There has been much discussion recently about how Rule 702 is in need of a tune-up to better guide district courts’ gatekeeping. More about that soon.
But a case now pending before the Supreme Court, Monsanto Company v. Hardeman, No. 21-241, demonstrates that it’s not always the fault of the district courts. (Disclaimer: This firm (and this author) filed an amicus brief supporting certiorari.) Sometimes it’s about a lack of stewardship at the circuit level. Absent direct and unequivocal guidance from the Supreme Court, appellate courts call the tune, and the district courts are required to follow it. And in the interstices, district judges read the tea leaves and try to follow the circuit court’s leads and signals. No one likes to get reversed. Even if the district judges think the circuit has gotten it wrong, they honor the hierarchy and follow the commands of stare decisis, human nature and common sense.
That tension is highlighted in the RoundUp MDL, and the district court and Ninth Circuit opinions in Hardeman, a bellwether case. 997 F.3d 941 (9th Cir. 2021). We’ve written before about the RoundUp litigation (here and here), including the district court’s pretrial decision largely allowing plaintiffs’ experts’ general and specific causation testimony in Hardeman.
The district court’s decision was not exactly a ringing endorsement. The district court observed – repeatedly – that it was bound to follow the Ninth Circuit case law and apply the circuit’s proposition that Rule 702 embodies a “liberal thrust.” Distinguishing several sister circuit precedents, the court emphasized the Ninth Circuit’s heightened tolerance for “borderline opinions.” Compelled by this preference for admissibility, the district court admitted causation opinions while candidly conceding they would likely be excluded in other circuits.
That tacit criticism of the case law qualifies as a vast understatement. In several prior opinions, most notably Messick v. Novartis Pharms. Corp., 747 F.3d 1193 (9th Cir. 2014) and Wendell v. GlaxoSmithKline LLC, 858 F.3d 1227 (9th Cir. 2017), the Ninth Circuit has set Rule 702 standards that ignore the rule, the consensus interpretations of the rule in other circuits, and Supreme Court authority. [Double disclaimer: This firm (and this author) represented GlaxoSmithKline in Wendell.] When it comes to Rule 702 and causation testimony in products liability and toxic tort cases, the Ninth Circuit has proven to be … problematic.
Ironically, the problem starts with the circuit’s methodology – its manner of reviewing rulings that exclude expert causation testimony as unreliable. Recall that in General Elec. Co. v. Joiner, 522 U.S. 136 (1997) (discussed previously here), the Supreme Court (1) held that appellate courts must apply deferential review to district court gatekeeping determinations, reversing only for abuse of the gatekeeper’s wide discretion; and (2) rejected the 11th Circuit’s practice of subjecting exclusions, especially if outcome-determinative, to a “harder look” (“particularly stringent” review).
Like a hoary cinematic trope, the Ninth Circuit seems to exist in an alternate universe, one where the 11th Circuit’s hard look standard was endorsed rather than definitively rejected in Joiner. It applies the abuse of discretion standard in two-step fashion, reviewing de novo whether the district court applied the “correct legal standard” or “construction or interpretation” of the rule, “including whether particular evidence falls within the scope of [the rule].” Only if the lower court correctly applied the rule does the court defer to its ultimate decision.
It is well-established that discretion is abused if the district court applies the “wrong legal standard.” But there is a vast difference between evaluating whether the lower court applied the correct rule and evaluating whether it applied the rule correctly, i.e., the way the reviewing court would.
The key question: At what level of generality/specificity does one characterize the “legal standard”? At a high level of generality, the correct legal standard is Rule 702 and its construction is essentially recognition of the gatekeeping function and the standards enumerated in Rule 702 (sufficient basis in facts or data and the product of reliable application of reliable principles and methods). In contrast, at a high level of specificity lies what the Ninth Circuit does – micromanage the district court’s application of the reliability requirements to the record. Where’s the deference in that? The Ninth Circuit’s method of second-guessing district court gatekeeping honors neither the letter nor the spirit of Joiner’s mandate of deferential review.
And as both Messick and Wendell demonstrate, this de novo review seems to have an extra gear where the district court has excluded the expert’s causation opinion. Where the expert’s testimony has been admitted – as in Hardeman – the review is, palpably, far more forgiving. This differential review ethos flows from the Ninth Circuit’s broad view that Rule 702 favors admissibility and excludes only “junk science” or “nonsense opinions.”
This reading violates Joiner in another way. In rejecting the 11th Circuit’s “hard look” practice, the Supreme Court, at least implicitly, also invalidated its basis – that Rule 702 establishes a “preference for admissibility.” The Court unequivocally held that the appellate courts may not “categorically distinguish between rulings allowing expert testimony and rulings that disallow it.” And Daubert had previously held that the proponent bears the burden of establishing reliability under Rule 104(a), by a preponderance of the evidence.
Accordingly, the Ninth Circuit’s Rule 702 jurisprudence violates Joiner in at least three ways: It fails to apply deference to gatekeeping rulings, it fails to review exclusions of expert testimony with equal dignity and deference, and it explicitly applies Rule 702 with a discredited preference for admission. Though Daubert did refer to the Federal Rules having a “liberal thrust,” it did not suggest this was an operative standard governing application; the High Court was merely contrasting the “rigid” and “austere” approach of the Frye rule it was replacing.
The Ninth Circuit also carves its own path in more specific ways. It is Rule 702 canon, and built into the rule’s architecture, that the expert’s qualifications and the reliability of their testimony are two separate requirements. “A supremely qualified expert cannot waltz into the courtroom and render opinions unless those opinions are … reliable and relevant.” Clark v. Takata Corp., 132 F.3d 750, 759 n.5 (7th Cir. 1999). But the Ninth Circuit reversed exclusions in Wendell and Messick in large part because the experts were highly qualified. Similarly, even though clinical practice has little to do with studying causal relationships between chemical exposures and poorly understood disease processes – diagnosis and treatment is not etiology – the court gave great weight to clinical experience and application of the “art” of differential diagnosis, because “medicine partakes of art as well as science.”
Causation, however, is not “art” – the inherently subjective, creative expression of the artist’s skill and imagination. Even as to specific causation, it is closer to the opposite, hard science: the objective pursuit and discovery of verifiable/falsifiable scientific truths based on data acquired through rigorous regimented investigation, application of accepted scientific principles and methods, and guided inference. The Ninth Circuit’s facile characterization of causal investigation as “art” improperly exalts the subjective impressions of the expert.
That’s the problem, in a nutshell. Too much deference to experts, too little deference to district judges and their concerns over the reliability of the expert’s work and work product.
Given the standards and circumstances, the district court in Hardeman had relatively little choice but to admit the causation testimony, even though it lacked any solid scientific basis and, in fact, defied the scientific consensus.
So, how do you solve a problem like the Ninth Circuit? Grant certiorari and reverse.
And, perhaps, amend and strengthen the language of Rule 702. But that’s a solution for another post.