from the Arizona courts or the Arizona legislature that such an extension would be desirable.  We have limited discretion in a diversity case ‘to adopt untested legal theories brought under the rubric of state law.’”).

King, 113 F.3d at 97.  Expanding state law to allow new tort claims or adopted untested legal theories is for state legislatures and sometimes state courts, but not for federal courts, MDL or otherwise.