31 F.4th 651 (9th Cir. 2022)

From Casetext: Smarter Legal Research

Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC

United States Court of Appeals, Ninth Circuit.

Apr 8, 2022

31 F.4th 651 (9th Cir. 2022)

Summary

holding that Rule 23 permits "certification of a class that potentially includes more than a de minimus number of uninjured class members"

Summary of this case from Davis v. Lab. Corp. of Am. Holdings

See 6 Summaries

Opinion

No. 19-56514

04-08-2022

OLEAN WHOLESALE GROCERY COOPERATIVE, INC., Beverly Youngblood, Pacific Groservice, Inc., DBA Pitco Foods, Capitol Hill Supermarket, Louise Ann Davis Matthews, James Walnum, Colin Moore, Jennifer A. Nelson, Elizabeth Davis-berg, Laura Childs; Nancy Stiller; Bonnie Vanderlaan; Kristin Millican; Trepco Imports and Distribution, Ltd.; Jinkyoung Moon; Corey Norris; Clarissa Simon; Amber Sartori; Nigel Warren ; Amy Joseph; Michael Juetten; Carla Lown; Truyen Ton-Vuong, AKA David Ton; A-1 Diner; Dwayne Kennedy; Rick Musgrave; Dutch Village Restaurant; Lisa Burr; Larry Demonaco; Michael Buff; Ellen Pinto; Robby Reed; Blair Hysni; Dennis Yelvington; Kathy Durand Gore; Thomas E. Willoughby III; Robert Fragoso; Samuel Seidenburg; Janelle Albarello; Michael Coffey; Jason Wilson; Jade Canterbury ; Nay Alidad; Galyna Andrusyshyn; Robert Benjamin; Barbara Buenning; Danielle Greenberg ; Sheryl Haley; Lisa Hall; Tya Hughes; Marissa Jacobus; Gabrielle Kurdt; Erica Pruess; Seth Salenger; Harold Stafford ; Carl Lesher; Sarah Metivier Schadt; Greg Stearns; Karren Fabian; Melissa Bowman; Vivek Dravid; Jody Cooper; Danielle Johnson; Herbert H. Kliegerman; Beth Milliner; Liza Milliner; Jeffrey Potvin; Stephanie Gipson; Barbara Lybarger; Scott A. Caldwell; Ramon Ruiz; Thyme Cafe & Market, Inc.; Harvesters Enterprises, LLC; Affiliated Foods, Inc.; Piggly Wiggly Alabama Distributing Co., Inc.; Elizabeth Twitchell; Tina Grant; John Trent ; Brian Levy; Louise Adams; Marc Blumstein; Jessica Breitbach; Sally Crnkovich; Paul Berger; Sterling King; Evelyn Olive ; Barbara Blumstein; Mary Hudson; Diana Mey; Associated Grocers of New England, Inc.; North Central Distributors, LLC; Cashwa Distributing Co. Of Kearney, Inc.; Urm Stores, Inc.; Western Family Foods, Inc.; Associated Food Stores, Inc.; Giant Eagle, Inc. ; Mclane Company, Inc.; Meadowbrook Meat Company, Inc.; Associated Grocers, Inc. ; Bilo Holding, LLC; WinnDixie Stores, Inc.; Janey Machin; Debra L. Damske; Ken Dunlap; Barbara E. Olson; John Peychal; Virginia Rakipi; Adam Buehrens; Casey Christensen; Scott Dennis; Brian Depperschmidt; Amy E. Waterman; Central Grocers, Inc.; Associated Grocers of Florida, Inc. ; Benjamin Foods LLC; Albertsons Companies LLC; H.E. Butt Grocery Company ; Hyvee, Inc. ; The Kroger Co. ; Lesgo Personal Chef LLC ; Kathy Vangemert; Edy Yee; Sunde Daniels; Christopher Todd; Publix Super Markets, Inc. ; Wakefern Food Corp.; Robert Skaff; Wegmans Food Markets, Inc. ; Julie Wiese; Meijer Distribution, Inc.; Daniel Zwirlein; Meijer, Inc. ; Supervalu Inc.; John Gross & Company; Super Store Industries; W Lee Flowers & Co Inc.; Family Dollar Services, LLC; Amy Jackson; Family Dollar Stores, Inc.; Katherine Mcmahon ; Dollar Tree Distribution, Inc.; Jonathan Rizzo; Greenbrier International, Inc.; Joelyna A. San Agustin; Alex Lee, Inc. ; Rebecca Lee Simoens; Big Y Foods, Inc. ; David Ton; Kvat Food Stores, Inc., DBA Food City; Affiliated Foods Midwest Cooperative, Inc.; Merchants Distributors, LLC; Brookshire Brothers, Inc.; Schnuck Markets, Inc. ; Brookshire Grocery Company ; Kmart Corporation ; Certco, Inc.; Rushin Gold, LLC, DBA The Gold Rush; Unified Grocers, Inc.; Target Corporation; Simon-Hindi, LLC ; Fareway Stores, Inc. ; Moran Foods, LLC, DBA Save-A-Lot; Woodman's Food Market, Inc. ; Dollar General Corporation; Sam's East, Inc.; Dolgencorp, LLC; Sam's West, Inc.; Krasdale Foods, Inc.; Walmart Stores East, LLC ; CVS Pharmacy, Inc.; Walmart Stores East, LP; Bashas' Inc. ; Wal-mart Stores Texas, LLC ; Marc Glassman, Inc. ; Wal-mart Stores, Inc.; 99 Cents Only Stores ; Jessica Bartling; Ahold U.S.A., Inc.; Gay Birnbaum; Delhaize America, LLC ; Sally Bredberg; Associated Wholesale Grocers, Inc.; Kim Craig; Maquoketa Care Center ; Gloria Emery; Erbert & Gerbert's, Inc. ; Ana Gabriela Felix Garcia; Janet Machen; John Frick; Painted Plate Catering ; Kathleen Garner ; Robert Etten; Andrew Gorman; Groucho's Deli of Five Points, LLC; Edgardo Gutierrez; Groucho's Deli of Raleigh ; Zenda Johnston; Sandee's Catering; Steven Kratky; Confetti's Ice Cream Shoppe; Kathy Lingnofski; End Payer Plaintiffs; Laura Montoya; Kirsten Peck; John Pels; Valerie Peters; Elizabeth Perron ; Audra Rickman; Erica C. Rodriguez, Plaintiffs-Appellees, and Jessica Decker, Joseph A. Langston, Sandra Powers, Grand Supercenter, Inc., The Cherokee Nation, US Foods, Inc., Sysco Corporation, Gladys, LLC, Spartannash Company, Bryan Anthony Reo, Plaintiffs, v. BUMBLE BEE FOODS LLC ; StarKist Co.; Dongwon Industries Co., Ltd., Defendants-Appellants, and King Oscar, Inc.; Thai Union Frozen Products PCL; Del Monte Foods Company; Tri Marine International, Inc.; Dongwon Enterprises; Del Monte Corp.; Christopher D. Lischewski; Lion Capital (Americas), Inc.; Big Catch Cayman LP, AKA Lion/Big Catch Cayman LP ; Francis T Enterprises ; Glowfisch Hospitality ; Thai Union North America, Inc., Defendants.

Gregory G. Garre (argued), Samir Deger-Sen, and Shannon Grammel, Latham & Watkins LLP, Washington, D.C.; Christopher S. Yates, Belinda S. Lee, and Ashley M. Bauer, Latham & Watkins LLP, San Francisco, California; for Defendants-Appellants StarKist Co. and Dongwon Industries Co. Ltd. Christopher L. Lebsock (argued), Michael P. Lehmann, Bonny E. Sweeney, and Samantha J. Stein, Hausfeld LLP, San Francisco, California, for Plaintiffs-Appellees Direct Purchaser Plaintiff Class. Jonathan W. Cuneo (argued), Joel Davidow, and Blaine Finley, Cuneo Gilbert & Laduca LLP, Washington, D.C., for Plaintiffs-Appellees Commercial Food Preparer Plaintiff Class. Thomas H. Burt (argued), Wolf Haldenstein Adler Freeman & Herz LLP, New York, New York; Betsy C. Manifold, Rachele R. Byrd, Marisa C. Livesay, and Brittany N. DeJong, Wolf Haldenstein Adler Freeman & Herz LLP, San Diego, California; for Plaintiffs-Appellees End Payer Plaintiff Class. Corbin K. Barthold and Cory L. Andrews, Washington, D.C., for Amicus Curiae Washington Legal Foundation. Ashley C. Parrish and Joshua N. Mitchell, King & Spalding LLP, Washington, D.C.; Steven P. Lehotsky, Jonan D. Urick, Daryl Joseffer, and Jennifer B. Dickey, United States Chamber Litigation Center; Anne M. Voigts, Quyen L. Ta, and Suzanne E. Nero, King & Spalding LLP, San Francisco, California; Kelly Perigoe, King & Spalding LLP, Los Angeles, California; Christopher A. Mohr, Software & Information Industry Association, Washington, D.C.; Jeanine Poltronieri, Internet Association, Washington, D.C.; for Amici Curiae Chamber of Commerce of the United States of America, Software Information Industry Association, and Internet Association. Randy M. Stutz, American Antitrust Institute, Washington, D.C.; Professor Joshua P. Davis, University of San Francisco School of Law, San Francisco, California; Ellen Meriwether, Cafferty Clobes Meriwether & Sprengal, Media, Pennsylvania; for Amicus Curiae American Antitrust Institute. Scott L. Nelson and Allison M. Zieve, Public Citizen Litigation Group, Washington, D.C., for Amicus Curiae Public Citizen Inc. Jocelyn D. Larkin, Lindsay Nako, and David S. Nahmias, Impact Fund, Berkeley, California, for Amici Curiae Impact Fund, Bet Tzedek, California Rural Legal Assistance Foundation, Centro Legal de la Raza, Legal Aid at Work, and Public Counsel. Karla Gilbride, Washington, D.C., as and for Amicus Curiae Public Justice P.C. Deborah A. Elman and Chad Holtzman, Garwin Gerstein & Fisher LLP, New York, New York; Warren T. Burns and Kyle K. Oxford, Burns Charest LLP, Dallas, Texas; Robert S. Kitchenoff, President; Lin Y. Chan, Vice President, Committee to Support the Antitrust Laws, Washington, D.C.; for Amicus Curiae Committee to Support the Antitrust Laws. Jonathan F. Cohn, Joshua J. Fougere, and Jacquelyn E. Fradette, Sidley Austin LLP, Washington, D.C., for Amicus Curiae Consumer Healthcare Products Association.

IKUTA, Circuit Judge

Christopher L. Lebsock (argued), Michael P. Lehmann, Bonny E. Sweeney, and Samantha J. Stein, Hausfeld LLP, San Francisco, California, for Plaintiffs-Appellees Direct Purchaser Plaintiff Class.

Jonathan W. Cuneo (argued), Joel Davidow, and Blaine Finley, Cuneo Gilbert & Laduca LLP, Washington, D.C., for Plaintiffs-Appellees Commercial Food Preparer Plaintiff Class.

Thomas H. Burt (argued), Wolf Haldenstein Adler Freeman & Herz LLP, New York, New York; Betsy C. Manifold, Rachele R. Byrd, Marisa C. Livesay, and Brittany N. DeJong, Wolf Haldenstein Adler Freeman & Herz LLP, San Diego, California; for Plaintiffs-Appellees End Payer Plaintiff Class.

Corbin K. Barthold and Cory L. Andrews, Washington, D.C., for Amicus Curiae Washington Legal Foundation.

Ashley C. Parrish and Joshua N. Mitchell, King & Spalding LLP, Washington, D.C.; Steven P. Lehotsky, Jonan D. Urick, Daryl Joseffer, and Jennifer B. Dickey, United States Chamber Litigation Center; Anne M. Voigts, Quyen L. Ta, and Suzanne E. Nero, King & Spalding LLP, San Francisco, California; Kelly Perigoe, King & Spalding LLP, Los Angeles, California; Christopher A. Mohr, Software & Information Industry Association, Washington, D.C.; Jeanine Poltronieri, Internet Association, Washington, D.C.; for Amici Curiae Chamber of Commerce of the United States of America, Software Information Industry Association, and Internet Association.

Randy M. Stutz, American Antitrust Institute, Washington, D.C.; Professor Joshua P. Davis, University of San Francisco School of Law, San Francisco, California; Ellen Meriwether, Cafferty Clobes Meriwether & Sprengal, Media, Pennsylvania; for Amicus Curiae American Antitrust Institute.

Scott L. Nelson and Allison M. Zieve, Public Citizen Litigation Group, Washington, D.C., for Amicus Curiae Public Citizen Inc.

Jocelyn D. Larkin, Lindsay Nako, and David S. Nahmias, Impact Fund, Berkeley, California, for Amici Curiae Impact Fund, Bet Tzedek, California Rural Legal Assistance Foundation, Centro Legal de la Raza, Legal Aid at Work, and Public Counsel.

Karla Gilbride, Washington, D.C., as and for Amicus Curiae Public Justice P.C.

