Olean Wholesale Grocery Cooperative, Inc. V. Bumble Bee Foods LLC

Ninth Circuit Offers Guidance to Trial Courts in Evaluating Ancillary Predominance Issues for Purposes of Rule 23(b)(3)

Jonathan L. Goldberg

Recent Case: Olean Wholesale Grocery Cooperative, Inc. v. Bumble Bee Foods LLC, 31 F.4th 651 (9th Cir. 2022) (en banc)

The Ninth Circuit Court of Appeals recently held in an en banc ruling that district courts wield significant discretion when deciding whether to certify a class action containing potentially uninjured class members. The opinion rejected a “de minimis” rule, which, according to Defendants, other circuits adopted. The court properly focused on Rule 23’s broad text and the class mechanism’s core efficiency goals. However, the Ninth Circuit prematurely addressed the de minimis issue because its opinion reaffirmed the district court’s finding that each plaintiff was similarly situated. Still, in concluding Defendants failed to demonstrate a fatal dissimilarity within the class, the en banc panel effectively reasoned that opposition to predominance at class certification must attack evidence’s relevancy as to each class member rather than its sufficiency in proving the class claims.

Introduction

The Supreme Court recently confirmed that uninjured plaintiffs may not recover damages from class action judgments but left open questions of how that rule might affect a trial court’s class certification decision.²2. See TransUnion LLC v. Ramirez, 141 S. Ct. 2190, 2208, 2208 n.4 (2021) (holding Article III requires class members to have standing to recover damages but declining to answer the “distinct question whether every class member must demonstrate standing before a court certifies a class”). In Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC,³3. 31 F.4th 651 (9th Cir. 2022) (en banc). the Ninth Circuit provided crucial guidance to trial courts struggling to apply TransUnion’s holding.

Class action lawsuits depart from the usual rule that only named parties conduct litigation.⁴4. Califano v. Yamasaki, 442 U.S. 682, 700–01 (1979). Federal Rule of Civil Procedure 23 outlines the strict conditions a class must meet to ensure aggregate litigation proceeds fairly, both for absent class members and defendants, and advances judicial economy.⁵5. See Amgen Inc. v. Connecticut Ret. Plans & Tr. Funds, 568 U.S. 455, 470 (2013) (citing Richard A. Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L. Rev. 97, 107 (2009)) (confirming predominance tests whether a court can resolve dissimilarities among class members in a manner that is not “inefficient or unfair”). To recover damages for themselves and the absent class members they represent, putative class representatives usually certify their class under Rule 23(b)(3).⁶6. See Richard A. Nagareda, Robert G. Bone, Elizabeth Chamblee Burch & Patrick Wooley, The Law of Class Actions and Other Aggregate Litigation 274–75 (3d ed. 2020) (suggesting that plaintiffs are extremely unlikely to recover monetary damages through a (b)(2) class after Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338 (2011)). Rule 23(b)(3) demands that class litigation be superior to other adjudicatory methods and that common questions of law or fact predominate over individual questions.⁷7. Fed. R. Civ. P. 23(b)(3). Parties frequently target the predominance requirement to challenge motions for class certification.⁸8. See, e.g., Amchem Prods., Inc. v. Windsor, 521 U.S. 591 (1997) (class settlement); Halliburton Co. v. Erica P. John Fund, Inc. (Halliburton II), 573 U.S. 258 (2014) (securities fraud); Amgen, 568 U.S. at 466 (securities fraud); Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 454 (2016) (Fair Labor Standards Act). In short, predominance asks whether common questions of law or fact are both central to the litigation and more prevalent or important than individual ones.⁹9. Tyson Foods, 577 U.S. at 453–54 (2016) (“The predominance inquiry ‘asks whether the common, aggregation-enabling, issues in the case are more prevalent or important than the non-common, aggregation-defeating, individual issues’ [and whether] ‘one or more of the central issues in the action are common to the class and can be said to predominate . . . .’” (first quoting 2 W. Rubenstein, Newberg on Class Actions § 4:50 (5th ed. 2012); then quoting 7AA Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure § 1778 (3d ed. 2005))).

The Olean defendants contended that the plaintiff class included more than a de minimis number of uninjured members, prompting many individual questions, preventing common issues from predominating, and automatically precluding certification.¹⁰10. A question is common because its answer resolves a central issue in each class member’s claim. Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 31 F.4th 651, 663 (9th Cir. 2022) (en banc) (quoting Wal-Mart, 564 U.S. at 350). By contrast, an individual question requires different evidence to prove each class member’s claim. Id. (citing Tyson Foods, 577 U.S. at 453). Uninjured plaintiffs within a class may raise individual questions because a court must determine “which ones” are injured and “which ones” are not. See Tyson Foods, 577 U.S. at 464–66 (Roberts, C.J., concurring) (finding decertification appropriate where a district court cannot identify the uninjured plaintiffs within the class); cf. Olean, 31 F.4th at 681–82 & n.31 (discussing mini-trials to identify each plaintiff’s damages award). Defendants also argued the D.C.¹¹11. In re Rail Freight Fuel Surcharge Antitrust Litig. (Rail Freight II), 934 F.3d 619, 624–25 (D.C. Cir. 2019) (discussing a “six-percent upper limit” on the number of uninjured class members in a certified class). and First¹²12. See In re Asacol Antitrust Litig., 907 F.3d 42, 47, 51–58 (1st Cir. 2018) (intimating that 10% exceeds the de minimis boundary). Circuits already adopted a “de minimis” rule and urged the Ninth Circuit to hold similarly.¹³13. Defendants-Appellants’ Supplemental En Banc Brief at 19, 31 F.4th 651 (9th Cir. 2022) (No. 3:15-MD-026770-JLS-MDD), 2021 WL 4126353, at *19. Cf. Olean, 31 F.4th at 666 n.9 (9th Cir. 2022) (discussing the argument but not directly attributing it to defendants); id. at 692 (Lee, J., dissenting) (same). In rejecting a per se de minimis standard, the en banc court held that a district court is in the best position to determine whether individual questions, including those regarding class members’ injury, will overwhelm common ones.¹⁴14. Olean, 31 F.4th at 669. In other words, the district court’s decision to certify the class fell within the broad range of permissible conclusions that a class certification appeal’s abuse of discretion standard affords.¹⁵15. Id. (quoting Hung Lam v. City of San Jose, 869 F.3d 1077, 1084 (9th Cir. 2017) (quoting Kode v. Carlson, 596 F.3d 608, 612 (9th Cir. 2010))).

