Extracted from Law360
The U.S. Court of Appeals for the Ninth Circuit's July 25 decision in Noohi v. Johnson & Johnson Consumer Inc. addresses whether an expert who proposes a damages model in a putative class action must run the model before a class may be certified.[1]
According to the Noohi court, perhaps not. Noohi appears to hold that unexecuted damages models may, in certain circumstances, suffice for purposes of class certification under Rule 23.
To the extent it so holds, the decision has the potential to create judicial inefficiencies and crippling uncertainties for class action defendants.
As this article explores, parties and courts alike benefit from certifying class actions only upon a finding that an expert's executed model is appropriate, accurate and reliable.
The plaintiff in Noohi alleged that Johnson & Johnson's Neutrogena "Oil-Free Face Moisturizer for Sensitive Skin" was not oil-free as advertised because it contained ethylhexyl palmitate and soybean sterols.
The plaintiff moved for class certification, relying on a materials science expert and an econometrics expert. Johnson & Johnson opposed class certification and challenged both experts, but the U.S. District Court for the Central District of California certified the class over those objections.
On appeal under Rule 23(f), the Ninth Circuit rejected the arguments that the econometrics expert's proposed damages model was too underdeveloped and affirmed the district court's holdings.
In so doing, the Ninth Circuit asserted that:
class action plaintiffs may rely on a reliable though not-yet-executed damages model to demonstrate that damages are susceptible to common proof so long as the district court finds that the model is reliable and, if applied to the proposed class, will be able to calculate damages in a manner common to the class at trial.
The court also underscored that only a "limited Daubert analysis may be necessary" at class certification and that a "full-blown Daubert assessment ... would be premature."
Johnson & Johnson argued the econometrics expert did not meet the Daubert standard, derived from the 1993 U.S. Supreme Court decision in Daubert v. Merrell Dow Pharmaceuticals Inc., because the expert had not finalized the survey design or selected the final survey population.
The Ninth Circuit rejected this argument because the expert "had designed the survey methodology and identified target respondent populations," and the "speculative possibility" that the expert might slip up in executing the model was insufficient to defeat class certification.
The Noohi court emphasized that the "plaintiff may not avoid ultimate scrutiny of the admissibility of their experts' final opinions simply by declining to develop those opinions in advance of class certification," and the defendants "must be given the opportunity to test the admissibility and reliability" of the model once executed.
The Supreme Court has not opined on whether an unexecuted damages model may support a motion for class certification, but it has commented on whether a full Daubert analysis is required of experts whose opinions are relevant to the question of certification.
In 2011, in Wal-Mart Stores Inc. v. Dukes, the Supreme Court cast doubt on the proposition asserted by a district court judge that Daubert did not apply at the certification stage of class action proceedings.[2]
In 2013, in Comcast Corp. v. Behrend, another putative class action involving a proposed damages model, the Supreme Court emphasized that it may be necessary to probe behind the pleadings before coming to rest on the certification question.[3]
The high court did not address the issue of unexecuted damages models, finding that the plaintiffs' damages model did not adequately isolate the damages it measured to those properly sought on a classwide basis — creating a mismatch between the damages measured and those the plaintiffs could seek.
Following the Supreme Court's comments in the cases noted above, lower courts have split on the question of whether Daubert applies at the certification stage.
The Third, Fifth, Sixth, Seventh and Eleventh Circuits have required that expert evidence that is relevant or important to the predominance inquiry must pass a full Daubert analysis before a class may be certified.
The U.S. Court of Appeals for the Eighth Circuit, on the other hand — in a position that appears to depart from Supreme Court jurisprudence — has held that a district court need not decide at the class certification stage whether the evidence will ultimately be admissible, even if the court must ultimately determine if the expert evidence is admissible at a later junction.
To the extent Noohi — or any other case — requires less than a thorough Daubert analysis for class certification, it is at odds with the Supreme Court's jurisprudence regarding the plaintiff's burden at class certification.
As the court has observed:
the class action is an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only. In order to justify a departure from that rule, a class representative must be part of the class and possess the same interest and suffer the same injury as the class members.
In other words, permitting class certification before knowing whether class members and their named representative have the same injury that can be addressed with common proof of damages departs from the policy underlying the entire existence of the class action mechanism.
The problem is not only theoretical, and courts have acknowledged the enormous practical pressure that class certification creates for defendants.
In In re: Rhone-Poulenc Rorer Inc., the U.S. District Court for the Northern District of Illinois had certified a class action in 1994 that was brought by hemophilia patients who were allegedly infected with HIV from using certain manufacturers' blood products.[4]
At the time, there was no route to appeal the certification of the class, so the defendants sought a writ of mandamus from the U.S. Court of Appeals for the Seventh Circuit to decertify the class.
Despite noting that mandamus "is issued only in extraordinary cases," the appeals court granted the defendants' petition, in large part because it acknowledged the "intense pressure to settle" that came with the certification and, consequently, a single roll of the dice with a jury as opposed to trying each case separately.
The court even noted that U.S. Circuit Judge Henry Friendly in "Federal Jurisdiction: A General View" had once "called settlements induced by a small probability of an immense judgment in a class action 'blackmail settlements.'" It was precisely this concern about settlement pressure that influenced the decision to adopt Rule 23(f) in 1998.
The amendment established a discretionary process for interlocutory appeals of class certification decisions, and as the advisory committee for Rule 23(f) noted, "[a]n order granting certification ... may force a defendant to settle rather than incur the costs of defending a class action and run the risk of potentially ruinous liability."
Similarly, in 1999 in Blair v. Equifax Check Services, the Seventh Circuit determined that:
a grant of class status can put considerable pressure on the defendant to settle, even when the plaintiff's probability of success on the merits is slight ... This interaction of procedure with the merits justifies an earlier appellate look. By the end of the case it will be too late.[5]
To the extent Noohi allows class certification without concrete evidentiary support, it conflicts with the history of Rule 23(f) and cases such as Rhone-Poulenc.
An additional problem with Noohi is efficiency. If Noohi means that plaintiffs can throw out any generally accepted methodology but not actually perform the analysis, then they could be flat wrong that a class should be certified.
For example, what if the expert's model in Noohi turns out not to be feasible, determines there are no damages, or finds no common proof of damages? Then there was never any evidence to support class certification in the first place.
Noohi says that defendants "must be given the opportunity to test the admissibility and reliability" of an expert's model "once it has been fully executed."
But this conclusion unnecessarily and, for defendants, unjustly creates two class certification phases — one where plaintiffs simply propose a damages model and a second where plaintiffs actually run a damages model to assess potential damages. This outcome is at odds with Rule 23(f).
For judicial and party efficiency, a proposed methodology for determining damages in a class action should be required to be run and addressed under Daubert "at the earliest practicable time," as Rule 23(f) requires.
It is difficult to conceive of other expert testimony where a proposed opinion rather than an actual opinion is permissible, and Noohi offers no countervailing reasons, certainly none strong enough to overcome the strong policy reasons for requiring a defensible, executed model at the class certification stage.
[1] Noohi v. Johnson & Johnson Consumer Inc., No. 23-55190, 2025 WL 2089582 (9th Cir. July 25, 2025).
[2] Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350 (2011).
[3] Comcast Corp. v. Behrend, 569 U.S. 27 (2013).
[4] In re Rhone-Poulenc Rorer, Inc., 51 F.3d 1293, 1298 (7th Cir. 1995).
[5] Blair v. Equifax Check Services,181 F.3d 832, 834 (7th Cir. 1999).