Katherine Weaver1Katherine Weaver is a J.D. Candidate at the University of Chicago Law School, Class of 2024. She thanks Michael T. Brody for his expertise and insight, as well as Matthew Makowski, Renic Sloan, Virginia Robinson, and the University of Chicago Law Review Online team.
* * *
In April of 2022, the Ninth Circuit, sitting en banc, handed down the latest decision in its “packaged tuna antitrust saga,” with implications that have the potential to usher in a sea change to class action practice. The lengthy opinion addressed a number of unsettled questions pertinent to class action jurisprudence, but one particularly salient holding—and the topic of this Case Note—was the majority’s refusal to adopt a per se rule prohibiting class certification when the putative class contains more than a de minimis number of uninjured class members. In so doing, the majority potentially created a circuit split.
Because of the significant attention the decision received, as well as its high stakes for class action litigants, many predicted the Supreme Court would grant review. But in November the Court declined to hear the case. Accordingly, for now, lower courts and practitioners are left to navigate this divisive issue without important Supreme Court guidance.
This Case Note begins by providing a brief overview of the basic principles of class certification animating the Olean decision. It then describes the potential circuit split arising from the majority’s opinion. This Case Note concludes by suggesting that lower courts grappling with this issue in the future should take into account the Seventh Circuit’s opinion in Mussat v. IQVIA, Inc. (2020) conceiving of a class as an “entity.”
I. Rule 23 and Class Certification Requirements
Class actions are governed by Federal Rule of Civil Procedure 23. Perhaps in response to mounting criticism over the class action device, the Supreme Court has recognized and strictly enforced important new guardrails for class certification over the past decade. Specifically, where class certification depends on contested facts, the Supreme Court now mandates that a trial court engage in a “rigorous analysis” to ensure that the prerequisites of Rule 23 have been satisfied before certifying a class. This analysis, the Supreme Court has noted, will sometimes “entail some overlap with the merits of the plaintiff’s underlying claim.” Wal-Mart Stores, Inc. v. Dukes (2011) marked the onset of this new understanding in class certification practice, which was later cemented in Comcast Corp. v. Behrend (2013).
Perhaps the greatest hurdle that plaintiffs face at class certification in damages actions, then, is affirmatively demonstrating satisfaction of Rule 23(b)(3)’s “predominance” requirement.2Rule 23(b)(1) and (b)(2) allow for alternate methods of certification that do not have a “predominance” requirement, but actions seeking damages must comply with Rule 23(b)(3) and thereby satisfy predominance. Rule 23(b)(3) is by far the most commonly invoked basis for class certification. To meet this requirement, plaintiffs must show that questions of law or fact common to the class members—i.e., issues which may be resolved by common evidence “in one stroke”—will predominate over issues specific to individual class members. Wal-Mart and Comcast make clear that putative class representatives must actually prove—not simply plead—that their proposed class satisfies each requirement of Rule 23. Some courts have thus hesitated to certify putative classes where statistical evidence proffered at the class certification stage indicates the presence of more than a few uninjured class members. Undergirding the courts’ logic in these instances was the worry that individualized inquiries into which plaintiffs suffered an injury will inevitably predominate over other common questions of law or fact.
II. Olean and a Potential Circuit Split
While the proper treatment of uninjured class members has long puzzled the courts, the Ninth Circuit’s recent decision in Olean v. Bumble Bee Foods LLC (2022) appears to have created a circuit split over whether certification is permissible when the putative class potentially includes more than a de minimis number of uninjured plaintiffs.
The First and D.C. Circuits have both suggested that a class including more than a de minimis number of uninjured class members may not be certified, citing the aforementioned Rule 23(b)(3) predominance concerns. Specifically, in In re Asacol Antitrust Litigation (2018), a case concerning antitrust claims against a drug manufacturer, the First Circuit held that because 10% of the members of a proposed class were uninjured, the putative class did not satisfy the predominance requirement. The court explained that “this is not a case in which a very small absolute number of class members might be picked off in a manageable individualized process at or before trial. . . . The need to identify [uninjured] individuals will predominate and render an adjudication unmanageable.”