The court properly resolved an issue percolating in class action jurisprudence through careful attention to Rule 23’s text and the class mechanism’s core efficiency goals. Furthermore, the decision follows a broad trend—developing as judicial experience with class actions grows—of increasing deference towards trial courts at the class certification stage. Nevertheless, the en banc panel should never have reached the issue. The opinion’s logic renders the holding advisory by concluding each plaintiff could rely upon their expert’s report to prove class-wide antitrust impact—i.e., that defendants injured each and every class member by causing them to pay for tuna at supra-competitive prices. Still, the appellate tribunal faithfully applied Supreme Court precedent on a frequently confusing aspect of the predominance inquiry, carefully distinguishing between evidentiary issues of relevancy (whether there is a “fatal dissimilarity”) and sufficiency or persuasiveness (whether there is a “fatal similarity”).¹⁶16. See Nagareda, supra note 4, at 131 (arguing courts should address fatal dissimilarities between class members at certification and address fatal similarities, such as a failure of proof, at summary judgment).

I. Background

A. The District Court’s Class Certification Order

Following a 2015 Department of Justice antitrust investigation, various plaintiffs (collectively “Tuna Purchasers”) filed suit against Bumble Bee, StarKist, Chicken of the Sea (“COSI”), and their parent corporations (collectively “Tuna Suppliers”), alleging the corporations conspired to fix tuna prices in violation of federal and state antitrust laws.¹⁷17. The Tuna Purchasers allege the Tuna Suppliers engaged in a price-fixing conspiracy from November 2010 to at least December 31, 2016 and further claim the conspiracy forced them to pay supra-competitive prices for the Tuna Suppliers’ products. Olean, 31 F.4th at 661–62. Soon thereafter, the Department of Justice (“DOJ”) entered notice of a pending investigation into the packaged tuna industry for similar violations of the antitrust laws.¹⁸18. Id. at 661; see In re Packaged Seafood Prods. Antitrust Litig., 332 F.R.D. 308, 317 (S.D. Cal. 2019) (“Shortly after the commencement of this action, the U.S. Department of Justice (‘DOJ’) noticed the Court of pending investigations of the Defendants. Since that time, Defendants and individual employees have pled guilty and the DOJ has entered multiple indictments.”), vacated and remanded sub nom. Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 993 F.3d 774 (9th Cir. 2021), on reh’g en banc, 31 F.4th 651 (9th Cir. 2022), and aff’d sub nom. Olean, 31 F.4th at 661. During the ongoing civil litigation, the DOJ filed multiple indictments alleging a criminal price-fixing conspiracy in the industry from around November 2011 to December 2013.¹⁹19. Olean, 31 F.4th at 661–62; see also In re Packaged Seafood Prod. Antitrust Litig., 332 F.R.D. at 317. Bumble Bee, StarKist, and three industry executives ultimately pled guilty to the conspiracy; a jury convicted Bumble Bee’s former CEO, and COSI cooperated with the DOJ, admitting to price fixing in exchange for leniency.²⁰20. Olean, 31 F.4th at 662. By the end of 2015, the Judicial Panel on Multidistrict Litigation consolidated the civil complaints in the Southern District of California.²¹21. In re Packaged Seafood Prods. Antitrust Litig., 332 F.R.D. at 316. Judge Janis Lynn Sammartino divided the Tuna Purchasers into four tracks: (1) plaintiffs who filed suit individually against the Tuna Suppliers (“DAPs”); (2) direct purchasers, such as nationwide retailors or regional grocery stores (“DPPs”); (3) indirect purchasers who bought bulk-sized products for prepared food or resale (“CFPs”); and (4) individual end purchasers (“EPPs”).²²22. Id. at 316–17. The DPPs and EPPs include plaintiffs who purchased packaged tuna between June 1, 2011 and July 1, 2015, but the CFPs include plaintiffs who purchased tuna products from June 2011 through December 2016. Olean, 31 F.4th at 662.

The latter three groups moved for class certification in 2018 under Rule 23(b)(3). The Tuna Suppliers opposed the motion, arguing individual questions predominated over common ones because the DPPs’ expert, Dr. Russell Mangum, could not demonstrate a common class-wide antitrust impact.²³23. Olean, 31 F.4th at 673. Each of the three plaintiff subclasses employed their own expert to establish antitrust impact through qualitative and quantitative analyses. Id. at 662. However, this Case Comment will focus on the DPPs’ class certification, the center of each opinion. Dr. Mangum constructed a multiple regression model to assess whether the price-fixing conspiracy subjected each DPP to an overcharge.²⁴24. Id. at 671. To do so, Dr. Mangum pooled the Tuna Suppliers’ actual sales transaction data during benchmark periods before and after the conspiracy, identified a number of variables that could affect the price of tuna—like product characteristics, input costs, consumer type, consumer preferences and demand, etc.—and recorded the model’s results.²⁵25. Id. The model showed “the DPPs paid 10.28 percent more for tuna during the conspiracy period than they did during the benchmark periods.”²⁶26. Id. To further support this finding, Dr. Mangum conducted four robustness checks,²⁷27. Dr. Mangum (1) evaluated the overcharge to each defendant, (2) changed the model to assess the overcharge for different products with different characteristics, (3) altered the model to evaluate overcharge by customer types, and (4) used the output of the pooled regression model to predict the but-for prices paid by the DPP class. Id. at 672. According to Dr. Mangum, each robustness check confirmed the conspiracy generated higher prices for all or nearly all DPPs. Id. and the final one indicated that 94.5 percent of the DPPs purchased at least one product at a supra-competitive rate.²⁸28. Id. To be clear, despite the regression’s result, Dr. Mangum concluded the Tuna Suppliers injured each DPP: The robustness check was one basis for a conclusion that rested on additional “correlation tests, the record evidence and the guilty pleas and admissions entered in [the] case.” See id. at 676.