Deborah A. Elman and Chad Holtzman, Garwin Gerstein & Fisher LLP, New York, New York; Warren T. Burns and Kyle K. Oxford, Burns Charest LLP, Dallas, Texas; Robert S. Kitchenoff, President; Lin Y. Chan, Vice President, Committee to Support the Antitrust Laws, Washington, D.C.; for Amicus Curiae Committee to Support the Antitrust Laws.

Jonathan F. Cohn, Joshua J. Fougere, and Jacquelyn E. Fradette, Sidley Austin LLP, Washington, D.C., for Amicus Curiae Consumer Healthcare Products Association.

Before: Andrew J. Kleinfeld, Sidney R. Thomas, Susan P. Graber, William A. Fletcher, Ronald M. Gould, Richard A. Paez, Consuelo M. Callahan, Sandra S. Ikuta, Paul J. Watford, Michelle T. Friedland and Kenneth K. Lee, Circuit Judges.

Dissent by Judge Lee

IKUTA, Circuit Judge: The primary suppliers of packaged tuna in the United States appeal the district court's order certifying three classes of tuna purchasers who allege the suppliers violated federal and state antitrust laws. The main issue on appeal is whether the purchasers' statistical regression model, along with other expert evidence, is capable of showing that a price-fixing conspiracy caused class-wide antitrust impact, thus satisfying one of the prerequisites for bringing a class action under Rule 23(b)(3) of the Federal Rules of Civil Procedure. Because the district court did not abuse its discretion in concluding that Rule 23(b)(3) was satisfied, we affirm.

Bumble Bee, StarKist, and Chicken of the Sea (COSI), and their parent companies are the largest suppliers of packaged tuna in the United States (referred to collectively as the "Tuna Suppliers"). Their products include packaged tuna sold to direct purchasers like Costco and Walmart, and food-service-size tuna products sold to various distributors for resale. Together, the Tuna Suppliers sell over 80 percent of the packaged tuna in the country.

As a result of Appellant Bumble Bee Foods LLC's bankruptcy proceeding, appellate proceedings against Bumble Bee Foods have been held in abeyance due to the automatic stay imposed by 11 U.S.C. § 362. Dkt. No. 51.

In late 2015, the United States Department of Justice (DOJ) opened an investigation into the packaged tuna industry for violations of federal antitrust law. The DOJ investigation uncovered evidence of a price-fixing scheme among the Tuna Suppliers, which led the DOJ to enter multiple indictments alleging a criminal conspiracy to fix prices of canned tuna for the period from approximately November 2011 through December 2013. Bumble Bee, StarKist, and three tuna industry executives pleaded guilty to the conspiracy. Bumble Bee's former CEO was convicted by a jury of a conspiracy to fix prices. COSI cooperated with the DOJ and admitted to price fixing in exchange for leniency.

Plea Agreement, United States v. Bumble Bee Foods LLC , No. 3:17-cr-00249-EMC (N.D. Cal. Aug. 2, 2017), ECF No. 32; Plea Agreement, United States v. Worsham , No. 3:16-cr-00535-EMC (N.D. Cal. Mar. 15, 2017), ECF No. 14; Plea Agreement, United States v. Cameron , No. 3:16-cr-00501-EMC (N.D. Cal. Jan. 25, 2017), ECF No. 18; Plea Agreement, United States v. Hodge , No. 3:17-cr-00297-EMC (N.D. Cal. June 28, 2017), ECF No. 13; Plea Agreement, United States v. StarKist Co. , No. 3:18-cr-00513-EMC (N.D. Cal. Nov. 14, 2018), ECF No. 24.

A number of purchasers of the Tuna Suppliers' products (referred to collectively as the "Tuna Purchasers") filed putative class actions against the Tuna Suppliers alleging violations of various federal and state antitrust laws. The Tuna Purchasers alleged that the Tuna Suppliers engaged in a conspiracy from November 2010 through at least December 31, 2016 to fix prices of tuna, along with other collusive activities in furtherance of the price-fixing conspiracy. The Tuna Purchasers alleged that they were damaged by the conspiracy because they paid supra-competitive prices for the Tuna Suppliers' products.

Supra-competitive prices are those prices elevated "above competitive levels" by a market participant who "exercise[s] [its] market power" to do so. ABA Section of Antitrust Law, Econometrics: Legal, Practical, and Technical Issues 252 (2d ed. 2014) ("Econometrics ").

The Tuna Purchasers' actions were consolidated in a multidistrict litigation pretrial proceeding in the Southern District of California. The Tuna Purchasers consist of three putative subclasses: (i) direct purchasers of the Tuna Suppliers' products, such as nationwide retailers and regional grocery stores, who purchased packaged tuna between June 1, 2011 and July 1, 2015 (the "DPPs"); (ii) indirect purchasers of the Tuna Suppliers' products who bought bulk-sized tuna products between June 2011 and December 2016 for prepared food or resale (the "CFPs"); and (iii) individual end purchasers who bought the Tuna Suppliers' products between June 1, 2011 and July 1, 2015 for personal consumption (the "EPPs").

In 2018, the Tuna Purchasers moved to certify the three subclasses under Rule 23 of the Federal Rules of Civil Procedure to proceed as a class action. See Fed. R. Civ. P. 23(a), (b)(3). To demonstrate class-wide antitrust impact, each subclass proffered evidence from a different economist, each of whom employed substantially similar methodologies, to show that each member of the subclasses had paid an overcharge caused by the Tuna Suppliers' conspiracy. The Tuna Suppliers contested this expert evidence through their own economists. The district court held a three-day evidentiary hearing on the certification motion, and heard substantial testimony from each expert witness. In July 2019, the district court certified all three subclasses.

The Tuna Suppliers timely appealed, and a panel of this court vacated the district court's order and remanded. See Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC , 993 F.3d 774, 794 (9th Cir. 2021), reh'g en banc granted , 5 F.4th 950 (9th Cir. 2021). We took the case en banc to consider whether the district court erred in finding that each subclass satisfied the requirement that "questions of law or fact common to class members predominate over any questions affecting only individual members." Fed. R. Civ. P. 23(b)(3).

We have jurisdiction under 28 U.S.C. § 1292(e) and Rule 23(f) of the Federal Rules of Civil Procedure. We review the decision to certify a class and "any particular underlying Rule 23 determination involving a discretionary determination" for an abuse of discretion. Yokoyama v. Midland Nat'l Life Ins. Co. , 594 F.3d 1087, 1091 (9th Cir. 2010). We review the district court's determination of underlying legal questions de novo, id. , and its determination of underlying factual questions for clear error, see Ruiz Torres v. Mercer Canyons Inc. , 835 F.3d 1125, 1132 (9th Cir. 2016). The Supreme Court has indicated that a court's determination regarding what a statistical regression model may prove or is capable of proving is not a question of fact, even though there may be disputed issues of fact raised by "the data contained within an econometric model." Comcast Corp. v. Behrend , 569 U.S. 27, 36 n.5, 133 S.Ct. 1426, 185 L.Ed.2d 515 (2013). Accordingly, we review the district court's determination that a statistical regression model, along with other expert evidence, is capable of showing class-wide impact, thus satisfying one of the prerequisites of Rule 23(b)(3) of the Federal Rules of Civil Procedure, for an abuse of discretion. See Yokoyama , 594 F.3d at 1091.

Rule 23 provides a procedural mechanism for "a federal court to adjudicate claims of multiple parties at once, instead of in separate suits." Shady Grove Orthopedic Assocs., P.A. v. Allstate Ins. Co. , 559 U.S. 393, 408, 130 S.Ct. 1431, 176 L.Ed.2d 311 (2010). As a claims-aggregating device, Rule 23 "leaves the parties' legal rights and duties intact and the rules of decision unchanged," id. , and it does not affect the substance of the claims or plaintiffs' burden of proof, see 28 U.S.C. § 2072(b).

To take advantage of Rule 23's procedure for aggregating claims, plaintiffs must make two showings. First, the plaintiffs must establish "there are questions of law or fact common to the class," as well as demonstrate numerosity, typicality and adequacy of representation. Fed. R. Civ. P. 23(a). A common question "must be of such a nature that it is capable of classwide resolution—which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke." Wal-Mart Stores, Inc. v. Dukes , 564 U.S. 338, 350, 131 S.Ct. 2541, 180 L.Ed.2d 374 (2011). By contrast, an individual question is one where members of a proposed class will need to present evidence that varies from member to member. See Tyson Foods, Inc. v. Bouaphakeo , 577 U.S. 442, 453, 136 S.Ct. 1036, 194 L.Ed.2d 124 (2016).

Rule 23(a) provides:

Prerequisites. One or more members of a class may sue or be sued as representative parties on behalf of all members only if:

(1) the class is so numerous that joinder of all members is impracticable;

(2) there are questions of law or fact common to the class;

(3) the claims or defenses of the representative parties are typical of the claims or defenses of the class; and

(4) the representative parties will fairly and adequately protect the interests of the class.

Second, the plaintiffs must show that the class fits into one of three categories. See Fed. R. Civ. P. 23(b). To qualify for the third category, Rule 23(b)(3), the district court must find that "the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy." Fed. R. Civ. P. 23(b)(3). "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." Tyson Foods , 577 U.S. at 453, 136 S.Ct. 1036 (cleaned up). The requirements of Rule 23(b)(3) overlap with the requirements of Rule 23(a) : the plaintiffs must prove that there are "questions of law or fact common to class members" that can be determined in one stroke, see Wal-Mart , 564 U.S. at 349, 131 S.Ct. 2541, in order to prove that such common questions predominate over individualized ones, see Tyson Foods , 577 U.S. at 453–54, 136 S.Ct. 1036. Therefore, courts must consider cases examining both subsections in performing a Rule 23(b)(3) analysis.

Rule 23(b)(3) provides in pertinent part:

A class action may be maintained if Rule 23(a) is satisfied and if ... (3) the court finds that the questions of law or fact common to class members predominate over any questions affecting only individual members, and that a class action is superior to other available methods for fairly and efficiently adjudicating the controversy.

Before it can certify a class, a district court must be "satisfied, after a rigorous analysis, that the prerequisites" of both Rule 23(a) and 23(b)(3) have been satisfied. Gen. Tel. Co. of Sw. v. Falcon , 457 U.S. 147, 161, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982) ; Comcast , 569 U.S. at 35, 133 S.Ct. 1426. "[P]laintiffs wishing to proceed through a class action must actually prove —not simply plead—that their proposed class satisfies each requirement of Rule 23, including (if applicable) the predominance requirement of Rule 23(b)(3)," and must carry their burden of proof "before class certification." Halliburton Co. v. Erica P. John Fund, Inc. , 573 U.S. 258, 275–76, 134 S.Ct. 2398, 189 L.Ed.2d 339 (2014).

We have not yet prescribed the plaintiffs' burden for proving that the prerequisites of Rule 23 are satisfied. In the absence of direction from Congress or the Constitution, it is up to the court to prescribe the burden of proof. See Herman & MacLean v. Huddleston , 459 U.S. 375, 389–90, 103 S.Ct. 683, 74 L.Ed.2d 548 (1983). To do so, we must consider both the allocation of "the risk of error between the litigants" and "the relative importance attached to the ultimate decision." Id. at 389, 103 S.Ct. 683 (quoting Addington v. Texas , 441 U.S. 418, 423, 99 S.Ct. 1804, 60 L.Ed.2d 323 (1979) ). The preponderance of the evidence standard allows both parties to "share the risk of error in roughly equal fashion," id. at 390, 103 S.Ct. 683 (quoting Addington , 441 U.S. at 423, 99 S.Ct. 1804 ), while "[a]ny other standard expresses a preference for one side's interests," id. Therefore, the preponderance of the evidence standard is "generally applicable in civil actions." Id. By contrast, the Court has "required proof by clear and convincing evidence where particularly important individual interests or rights are at stake," such as termination of parental rights or involuntary commitment proceedings. Id. at 389, 103 S.Ct. 683.

Applying this test here, the balance of interests in this case favors prescribing the preponderance of the evidence standard. The Supreme Court has made clear that Rule 23 is consistent with the Rules Enabling Act and does not "abridge, enlarge or modify any substantive right." Shady Grove , 559 U.S. at 406–07, 130 S.Ct. 1431 (citing 28 U.S.C. § 2072(b) ). Rule 23 does not "change plaintiffs' separate entitlements to relief nor abridge defendants' rights" and, instead, alters "only how the claims are processed." Id. at 408, 130 S.Ct. 1431. Therefore, the Supreme Court has concluded that the authorization of class actions is substantively neutral, even though it may expose defendants to the imposition of aggregate liability. Id. Because the application of Rule 23 to certify a class does not alter the defendants' rights or interests in a substantive way, there is no basis for applying a heightened standard of proof beyond the traditional preponderance standard. We therefore join our sister circuits in concluding that plaintiffs must prove the facts necessary to carry the burden of establishing that the prerequisites of Rule 23 are satisfied by a preponderance of the evidence.

