Photo of Bexis

In Puerto Rico v. Franklin-California Tax-Free Trust, 579 U.S. 115 (2016) (initially discussed here), the Supreme Court drove a stake through the heart of the misbegotten “presumption against preemption” in express preemption cases.

[B]ecause the statute contains an express pre-emption clause, we do not invoke any presumption against pre-emption but instead focus on the plain wording of the clause, which necessarily contains the best evidence of Congress’ pre-emptive intent.

Id. at 125 (citations and quotation marks omitted).  As we also discussed, this abolition has been recognized as generally applicable by every Court of Appeals in the country, save the Third Circuit.

We have applauded this development, but we have also warned against “zombie” presumption against preemption decisions – courts that do the same thing sub rosa, by quoting and following language from pre-PR v. Franklin cases while simply omitting the dirty word “presumption.”  That post criticized Mata v. Allupick, Inc., 2022 WL 1541294, at *2 (N.D. Ala. May 16, 2022), for the foible of relying on quotes from earlier presumption-based cases, but with the P-word excised.

Mata was from the Eleventh Circuit, and two months after that decision, the en banc Eleventh Circuit joined the abolitionist movement recognizing that the presumption against preemption was no more.  See Carson v. Monsanto Co., 72 F.4th 1261 (11th Cir. 2023) (“Carson I”).  Carson I recognized that PR v. Franklin, “abrogated” earlier Supreme Court decisions that had applied a presumption in express preemption cases:

Express preemption turns primarily on “the language of the pre-emption statute and the statutory framework surrounding it.”  Medtronic, Inc. v. Lohr, 518 U.S. 470, 486 (1996) (citation and internal quotation marks omitted), abrogated in part on other grounds by Puerto Rico v. Franklin Cal. Tax-Free Tr., 579 U.S. 115 (2016).  Where Congress has enacted an express-preemption provision, we identify the state law that it preempts according to ordinary principles of statutory interpretation, and no presumption against preemption applies.  See Franklin Cal. Tax-Free Tr., 579 U.S. at 125.

72 F.4th at 1267.  Carson I accordingly overturned a panel decision that had interposed a “force of law” requirement to avoid the statute’s express preemption clause.  Id. at 1267-68.  That requirement was “inapposite”; only applying to implied preemption.  Id. at 1267.  Agency force of law is not needed to activate the Supremacy Clause where Congress did so in the relevant statute. Thus, it “does not extend to express-preemption cases, where, as we have explained, the meaning of the express-preemption provision—not conflicting federal and state legal obligations—triggers preemption.”  Id. at 1268.

Rather than decide the preemption issue itself, however, Carson I remanded the matter to the anti-preemption appellate panel that had decided that issue erroneously the first time around.

That turned out to be a big mistake.

On remand, the original Carson 3-judge panel again rejected preemption.  In so doing, Carson v. Monsanto Co., 92 F.4th 980 (11th Cir. 2024) (“Carson II”), let loose the biggest zombie presumption against preemption decision that we have yet seen.  The Carson litigation isn’t about prescription medical products, but rather about a herbicide – Roundup.  Regardless, defense counsel in the Carson litigation, and we hope the en banc court, needs to round up and extinguish this zombie before it runs amok in the Eleventh Circuit.

Here’s how Carson II created the zombie.

The Roundup litigation involves the Federal Insecticide, Fungicide, and Rodenticide Act (“FIFRA”), which has an express preemption clause almost verbatim identical to the FDCA provision (21 U.S.C. §360k(a)) protecting medical devices − albeit limited to labeling.  Section 136v(b) mandates that states may “not impose or continue in effect any requirements for labeling or packaging in addition to or different from those required under this subchapter” (emphasis added).  Except for being in reverse order – “in addition to or different from” versus “different from or in addition to” − both statutes share the same basic preemption language.  Carson II gave lip service to the abolition of the presumption against preemption, 92 F.4th at 989, but that was all.  Rather than apply the express terms of the statute, it turned to one of those pre-PR v. Franklin cases, that, with respect to the erstwhile “presumption,” has been (in Carson I‘s terms) “abrogated.”  Bates v. Dow Agrosciences LLC, 544 U.S. 431 (2005).  See Carson II, 92 F.4th at 990 (“To determine whether FIFRA preempts state requirements that go beyond mere duplication of FIFRA’s requirements, we turn to Bates”).

After PR v. Franklin, as applied to FIFRA in Carson I, the panel should not have simply have “turned to” Bates, because Bates was practically marinated in the now-abolished “presumption against preemption.”  “[W]e have long presumed that Congress does not cavalierly pre-empt state-law causes of action.”  544 U.S. at 449 (quoting what Carson I recognized as the “abrogated” Lohr presumption against preemption discussion).  Thus, from among “plausible alternative reading[s]” of FIFRA’s preemption clause, Bates found “a duty to accept the reading that disfavors pre-emption.”  Id.  Thus, Bates followed a −

basic presumption against pre-emption.  If Congress had intended to deprive injured parties of a long available form of compensation, it surely would have expressed that intent more clearly.