The Tuna Suppliers’ rebuttal expert to the DPPs, Dr. John Johnson, advanced two areas of critique: (1) Dr. Mangum inappropriately pooled direct purchaser data for his model, papering over differences among class members, such as disparities in bargaining power or negotiating tactics;²⁹29. Dr. Johnson pointed to several empirics to support this argument. A Chow test, a commonly employed statistical tool to assess whether data can be pooled, counseled against data pooling. See id. at 673. Further, Dr. Mangum’s model could not find statistically significant results for twenty-eight percent of the direct purchaser class, so Dr. Johnson argued the plaintiffs could not rely on the model to demonstrate class-wide impact. Id. and (2) Dr. Mangum’s model contained various errors that undermined its validity, including the use of an improper cost index.³⁰30. First, Dr. Mangum’s model outputted false positives, including those who purchased tuna products from non-defendants (non-conspiring tuna producers). See id. at 674. Second, Dr. Mangum’s model did not match the time periods listed in the plaintiff’s complaint. Id. Third, Dr. Mangum used a cost index rather than the Tuna Suppliers’ actual accounting cost. Id. Further, the Tuna Suppliers argued Dr. Johnson’s superior report indicated around twenty-eight percent of the class was uninjured.³¹31. Id. at 680.

The district court certified the class after carefully evaluating Dr. Johnson’s critiques and Dr. Mangum’s rebuttal.³²32. Id. at 662, 675–76. First, the district court found Dr. Mangum’s pooled model to be acceptable. Dr. Mangum’s model included statistically insignificant results as to some direct purchasers because those class members completed too few transactions to provide significant results, but this data issue had no bearing on a direct purchaser’s ability to rely on the model as evidence of impact. Id. at 675. For instance, general evidence that the Tuna Suppliers inflated prices through their conspiracy supported the inference that all direct purchasers were similarly situated. See id. at 674. Second, while the court acknowledged the Chow Test should be taken seriously, its opinion reiterated Dr. Mangum’s assertions that Dr. Johnson designed the Chow Tests to fail by including too many coefficients and observations and concluded that Dr. Mangum’s testimony gave “persuasive reasons, grounded in economic theory, for why a pooled model [was] appropriate” despite the concerning Chow Test results. In re Packaged Seafood, 332 F.R.D. 308, 225 (S.D. Cal. 2019); see also id. at 325 n.9 (offering examples of “multiple courts [that] have addressed instances where a pooled regression model failed a Chow Test, yet still accepted those models”). The trial judge first found each plaintiff was similarly situated and, therefore, able to rely upon Dr. Mangum’s report as well as other evidence—i.e., guilty pleas, market characteristics, and record evidence—to prove a common antitrust impact.³³33. In re Packaged Seafood, 332 F.R.D. at 324. The court concluded that the Tuna Suppliers’ remaining criticisms were “serious and could be persuasive to a finder of fact” but ultimately “beyond the scope of” the certification motion because they merely attacked Dr. Mangum’s persuasiveness rather than his capability of establishing impact for each class member.³⁴34. Id. at 328. The district court still rejected Dr. Johnson’s additional critiques. First, Dr. Mangum included purchases from non-defendant tuna suppliers because the conspiracy had an “umbrella effect” that raised non-colluding tuna suppliers’ prices. Olean, 31 F.4th at 676. Second, Dr. Mangum’s choice to narrow the time frame added to the report’s credibility by improving its accuracy. Id. Third, the court accepted Dr. Mangum’s arguments that cost indexes were preferable for determining competitive market prices as well as his conclusion that defendant-specific costs confirmed the pooled model’s results in any event. Id. at 675–76.

B. The Ninth Circuit’s Three-Judge Panel

The Tuna Suppliers appealed, and the Ninth Circuit’s three-judge panel vacated and remanded. The circuit panel concluded each class member could rely upon Dr. Mangum’s model to establish antitrust impact.³⁵35. See Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 993 F.3d 774, 790 (9th Cir. 2021), reh’g en banc granted, 5 F.4th 950 (9th Cir. 2021), and on reh’g en banc, 31 F.4th 651 (9th Cir. 2022). However, the court found the trial judge abused its discretion in certifying the class without resolving the experts’ competing conclusions on the number of uninjured plaintiffs within the class.³⁶36. Id. at 793. Even though the issue of the experts’ persuasiveness overlaps with the merits of plaintiffs’ claims, the court held that more than a de minimis number of uninjured class members would raise too many individual questions and defeat predominance.³⁷37. Id. at 794. The court precedent analysis “suggest[s] that 5% to 6% constitutes the outer limits of a de minimis number.” Id. at 792 (quoting Rail Freight II, 934 F.3d 619, 624–25 (D.C. Cir. 2019). However, the panel insisted it did “not adopt a numerical or bright-line rule” but only held “that 28% would be out-of-bounds.” Id. at 793. As a result, the district court should have weighed the persuasiveness of each expert report, entered findings on the number of uninjured class members,³⁸38. See Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350–51 (2011) (quoting Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 161 (1982)) (holding that Rule 23 requires a party seeking class certification to affirmatively demonstrate compliance with the Rule, a standard that will frequently require courts to engage in a “rigorous analysis” at certification that overlaps with the merits of the moving party’s claims). and only have certified the class if it contained fewer than a de minimis number of uninjured plaintiffs. Judge Andrew D. Hurwitz broke from the panel. He concurred the trial court should have resolved the factual dispute relating to uninjured class members before certification but dissented from the panel’s decision to adopt a de minimis standard.³⁹39. Olean, 993 F.3d at 794 (Hurwitz, J., concurring in part and dissenting in part). According to Judge Hurwitz, predominance asks not about the number of uninjured class members but whether a district court may “economically” separate uninjured plaintiffs from the class, a determination best left to the trial court’s discretion.⁴⁰40. Id. at 794–95. After all, “Rule 23 certification is at bottom a trial management decision.”⁴¹41. Id. at 796.