See In re Lamictal Direct Purchaser Antitrust Litig. , 957 F.3d 184, 191 (3d Cir. 2020) ; In re Nexium Antitrust Litig. , 777 F.3d 9, 27 (1st Cir. 2015) ; Messner v. Northshore Univ. HealthSystem , 669 F.3d 802, 811 (7th Cir. 2012) ; Alaska Elec. Pension Fund v. Flowserve Corp. , 572 F.3d 221, 228 (5th Cir. 2009) ; Teamsters Loc. 445 Freight Div. Pension Fund v. Bombardier Inc. , 546 F.3d 196, 202 (2d Cir. 2008).

In carrying the burden of proving facts necessary for certifying a class under Rule 23(b)(3), plaintiffs may use any admissible evidence. See Tyson Foods , 577 U.S. at 454–55, 136 S.Ct. 1036 (explaining that admissibility of evidence at certification must meet all the usual requirements of admissibility and citing to Rules 401, 403, and 702 of the Federal Rules of Evidence ). Plaintiffs frequently offer expert evidence, including statistical evidence or class-wide averages, to prove that they meet the prerequisites of Rule 23(b)(3). See id. at 455, 136 S.Ct. 1036. Where, as here, a defendant did not raise a Daubert challenge to the expert evidence before the district court, the defendant forfeits the ability to argue on appeal that the evidence was inadmissible, but may still argue that the evidence is not capable of answering a common question on a class-wide basis. See Comcast , 569 U.S. at 32 n.4, 133 S.Ct. 1426 ; Tyson Foods , 577 U.S. at 458–59, 136 S.Ct. 1036.

In a class proceeding, defendants may challenge the reliability of an expert's evidence under Daubert v. Merrell Dow Pharmaceuticals, Inc. , 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), and Rule 702 of the Federal Rules of Evidence. See Tyson Foods , 577 U.S. at 459, 136 S.Ct. 1036 ; see also Comcast , 569 U.S. at 32 n.4, 133 S.Ct. 1426.

In order for the plaintiffs to carry their burden of proving that a common question predominates, they must show that the common question relates to a central issue in the plaintiffs' claim. See Wal-Mart , 564 U.S. at 349–50, 131 S.Ct. 2541. Therefore, "[c]onsidering whether ‘questions of law or fact common to class members predominate’ begins, of course, with the elements of the underlying cause of action." Erica P. John Fund, Inc. v. Halliburton Co. , 563 U.S. 804, 809, 131 S.Ct. 2179, 180 L.Ed.2d 24 (2011) (quoting Fed. R. Civ. P. 23(b)(3) ).

The claims at issue here are violations of section 1 of the Sherman Antitrust Act, 15 U.S.C. § 15, and California's Cartwright Act, Cal. Bus. & Prof. Code §§ 16700 et seq. The elements of a claim for such antitrust action are (i) the existence of an antitrust violation; (ii) "antitrust injury" or "impact" flowing from that violation (i.e., the conspiracy); and (iii) measurable damages. See Big Bear Lodging Ass'n v. Snow Summit, Inc. , 182 F.3d 1096, 1101–02 (9th Cir. 1999) ; In re Hydrogen Peroxide Antitrust Litig. , 552 F.3d 305, 311 (3d Cir. 2008), as amended (Jan. 16, 2009) (citing 15 U.S.C. § 15 ). "Antitrust injury" is "injury of the type the antitrust laws were intended to prevent and that flows from that which makes defendants' acts unlawful." Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc. , 429 U.S. 477, 489, 97 S.Ct. 690, 50 L.Ed.2d 701 (1977). Damages are measured only after each plaintiff has demonstrated that the defendant's conduct caused the plaintiff to suffer an antitrust injury. See In re Hydrogen Peroxide , 552 F.3d at 311.

The DPPs claim a violation of the Sherman Act, while the CFPs and the EPPs allege violations of California's antitrust law, the Cartwright Act, Cal. Bus. & Prof. Code, § 16700 et seq. The elements of a Cartwright Act claim are "(1) the formation and operation of the conspiracy, (2) the wrongful act or acts done pursuant thereto, and (3) the damage resulting from such act or acts." Marsh v. Anesthesia Servs. Med. Grp., Inc. , 200 Cal.App.4th 480, 132 Cal. Rptr. 3d 660, 670–71 (2011) (cleaned up). Because the analysis of a claim under the Cartwright Act "mirrors the analysis under federal [antitrust] law," we do not consider the Cartwright Act claims separately from the federal antitrust claims. County of Tuolumne v. Sonora Cmty. Hosp. , 236 F.3d 1148, 1160 (9th Cir. 2001).

Therefore, to prove there is a common question of law or fact that relates to a central issue in an antitrust class action, plaintiffs must establish that "essential elements of the cause of action," such as the existence of an antitrust violation or antitrust impact, are capable of being established through a common body of evidence, applicable to the whole class. Id. (cleaned up). Here, the Tuna Purchasers claim that they can establish the existence of antitrust impact through common proof.

In making the determinations necessary to find that the prerequisites of Rule 23(b)(3) are satisfied, the district court must proceed "just as the judge would resolve a dispute about any other threshold prerequisite for continuing a lawsuit." In re Initial Pub. Offerings Sec. Litig. , 471 F.3d 24, 42 (2d Cir. 2006), decision clarified on denial of reh'g , 483 F.3d 70 (2d Cir. 2007). This means that the court must make a "rigorous assessment of the available evidence and the method or methods by which plaintiffs propose to use the [class-wide] evidence to prove" the common question in one stroke. In re Hydrogen Peroxide , 552 F.3d at 312. In addition, the court must find that this common question (i.e., the "common, aggregation-enabling" issue) predominates over individual issues. Tyson Foods , 577 U.S. at 453, 136 S.Ct. 1036. The determination whether expert evidence is capable of resolving a class-wide question in one stroke may include "[w]eighing conflicting expert testimony" and "[r]esolving expert disputes," In re Hydrogen Peroxide , 552 F.3d at 323–24, where necessary to ensure that Rule 23(b)(3)'s requirements are met and the "common, aggregation-enabling" issue predominates over individual issues, Tyson Foods , 577 U.S. at 453, 136 S.Ct. 1036.

Not all expert evidence is capable of resolving a class-wide issue in one stroke. Cf. Dissent at 688–89. Courts have frequently found that expert evidence, while otherwise admissible under Daubert , was inadequate to satisfy the prerequisites of Rule 23. For instance, a class did not meet the prerequisites of Rule 23 where the expert evidence was inadequate to prove an element of the claim for the entire class, see Wal-Mart , 564 U.S. at 354, 356, 359, 131 S.Ct. 2541 (holding that class members failed to establish existence of common question with respect to Title VII claims because they "provide[d] no convincing proof of a companywide discriminatory pay and promotion policy"); where the damages evidence was not consistent with the plaintiffs' theory of liability, see Comcast , 569 U.S. at 35, 133 S.Ct. 1426 (holding that at the class certification stage, "any model supporting a plaintiff's damages case must be consistent with its liability case"); where the evidence contained unsupported assumptions, see In re New Motor Vehicles Canadian Exp. Antitrust Litig. , 522 F.3d 6, 29 (1st Cir. 2008) (criticizing the unsupported assumption that, absent the defendants' anti-competitive conduct, there would have been an influx of cars from Canada to United States sufficient to substantially decrease national prices); or where the evidence demonstrated nonsensical results such as false positives, i.e., injury to class members who could not logically have been injured by a defendant's conduct, see In re Rail Freight Fuel Surcharge Antitrust Litig.-MDL No. 1869 (Rail Freight I ), 725 F.3d 244, 252–55 (D.C. Cir. 2013) (vacating a certification order where the plaintiffs' expert evidence predicted that certain plaintiffs had been injured by a price-fixing conspiracy even though they operated under fixed-price contracts and were not exposed to overcharges caused by the conspiracy).

In determining whether the "common question" prerequisite is met, a district court is limited to resolving whether the evidence establishes that a common question is capable of class-wide resolution, not whether the evidence in fact establishes that plaintiffs would win at trial. While such an analysis may "entail some overlap with the merits of the plaintiff's underlying claim," Wal-Mart , 564 U.S. at 351, 131 S.Ct. 2541, the "[m]erits questions may be considered [only] to the extent [ ] that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied," Amgen Inc. v. Conn. Ret. Plans & Tr. Funds , 568 U.S. 455, 466, 133 S.Ct. 1184, 185 L.Ed.2d 308 (2013) ; see also Ellis v. Costco Wholesale Corp. , 657 F.3d 970, 983 n.8 (9th Cir. 2011). " Rule 23 grants courts no license to engage in free-ranging merits inquiries at the certification stage." Amgen , 568 U.S. at 466, 133 S.Ct. 1184.

A district court must also resolve disputes about historical facts if necessary to determine whether the plaintiffs' evidence is capable of resolving a common issue central to the plaintiffs' claims. For instance, in a case in which a nationwide class of plaintiff employees alleged nationwide discrimination by their employer, we held that a district court had to resolve factual disputes at certification regarding whether decisions regarding promotions were made at the local level or by upper management. See Ellis , 657 F.3d at 983–84 & n.7. We reasoned that if such decisions were made only at the local level, plaintiffs "would face an exceedingly difficult challenge in proving that there are questions of fact and law common to the nationwide class." Id. at 983–84. Nevertheless, the district court was not required to resolve factual disputes regarding ultimate issues on the merits, such as "whether women were in fact discriminated against" or whether the defendant "does in fact have a culture of gender stereotyping and paternalism." Id. at 983 ; see also id. at 983 n.8. Resolving such issues would "put the cart before the horse" by requiring plaintiffs to show at certification that they will prevail on the merits. Amgen , 568 U.S. at 460, 133 S.Ct. 1184.

The district court's findings at the certification stage "do not bind the fact-finder on the merits." In re Hydrogen Peroxide , 552 F.3d at 318.

Therefore, a district court cannot decline certification merely because it considers plaintiffs' evidence relating to the common question to be unpersuasive and unlikely to succeed in carrying the plaintiffs' burden of proof on that issue. See id. at 459–60, 133 S.Ct. 1184. Rather, Tyson Foods established the rule that if "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, Tyson Foods , 577 U.S. at 455, 136 S.Ct. 1036, then a district court may conclude that the plaintiffs have carried their burden of satisfying the Rule 23(b)(3) requirements as to that common question of law or fact. In Tyson Foods , for instance, the Court held that if the class members had pursued individual lawsuits, each could have relied on the expert evidence purporting to show how long it took to don and doff protective equipment. Tyson Foods , 577 U.S. at 456–57, 136 S.Ct. 1036. Accordingly, the Court concluded that such expert evidence was capable of answering a common question for the entire class in one stroke, and could reasonably sustain a jury verdict in favor of the plaintiffs, even though a jury could still decide that the evidence was not persuasive. Id. at 459–60, 136 S.Ct. 1036 ; see also id. at 457, 136 S.Ct. 1036 (explaining that the question whether the expert's "study was unrepresentative or inaccurate" was "itself common to the claims made by all class members"). The rule that the evidence need merely be capable of resolving a common question on a class-wide basis holds true whether the common question concerns an element of plaintiffs' claim, see Amgen , 568 U.S. at 468–69, 133 S.Ct. 1184 (materiality in a Rule 10b-5 action), or a fact that must be determined to establish liability, see Tyson Foods , 577 U.S. at 450, 136 S.Ct. 1036 (time spent donning and doffing protective equipment per week).

Senne v. Kansas City Royals Baseball Corp. referenced Tyson Foods 's rule that a district court may deny the use of admissible expert evidence to meet the requirements of Rule 23(b)(3) only if " ‘no reasonable juror’ could find it probative of whether an element of liability was met," and then stated in passing that "Tyson expressly cautioned that this rule should be read narrowly and not assumed to apply outside of the wage and hour context." 934 F.3d 918, 947 & n.27 (9th Cir. 2019) (citing Tyson Foods , 577 U.S. at 459–60, 136 S.Ct. 1036 ). But Tyson Foods contains no such limitation; rather, it declined to adopt "broad and categorical rules governing the use of representative and statistical evidence in class actions," and indicated that district courts should evaluate the sufficiency of plaintiffs' evidence on a case-by-case basis, depending on the purpose for which the expert evidence is being introduced and the underlying cause of action. 577 U.S. at 459–60, 136 S.Ct. 1036. Accordingly, we disapprove this dictum in Senne , 934 F.3d at 947 n.27.

Nor can a district court decline to certify a class that will require determination of some individualized questions at trial, so long as such questions do not predominate over the common questions. See Fed. R. Civ. P. 23(b)(3). "When one or more of the central issues in the action are common to the class and can be said to predominate, the action may be considered proper under Rule 23(b)(3) even though other important matters will have to be tried separately, such as damages or some affirmative defenses peculiar to some individual class members." Tyson Foods , 577 U.S. at 453, 136 S.Ct. 1036 (internal quotation marks omitted). Thus, Halliburton concluded that so long as plaintiffs could show that their evidence is capable of proving the prerequisites for invoking the presumption of reliance (a key element in a securities class action) on a class-wide basis, the fact that the defendants would have the opportunity at trial to rebut that presumption as to some of the plaintiffs did not raise individualized questions sufficient to defeat predominance. 573 U.S. at 276, 134 S.Ct. 2398. "That the defendant might attempt to pick off the occasional class member here or there through individualized rebuttal does not cause individual questions to predominate." Id.