Id. at 449.  But this “intended to deprive” proposition involved implied preemption − Silkwood v. Kerr-McGee Corp., 464 U.S. 238, 251 (1984), that the Supreme Court (two years after Silkwood) ruled was inapplicable to express preemption cases – since express preemption clauses have precisely that function.  See Riegel v. Medtronic, Inc., 552 U.S. 312, 326 (2008) (rejecting dissent’s reliance on Silkwood; precluding “judicial recourse . . . is exactly what a pre-emption clause . . . does by its terms”).  An express preemption clause precludes Silkwood’s airy speculation about unstated congressional intent.  “The operation of a law enacted by Congress need not be seconded by a committee report on pain of judicial nullification.”  Riegel, 552 U.S. at 326 (citation omitted).

Having explained why Bates is no longer good law in disfavoring preemption where (as here) Congress has enacted an express preemption provision, we return to Carson II. In Carson II, the plaintiff’s principal claim was that the defendant “should have included a warning about [the product’s] potentially carcinogenic effects on its label,” 92 F.4th at 991, even though the relevant federal regulator (the EPA) did not require any such warning.  If that allegation had been asserted against a PMA medical device claim, under the essentially identical wording of the FDCA’s preemption clause, it would be preempted, since a common-law claim demanding an unapproved warning would be both “different” and “in addition” to the federally required warning.  E.g., Riegel, 552 U.S. at 329 (“a state common-law requirement for additional warnings” is “surely” preempted).

By citing Bates, as supposedly “preserv[ing] a broad role for state regulation,” 92 F.4th at 991, Carson II turned the plaintiff’s demand for a warning found nowhere on the approved label into a “parallel claim.”  By calling it “misbranding,” Carson II transformed essentially anything a plaintiff might allege about a warning’s claimed inadequacy into a purportedly “parallel” FIFRA violation.  Id. at 991-92.  In the FDCA context, we’ve decried similar abuse of broad misbranding language, for the same basic reason (sub rosa disinterment of the presumption against preemption), in OTC drug litigation.

How did Carson II accomplish this?

Any state-law labeling requirement, no matter how entirely different from what the EPA approved, survives because FIFRA “effectively imposes a strict-liability standard.”  Id. at 991.  State common-law, by contrast, is “narrower” because it imposes only a “knows or reasonably should have known” standard.  Id. at 992.  Carson II then put the rabbit in the hat using Bates.  “Different” as used in the FIFRA preemption clause doesn’t really mean what it says – that is, barring claims that are actually “different” − because anything “narrower” escapes preemption under Bates. Or, quoting directly from Carson II:

[T]he Supreme Court has explained that “state law need not explicitly incorporate FIFRA’s standards as an element of a cause of action in order to survive pre-emption.” Rather, so long as the state-law duty parallels or is “fully consistent” with FIFRA, FIFRA does not preempt it. . . .  If anything, Georgia common law about failure-to-warn claims imposes less of a duty on pesticide manufacturers than FIFRA. . . .  Because Carson’s state failure-to-warn claim is “fully consistent with” or even narrower than federal requirements, FIFRA does not expressly preempt that claim. After all, as the Supreme Court has reasoned, “[w]hile such a narrower requirement might be ‘different from’ ” FIFRA’s requirements “in a literal sense,” that would be “a strange reason for finding pre-emption of a state rule insofar as it duplicates” FIFRA.  So FIFRA does not expressly preempt “narrower” state requirements.

92 F.4th at 992 (all quotations are to Bates).

So, relying on the (unnamed) presumption as applied in Bates, “different” and “addition” only mean not “narrower,” rather than the actual words Congress used.

Got that?  Because of the presumption against preemption-based rationale in Bates, any state common-law warning claim – no matter how disparate – is automatically not preempted because state common law claims are inherently “narrower” than FIFRA’s requirements, and anything “narrower” cannot be preempted due to Bates’ refusal to read FIFRA’s preemption language “in addition to or different from” literally due to a now-abolished presumption against preemption.  That’s pretzel logic if we’ve ever seen it.

Thus a zombie presumption against preemption now stalks the Eleventh Circuit.  To finish the job of emasculating FIFRA preemption, Carson II next returned to Bates to re-import through a side door the very same “force-of-law” inquiry the Carson I had rejected.

To establish whether a particular Agency action amounts to a “requirement” under FIFRA, we must determine whether that Agency action carries the force of law.  If it is not “a rule of law that must be obeyed,” then as the Supreme Court has directed, it is not a “requirement.”  Bates, 544 U.S. at 445, 125 S.Ct. 1788.  So though we need not perform a threshold force-of-law analysis before defining the scope of FIFRA’s preemption, we must do that analysis to determine whether an Agency action qualifies as a “requirement.”

92 F.4th at 993.

We’ve read a lot of judicial opinions in our time, but we can’t recall any remand decision that so thoroughly ignores the decision that remanded it as Carson II did to Carson I.  At every turn Carson II reflected the walking dead − the influence of the presumption against preemption in Bates from beyond its PR v. Franklin grave.