C. The Ninth Circuit’s En Banc Resolution

The Ninth Circuit then vacated the panel’s decision⁴²42. Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 5 F.4th 950, 952 (9th Cir. 2021). and reheard the case en banc.⁴³43. Olean, 31 F.4th at 662. Writing for a 9-2 majority, Judge Sandra Segal Ikuta rejected the “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.”⁴⁴44. Id. at 669. The panel conceded that “[w]hen individualized questions relate to the injury status of class members, Rule 23(b)(3) requires that the court determine whether individualized inquiries about such matters would predominate over common questions.”⁴⁵45. Id. at 668. But any numerical rule would go too far. First, the court analogized classes with uninjured plaintiffs to class actions that require individual proof of damages.⁴⁶46. Id. at 668–69. Both the Ninth Circuit⁴⁷47. See, e.g., Vaquero v. Ashley Furniture Indus., Inc., 824 F.3d 1150, 1155 (9th Cir. 2016) (“Our precedent is well settled on this point. . . . [T]he need for individualized findings as to the amount of damages does not defeat class certification.”); Blackie v. Barrack, 524 F.2d 891, 905 (9th Cir. 1975) (“The amount of damages is invariably an individual question and does not defeat class action treatment.”). and Supreme Court⁴⁸48. Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 453–54 (2016) (citing Wright & Miller, supra note 8, § 1778) (noting that individual questions like damages and affirmative defenses do not defeat predomination). permit district courts to certify classes despite the need for individualized damages assessments at trial, “a conclusion implicitly based on the determination that such individualized issues do not predominate over common ones.”⁴⁹49. Olean, 31 F.4th at 669. Second, the majority determined a de minimis rule to be inconsistent with Rule 23’s text, “which requires only that the district court determine after rigorous analysis whether the common question predominates over any individual questions, including individualized questions about injury or entitlement to damages.”⁵⁰50. Id. (citing Fed. R. Civ. P. 23(b)(3)). Rule 23(b)(3)’s general language points towards decisionmaking on a “case-by-case basis, rather than . . . a per se rule.”⁵¹51. Id. at 669 n.13. The court also rejected the dissent’s policy arguments as atextual. “[W]e are bound to apply Rule 23(b)(3) as written, regardless of policy preferences.” Id. Finally, the panel asserted that a district court “is in the best position to determine whether individualized questions” predominate over common ones.⁵²52. Id. at 669. The opinion abandoned the previous panel’s strong justifications for this position, stating only that the abuse of discretion standard permits district courts to rule within a “wide range of permissible outcomes.”⁵³53. Id. (quoting Hung Lam v. City of San Jose, 869 F.3d 1077, 1084 (9th Cir. 2017) (quoting Kode v. Carlson, 596 F.3d 608, 612 (9th Cir. 2010))). However, the en banc panel also noted the presence of uninjured class members may indicate the class is fatally overbroad, an issue a district court may resolve sua sponte with its inherent authority to manage the class action.⁵⁴54. See id. at 669 n.14 (“[A] court must consider whether the possible presence of uninjured class members means that the class definition is fatally overbroad.”); see also id. at 666 (“In such a case, the court may redefine the overbroad class to include only those members who can rely on the same body of common evidence to establish the common issue.”); Tobias Barrington Wolff, Discretion in Class Certification, 162 U. Pa. L. Rev. 1897, 1925 (2014) (“[T]he discretionary power that federal courts possess to reshape the boundaries and composition of the class is continuous with their power to decide whether to certify at all.”).

The court then addressed the “central questions on appeal[:] . . . whether the expert evidence presented by the DPPs is capable of resolving this issue ‘in one stroke;’ and whether this common question predominates over any individualized inquiry.”⁵⁵55. Olean, 31 F.4th at 670 (citation omitted). It found the district court did not abuse its discretion in concluding so.⁵⁶56. Id. at 670. The majority rigorously analyzed both Dr. Mangum’s and Dr. Johnson’s reports and the district court’s handling of the expert’s disagreements.⁵⁷57. Id. at 670–77. The en banc panel held that the district court appropriately addressed Dr. Johnson’s arguments and considered unrebutted record evidence, such as prior guilty pleas.⁵⁸58. Id. at 676. For more detail on the arguments made by both experts, see supra notes 26–33 and accompanying text. The trial court’s recognition that Dr. Johnson’s arguments may prove persuasive at trial did not detract from its ultimate conclusion that “Dr. Mangum’s evidence was capable of showing class-wide impact.”⁵⁹59. Id. at 676. At bottom, “‘each class member could have relied on [the plaintiffs’ evidence] to establish liability if he or she had brought an individual action,’ and the evidence ‘could have sustained a reasonable jury finding’ on the merits of a common question.’”⁶⁰60. Id. at 667 (quoting Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 455 (2016)) (alterations in original).

The court then rejected the Tuna Suppliers’ primary arguments. Their main argument was that the regression model used averaging assumptions to “paper over” or mask the individual differences in class members’ bargaining power and negotiation tactics.⁶¹61. Id. at 677. The court affirmed that regressions models using averaging assumptions are not inherently suspect but rather a commonly used econometric tool.⁶²62. Id. Then, the majority rejected the Tuna Suppliers’ attempt to establish a fatal dissimilarity between Plaintiffs. Even if some DPPs negotiated their tuna prices with greater bargaining power than that of their peers, a conspiracy would logically and plausibly impact all purchasers by inflating the baseline for price negotiations.⁶³63. Id. at 677–78 (quoting In re Urethane Antitrust Litig., 768 F.3d 1245, 1254–55 (10th Cir. 2014)). The court noted Dr. Mangum concluded the largest retailers—those that should have the most bargaining power, such as Wal-Mart—still paid supracompetitive prices.⁶⁴64. Id. at 678. At most, the Tuna Suppliers’ argument suggested DPPs have different damages. But, “[w]hile individualized differences among the overcharges imposed on each purchaser may require a court to determine damages on an individualized basis, . . . such a task would not undermine the regression model’s ability to provide evidence of common impact.”⁶⁵65. Id. at 679. The majority reiterated that individualized damages do not threaten predominance. Id. With respect to impact, all DPPs were similarly situated.

