As we witness increased market concentration and dominant players in many important industries, antitrust law and its enforcement are under attack as weak and ineffectual. Many in Congress want to strengthen antitrust law and make it more relevant to today’s business environment, including the increasing number of clear or near monopolies. So now is an awkward time for judicial decisions making it more difficult for antitrust claimants to secure class certification. The Ninth Circuit U.S. Court of Appeals has done just that in Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 993 F.3d 774 (9th Cir. 2021). The good news is that the panel, at the suggestion of an undisclosed Circuit’s judge, has invited attorneys to argue whether the full court should reexamine the decision. The sole issue on rehearing would be whether Federal Rule of Civil Procedure 23(b)(3) requires a district court to find that “no more than a ‘de minimis’ number of class members are uninjured before certifying a class.”
It should be noted that the Olean decision is not all bad. To the contrary, the Court found “no issue with Plaintiffs’ use of averaging assumptions in its regression models”; determined that even if individualized differences existed, “a higher initial list price as a result of Defendants’ price-fixing scheme could have raised the baseline price at the start of negotiations and could have affected the range of prices that resulted from negotiation”; and, reiterated the principle that the presence of individualized damages cannot, by itself, defeat class certification under Rule 23(b)(3).
But in a prelude to what was coming, the Court also determined that the burden of proof of predominance requires the District Court to resolve factual disputes by a preponderance of the evidence and that the question for the District Court is not whether common issues could predominate at trial but whether they actually do, before certifying the class.
Then, the panel went off the rails by interpreting Rule 23(b)(3) to require a proposed class to be defined in a way that would exclude class members who did not suffer legal injury caused by a defendant’s wrongdoing. According to the decision, a class that includes more than a de minimis number of uninjured members infers that issues common to the class do not predominate over individual issues, so a district court may not certify the class. While it didn’t draw a bright line, the decision suggests the de minimis threshold would top out at about 5% or 6% of class members, while 12.7% would likely be too many.
The relevant requirement of Federal Rule of Civil Procedure 23(b)(3) is that “the questions of law or fact common to class members predominate over any questions affecting only individual members.” But not a single aspect of the language of the rule suggests that the presence of some number of uninjured parties within the proposed class should determine whether “questions affecting only individual members” predominate.
Focusing on how many uninjured class members might be included in the class definition ignores the volume of commerce in question.
While commonly applied, Rule 23(b)(3) predominance is not precisely defined in the case law, even in cases focusing on its application. Most often it is considered a heightened form of the commonality required by Rule 23(a)(2) combined with aspects of manageability to determine whether the case and its proof will mostly be about issues common to class members rather than individualized ones, often referred to as “cohesiveness.” [i] The Merriam-Webster Dictionary defines predominance as “having superior strength, influence, or authority” and “being most frequent or common.” So does Black’s Law Dictionary. Under the plain language of Rule 23, therefore, something akin to a clear majority should satisfy predominance. Ninth Circuit precedents hold that predominance is not a matter of nose-counting, but rather is an assessment of “whether proposed classes are sufficiently cohesive to warrant adjudication by representation.” [ii] Yet, Olean found that if “only” 83.3% of class members suffered injury the court should deny certification on grounds that common issues lack predominance. Nose-counting?
It is conceivable that while 12.7% of class members appear uninjured, the volume of commerce accounted for by those parties could be de minimis relative to the total commerce affected by defendants’ violation. We say this not to suggest substituting an “affected commerce” test for the de minimis test articulated in the opinion, but to highlight the arbitrary nature of the Olean court’s focus on the number of class members.
In issuing its ruling, the Ninth Circuit reversed the District Court for the Southern District of California’s certification of three classes of purchasers of packaged tuna, which globally is an $8 billion industry. The District Court reserved for the jury the parties’ dispute over number of uninjured class members, which was error, the appeals court said, because the District Court itself must decide whether or not the predominance requirement is met. The appeals court therefore vacated class certification based on its view that predominance cannot be found if more than a de minimis number of uninjured class members falls within the class definition.
To support class certification, the plaintiffs submitted the kind of econometric evidence frequently used to show common impact and to estimate the anti-competitive overcharge in price-fixing cases — the amount by which prices during the violation period exceeded prices that would have prevailed in the absence of the defendants' unlawful collusion. Their economists estimated reduced-form regression models of wholesale prices as a function of a series of relevant explanatory variables over different time periods, using transactional data provided by defendants and other data on product characteristics and other relevant factors. The direct purchasers' regression showed that prices during the violation period were, on average, 10.28% higher than the competitive price during the baseline periods without price-fixing. A second regression analysis showed that 1,111 out of 1,176 class members, or 94.5%, experienced the 10.28% overcharge.