When 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. See Cordoba v. DIRECTV, LLC , 942 F.3d 1259, 1277 (11th Cir. 2019). In an analogous context, we have held that a district court is not precluded from certifying a class even if plaintiffs may have to prove individualized damages at trial, a conclusion implicitly based on the determination that such individualized issues do not predominate over common ones. Vaquero v. Ashley Furniture Indus., Inc. , 824 F.3d 1150, 1155 (9th Cir. 2016) ; see also Pulaski & Middleman, LLC v. Google, Inc. , 802 F.3d 979, 988 (9th Cir. 2015) ; In re Urethane , 768 F.3d 1245, 1255 (10th Cir. 2014) ("The presence of individualized damages issues" does not preclude a court from certifying a class because "[c]lass-wide proof is not required for all issues").

Because the Supreme Court has clarified that "[e]very class member must have Article III standing in order to recover individual damages," TransUnion LLC v. Ramirez , ––– U.S. ––––, 141 S. Ct. 2190, 2208, 210 L.Ed.2d 568 (2021), Rule 23 also requires a district court to determine whether individualized inquiries into this standing issue would predominate over common questions, see Cordoba , 942 F.3d at 1277.

Therefore, we 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. Dissent at 691–92. This position is inconsistent with Rule 23(b)(3), 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. See Fed. R. Civ. P. 23(b)(3).

The dissent focuses on policy reasons why district courts should refrain from certifying classes that may include more than a de minimis number of uninjured class members. Dissent at 685–86, 690–91, 691–92. But we are bound to apply Rule 23(b)(3) as written, regardless of policy preferences. And contrary to the dissent's assertion, our conclusion that courts must apply Rule 23(b)(3) on a case-by-case basis, rather than rely on a per se rule that a class cannot be certified if it includes more than a de minimis number of uninjured class members, is consistent with the approach taken by our sister circuits. Dissent at 692. Neither of the two cases cited by the dissent, In re Rail Freight Fuel Surcharge Antitrust Litig.-MDL No. 1869 (Rail Freight II ), 934 F.3d 619 (D.C. Cir. 2019) and In re Asacol Antitrust Litig. , 907 F.3d 42 (1st Cir. 2018), adopted a per se rule. Rather, based on the particular facts of the cases before them, our sister circuits held that Rule 23(b)(3)'s predominance requirement is not satisfied when the need to identify uninjured class members "will predominate and render an adjudication unmanageable." In re Asacol Antitrust Litig. , 907 F.3d at 53–54 ; see also Rail Freight II , 934 F.3d at 625 (holding that a 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 determine the common question of whether class members were injured).

A district court is in the best position to determine whether individualized questions, including those regarding class members' injury, "will overwhelm common ones and render class certification inappropriate under Rule 23(b)(3)." Halliburton , 573 U.S. at 276, 134 S.Ct. 2398 ; see also Ruiz Torres , 835 F.3d at 1137 (stating that "the district court is well situated to winnow out" a fortuitously non-injured subset of class members). We "uphold a district court's determination that falls within a broad range of permissible conclusions." 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) ). III

Nevertheless, a court must consider whether the possible presence of uninjured class members means that the class definition is fatally overbroad. When "a class is defined so broadly as to include a great number of members who for some reason could not have been harmed by the defendant's allegedly unlawful conduct, the class is defined too broadly to permit certification." Messner , 669 F.3d at 824 ; see also Mazza v. Am. Honda Motor Co. , 666 F.3d 581, 596 (9th Cir. 2012) (holding that the class definition in a false advertising action was fatally overbroad where many members learned that the advertising was misleading before purchase or had never been exposed to the allegedly misleading advertisements); In re Asacol , 907 F.3d at 55–58 (holding that the class did not meet Rule 23(b)(3) requirements because the plaintiffs' evidence showed that thousands of plaintiffs who were loyal to brand-name drugs would not have purchased the generic drugs that were the subject of the price-fixing conspiracy). 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. See, e.g., Mazza , 666 F.3d at 596 (holding that false advertising "class must be defined in such a way as to include only members who were exposed to advertising that is alleged to be materially misleading"). A court may not, however, create a "fail safe" class that is defined to include only those individuals who were injured by the allegedly unlawful conduct. See Ruiz Torres , 835 F.3d at 1138 n.7 (internal quotation marks omitted). "Such a class definition is improper because a class member either wins or, by virtue of losing, is defined out of the class and is therefore not bound by the judgment." Messner , 669 F.3d at 825. But, ultimately, the problem of a potentially "over-inclusive" class "can and often should be solved by refining the class definition rather than by flatly denying class certification on that basis." Id.

We now turn to the Tuna Suppliers' arguments and consider them in light of this legal framework. We begin with the DPP class, which is the focus of the Tuna Suppliers' arguments. In order to prevail on their antitrust claim, the DPP class must prove that the Tuna Suppliers engaged in a conspiracy (an antitrust violation), which resulted in antitrust impact in the form of higher prices paid by each member of the class, which in turn led to measurable damages. The question whether each member of the DPP class suffered antitrust impact "is central to the validity of each one of the [DPP] claims." Wal-Mart , 564 U.S. at 350, 131 S.Ct. 2541. The central questions on appeal are whether the expert evidence presented by the DPPs is capable of resolving this issue "in one stroke," id. , and whether this common question predominates over any individualized inquiry. We conclude that the district court did not abuse its discretion in certifying the class.

The centerpiece of the DPPs' claim that each member of the class suffered antitrust impact is economist Dr. Russell Mangum's expert testimony and report. According to his testimony and report, Dr. Mangum reviewed a comprehensive range of available information to develop an understanding of the nature of the market at issue and the details of the Tuna Suppliers' price-fixing conspiracy. That information included court filings, the Tuna Suppliers' guilty pleas, discovery materials such as the Tuna Suppliers' business records concerning their sales of packaged tuna, deposition testimony, publicly available information regarding the tuna industry, and data regarding supply and demand factors that affect the manufacture, sale and consumption of packaged tuna such as raw material prices and details about customer preferences.

After examining the economic structure of the tuna market and the available record evidence concerning the Tuna Suppliers' behavior, Dr. Mangum determined that the packaged tuna market was conducive to price-fixing, given the Tuna Suppliers' dominance in the market, the attendant barriers to entry for competitors, the Tuna Suppliers' use of price lists for their products, and other characteristics of the packaged tuna industry. According to Dr. Mangum, these findings supported a baseline economic theory that the Tuna Suppliers' collusive behavior would affect the DPPs on a class-wide basis. Dr. Mangum then used a number of different econometric tools to evaluate whether quantitative evidence supported this theory.

Econometrics is "the application of statistical methods to economic data ... to draw inferences about economic relationships from observed data on market outcomes [i.e., price], even when those outcome are the result of complex interactions among numerous economic forces." Econometrics at 1.

Dr. Mangum first performed a pricing correlation test, which demonstrated that the prices of the Tuna Suppliers' products moved up or down together regardless of product or customer type, and thus supported the proposition that the Tuna Suppliers' collusion had a common, supra-competitive impact on their prices. Based on this evidence, Dr. Mangum concluded that the Tuna Suppliers' collusion would result in higher prices that would affect direct purchasers on a class-wide basis, which was consistent with his original theory. This finding is also consistent with "the prevailing [economic] view [that] price-fixing affects all market participants, creating an inference of class-wide impact even when prices are individually negotiated." In re Urethane , 768 F.3d at 1254.

To further explore whether the DPPs were subject to an overcharge caused by the price-fixing conspiracy (rather than by other variables that could affect prices) on a class-wide basis, Dr. Mangum constructed a statistical model using a multiple regression analysis. Regression analyses are used to determine "the relationship between an unknown [dependent] variable [such as price] and one or more independent variables [e.g., transaction characteristics, and supply and demand factors] that are thought to impact the dependent variable." Id. at 1260 (quotation marks omitted) (citing Michael J. Saks, et al., Reference Manual on Scientific Evidence 179, 181 (2d ed. 2000)). If a regression model uses "appropriate independent [or explanatory] variables," it can test and isolate the extent to which the actual prices paid by plaintiffs are higher because of a defendant's collusive behavior. Id. Assuming Dr. Mangum's regression model met this standard, it could provide further evidence that the DPPs were impacted by the Tuna Suppliers' collusion on a classwide basis.

In simple terms, Dr. Mangum first aggregated (or "pooled") the actual tuna sale transaction data for the Tuna Suppliers' sales to the DPPs during both the alleged conspiracy period and during benchmark periods before and after the conspiracy. Dr. Mangum then identified a number of variables (referred to as independent or explanatory variables) that could affect the price of tuna, including product characteristics, input costs, customer type, and variables related to consumer preference and demand, such as disposable income, seasonal effects, and geography. The model then isolated (or "controlled for") the effect of these explanatory variables on the prices paid by DPPs, which allowed the model to isolate the effect that the conspiracy by itself had on the prices paid by DPPs. When all the tuna sale transactions were aggregated, and the explanatory variables (other than the price-fixing conspiracy) were controlled for, the model showed that the DPPs paid 10.28 percent more for tuna during the conspiracy period than they did during the benchmark periods. Dr. Mangum labeled this 10.28 percent as the "overcharge," meaning the common amount paid by the DPPs resulting from the collusive behavior alone. This result was statistically significant, meaning that there was a less than five percent chance that the higher prices during the price-fixing conspiracy was a product of chance. Thus, by isolating the common overcharge amount, Dr. Mangum's regression model was further confirmation of his theory that the Tuna Suppliers' collusion had a class-wide effect.

The dissent argues that Dr. Mangum's opinion is not persuasive because large retailers have bargaining power and can extract price discounts, promotional credits, and rebates. Dissent at 689–90. As the dissent concedes, Dissent at 690 & n. 5, Dr. Mangum took these issues into account (to the extent that the Tuna Suppliers provided relevant data). After doing so, Dr. Mangum ran the regression model using both gross and net prices and determined that his regression model continued to produce a statistically significant overcharge. Dr. Mangum therefore reasoned that discounts and promotions did not affect his pooled model or his conclusion of class-wide impact. Although the dissent argues that (in the dissent's view) Dr. Mangum did not consider price discounts, promotional credits, and rebates "adequately," Dissent at 690 & n. 5, the persuasiveness of Dr. Mangum's analysis is not at issue at this phase of the proceeding.

Dr. Mangum performed several tests (which he referred to as "robustness checks") to confirm that his regression model was an appropriate tool to be used by the entire DPP class to show common impact. These tests were used to confirm the reliability of the model, and, according to Dr. Mangum, the test results supported his ultimate conclusion that the model could be used to show class-wide injury. First, Dr. Mangum changed the model to evaluate the overcharge specific to each individual defendant. The results showed that prices were still elevated above competitive levels during the collusion period. Second, Dr. Mangum changed the model to evaluate the overcharge specific to certain products with different characteristics, such as fish type and package type. These tests showed that each type of product tested was impacted to a similar degree. Third, Dr. Mangum changed the model to evaluate the overcharge based on customer types. This test showed that there were large, statistically significant overcharges for every customer type. These robustness checks confirmed Dr. Mangum's theory that the DPPs paid an overcharge during the conspiracy period. Finally, Dr. Mangum used the output of the pooled regression model to predict the but-for prices (i.e., what the price of tuna during the conspiracy period would have been without the overcharge caused by the conspiracy), and compared these predicted but-for prices to the actual prices paid by the DPP class. This comparison showed that 94.5 percent of the purchasers had at least one purchase above the predicted but-for price, which again provided further evidence that the conspiracy had a common impact on all or nearly all the members of the DPP class. Dr. Mangum therefore concluded that his aggregated regression model provided econometric evidence that the conspiracy resulted in higher prices paid by all or nearly all DPPs. According to Dr. Mangum, the results were strong evidence of common, class-wide antitrust impact.

Direct purchasers were grouped into categories called customer types, which included Retail, Club, Special Market, Food Service, Mass Merchandise, Discount, and e-Commerce.

According to Dr. Mangum, the purpose of this robustness test was to demonstrate that his regression model was sound. Contrary to the dissent's assertion, Dissent at 686, Dr. Mangum did not "suggest" that 5.5 percent of the class were uninjured. Rather, Dr. Mangum concluded that each class member was injured by supra-competitive prices, and used a different methodology for calculating damages for each member of the class. See infra at n.19. The Tuna Suppliers do not develop the argument that the results of this robustness test preclude certification of the class as currently defined. Therefore, we do not address this issue here. See United States v. Sineneng-Smith , ––– U.S. ––––, 140 S. Ct. 1575, 1578, 206 L.Ed.2d 866 (2020) ; see also Tyson Foods , 577 U.S. at 460, 136 S.Ct. 1036 (declining to reach a similar issue).