Finally, the court dismissed the Tuna Suppliers’ complaint that the district court refused to resolve the parties’ dispute on the number of uninjured class members. The majority first clarified the Tuna Suppliers’ argument was premised on a misreading of Dr. Johnson’s report.⁶⁶66. Id. at 680. The Tuna Suppliers read Dr. Johnson’s report to suggest twenty-eight percent of the DPP class was uninjured. Id. However, Dr. Johnson’s test was only an attempt to undermine the confidence in Dr. Mangum’s model because it did not produce statistically significant results. Id. The court held the statistic did not support the Tuna Suppliers’ underlying claim. Id. The district court resolved this dispute as well. Id. at 681 (“[T]he district court determined that Dr. Mangum’s pooled regression model was capable of showing that the DPP class members suffered antitrust impact on a class-wide basis, notwithstanding Dr. Johnson’s critique.”). Then, the opinion confirmed neither expert’s report raised individual inquiries into the class members’ injuries. The trial court already concluded each DPP’s bargaining power was immaterial to a finding of common price impact, and the Tuna Suppliers provided no other factual or legal grounds to distinguish between individual class members.⁶⁷67. Id. at 681 (“The district court fulfilled its obligation to resolve the disputes raised by the parties in order to satisfy itself that the evidence proves the prerequisites for Rule 23(b)(3), which is that the evidence was capable of showing that the DPPs suffered antitrust impact on a class-wide basis.”); see also Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 622–23 (1997) (“The predominance requirement . . . trains on the legal or factual questions that qualify each class member’s case as a genuine controversy . . . .”). Thus, each class member was similarly situated. In other words, evidence relevant to one class member would be relevant to them all. The Tuna Suppliers’ remaining arguments simply attacked the expert report’s persuasiveness, a determination for the jury at trial.⁶⁸68. Olean, 31 F.4th at 681.

If the jury found that Dr. Mangum’s model was reliable, then the DPPs would have succeeded in showing antitrust impact on a class-wide basis, an element of their antitrust claim. On the other hand, if the jury were persuaded by Dr. Johnson’s critique, the jury could conclude that the DPPs had failed to prove antitrust impact on a class-wide basis.⁶⁹69. Id.

Judge Kenneth K. Lee dissented.⁷⁰70. The majority’s opinion briefly addressed the CFP and EPP classes in its conclusion. The en banc panel held the district court did not abuse its discretion in certifying both classes. Id. He first stressed the importance of a rigorous analysis at class certification to prevent in terrorem settlements.⁷¹71. Id. at 691 (Lee, J., dissenting) (“‘[W]hen damages allegedly owed to tens of thousands of potential claimants are aggregated and decided at once, the risk of an error will often become unacceptable. Faced with even a small chance of devastating loss, defendants will be pressured into settling questionable claims.’” (quoting AT&T Mobility LLC v. Concepcion, 563 U.S. 333, 350 (2011))). Next, he argued the district court did not resolve the dueling experts’ opinions on the presence of uninjured class member, asserting class certification demands the moving party prove Rule 23’s prerequisites by a preponderance of the evidence after a rigorous analysis.⁷²72. Id. at 687–88. The dissent took issue with the majority’s attempt to “wave[] away” the differences in DPPs’ “negotiating power,” which would have permitted certain retailers to extract rebates or promotional concessions and thereby push tuna prices below competitive levels.⁷³73. Id. at 690. At the very least, according to the dissent, the “only way” to “find out if Wal-Mart and other major retailers suffered any injury” would be to conduct a “highly individualized analys[is]” that defeats predominance.⁷⁴74. Id. Finally, the dissent stressed that the court’s rejection of a de minimis rule would generate a circuit split.⁷⁵75. Id. at 691. However, the majority denies the creation of a circuit split. Id. at 669 n.13 (arguing neither case adopted a per se rule but held that based on the particular facts in those disputes, the “need to identify uninjured class members” would “render an adjudication unmanageable” (quoting In re Asacol Antitrust Litig., 907 F.3d 42, 53–54 (1st Cir. 2018))).

According to the dissent, both the D.C. and First Circuits settled on a de minimis rule. The D.C. Circuit stated “5% to 6% constitutes the outer limits of a de minimis number” of uninjured class members,⁷⁶76. Id. at 692 (quoting In re Rail Freight Fuel Surcharge Antitrust Litig. (Rail Freight II), 934 F.3d 619, 625 (D.C. Cir. 2019)). and the “First Circuit suggested that ‘around 10%’ of uninjured class members marks the de minimis border.”⁷⁷77. Id. (quoting Asacol, 907 F.3d at 47). But, the Ninth Circuit majority properly denied the creation of a circuit split.⁷⁸78. Id. at 699 n.13. Both sister circuits defined de minimis “in functional terms”⁷⁹79. Asacol, 907 F.3d at 54 (quoting In re Nexium Antitrust Litig., 777 F.3d 9, 30 (1st Cir. 2015)). and concluded that the need to identify uninjured class members precluded predominance based on the “nuanced”⁸⁰80. Rail Freight II, 934 F.3d at 625. and “particular facts of the cases before them.”⁸¹81. Olean, 31 F.4th at 699 n.13. While the D.C. and First Circuits contemplated per se boundaries, their opinions ultimately hinged on whether the trial court could employ a “mechanism that can manageably remove uninjured persons from the class in a manner that protects the parties’ rights.”⁸²82. Asacol, 907 F.3d at 54; see Rail Freight II, 934 F.3d at 625 (holding the district court did not abuse its discretion in denying class certification where the plaintiffs “proposed no ‘further way’—short of full-blown, individual trials—‘to reduce this number and segregate the uninjured from the truly injured’” (citation omitted)); see also Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 462 (2016) (Roberts, C.J., concurring) (suggesting the class jury verdict should not stand if the district court cannot “fashion a method for awarding damages only to those class members who suffered an actual injury”). Thus, the en banc panel’s focus on discretion largely accorded with their sister circuits’ case law.