Next, in In re Rail Freight Fuel Surcharge Antitrust Litigation (2019), a case involving an alleged price-fixing conspiracy for fuel surcharges among freight railroads, the D.C. Circuit affirmed the district court’s denial of certification to a class that could not prove an injury to 12.7% of its members. Although the D.C. Circuit did not expressly acknowledge a specific de minimis threshold, it concluded that “for the sake of argument” a class containing 12.7% uninjured class members would not meet that bar. Instead, it suggested that 5% to 6% uninjured class members would constitute the outer limits of an appropriate de minimis threshold.
In April 2022, the Ninth Circuit, sitting en banc, arguably diverged from its sister circuits. The Olean plaintiffs are grocery stores, commercial food preparers, and individual consumers. Together they sued the three largest domestic packaged tuna producers—Tri Union Seafoods, Bumble Bee Foods, and Starkist—under state and federal antitrust law for conspiring to fix packaged tuna prices. One of the issues before the district court concerned a “battle of the experts.” Specifically, the plaintiffs’ expert introduced a statistical model suggesting that approximately 5.5% of the putative class members did not suffer an antitrust injury, while defendants’ expert put forth a similar model indicating that approximately 28% of the class members could not demonstrate an antitrust injury. The district court left it to the jury to determine whose model was ultimately correct, noting that at the class certification stage the only concern was whether plaintiffs’ model was capable of showing impact to nearly all of the class—not that it in fact showed that the injury occurred.
On appeal, a panel of the Ninth Circuit held that the district court should have resolved the factual dispute regarding uninjured class members before certification. “Statistical evidence,” the panel remarked, “is not a talisman. Courts must still rigorously analyze the use of evidence to test its reliability and to see if the statistical model does in fact mask individualized differences.” Not long after the panel’s decision, however, a majority of Ninth Circuit judges voted to rehear the case en banc and found that the district court had not abused its discretion in declining to resolve this “battle of the experts.” According to the majority, the district court had properly inquired whether the evidence established that the common question of antitrust injury was capable of class-wide resolution. There was no need to decide “whether the evidence in fact established that plaintiffs would win at trial” (emphasis added). Next, and relevant here, the en banc majority went on to “reject the dissent’s argument that Rule 23 does not permit the certification of a class that potentially includes more than a de minimis number of uninjured class members.”
In dissent, Judge Lee argued that the majority’s holding on the de minimis rule flouted the plain text of Rule 23, common sense, and the D.C. and First Circuit case law discussed above. Judge Lee notably spilled much ink on the policy implications of punting the resolution of the expert battle to trial, highlighting that class actions nearly always settle because the costs and risks of litigation are too high for defendants. The majority’s decision, Judge Lee proclaimed, would incentivize plaintiffs’ counsel to construct “monstrously oversized classes” stuffed with uninjured class members. In creating such classes, Judge Lee warned, plaintiffs would “inflate the potential liability” for defendants in order “to pressure and extract settlements.”
III. The Class as an “Entity” for Class Certification
Even though certiorari has been denied, this divisive issue is not going to go away. The Seventh Circuit’s recent decision in Mussat v. IQVIA, Inc. (2020) suggests an approach for courts trying to resolve future conflict over the proper treatment of uninjured class members. Specifically, courts facing the issue of uninjured class members should draw on the Mussat majority’s logic, which conceptualized a class as an “entity.” To be sure, this conception best applies where aggregate damages can be calculated on a class-wide basis, as was true in Olean, Asacol, and Rail Freight. In these cases, this Part will argue, the presence of uninjured members does not interfere with predominance.
In Mussat, the Seventh Circuit addressed the question of whether the personal jurisdiction principles addressed in Bristol-Myers Squibb Co. v. Superior Court of Cal. (2017) apply to nationwide class actions filed in federal court under a federal statute. There, Florence Mussat, an Illinois physician, brought a putative class action against pharmaceutical company IQVIA under the Telephone Consumer Protection Act on behalf of all persons in the United States who had received unsolicited faxes from IQVIA in the four previous years. IQVIA moved to strike the class definition, arguing that the district court did not have personal jurisdiction over the non-Illinois members of the proposed nationwide class. In conceptualizing the class as one litigating entity, as opposed to an aggregation of individual claims, Judge Wood noted that the district court had personal jurisdiction over the class as a whole, despite not technically having personal jurisdiction over a few of the absent class members. The relevant inquiry, she stated, is whether the “named representatives . . . [are] able to demonstrate either general or specific jurisdiction” (emphasis added). Moreover, noted Judge Wood, the fact that absent class members are not considered in subject matter jurisdiction or venue analyses cautioned against making personal jurisdiction analysis an exception.