Although the regression model primarily served as evidence of class-wide antitrust impact, Dr. Mangum used the overcharge derived from the regression model to estimate class-wide damages. This estimate was developed by multiplying the overcharge estimate of 10.28 percent by the appropriate sales volume for the defendants, adjusted by several pertinent factors. Dr. Mangum used the same method to estimate damages for each of the class representatives identified in the complaint. The Tuna Suppliers do not challenge Dr. Mangum's damages methodology. Thus, the dissent's contention that the court has created a "sweeping rule that gives a free pass to the intractable problem of highly individualized damages analyses" misses the mark. Dissent at 690.

In sum, Dr. Mangum's findings about the tuna market and the Tuna Suppliers' collusive behavior, his pricing correlation test, his regression model, and his robustness checks all confirmed his theory that the conspiracy resulted in substantial price impacts, and that the impact was common to the DPPs during the collusion period.

The Tuna Suppliers attacked Dr. Mangum's expert report on multiple fronts, but primarily relied on their rebuttal expert, economist Dr. John Johnson, who made multiple criticisms of Dr. Mangum's methodology. The essence of Dr. Johnson's critique was that it was not statistically appropriate to use a pooled regression model for transactions in the tuna market, given the multiple individualized differences among class members, such as disparities in negotiating tactics and bargaining power. Dr. Mangum's use of pooled data, Dr. Johnson alleged, masked these individual differences among class members. Thus, Dr. Johnson claimed, Dr. Mangum's conclusion that the conspiracy had a class-wide impact based on a uniform overcharge did not reflect the real world.

Dr. Johnson supported this allegation on several grounds. First, Dr. Johnson claimed that a statistical tool called a Chow test shows that the data relating to tuna transactions should not be pooled due to individual differences in each purchaser's transactions. Second, Dr. Johnson criticized Dr. Mangum's calculation that 94.5 percent of DPPs whose transactional data were included in the model had at least one purchase at a price above the predicted but-for price. According to Dr. Johnson, this calculation was misleading because it was premised on what Dr. Johnson characterized as the faulty assumption that all direct purchasers paid the same 10.28 percent overcharge throughout the proposed class period. Instead, Dr. Johnson performed his own test of Dr. Mangum's model. As part of this test, Dr. Johnson changed the model to evaluate overcharge based on each individual customer. According to Dr. Johnson, the test showed that of the 604 direct purchasers who bought from the Tuna Suppliers during the proposed class period, the model did not estimate a positive and statistically significant overcharge (attributable to the conspiracy) for 169 direct purchasers (or 28 percent). Therefore, Dr. Johnson argued that the plaintiffs could not rely on the model to demonstrate class-wide impact of the conspiracy. Dr. Johnson made several additional critiques in arguing that Dr. Mangum's model was not capable of demonstrating class-wide impact. First, Dr. Johnson argued that Dr. Mangum's model showed false positives. According to Dr. Johnson, an application of Dr. Mangum's regression model showed that several DPP class members had paid an overcharge when they purchased tuna products from non-defendants, i.e., tuna suppliers who had not participated in the conspiracy. Second, Dr. Johnson attacked the reliability of Dr. Mangum's model because Dr. Mangum's model selected time periods that did not precisely match the class periods in the DPPs' complaint. Finally, Dr. Johnson criticized Dr. Mangum's use of a cost index (a calculated measure of costs for all the Tuna Suppliers) as one of the explanatory variables in his model, rather than using actual accounting cost data. According to Dr. Johnson, the use of a cost index inappropriately assumed that the Tuna Suppliers' costs responded in a like way to supply and demand factors.

A Chow test is a statistical test designed to "determine whether it is appropriate to pool potential subgroups when estimating the average effect of the alleged conspiracy." Econometrics at 358.

Dr. Johnson's test attempted to show that Dr. Mangum's model was flawed because 169 direct purchasers could not rely on the model to show antitrust impact due to the fact (as Dr. Mangum subsequently explained) that some purchasers had no or too few transactions during the pre-collusion benchmark period to generate statistically significant results. Contrary to the dissent's claim, Dissent at 686, Dr. Johnson did not show that 28 percent of the class potentially suffered no injury. See infra Section IV.B.

In rebuttal, Dr. Mangum rejected Dr. Johnson's premise that a pooled, aggregated model was inappropriate to use in this case. Dr. Mangum explained that his technique was a well-known and well-accepted method for examining antitrust impact in markets with individualized differences among purchasers. According to Dr. Mangum, both of the bases for Dr. Johnson's challenges to the use of a pooled regression model failed. First, Dr. Mangum claimed that a Chow test should not be used in the manner employed by Dr. Johnson in his report. According to Dr. Mangum, Dr. Johnson's Chow test was "designed to fail," meaning that in this context, the test results would always show that the data relating to tuna transactions should not be pooled. Second, Dr. Mangum asserted that the record contained insufficient transaction data for Dr. Johnson's test of the regression model to yield meaningful results. For example, Dr. Mangum acknowledged that the model, as changed by Dr. Johnson to consider purchasers on an individual basis, could not estimate a positive and statistically significant overcharge for 169 direct purchasers. But according to Dr. Mangum, no regression model could yield a statistically significant estimate for many of those 169 direct purchasers on such an individual purchaser-by-purchaser basis, because 61 of those purchasers did not make any purchases during the benchmark periods, and many of the other purchasers had not undertaken a sufficient number of transactions during either the benchmark periods or collusion period to yield statistically significant results. And logically, Dr. Mangum asserted, given the evidence that the defendants were able to inflate prices generally through the conspiracy, that the tuna market was susceptible to collusion, and that the model showed a robust, statistically significant impact of the price-fixing scheme on the tuna market, even the DPP class members for whom Dr. Johnson's test did not yield a positive, statistically significant overcharge should be able to rely on the pooled regression model as evidence of impact. Therefore, according to Dr. Mangum, Dr. Johnson erred in concluding that the regression model had no relevance for that 28 percent of class members.

Dr. Mangum also rebutted Dr. Johnson's additional critiques. With respect to Dr. Johnson's claim that the regression model yielded false positives, Dr. Mangum explained that overcharges imposed by non-defendant tuna suppliers (who were not part of the conspiracy) were not false positives but were caused by the "umbrella effect." This term refers to an economic observation that when many suppliers engage in a conspiracy to raise prices, non-conspirators may raise their prices to supra-competitive levels because of the conspirator's dominant market power. See ABA Section of Antitrust Law, Proving Antitrust Damages: Legal & Economic Issues 226 (2d ed. 2010). Dr. Mangum also argued that Dr. Johnson's claim of false positives was based on an erroneous analysis of the tuna market. According to Dr. Mangum, Dr. Johnson incorrectly claimed that two of the individual DPPs (Sysco and U.S. Foods) purchased tuna from non-defendant suppliers because both of those class members actually purchased tuna that was produced by the Tuna Suppliers and merely sold through a middleman. Dr. Mangum defended his selection of time periods relating to the model, claiming he narrowed the class period based on his analysis of the evidence in the case. Finally, Dr. Mangum rejected Dr. Johnson's critique of his use of cost indexes. Dr. Mangum asserted that costs indexes were statistically superior to using individual cost accounting data. He noted that one of the robustness tests he performed on the data showed that using defendant-specific cost structures confirmed the results of the pooled model. And he asserted that it was preferable to use his cost index for determining competitive market prices based on market supply and demand conditions, rather than relying on cost data derived from the Tuna Suppliers' individual approaches to cost accounting.

In considering whether the DPPs' evidence was capable of establishing antitrust impact for the class as a whole, the district court reviewed Dr. Mangum's expert testimony and report, the rebuttal testimony and report by Dr. Johnson, and Dr. Mangum's reply, and then addressed the parties' disputes. In doing so, the district court did not make any legal or factual error.

First, the district court considered Dr. Johnson's argument that Dr. Mangum's pooled regression model masked differences between purchasers, and that when the overcharge is determined for individual DPP class members the model did not show a positive, statistically significant impact for some 28 percent of the class. After reviewing each of the experts' analyses, the district court credited Dr. Mangum's rebuttal of Dr. Johnson's critique. Even if the model (when modified by Dr. Johnson to evaluate individual purchasers) did not yield a positive, statistically significant overcharge for some purchasers who had no or too few transactions during the pre-collusion benchmark period, the district court concluded that those purchasers could still rely on the pooled regression model as evidence of the conspiracy's impact on similarly situated class members. The court further noted that other evidence in the record, including the guilty pleas and market characteristics, showed that class members suffered a common impact.

The district court also considered Dr. Johnson's argument that the Chow test showed that Dr. Mangum's model cannot be applied to all defendants. The court acknowledged that failure of a statistical test used to determine whether a regression is appropriate should be taken seriously, and could lead a court to reject the model at the class certification stage as not capable of providing class-wide proof. But it also noted that most regressions models will fail one or more tests if enough are run, even if the model itself is statistically sound. Because there was a rational basis for Dr. Mangum's use of the pooled regression model to demonstrate class-wide impact, the court concluded the failure of the Chow test did not require the court to reject the model.

The district court rejected Dr. Johnson's additional arguments. With respect to Dr. Johnson's claim that the false positives in Dr. Mangum's model rendered the model unreliable, the court credited Dr. Mangum's explanation that the false positives could be explained by the umbrella effect and that Dr. Johnson had erroneously concluded that some tuna was supplied by non-defendants when in fact the tuna was supplied by defendants. The district court also addressed the dispute over Dr. Mangum's selection of the time period for the class, and concluded that Dr. Mangum's narrowing of the time frame bolstered the reliability of the model. Finally, the district court rejected Dr. Johnson's critique of Dr. Mangum's use of a cost index, rather than actual accounting cost data. The court credited Dr. Mangum's explanation as to why the use of such an index provided more reliable results than actual cost accounting data, and concluded that his use of a cost index did not undermine the reliability of his methodology or model.

After resolving each dispute between the experts, the district court acknowledged that the defendants' critique of Dr. Mangum's model could be persuasive to a jury at trial. But the district court recognized that at this stage of the proceedings, its task was to determine whether Dr. Mangum's evidence was capable of showing class-wide impact, not to reach a conclusion on the merits of the DPPs' claims. After weighing the evidence put forth by the DPPs, including the regression model, the correlation tests, the record evidence and the guilty pleas and admissions entered in this case, the district court concluded there was sufficient evidence to show common questions predominated as to common impact. Therefore, it ruled that this prerequisite to Rule 23(b)(3) was met.

We conclude that the district court did not abuse its discretion in reaching this conclusion. The court conducted a rigorous analysis of the expert evidence presented by the parties. The district court did not err legally or factually in concluding that Dr. Mangum's pooled regression model, along with other evidence, is capable of answering the question whether there was antitrust impact due to the collusion on a class-wide basis, thus satisfying this prerequisite of Rule 23(b)(3).

We now turn to the Tuna Suppliers' claims that the district court abused its discretion in determining that the evidence presented by the DPPs proved: (1) that the element of antitrust impact is capable of being established class-wide through common proof, and (2) that this common question predominates over individual questions.

The Tuna Suppliers do not challenge the district court's gatekeeping function under Daubert , to ensure that Dr. Mangum's evidence was not "statistically inadequate or based on implausible assumptions." Tyson Foods , 577 U.S. at 459, 136 S.Ct. 1036. And contrary to the dissent's argument, Dissent at 687–88, the district court did not merely determine that Dr. Mangum's evidence was admissible under Daubert. Rather, it subjected the evidence to a rigorous examination with full consideration of Dr. Johnson's critique. Therefore, the dissent's assertion that the district court committed the same error as the district court in Ellis is misplaced. Dissent at 687–88.

The Tuna Suppliers' main argument is that the district court abused its discretion in determining that Dr. Mangum's model is capable of proving common impact for all class members. According to the Tuna Suppliers, Dr. Mangum's evidence is not a permissible method of proving class-wide liability because the regression model uses "averaging assumptions," meaning that the model assumes that all DPPs were overcharged by the same uniform percentage (10.28 percent). These averaging assumptions, according to the Tuna Suppliers, "paper over" individualized differences among class members. Because the tuna market is characterized by individualized negotiations and different bargaining power among the purchasers, the Tuna Suppliers claim it is fundamentally impossible to show common proof of injury. To support this argument, the Tuna Suppliers note that the DPPs who pursued their antitrust claims individually did not rely on a pooled regression model but used actual cost data and claimed an individualized overcharge rate. Given the nature of the tuna market, the Tuna Suppliers conclude, Dr. Mangum's model cannot meet the prerequisites of Rule 23(b)(3).

To the extent that the Tuna Suppliers argue that pooled regression models involve improper "averaging assumptions" and therefore are inherently unreliable when used to analyze complex markets, we disagree. In antitrust cases, regression models have been widely accepted as a generally reliable econometric technique to control for the effects of the differences among class members and isolate the impact of the alleged antitrust violations on the prices paid by class members. See, e.g., Econometrics at 1. Further, Tyson Foods rejected any categorical exclusion of representative or statistical evidence. 577 U.S. at 459–60, 136 S.Ct. 1036. Therefore, any categorical argument that a pooled regression model cannot control for variables relating to the individualized differences among class members must be rejected.