The court addressed several questions relating to the threshold question of “whether Plaintiffs' representative evidence can be used to establish predominance.” These are: (1) whether the same evidence could be used to establish liability in a class member's individual suit, (2) whether the representative evidence sufficiently links plaintiffs' injuries to their theory of antitrust liability, and (3) the use of averaging assumptions does not categorically defeat predominance.
Expanding on the use of regression analysis, the court said it had “no issue” with the direct purchaser plaintiffs' use of “averaging assumptions in its regression models.” Their expert averaged the overcharge calculation using Bumble Bee Foods’ own data, then used that average in a regression model to calculate what percentage of the class was impacted. “Presuming the reliability of Plaintiffs’ statistical methodology ... the representative evidence can show that virtually all class members suffered an injury due to Defendants’ alleged wrongdoing,” the Court of Appeals concluded. The panel also reiterated a principle from its 2013 ruling in Leyva v. Medline Indus., 716 F.3d 510 (9th Cir. 2013), that the presence of individualized damages cannot, by itself, defeat class certification under Rule 23(b)(3).
The court stressed throughout its decision that the District Court must resolve factual disputes over predominance by a preponderance of the evidence and that the question for the District Court is not whether common issues could predominate at trial but whether they actually do, before certifying the class.
In the end, the panel found in favor of a de minimis requirement, which we believe is a remarkable repudiation of Rule 23’s design. The potential number of class members, a metric which the court relied on, is an arbitrary one. Olean’s de minimis test also contradicts the well-settled principle, expressly acknowledged by the court that variations in the extent of each class member’s injury does not defeat predominance. By requiring no more than a small number of class members that potentially have suffered no injury, the new de minimis rule needlessly distinguishes between class members that have suffered minor or trivial injury from those that suffered no injury at all. The presence of individualized damages does not defeat predominance – and, therefore, certification – so it is not at all clear why the presence of members whose damages are zero should, even if they comprise a large minority. Historically such issues were best handled as an administrative issue in the claims process.
Antitrust laws are remedial; deterrence is an important underlying policy.
The law has previously effectuated the goals of remediation and deterrence by means of liberal interpretations in favor of class members. The upshot of the Ninth Circuit’s decision is not to protect against recovery by uninjured parties, but to penalize the clear majority of class members by denying certification.
U.S. Circuit Judge Andrew Hurwitz observed in his dissent that the critical question is not “what percentage of class members is injured, but rather whether the district court can economically ‘winnow out’ uninjured plaintiffs to ensure they cannot recover for injuries they did not suffer. … If the district court can ensure that uninjured plaintiffs will not recover, their mere presence in the putative class does not mean that common issues will not predominate.” The presence of uninjured class members, he wrote, merely reflects the possibility that “an injurious course of conduct may sometime fail to cause injury.”
Judge Hurwitz nailed it.
It is often difficult or impossible for class action plaintiffs in antitrust cases to define the proposed class in a way that includes only class members who suffered recoverable damages without violating the prohibition against “fail-safe” class definitions.[iii] When this occurs, defendants are quick to attack certification on the grounds that the proposed class includes uninjured class members who would lack standing were they to bring a claim individually. Courts recently have responded to such arguments by denying class certification where “too many” class members potentially suffered no injury. It is often not clear, however, whether a given proportion of potentially uninjured class members is too high or why the presence of such parties should defeat class certification at all.
In the world of track and field, hurdlers must clear a series of equally high obstacles just to finish – not even win – a race. If the rules suddenly add a hurdle in lane two that is higher than the rest, only the nimblest of sprinters would keep pace with the rest. The Ninth Circuit has constructed an additional and higher hurdle for plaintiffs to overcome, awarding them with a harsher punishment even than those who cause antitrust injury.
Competition law is already facing criticism that it is weak and ineffectual. If it is left to stand, the Ninth Circuit’s decision in Olean v. Bumble Bee Foods will only further undermine the efficacy of American antitrust law. En banc review is warranted. We are encouraged that the panel has called for briefs on the question and are hopeful that it will see what damage the ruling will do to our nation’s fair market goals.
[i] See, e.g., Wal-Mart v. Dukes, 564 U.S. 338 (2011), Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 455 (2013), and Comcast Corp. v. Behrend, 569 U.S. 27 (2013).
[ii] See, e.g. Jimenez v. Allstate Ins. Co., 765 F.3d 1161, 1165 (9th Cir. 2014); Vinole v. Countrywide Home Loans, Inc., 571 F.3d 935, 944 (9th Cir. 2009)
[iii] See, e.g. Vizcaino v. U.S. District Court for the Western District of Washington, 173 F.3d 713 (9th Cir. 1999), amended, 184 F.3d 1070 (9th Cir. 1999); Kamar v. RadioShack Corp., 375 F. App’x 734, 736–37 (9th Cir. 2010); Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 825 (7th Cir. 2012) (describing fail-safe class definition as “heads I win; tails you lose”).
Edited by Tom Hagy for MoginRubin LLP.