III. Implications of the Ruling

A. Predominance as Efficient Aggregation

In rejecting the de minimis standard, the Ninth Circuit adhered to Rule 23’s broad textual commands to advance the class mechanism’s driving goal of judicial economy.⁸³83. See Samuel Issacharoff, Rule 23 and the Triumph of Experience, 84 Duke L.J. 161, 168 (2021) (arguing a “simple ‘light touch’ textual reading shows that the words [of Rule 23] point to concerns about the overall administration of justice, measured in terms of the substantive results of aggregate litigation rather than the nature of the rights-holder”). Rule 23(b)(3) asks whether common questions predominate over individual ones, making no statement on specific characteristics⁸⁴84. Compare Fed. R. Civ. P. 23(b)(3) (defining predominance generally), with id. 23(b)(3)(A)–(D) (detailing the four factors courts must consult to conclude a class action is superior to other adjudicatory methods). that influence such an analysis.⁸⁵85. Standard canons of statutory construction counsel against limiting general language. See Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 101 (2012) (“Without some indication to the contrary, general words . . . are to be accorded their full and fair scope. They are not to be arbitrarily limited.”). Without specific guidance, the Supreme Court has been hesitant to adopt hard rules based on general language.⁸⁶86. See, e.g., Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 625 (1997) (noting that while the Advisory Committee for Rule 23’s 1966 revision cautioned mass accidents are “ordinarily not appropriate” for class litigation, “the text of the Rule does not categorically exclude mass tort cases from class certification, and District Courts, since the late 1970’s, have been certifying such cases in increasing number”); Goldman Sachs Grp., Inc. v. Ark. Tchr. Ret. Sys., 141 S. Ct. 1951, 1960 (2021) (addressing “whether the generic nature of a misrepresentation is relevant to price impact” findings at class certification and concluding “courts ‘should be open to all probative evidence on that question—qualitative as well as quantitative—aided by a good dose of common sense’” (citation omitted)). For instance, in Tyson Foods v. Bouaphakeo, the Supreme Court rejected a “broad” and “categorical” rule forbidding plaintiffs from using representative evidence to establish predominance, holding such a rule would make “little sense” because evidence’s permissibility turns on a specific case’s cause of action.⁸⁷87. Tyson Foods, 577 U.S. at 454–55. Rule 23’s purposefully general language⁸⁸88. See Samuel Issacharoff & Peter Zimroth, An Oral History of Rule 23: An Interview with Professor Arthur Miller, 74 N.Y.U. Ann. Surv. Am. L. 105, 117 (2018) (quoting Professor Miller, noting predominance and superiority were meant to ensure (b)(3) classes were a “true efficiency economy win,” but also confirming those “[w]ords . . . were like silly putty that could be molded in any way by a judge in a particular context”). affords trial courts ample latitude to certify, or decline to certify, class proposals based on whether aggregation may materially advance the litigation before them in a fair and efficient manner.⁸⁹89. See Amchem, 521 U.S. at 615 (stating that predominance and superiority were added for efficiency and fairness, among other considerations); Fed. R. Civ. P. 23 advisory committee’s note to 1966 amendment (“It is only where this predominance exists that economies can be achieved by means of the class-action device.”); see also Am. L. Inst., Principles of the Law of Aggregate Litigation § 2.02(a)(1) (2010) (authorizing aggregate treatment where such adjudication would “materially advance” litigation “in a manner . . . so as to generate significant judicial efficiencies”). Though the dissent attempted to interpret a de minimis rule as enforcing that policy, the dissenting judges provided no assurance such a rule best economizes judicial procedure.⁹⁰90. Olean, 31 F.4th at 692 (Lee, J., dissenting) (claiming that “allowing more than a de minimis number of uninjured class members tilts the playing field in favor of plaintiffs”). Instead, the dissent erred on the side of preventing “oversized classes,”⁹¹91. Id. but the Rules Committee added (b)(3) certification in the 1966 revision precisely to help vindicate the rights of people “who individually would be without effective strength to bring their opponents into court at all.”⁹²92. Amchem, 521 U.S. at 617 (quoting Benjamin Kaplan, A Prefatory Note, 10 B.C. Ind. & Com. L. Rev. 497, 497 (1969)) (discussing how class actions can be a tool for those with smaller damages claims to still obtain relief). The majority’s critical move is to train the predominance inquiry on how a judge will resolve the issue of uninjured class members at trial. After all, “Rule 23 certification is at bottom a trial management decision; it simply allows the class litigation to continue under the district court’s ongoing supervision.”⁹³93. Olean Wholesale Grocery Coop. v. Bumble Bee Foods LLC, 993 F.3d 774, 796 (9th Cir. 2021) (Hurwitz, J., concurring in part and dissenting in part), aff’d on reh’g en banc, 31 F.4th 651 (9th Cir. 2022). So long as a district court can fairly and efficiently “winnow out” a “non-injured subset of class members,”⁹⁴94. Olean, 31 F.4th at 669. common questions should predominate, even if the number of uninjured plaintiffs or percentage of the class appears to be more than de minimis.⁹⁵95. The Supreme Court recently concluded that 6,332 class members, in a class of 8,185 plaintiffs, did not suffer an injury in fact. TransUnion LLC v. Ramirez, 141 S. Ct. 2190, 2214 (2021). The Court remanded on the issue of typicality but made no mention of predominance. Id. Even though the Court concluded most of the class was uninjured, neither the Supreme Court nor the district court had a difficult time separating class members based on the injury-defining characteristic of whether TransUnion provided their tainted credit reports to third parties.