See, e.g., Kleen Prod. LLC v. Int'l Paper Co. , 831 F.3d 919, 929 (7th Cir. 2016) ; In re Urethane , 768 F.3d at 1263 ; Cordes & Co. Fin. Servs. v. A.G. Edwards & Sons, Inc. , 502 F.3d 91, 97, 107 (2d Cir. 2007) ; Hemmings v. Tidyman's Inc. , 285 F.3d 1174, 1188–89 (9th Cir. 2002) ; In re Linerboard Antitrust Litig. , 305 F.3d 145, 153 (3d Cir. 2002).

Although the Tuna Suppliers refer to Dr. Mangum's regression model as "representative evidence," that term is imprecise. As explained in Tyson Foods , representative evidence generally refers to a sample that represents the class as a whole. See 577 U.S. at 454–55, 136 S.Ct. 1036. Thus, Tyson Foods concluded that each individual in a class could rely on exemplars of persons donning and doffing protective equipment to prove the amount of time each spent donning and doffing; this sample was claimed to be representative of all members of the class. See id. By contrast, a regression model analyzes available data to determine the degree to which a known variable, such as collusion, affected an unknown variable, such as price, while eliminating the effect of other variables.

To the extent the Tuna Suppliers and the dissent raise the more focused argument that, in this case, the model's output (estimating that the Tuna Suppliers' conspiracy resulted in a 10.28 percent overcharge for the entire class) cannot plausibly serve as common evidence for all class members given the individualized differences among those class members, we again disagree. It is not implausible to conclude that a conspiracy could have a class-wide impact, "even when the market involves diversity in products, marketing, and prices," especially "where, as here, there is evidence that the conspiracy artificially inflated the baseline for price negotiations." In re Urethane , 768 F.3d at 1254–55. As the Tenth Circuit explained, a district court could reasonably conclude "that price-fixing would have affected the entire market, raising the baseline prices for all buyers." Id. at 1255. In other words, it is both logical and plausible that the conspiracy could have raised the baseline prices for all members of the class by roughly ten percent. The district court did not abuse its discretion in so concluding.

To the extent the Tuna Suppliers challenged the model's inputs, the district court considered and rejected Dr. Johnson's critique that some of the model's inputs (i.e., the use of a cost index and Dr. Mangum's selection of time periods) rendered the model incapable of demonstrating class-wide impact. Cf. In re Lamictal , 957 F.3d at 194 (holding that the district court abused its discretion in certifying class because it failed to scrutinize each expert's data).

The dissent argues that Dr. Mangum's expert opinion "flies against common sense and empirical evidence," because large retailers like Walmart likely would have used their bargaining power to negotiate lower prices, and thus may not have paid higher prices because of the Tuna Suppliers' collusion. Dissent at 689. But the district court is not free to prefer its own views about the economics of the tuna market over the statistical evidence submitted by the plaintiffs, and here the regression model controlled for the variables identified by the dissent. Indeed, Dr. Mangum provided an individualized overcharge estimate for Walmart when he changed the model to evaluate the overcharge based on customer types. This test showed that Walmart paid statistically significant overcharges because of the conspiracy. Provided that the evidence is admissible and, after rigorous review, determined to be capable of establishing antitrust impact on a class-wide basis, it is for the jury, not the court, to decide the persuasiveness of Dr. Mangum's evidence in light of "common sense and empirical evidence."

The Tuna Suppliers rely on In re New Motor Vehicles Canadian Export Antitrust Litigation , 522 F.3d 6 (1st Cir. 2008), for the proposition that a market involving individualized negotiations is inherently incompatible with common impact. This reliance is misplaced. In New Motor Vehicles , plaintiffs raised a "novel and complex" theory of how consumers were injured by defendants' alleged horizontal conspiracy to discourage imports of lower-cost cars from Canada into the United States. Id. at 27. Plaintiffs' theory proceeded in two steps: (1) "but for the defendants' illegal stifling of competition," manufacturers would have set lower prices to compete with Canadian imports; and (2) because the manufacturers did not do so, consumers paid higher retail prices. Id. The First Circuit rejected this theory because plaintiffs failed to demonstrate they had an approach for proving either step. For the first step, plaintiffs had not shown how they would establish that but for the horizontal conspiracy, enough lower-priced Canadian cars would flood into the American market so as to cause manufacturers to decrease their prices. Id. As for the second step, the plaintiffs had not proved their damages model was capable of showing "which consumers were impacted by the alleged antitrust violation and which were not." Id. at 28. In this regard, the plaintiffs relied on an inference that "any upward pressure on national pricing would necessarily raise the prices actually paid by individual consumers." Id. at 29. But the First Circuit rejected this inference because "[t]oo many factors play into an individual negotiation to allow an assumption—at least without further theoretical development —that any price increase or decrease will always have the same magnitude of effect on the final price paid." Id. at 29 (emphasis added). The court contrasted the plaintiffs' unsupported inference with cases allowing "a presumption of class-wide impact in price-fixing cases when ‘the price structure in the industry is such that nationwide the conspiratorially affected prices at the wholesale level fluctuated within a range which, though different in different regions, was higher in all regions than the range which would have existed in all regions under competitive conditions.’ " Id. (quoting In re Linerboard Antitrust Litig. , 305 F.3d 145, 151 (3d Cir. 2002) ). Despite rejecting the plaintiffs' theory at an early stage of the case, the court did not rule out certification of a class but instead concluded that "more work remained to be done in the building of plaintiffs' damages model and the filling out of all steps of plaintiffs' theory of impact." Id.

As a threshold matter, the First Circuit held in New Motor Vehicles that the district court lacked federal jurisdiction over the plaintiffs' claims, but went on to provide its thoughts on certification of the class in the event the district court exercised its discretion to exert supplemental jurisdiction over the state damages claims. 522 F.3d at 17.

As this explanation of the case makes clear, New Motor Vehicles ' analysis is not applicable here. First, the DPPs' price-fixing theory is not "novel" or "complex." Id. at 27. Rather than adopting a theory requiring multiple speculative steps, the DPPs have a simple one-step theory: the Tuna Suppliers conspired to raise tuna prices, resulting in higher prices for all buyers. Second, while the plaintiffs in New Motor Vehicles had not provided a thorough explanation or developed a model showing how they would establish their theory, id. at 29, the DPPs have already offered well-developed expert testimony and regression modeling supporting common impact. The other cases relied on by the Tuna Suppliers are equally inapposite. See, e.g., Blades v. Monsanto Co. , 400 F.3d 562, 572 (8th Cir. 2005) (affirming denial of class certification because evidence of a conspiracy to raise prices, without more, could not demonstrate impact across highly localized and highly individualized markets for hundreds of seed varieties, and the plaintiffs had not offered a common method of showing injury); Robinson v. Tex. Auto. Dealers Ass'n , 387 F.3d 416, 423 (5th Cir. 2004) (reversing class certification where the plaintiffs lacked a plausible theory of how the challenged conduct had consistently affected purchase prices).

The Tuna Suppliers also argue that because the individual plaintiffs pursuing their own antitrust claims showed overcharges both above and below the overcharge indicated by Dr. Mangum's model, a uniform 10.28 percent overcharge is implausible. We also reject this argument, because it improperly conflates the question whether evidence is capable of proving an issue on a class-wide basis with the question whether the evidence is persuasive. A lack of persuasiveness is not fatal at certification. See Amgen , 568 U.S. at 459–60, 133 S.Ct. 1184. For purposes of determining whether each member of the DPP class can rely on the model to prove antitrust impact, it is irrelevant whether actual sales data shows a specific class member was overcharged by more or less than 10.28 percent. Rather, the question is whether each member of the class can rely on Dr. Mangum's model to show antitrust impact of any amount. The district court did not abuse its discretion in finding that each member could. While individualized differences among the overcharges imposed on each purchaser may require a court to determine damages on an individualized basis, see supra Section III.C, such a task would not undermine the regression model's ability to provide evidence of common impact. Accordingly, we reject the Tuna Suppliers' argument that the regression model could not sustain liability in individual proceedings. Rather, "each class member could have relied on [the model] to establish liability if he or she had brought an individual action." See Tyson Foods , 577 U.S. at 455, 136 S.Ct. 1036. We therefore conclude that the district court did not err legally or factually in concluding that Dr. Mangum's pooled regression model does not fail on any of the grounds raised by the Tuna Suppliers.

The Tuna Suppliers do not "specifically and distinctly" raise the argument that the district court abused its discretion in resolving challenges to the inputs to the model which were raised below, such as Dr. Mangum's choice of benchmark period and use of cost indexes, so that argument is deemed forfeited on appeal. United States v. Kama , 394 F.3d 1236, 1238 (9th Cir. 2005).

The Tuna Suppliers and the dissent next contend that the district court erred by failing to resolve a dispute between the parties as to whether 28 percent of the class did not suffer antitrust impact. Instead of resolving the dispute between the parties' experts, the Tuna Suppliers claim, the district court improperly shifted the critical inquiry to the jury. In other words, the Tuna Suppliers argue that to satisfy Rule 23(b)(3)'s predominance requirement, plaintiffs must prove that all or nearly all class members were in fact injured by the alleged conspiracy, i.e., suffered antitrust impact.

Because the Tuna Suppliers' primary argument on appeal is that the DPPs failed to prove class-wide antitrust impact, we understand the Tuna Suppliers' reference to injury as referring to antitrust impact, an element of the class antitrust claims, not that the class members would not be able to prove that they suffered monetary damages.

In raising this argument, the Tuna Suppliers focus on Dr. Johnson's critique of Dr. Mangum's model, which stated that when he tested Dr. Mangum's model by changing it to evaluate the overcharge specific to each individual member of the DPP class, the test showed that 28 percent of the DPPs could not rely on the model to show an overcharge attributable to the conspiracy. According to the Tuna Suppliers, this evidence indicated that 28 percent of the DPP class did not suffer antitrust impact. And in district court, the Tuna Suppliers argued that "28% of a class—nearly one-third—far exceeds the de minimis number of uninjured class members that some courts have permitted in certifying a class." Therefore, the Tuna Suppliers argue that the class should not have been certified. Further, the Tuna Suppliers argue that the existence of a large number of uninjured class members raises a question as to whether the class has Article III standing. The Tuna Suppliers contend that because the class cannot be certified (and there are Article III issues) if Dr. Johnson's analysis is correct, the district court abused its discretion in failing to resolve the dispute regarding whether Dr. Johnson's conclusions about Dr. Mangum's model were correct.

We disagree. First, the Tuna Suppliers and the dissent mischaracterize the import of Dr. Johnson's critique. Dr. Johnson did not make a factual finding that 28 percent of the DPP class or 169 class members were uninjured. Instead, Dr. Johnson's test was aimed at undermining confidence in Dr. Mangum's pooled regression model, because class members with no or limited transactions during the benchmark period could not rely on the model to show that they suffered overcharges. At most, this critique supports the more attenuated argument that Dr. Mangum's model is unreliable, or would be unpersuasive to a jury. But the district court considered and resolved this methodological dispute between the experts in favor of Dr. Mangum by crediting his rebuttal that even class members with limited transactions during the class period can rely on the pooled regression model as evidence of impact on similarly situated class members. In other words, the 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. This was all that was necessary at the certification stage. The DPP class did not have to "first establish that it will win the fray" in order to gain certification under Rule 23(b)(3). Amgen , 568 U.S. at 460, 133 S.Ct. 1184. Nor is this a case such as Ellis , in which the court had to resolve a dispute regarding an issue of historical fact in order to determine whether the challenged discriminatory conduct could affect a class as a whole. See 657 F.3d at 983. There is no factual dispute that the Tuna Suppliers engaged in a price-fixing scheme affecting the entire packaged tuna industry nation-wide.

The district court's conclusion that the Tuna Suppliers could present Dr. Johnson's critique at trial did not improperly shift the burden of determining whether the Rule 23(b)(3) prerequisites were met to the jury. See Amgen , 568 U.S. at 459–60, 466, 133 S.Ct. 1184. 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. "Reasonable minds may differ as to whether the [overcharge Dr. Mangum] calculated is probative" as to all purchasers in the class, but that is a question of persuasiveness for the jury once the evidence is sufficient to satisfy Rule 23. See Tyson Foods , 577 U.S. at 459, 136 S.Ct. 1036.

The Tuna Suppliers do not "specifically and distinctly" develop the argument that the district court failed to resolve the parties' dispute as to whether the evidence generated false positives. Kama , 394 F.3d at 1238. In any event, as explained above, Dr. Mangum rebutted these critiques by reference to the umbrella effect, and by claiming that Dr. Johnson's analysis was itself flawed because Dr. Johnson thought DPP class members had purchased non-defendant tuna, when they actually purchased tuna supplied by defendants. The district court did not abuse its discretion in resolving this issue by crediting Dr. Mangum's rebuttal.