Adding insult to EPA on top of the injury it inflicted on the plain language of FIFRA’s preemption clause, Carson II went on to oust EPA product approvals from preemption entirely.  The EPA’s approvals were supposedly not “requirements” because “Agency approvals provide only ‘prima facie evidence,’ not conclusive proof, that a pesticide is not misbranded” since “the Agency can later retract its approval.” Id. at 993 (quoting 7 U.S.C. §136a(f)(2)).  “Since the Agency’s determination is neither conclusive nor irrevocable, it would make little sense to deem it a “requirement” on equal footing with FIFRA’s prohibition on misbranding.”  Id. (citation omitted).

Thus, by using Bates to impose a presumption against preemption sub silentio, Carson II:  (1) construed any common-law warning claim, no matter how divergent from the product’s actual approved warnings, as not “in addition to or different from” that warning; and (2) deprived the EPA’s product approval of any preemptive force at all, because it wasn’t a “requirement.”

Is the Carson II zombie likely to eat the brains of FDCA preemption clauses, too?

We don’t think so − ironically because to distinguish Riegel, Carson II had to violate another of the Supreme Court’s holdings from Riegel itself.  Again, we’ll explain.

Emphasizing the nearly identical express preemption clauses that govern both medical devices and  FIFRA, the defendant in Carson II relied heavily on Riegel’s recognition of broad preemption of product liability claims under the relevant FDCA preemption clause in §360k(a).  92 F.4th at 993-94.  In reaching its preemption friendly result, Riegel flatly rejected a plaintiff-side argument that claimed the meaning of the word “requirement” could vary by statute.  Instead, Riegel declared that “Congress is entitled to know what meaning this Court will assign to terms regularly used in its enactments.”  552 U.S. at 324.  Any state “damages” award “is designed to be[] a potent method of governing conduct and controlling policy.”  Id. (citation and quotation marks omitted).

To avoid these Riegel-based arguments about what constitutes a “requirement,” Carson II contradicted these two key aspects of Riegel.  First, contrary to Riegel’s opposite holding, Carson II denied that state-law litigation imposes mandatory “requirements” through damage awards – rather, the pressure of tort liability may merely “lead” either the regulated defendant or the agency itself to “decide that revised labels are required in light of the litigation.”  92 F.4th at 995 (quoting, of course, Bates).  Second, and equally contrary to Riegel, Carson II held that what constitutes a preemptive “requirement” varied depending on  the “context” of the two “different” statutes, with the FDCA being more “rigorous,” and thus more preemptive, than the essentially identical preemption language in FIFRA:

[W]hile the preemption provisions are similar, we must read them in context.  The statutes’ distinct approval processes confirm this significant difference.  Premarket approval under the [Medical Device] Amendments represents a “rigorous” conclusion that a device is safe and effective. . . .  By contrast, the [EPA’s] approval of a pesticide’s registration serves as only “prima facie evidence” that the pesticide complies with FIFRA’s requirements. . . .  What’s more, the [MDA] preemption provision expressly contemplates device-specific application, as it preempts requirements “with respect to a device.” FIFRA, on the other hand, contains no such limitation − it imposes only “general standards.” And different federal statutes and regulations may lead to different preemption results.  Given the differences between FIFRA and the [FDCA’s] statutory schemes, Riegel does not control here.

Carson II, 92 F.4th 995 (citations and quotation marks omitted).

This aspect of Carson II reminds us of the Vietnam-era quotation, “We had to destroy the village in order to save it.”  To save Riegel-based preemption of medical device claims from the zombie it created, Carson II found it necessary to destroy the rationale of Riegel itself – by once again relying on the pre-Riegel and pre-PR v. Franklin decision in Bates to resurrect propositions about the meaning of “requirement” that Riegel flatly rejected.

There’s more we could criticize about Carson II – particularly its rejection of implied preemption on the “Mouse Trap game” possibility that the relevant government agency could always change its mind, 92 F.4th at 998-99, which the Supreme Court rejected in PLIVA, Inc. v. Mensing, 564 U.S. 604, 619 (2011) – but we’ll stop here.  Carson II is a paradigmatic zombie presumption against preemption case, since it is 100% dependent on the presumption-based torturing of the relevant preemption language that initially occurred in BatesCarson II also illustrated why PR v. Franklin was right to abolish that presumption, because the FIFRA preemption clause as construed in Carson II no longer meant anything close to its plain statutory language.  A state-law cause of action demanding a warning that was never approved (and indeed had been rejected) by the relevant agency at the time of the claimed product use simply cannot be anything other than “in addition to or different from” the agency’s “requirements.”

Not surprisingly, the defendant has again sought en banc review. One can only hope that the en banc Eleventh Circuit has the fortitude to reverse this Orwellian statutory result yet again, and thereby to finish off, once and for all, the zombie presumption against preemption, before it eats any more judicial brains.