Although Judge Wood devoted eight of the twelve pages of her opinion to thoughtful analysis of how the case at bar diverged from Bristol-Myers, pragmatic concerns likely influenced her ultimate holding. If Judge Wood had held otherwise—i.e., that federal courts cannot assert jurisdiction over absent class members who have no connection to the forum in which the case was filed—nationwide class actions would be drastically curtailed.
Lower courts should strongly consider Judge Wood’s conception of the class as an “entity” when grappling with potential certification of a class containing uninjured class members . In this way, it should not matter if some, or more than a de minimis number, of the absent class members are uninjured—the relevant inquiry should focus on whether the named representative can demonstrate an injury.
That said, a caveat is necessary here. This approach is most relevant where the precise membership of the class does not bear on the total losses suffered. For example, in price-fixing antitrust actions, like those at the heart of the potential circuit split described above, plaintiffs nearly always put forth evidence of aggregate damages to the class as a whole. In these instances, the defendant’s liability typically depends on the average price overcharge resulting from the anticompetitive behavior multiplied by the defendant’s net sales of the product to class members over the class period. That a certain percentage of members in the defined purchaser class did not in fact pay the overcharge should therefore not matter for class certification and predominance—the number of uninjured members does not affect the defendant’s potential liability or the total amount of damages it owes to the class. Other common issues, such as those pertaining to the defendant’s alleged anticompetitive conduct, will therefore predominate at trial, as there will be no need to identify precisely which class members were injured at that stage of the litigation.
As such, Mussat’s conception of the class as an entity applies well in these outlined situations; if the class representative can show they suffered a harm, they can effectively litigate the case. Defendants are not prejudiced by the presence of uninjured members, as their potential liability is not “inflated,” as Judge Lee’s dissent feared. Finally, in these scenarios it is common practice for the district court to “winnow out” the uninjured members at the damages phase, thereby alleviating the concern that undeserving class members will be able to recover.
Of course, despite this approach’s cabined applicability, one could nevertheless raise a parade of horribles counterargument. To borrow Judge Lee’s terminology, would the Mussat approach nevertheless allow certification of a “monstrously oversized” class containing, for example, 80% uninjured class members? While this concern is well-taken, in practice, other elements of Rule 23 would likely be able to step in to prevent such a class from being certified. For example, Rule 23(b)(3) also requires that the class device be “superior to other available methods for fairly and efficiently adjudicating the controversy.” As such, a district court judge could fairly decide that certifying a class containing such an egregious number of uninjured members would not be the superior method of litigating the claims. Furthermore, the majority in Olean provides a different possibility. While the Olean majority warned of instances in which a class definition might be “fatally overbroad,” they noted that “ultimately, the problem of a potentially ‘over-inclusive’ class ‘can and often should be solved be refining the class definition rather than by flatly denying class certification on that basis.’” Accordingly, other safeguards should work to prevent these sorts of drastic edge cases, which constitute a clear misuse of the class device, from going forward.
Overall, while the policy concerns that Judge Lee highlights in his dissent are important, policy concerns regarding barriers to class action plaintiffs also carry great weight. Mandating that plaintiffs affirmatively prove injury to every class member at the class certification stage sets an extremely onerous evidentiary burden. Ultimately, using the Mussat conception of the class as an entity, where the presence of uninjured class members would not interfere with predominance, would have the benefit of harmonizing, in many instances, this area of class action practice with the analyses pertaining to jurisdiction and venue. Moreover, this approach would also ease the burdens on many class action plaintiffs, who already face myriad hurdles to certification.
* * *
Katherine Weaver is a J.D. Candidate at the University of Chicago Law School, Class of 2024. She thanks Michael T. Brody for his expertise and insight, as well as Matthew Makowski, Renic Sloan, Virginia Robinson, and the University of Chicago Law Review Online team.