Neither Dr. Mangum's pooled regression model nor Dr. Johnson's critique required individualized inquiries into the class members' injuries. 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. In neither case would the litigation raise individualized questions regarding which members of the DPP class had suffered an injury. Although such issues would have to be addressed at the damages stage, the dissent's argument that the district court here erred by failing to determine whether questions of individualized damages predominate, Dissent at 690, misses the mark. As noted above, the Tuna Suppliers have not argued that the complexity of damages calculations would defeat predominance here, and as previously explained, there is no per se rule that a district court is precluded from certifying a class if plaintiffs may have to prove individualized damages at trial.

Although Dr. Johnson argued that Dr. Mangum's pooled regression model was unreliable and so could not sustain a jury finding of antitrust injury to the entire DPP class, the evidence adduced at trial may nevertheless sustain a jury finding of antitrust injury to all or part of the class.

In any event, Dr. Mangum's proposal for calculating damages is a straightforward process of applying the class-wide overcharge to the Tuna Purchasers' net sales records. See supra n.19. That proposal does not give rise to a concern about individualized mini-trials to determine each class member's damage award. "That the defendant might attempt to pick off the occasional class member here or there through individualized rebuttal does not cause individual questions to predominate." Halliburton , 573 U.S. at 276, 134 S.Ct. 2398.

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 Tuna Purchasers have demonstrated that all class members have standing here. A plaintiff is required to establish the elements necessary to prove standing "with the manner and degree of evidence required at the successive stages of the litigation." Lujan v. Defs. of Wildlife , 504 U.S. 555, 561, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992). Here, the district court concluded that the DPPs' evidence was capable of establishing antitrust impact on a class-wide basis. Because antitrust impact—i.e., that the Tuna Suppliers' collusion had a common, supra-competitive impact on a class-wide basis—is sufficient to show an injury-in-fact traceable to the defendants and redressable by a favorable ruling, the Tuna Purchasers have adequately demonstrated Article III standing at the class certification stage for all class members, whether or not that was required. See TransUnion , 141 S.Ct. at 2208 n.4.

The Supreme Court expressly held open the question "whether every class member must demonstrate standing before a court certifies a class." TransUnion , 141 S.Ct. at 2208 n.4 (emphasis omitted). Outside the class action context, the Supreme Court has held that each plaintiff must demonstrate Article III standing in order to seek additional money damages and, therefore, a litigant must demonstrate Article III standing in order to intervene as a matter of right. Town of Chester v. Laroe Ests., Inc. , ––– U.S. ––––, 137 S.Ct. 1645, 1651, 198 L.Ed.2d 64 (2017). But the Supreme Court has long recognized that in cases seeking injunctive or declaratory relief, only one plaintiff need demonstrate standing to satisfy Article III. See, e.g., Baggett v. Bullitt , 377 U.S. 360, 366 n.5, 84 S.Ct. 1316, 12 L.Ed.2d 377 (1964) ; Rumsfeld v. F. for Acad. & Institutional Rts., Inc. , 547 U.S. 47, 52 n.2, 126 S.Ct. 1297, 164 L.Ed.2d 156 (2006) ; Horne v. Flores , 557 U.S. 433, 446–47, 129 S.Ct. 2579, 174 L.Ed.2d 406 (2009). We have likewise applied this rule where a class sought injunctive or equitable relief. See Bates v. United Parcel Serv., Inc. , 511 F.3d 974, 985 (9th Cir. 2007) (en banc). We therefore overrule the statement in Mazza that "no class may be certified that contains members lacking Article III standing," 666 F.3d at 594, which does not apply when a court is certifying a class seeking injunctive or other equitable relief. We do not overrule Mazza as to any other holding which remain good law.

Accordingly, we affirm the district court's certification of the DPP class.

We next turn to the Tuna Suppliers' arguments that the district court abused its discretion in determining that the evidence presented by the CFPs and EPPs was capable of proving the element of antitrust impact under California's Cartwright Act, thus satisfying the prerequisites of Rule 23(b)(3).

The CFP subclass includes individuals and commercial entities who purchased bulk sized packaged tuna (packages of 40 ounces or more) from six companies (direct purchasers) which had purchased the tuna from the Tuna Suppliers. The CFPs' theory of antitrust impact proceeds in two steps. First, the CFPs claim that the Tuna Suppliers' conspiracy resulted in the direct purchasers paying an overcharge. Second, the CFPs claim that the overcharge was passed on from the direct purchasers to the CFPs.

The CFPs supported this theory with the expert testimony and report of economist Dr. Michael Williams, who employed a methodology substantially similar to that employed by Dr. Mangum. Dr. Williams first conducted a regression analysis to determine the overcharge the CFPs' suppliers (i.e., the six direct purchasers) incurred because of the Tuna Suppliers' collusion. Like Dr. Mangum's analysis, Dr. Williams's regression analysis controlled for the effect of other variables that affected price in order to isolate the effect of the Tuna Suppliers' collusion. Dr. Williams concluded that COSI overcharged the CFPs' direct purchasers by 16.6 percent, StarKist by 18.2 percent, and Bumble Bee by 15.3 percent.

Next, Dr. Williams performed a separate regression analysis to determine if those overcharges passed through to the CFPs, and determined that the direct purchasers passed through 92 to 113 percent of their overcharge to the CFPs. Dr. Williams then performed two tests to verify that his estimates applied class-wide, both of which confirmed his theory.

To rebut Dr. Williams's analysis, the Tuna Suppliers relied on a critique by economist Dr. Linda Haider. Dr. Haider asserted that Dr. Williams erroneously assumed that all CFPs paid a common overcharge and that the same overcharge was passed through to the individual CFPs. Dr. Haider also contended that some of the CFP class members, such as food preparers and distributors, were not impacted because they could have passed through their overcharges to other purchasers downstream. Finally, Dr. Haider claimed that Dr. Williams's model was unreliable because it failed to account for non-defendant tuna purchased by the CFPs' direct purchasers.

The district court reviewed Dr. Williams's report and testimony as well as Dr. Haider's critiques, and after resolving the parties' disputes, concluded that Dr. Williams's methodology was valid and capable of resolving the antitrust impact issue in a single stroke, even though the Tuna Suppliers could raise the same critiques at trial to persuade the jury.

On appeal, the Tuna Suppliers argue that the district court abused its discretion in concluding that Dr. Williams's methodology satisfied Rule 23(b)(3)'s requirement of common proof of antitrust impact, because Dr. Williams erred in assuming that all direct purchasers were overcharged by the same percentage and that each class member was subject to the same pass-through rate. We disagree. As explained in Section IV, supra , a district court does not abuse its discretion in concluding that a regression model such as the one used by Dr. Williams may be capable of showing class-wide antitrust impact, provided that the district court considers factors that may undercut the model's reliability (such as unsupported assumptions, erroneous inputs, or nonsensical outputs such as false positives) and resolves disputes raised by the parties. The district court did so in this case, and therefore did not abuse its discretion in concluding that Dr. Williams's methodology was reliable and capable of showing class-wide impact.

We also reject the Tuna Suppliers' argument based on Dr. Haider's contention that some CFP class members may have passed on their overcharges to downstream purchasers. Dr. Haider claimed that the CFPs' ability to prove common impact was problematic because the impact of overcharges on class members who passed on their overcharges would be different from the impact on members who did not pass on such overcharges. The district court did not abuse its discretion in rejecting this argument on the ground that the Tuna Suppliers had not shown that determining whether or not those class members had passed overcharges down the distribution chain would overwhelm the common issues and require an individualized analysis. Therefore, the district court could reasonably conclude that the common question of antitrust impact predominated over individualized questions concerning a passed-on overcharge.

The EPP subclass contains individual consumers who purchased the Tuna Suppliers' products for personal consumption. Thus, like the CFPs, the EPPs are indirect purchasers whose theory of antitrust impact depends on two separate overcharges: first, an overcharge by the Tuna Suppliers to the direct purchasers (i.e., retail stores), and then an overcharge passed on to the EPPs. To carry their burden of showing they could establish class-wide overcharges through common proof, the EPPs offered the testimony of economist Dr. David Sunding, who employed a methodology substantially similar to that employed by Dr. Mangum and Dr. Williams.

Like Drs. Mangum and Williams, Dr. Sunding first conducted a regression analysis to isolate the impact of the collusion on the direct purchasers, which he concluded was an 8.1 percent overcharge from COSI, 4.5 percent from StarKist, and 9.4 percent from Bumble Bee. He then determined that the overcharges passed through to the EPP class members ranged from 65.3 to 135 percent with an estimated pass-through rate of 100 percent for the entire class. Dr. Sunding provided qualitative, quantitative and anecdotal evidence to support his assumption of a pass through rate for the entire class, including an examination of retail scanner data and the Tuna Suppliers' internal records.

Dr. Haider critiqued Dr. Sunding's methodology and findings on many of the same grounds as she criticized Dr. Williams's model and conclusions. She also made the additional criticisms that Dr. Sunding's methodology produced absurd results because it showed prices that made no economic sense, and that his model ignored, and therefore failed to control for, important factors like loss-leader and focal point pricing. The district court analyzed the evidence and the experts' disputes, and concluded that Dr. Sunding's report and testimony were capable of showing antitrust impact common to the class, for the same reasons explained in the court's analysis of Dr. Mangum's and Dr. Williams's models. The district court determined that Dr. Haider's additional critiques were based either on a misreading of Dr. Sunding's report, or her own miscalculations.

On appeal, the Tuna Suppliers argue only that Dr. Sunding's model and testimony was not capable of proving common impact for all class members because of its use of "averaging assumptions." This argument fails for the reasons explained above. See supra Section IV.A. Thus, the district court properly considered and rejected Dr. Haider's arguments, and determined that Dr. Sunding's methodology was capable of proving antitrust impact on a class-wide basis. That is enough to satisfy Rule 23(b)(3).

In a complex market such as the one at issue here, where different purchasers with different bargaining power purchased a range of products at different prices from different suppliers, commentators have raised reasonable questions whether statistical models are capable of resolving the issue of antitrust impact with common proof. See, e.g. , Michelle M. Burtis & Darwin V. Neher, Correlation and Regression Analysis in Antitrust Class Certification , 77 Antitrust L.J. 495, 518 (2011). But such statistical models and other evidence have been accepted as probative in a range of litigation contexts, and the Supreme Court has made clear that the permissibility of statistical evidence "turns not on the form a proceeding takes—be it a class or individual action—but on the degree to which the evidence is reliable in proving or disproving the elements of the relevant cause of action." Tyson Foods , 577 U.S. at 455, 136 S.Ct. 1036. Here the district court did not abuse its discretion in rigorously analyzing such statistical evidence, determining that it was not flawed in a manner that would make it incapable of providing class-wide proof, see supra Section III.C, concluding that the evidence was sufficient to sustain a jury verdict on the question of antitrust impact for the entire class, and preserving the defendants' ability to challenge the persuasiveness of such evidence at trial. We therefore affirm the district court's decision to certify the Tuna Purchasers' three subclasses under Rule 23(b)(3). Nevertheless, the Tuna Suppliers will have the opportunity to convince a jury that not all class members were overcharged due to their collusion.

AFFIRMED.

LEE, Circuit Judge, with whom KLEINFELD, Circuit Judge, joins, dissenting:

Over the past two decades, plaintiffs have notched over $103 billion in settlements from securities class actions alone. If we include other types of class actions—wage and hour, consumer lawsuits, antitrust disputes, and many others—that settlement amount almost certainly swells up by tens of billions of dollars more. These settlement sums are staggering because class action cases rarely go to trial. If trials these days are rare, class action trials are almost extinct. And it is no wonder why class actions settle so often: If a court certifies a class, the potential liability at trial becomes enormous, maybe even catastrophic, forcing companies to settle even if they have meritorious defenses.

See Securities Class Action Settlements—2019 Review and Analysis, Harvard Law School Forum on Corporate Governance, available at https://corpgov.law.harvard.edu/2020/03/11/securities-class-action-settlements-2019-review-and-analysis/ (last visited Oct. 21, 2021).

See, e.g. , Securities Class Action Filings, 2020 Year in Review, Cornerstone Research, at 18, available at https://www.cornerstone.com/Publications/Reports/Securities-Class-Action-Filings-2020-Year-in-Review (last visited Oct. 21, 2021) (noting only 11 securities class action cases tried to verdict in the past quarter century and only one tried since 2014).

That is why the Supreme Court has urged lower courts to "rigorous[ly]" scrutinize whether plaintiffs have met class certification requirements. See Wal-Mart Stores, Inc. v. Dukes , 564 U.S. 338, 351, 131 S.Ct. 2541, 180 L.Ed.2d 374 (2011). The majority opinion, however, allows the district court to certify a class, even though potentially about one out of three class members suffered no injury. But if defendants' econometrician expert is correct that almost a third of the class members may not have suffered injury, plaintiffs have not shown the predominance of common issues under Rule 23(b).