B. Discretion and Judicial Experience

Additionally, Olean solidifies the dominance of discretion at class certification, resulting from an accumulation of judicial experience with complex multi-district and class adjudications. As Professors Samuel Issacharoff and Arthur R. Miller explain, the past decade has seen judges certify classes that “would have given the Rules adopters grave pause.”⁹⁶96. See Issacharoff, supra note 82, at 163 (citing Issacharoff & Zimroth, supra note 87, at 125) (recounting his interview with Professor Arthur R. Miller and discussing how judicial experience influenced the settlement class’s development, culminating with In re National Football League Players Concussion Injury Litigation, 821 F.3d 410 (3d Cir. 2016)). Circuit courts, and even specific judges, that once viewed novel class proposals with skepticism abandoned their previous positions to embrace efficient aggregation. The rise of Rule 23(c)(4) issue classes provides an apt example. In the 1990s, a series of decisions erected barriers to certifying issue classes. Judge Richard Posner, writing for a Seventh Circuit panel in In re Rhone–Poulenc Rorer Inc.,⁹⁷97. 51 F.3d 1293 (7th Cir. 1995). rejected an attempt to certify a class only on a negligence element because the “desire to experiment with an innovative procedure” would possibly infringe upon the defendants’ Seventh Amendment rights to avoid re-examination of a jury’s decision.⁹⁸98. Id. at 1297, 1303. Similarly, the Fifth Circuit held a district court may not certify an issue class unless the “cause of action, as a whole, . . . satisf[ies] the predominance requirement.”⁹⁹99. Castano v. Am. Tobacco Co., 84 F.3d 734, 745 n.21 (5th Cir. 1996). But these barriers did not last long. Seven years after Rhone–Poulenc, Judge Posner upheld an issue class, stating issue class treatment “is appropriate and is permitted by Rule 23 when the judicial economy from consolidation of separate claims outweighs any concern with possible inaccuracies from their being lumped together in a single proceeding.”¹⁰⁰100. Mejdrech v. Met-Coil Sys. Corp., 319 F.3d 910, 911 (7th Cir. 2003). Posner made no mention of the Seventh Amendment but focused entirely on efficiency and accuracy. The Fifth Circuit also moved past the narrow view of issue class certification.¹⁰¹101. See Steering Comm. v. Exxon Mobil Corp., 461 F.3d 598, 603 (5th Cir. 2006) (noting trial bifurcation might eliminate “the obstacles preventing a finding of predominance”). The majority of circuit courts now take the broad view¹⁰²102. Russell v. Educ. Comm’n for Foreign Med. Graduates, 15 F.4th 259, 273–74, 273 n.6 (3d Cir. 2021) (writing that “the Second, Fourth, Sixth, Seventh, and Ninth Circuits” have adopted this view and that “[u]nder the broad view, courts apply the Rule 23(b)(3) predominance and superiority prongs after common issues have been identified for class treatment”), cert. denied, 142 S. Ct. 2706 (2022). and have eliminated the strict barriers that once completely precluded issue certification, reducing them into pieces of a multi-factor test that outline a district court’s wide discretion.¹⁰³103. Id. at 268 (listing nine factors that indicate when issue certification may be appropriate, including whether bifurcated proceedings risk re-examining a jury’s initial findings) cf. Issacharoff, supra note 82, at 176 (arguing the Third Circuit’s prior experience with class settlements allowed them to “give independent weight to the need for closure” in future cases). But, unlike the issue class’s story, the Ninth Circuit correctly resolved the dilemma of uninjured class members upon first impression. Rather than calcify class adjudication through a strict reading of Rule 23, the en banc panel left the decision to the district court’s sound discretion, acknowledging that the trial judge is in the best position to expend judicial resources efficiently and fairly.¹⁰⁴104. Olean Wholesale Grocery Coop. v. Bumble Bee Foods LLC, 31 F.4th 651, 669 (9th Cir. 2022) (en banc).

C. Predominance as Relevance

1. The Ninth Circuit’s Unnecessary Holding

Nevertheless, the Ninth Circuit should never have addressed this legal issue because the court’s conclusions were “clearly unnecessary to its resolution of the case, d[id] not affect its outcome in any manner, and constitute[d] an advisory opinion.”¹⁰⁵105. Spears v. Stewart, 283 F.3d 992, 998–99 (9th Cir. 2002) (Reinhardt, J., dissenting from denial of rehearing en banc) The Court has shared similar concerns. See Loc. 144 Nursing Home Pension Fund v. Demisay, 508 U.S. 581, 592 n.5 (1993) (describing dicta as language “uninvited, unargued, and unnecessary to the Court’s holdings”). Simply put, if a district court understood each class member to be similarly situated, it would not need to confront questions of how to handle a class containing both injured and uninjured class members.¹⁰⁶106. See Olean, 31 F.4th at 681 (noting that a jury’s findings as to the persuasiveness of Dr. Mangum’s report would not give rise to any individual issues regarding a class member’s injury status). Even here, the district court considered whether a de minimis standard would impact the case but did not develop the issue because it concluded Dr. Mangum’s report could establish class-wide impact.¹⁰⁷107. See In re Packaged Seafood Prods. Antitrust Litig., 332 F.R.D. 308, 323–24, 329 (S.D. Cal. 2019), vacated and remanded sub nom. Olean Wholesale Grocery Coop. v. Bumble Bee Foods LLC, 993 F.3d 774 (9th Cir. 2021), aff’d on reh’g en banc, 31 F.4th 651 (9th Cir. 2022). The en banc panel unequivocally affirmed that conclusion.¹⁰⁸108. Olean, 31 F.4th at 685. And, the Ninth Circuit’s refusal to resolve the related issue of whether each class member must prove Article III standing at certification renders the decision to address the de minimis question paradoxical.¹⁰⁹109. See id. at 682 (“We need not consider the Tuna Suppliers’ argument that the possible presence of a large number of uninjured class members raises an Article III issue, because . . . the district court concluded that the DPPs’ evidence was capable of establishing antitrust impact on a class-wide basis.”). It ultimately appears the court “reached out to address a novel, complex, and important issue in an advisory opinion.”¹¹⁰110. Spears, 283 F.3d at 1004.

2. Distinguishing Between Evidence’s Relevance and Sufficiency to Satisfy Predominance

Still, in addressing the predominance issues related to the parties’ experts, the Ninth Circuit navigated a confusing inquiry: whether the plaintiff must preliminarily prove antitrust impact or simply demonstrate that antitrust impact is capable of class-wide proof.¹¹¹111. See Nagareda et al., supra note 5, at 334 (discussing the thin line between the two different conceptions of the moving party’s burden at class certification). The Supreme Court in Amgen Inc. v. Connecticut Retirement Plans and Trust Funds clearly takes the latter side: Moving parties need not establish that they “will win the fray” but only that the class is cohesive enough to prompt predominating common questions of law or fact.¹¹²112. 568 U.S. 455, 460 (2013). Merits questions may overlap with the inquiry into whether the class is cohesive. In that situation, courts must engage in the trickier determination of which ancillary issues—for instance, factual questions about the nature of a product market¹¹³113. See, e.g., In re Hydrogen Peroxide Antitrust Litig., 552 F.3d 305, 316, 325 (3d Cir. 2008) (remanding and requiring the district court to resolve experts’ disputes as to hydrogen peroxide’s fungibility in the relevant market before certifying that plaintiffs can prove antitrust impact through common evidence). or a defendant’s particular business practices¹¹⁴114. See, e.g., Ellis v. Costco Wholesale Corp., 657 F.3d 970, 983–84 (9th Cir. 2011) (mandating the district court resolve factual disputes regarding whether local or upper management promoted individuals because plaintiffs would be unlikely to establish discrimination with common evidence if local managers promoted employees).—bear on the predominance inquiry.¹¹⁵115. Courts routinely engage in similar determinations when they address preliminary evidentiary questions of conditional relevance. See Fed. R. Evid. 104(a)–(b) (directing the court to “decide any preliminary question about whether . . . evidence is admissible”). Courts must first decide if the proffered evidence’s relevance “depends on whether a fact exists” and, if so, preliminarily rule on the existence of that fact. Id. 104(b). Similarly at certification, a trial court must determine if the putative class’s cohesion depends on whether a fact exists and, if so, resolve the factual question by a preponderance of the evidence. See Olean Wholesale Grocery Coop. v. Bumble Bee Foods LLC, 31 F.4th 651, 665 (9th Cir. 2022) (en banc) (holding “plaintiffs must prove the facts necessary” to satisfy Rule 23’s prerequisites “by a preponderance of the evidence”); Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 351–52 (2011) (directing lower courts to “resolve preliminary matters” at class certification); see also Jonah B. Gelbach, The Triangle of Law and the Role of Evidence in Class Action Litigation, 165 U. Pa. L. Rev. 1807, 1820 & n.62 (2017) (arguing courts at class certification should resolve disputes over “auxiliary assumption[s] necessary for counterfactual evidence to be probative” for each class member through Rule 104’s framework).

Olean presents such a case. To the dissent, every attack on Dr. Mangum’s report required the court’s attention because a jury might have believed Dr. Johnson’s argument that twenty-eight percent of the class was uninjured. But, the Ninth Circuit effectively distinguished between the Tuna Suppliers’ arguments that raised “fatal dissimilarit[ies]” and those that illustrated “fatal similarit[ies].”¹¹⁶116. Amgen, 568 U.S. at 470 (citing Nagareda, supra note 4, at 107). Fatal dissimilarities are those differences between the class members that “make use of the class-action device inefficient or unfair” because each class member may require individualized proof:¹¹⁷117. Id. Evidence would not be relevant to each class member.¹¹⁸118. Compare Wal-Mart, 564 U.S. at 356–57 (decertifying a class action because 1.5 million plaintiffs across thousands of stores managed by tens of thousands of managers were not similar enough for statistical regressions or sampling evidence to prove Wal-Mart discriminated against each plaintiff), with Amgen, 568 U.S. at 467 (holding proof of materiality is not needed at the certification stage because it is objective and applies to each member of the class), and Tyson Foods, Inc. v. Bouaphakeo, 577 U.S. 442, 459 (2016) (holding each plaintiff could rely upon a single study to recover under the Fair Labor Standards Act because “each employee worked in the same facility, did similar work, and was paid under the same policy”). The Tuna Suppliers only argued some plaintiffs were uninjured, and therefore dissimilar, because of their stronger bargaining power, but Dr. Mangum, the district court, and the Ninth Circuit adequately addressed and resolved the dispute on that ancillary issue.¹¹⁹119. See supra notes 60–64 and accompanying text (explaining how the Ninth Circuit concluded each plaintiff could establish antitrust injury with Dr. Mangum’s evidence). In contrast, the district court refused to enter any findings on many of Dr. Johnson’s general critiques—e.g., Dr. Mangum used inferior cost data—because those addressed a similarity: Evidence would not be sufficient for every class member.¹²⁰120. See In re Asacol Antitrust Litig., 907 F.3d 42, 54 (1st Cir. 2018) (“[P]laintiffs point to no such substantive law that would make an opinion that ninety percent of class members were injured both admissible and sufficient to prove that any given individual class member was injured.”). As the en banc aptly held, such a debate is best reserved for summary judgment and, ultimately, a jury.¹²¹121. Olean, 31 F.4th at 681.

Conclusion

The Ninth Circuit’s decision to reject a de minimis standard for uninjured class members is another step in acknowledging Rule 23’s core purpose of and the trial court’s superior position in efficiently managing complex cases and controversies. The de minimis standard is a proxy for Rule 23’s textual commands at best and a hindrance to district courts saddled with overwhelming dockets and weary plaintiffs at worst. Though the en banc panel should not have reached the de minimis question, its opinion still provided important guidance for district courts struggling to evaluate ancillary predominance issues. By properly distinguishing between disputes over dissimilarities—relevancy issues a court must address at certification—and similarities—sufficiency debates best reserved for summary judgment or a jury—the court ensured Rule 23 continues to serve as an important tool to redress democratic theft and preserve increasingly strained judicial resources.