The district court acknowledged the dueling experts' differing opinions on this crucial question but held that it would leave that issue for another day—at trial—because it involves a merits issue that a jury should decide. See In re Packaged Seafood Prods. Antitrust Litig. , 332 F.R.D. 308, 325–28 (S.D. Cal. 2019). But as a practical matter, that day will likely never come to pass because class action cases almost always settle once a court certifies a class. A district court thus must serve as a gatekeeper to resolve key issues implicating Rule 23 requirements—including whether too many putative class members suffered no injury—at the class certification stage. See Med. & Chiropractic Clinic, Inc. v. Oppenheim , 981 F.3d 983, 992 (11th Cir. 2020) (" Rule 23 makes clear that the district court in which a class action is filed operates as a gatekeeper").

Punting this key question until later amounts to handing victory to plaintiffs because this case will likely settle without the court ever deciding that issue. The refusal to address this key dispute now is akin to the NFL declining to review a critical and close call fumble during the waning minutes of the game unless and until the game reaches overtime (which, of course, will likely never occur if it does not decide the disputed call). Such a practice is neither fair nor true to the rule.

I thus respectfully dissent.

* * * * *

The U.S. Department of Justice's investigation revealed that the three largest domestic producers of packaged tuna colluded to try to inflate the prices of their products. This class action lawsuit soon followed the criminal indictment. Among the plaintiffs include the direct purchasers of the tuna products, ranging from multibillion dollar chain retailers to small mom-and-pop stores. Not surprisingly, some plaintiffs (such as Walmart) wield substantial negotiating leverage: They can demand lower prices or extract additional promotional credits or rebates that defray the offered price. In contrast, an owner of a bodega likely cannot demand even an audience with the tuna producers, let alone ask for lower prices or more promotional credits.

Despite the varying negotiating power among the plaintiffs, their expert, Dr. Russell Mangum III, concluded that the tuna producers overcharged the direct purchasers by an average of 10.28%. He also suggested that about 5.5% of the class may not have suffered an injury because of this price-fixing. In contrast, the defendants' expert, Dr. John Johnson, offered an analysis showing that potentially about 28% of the class members suffered no injury.

Faced with this gaping difference between the two experts' conclusions, the district court acknowledged that Dr. Johnson's "criticisms are serious." In re Packaged Seafood , 332 F.R.D. at 328. But it held that this question should be left for trial because Dr. Mangum's method was reliable under Daubert and "capable of showing" class-wide impact. Id. The majority agrees with the district court, ruling that a class can be certified—even if potentially one out of three members suffered no injury—because Plaintiffs' expert offered a method "capable" of measuring class-wide impact and the district court can winnow out those uninjured members later at trial. But the majority opinion conflicts with Rule 23's text, common sense, and precedent from other circuits.

I. The district court did not "rigorously" scrutinize the dueling experts' opinions about uninjured class members.

While around 10,000 class action lawsuits are filed annually , class actions are "an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only." Comcast Corp. v. Behrend , 569 U.S. 27, 33, 133 S.Ct. 1426, 185 L.Ed.2d 515 (2013) (quoting Califano v. Yamasaki , 442 U.S. 682, 700–01, 99 S.Ct. 2545, 61 L.Ed.2d 176 (1979) ). Rule 23 thus establishes stringent requirements for certifying a class.

Class Actions 2021 , Lexology, Jonathan D. Polkes and David J. Lender, eds., at 91 (2021).

Among the Rule 23 requirements, the plaintiff must show that "questions of law or fact common to class members predominate over any questions affecting only individual members." Fed. R. Civ. 23(b)(3). The word "common" means "belonging to or shared ... by all members of a group," while "predominate" means "to hold advantage in numbers or quantity." Rule 23(b)(3) thus requires that questions of law or fact be shared by all or substantially all members of the class.

"Common" and "predominance," Merriam-Webster Dictionary, available at www.merriam-webster.com/dictionary (last checked on Oct. 21, 2021).

The Supreme Court has also reminded us that Rule 23 does not establish a "mere pleading standard." Wal-Mart , 564 U.S. at 350, 131 S.Ct. 2541. Rather, plaintiffs must prove by a preponderance of the evidence that they have met the Rule 23 requirements. See id. ; Maj. Op. at 664–65. Rule 23 imposes a requirement on the trial court, too. A trial court can certify a class only after engaging in a "rigorous analysis" and determining that the plaintiff has satisfied Rule 23. Wal-Mart , 564 U.S. at 351, 131 S.Ct. 2541. And in conducting that "rigorous analysis," trial courts "[f]requently" must assess "the merits of the plaintiff's underlying claim" because the issues are often intertwined. Id.

Rule 23's "rigorous analysis" is different from "reliable" or "relevant." Ellis v. Costco Wholesale Corp. , 657 F.3d 970, 982–84 (9th Cir. 2011). A trial court must do more than just consider one side's expert opinion as "reliable" and then kick the can down the road until trial. Rather, it must dig into the weeds and decide the battle of dueling experts if their dispute implicates Rule 23 requirements.

Here, the two experts' contentions centered on Rule 23(b)(3)'s predominance requirement—whether it has been met if the defendants' expert concludes that potentially a significant number of putative class members were uninjured. Plaintiffs' expert argued that only about one out of twenty class members likely did not suffer an injury, while defendants' expert maintained it was potentially more than one out of four. The district court held that the plaintiffs' expert's opinion passed muster under Daubert but admitted that the defendants' expert offered "serious" criticism, too. The district court admirably analyzed this difficult issue but ultimately did not resolve it, ruling that a jury should decide it at trial.

Despite the detailed analysis of the district court, I believe it abused its discretion in committing the same error that we cautioned against in Costco. There, the two dueling experts offered contrasting opinions on whether Costco's alleged discrimination was regional or nationwide, which touched upon Rule 23(a)'s commonality requirement (i.e. , whether all the putative class members nationwide suffered discrimination). The trial court held the plaintiffs' expert was reliable under Daubert , and declined to decide which experts' opinion should prevail at the class certification stage. It then certified a class and ruled that this "battle of the experts" issue could be decided at trial because Costco's criticisms of the expert report "attack the weight of the evidence and not its admissibility." Id. at 982 (quoting district court opinion). But because that dispute implicated Rule 23(a)'s commonality requirement, we reversed the district court's certification order and directed it to address it at the class certification stage. As we put it, the trial court "confused" the Daubert standard's "reliable" requirement with the "rigorous analysis" standard for Rule 23. Id. at 982 ("Instead of judging the persuasiveness of the evidence presented, the district court seemed to end its analysis of the plaintiffs' evidence after determining such evidence was merely admissible."). Rather than "examining the merits [of the dispute between experts] to decide this issue," the trial court "merely concluded that, because both Plaintiffs' and Costco's evidence was admissible, a finding of commonality was appropriate." Id. at 984. That was error.

And that is exactly what happened here. The district court found plaintiffs' expert to be reliable under Daubert , but it also conceded that the defendants' expert offered a "serious" critique of plaintiffs' expert opinion. The district court ultimately held that resolving this "battle of the experts" was a merits issue. But the dispute over the number of uninjured class members overlaps with Rule 23(b)(3)'s predominance requirement as well as Rule 23(a)'s lower threshold commonality requirement. Simply put, a plaintiff cannot prove that common issues predominate if one out of three putative class members suffered no harm. Cf. Mazza v. Am. Honda Motor Co., Inc. , 666 F.3d 581, 596 (9th Cir. 2012) ("[T]he relevant class must be defined in such a way as to include only members who were [harmed by being] exposed to advertising that is alleged to be materially misleading."). If a large number of class members "in fact suffered no injury," identifying those class members "will predominate." In re Asacol Antitrust Litig. , 907 F.3d 42, 53–54 (1st Cir. 2018). Thus, the district court had to "examin[e] the merits" of this dispute between the experts, and not "merely conclude[ ] that" both expert reports are reliable and admissible. Costco , 657 F.3d at 984.

The majority holds that Dr. Mangum's estimate of a 10.2% "average" price inflation meets Rule 23's requirements because it shows a method "capable" of showing common antitrust impact. The majority appears to distinguish between (i) cases in which the class members "logically " could not have been harmed (because, for example, they were never exposed to the misleading advertisement) or there is insufficient evidence to support commonality, and (ii) cases like this one in which an expert holds that many class members in reality may not have suffered any harm, even if they theoretically could have. Maj. Op. 666–67, n.9. In the former scenario, the majority says that a class cannot be certified because logically there cannot be commonality under Rule 23 ; in the latter case, the majority appears to argue that it is a merits issue because a jury will need to assess the persuasiveness of the expert's opinion.

I believe that creates a false distinction. Nothing in our decision in Costco or the Supreme Court's opinion in Wal-mart creates such a difference. If the evidence presented implicates Rule 23 —as it does here—then the district court must decide whether the plaintiffs have "prove[n] that there are in fact ... common questions of law or fact," even if it means assessing the persuasiveness of the expert opinions. Wal-mart , 564 U.S. at 350–51, 131 S.Ct. 2541 (emphasis in original). In Costco , we chastised the district court for not "judging the persuasiveness of the evidence presented" and "end[ing] its analysis of the plaintiffs' evidence after determining such evidence was merely admissible." 657 F.3d at 982. If we had to refrain from deciding the persuasiveness of an expert opinion used to show commonality, a plaintiff could prevail on class certification by merely offering a well-written and plausible expert opinion. See West v. Prudential Sec., Inc. , 282 F.3d 935, 938 (7th Cir. 2002) (failure to resolve dueling experts "amounts to a delegation of judicial power to the plaintiffs, who can obtain class certification just by hiring a competent expert").

Admittedly, resolving a battle of dueling experts over highly technical issues may seem like a difficult job for a court. But that tough task is likely even more difficult and daunting for jurors. In the end, a "district judge may not duck hard questions by observing that each side has some support ... Tough questions must be faced and squarely decided, if necessary by holding evidentiary hearings and choosing between competing perspectives." Id. After reviewing the evidence, a district court must make findings of fact necessary for determining whether Rule 23's requirements have been met.

And here, the expert opinion offered by Plaintiffs to show commonality (though admissible) is not persuasive. The majority contends that the expert's model is capable of measuring class-wide impact through an "averaging assumption" of 10.2% price inflation from the price-fixing conspiracy. Put another way, the model assumes that almost all class members suffered an injury because the price-fixing would elevate the list price of tuna for everyone, even if individual class members ultimately paid different prices for the tuna. But the expert's assumption flies against common sense and empirical evidence. Powerful retailers (like Walmart) are not passive or ill-informed consumers; they will not sit still when faced with a price increase. They will fiercely negotiate the list price down, or more likely, demand promotional credits or rebates that offset any price increase. See R. Pandey, et al., Factors Influencing Organization Success: A Case Study of Walmart , International Journal of Tourism & Hospitality in Asia Pasific, Vol. 4, No. 2, June 2021. See also Gary Rivlin, Rigged: Supermarket Shelves for Sale , Center for Science in the Public Interest, September 2016, available at cspinet.org/Rigged (last visited January 4, 2021).

Major retailers wield significant power over manufacturing and food companies because they represent the major channel to distribute the food products. If a major retail chain refuses to carry a company's product after a pricing dispute, it can significantly affect that company's bottom line. As one case study put it, "Walmart has huge bargaining power since ... it is one of the largest distributors for manufacturing [sic]. For instance, 17% of the total sales of P&G and 38.7% of the total sales of CCA Industries rely on Walmart stores. Without Walmart, these businesses would be unable to operate." Pandey, supra page 10, at 120.

Large retailers can also extract rebate or promotional concessions from the companies by threatening to place their products at the bottom of the shelves or less-visited aisles where consumers are less likely to notice them. All told, large retailers use this power to "collect more than $50 billion a year in trade fees and discounts from food and beverage companies." Rivlin, supra page 10, at ii. And "[f]ood manufacturers pay these fees ... because they have no choice. The stores are the gatekeepers." Id. at 21.

None of this is to say that Wal-Mart and other retailers achieved those price discounts and promotional credits or rebates here. We simply do not know because Plaintiffs' expert did not adequately consider it. The only way we can find out if Wal-Mart and other major retailers suffered any injury (and if so, how much) would be if we conducted highly individualized analyses of each class member. But that would defeat the commonality requirement under Rule 23.

The majority cites the deposition testimony of Plaintiffs' expert to argue that he considered promotional credits and rebates. Maj. Op. 672, n.16. But the expert added the caveat that he did so only in instances that he "could reliably" calculate the data. He then conceded that he did not include "discount or promotional information" with much of the data but said that "I have done all that I could." He ultimately concluded that he could measure damages by relying on the average 10.2% "overcharge" analysis in his expert report.

The majority seemingly waves away this difference in negotiating power between the class members by relying on our oft-quoted language that the "need for individualized findings as to amount of damages does not defeat class certification." Maj. Op. 668–69 (citing Vaquero v. Ashley Furniture Indus., Inc. , 824 F.3d 1150, 1155 (9th Cir. 2016); Pulaski & Middleman, LLC v. Google, Inc. , 802 F.3d 979, 988 (9th Cir. 2015) ).

I believe our court has misconstrued that often-quoted language to create a sweeping rule that gives a free pass to the intractable problem of highly individualized damages analyses. And such a rule also conflicts with the Supreme Court's holding that a class action must be capable of being resolved in "one stroke." Wal-mart , 564 U.S. at 350, 131 S.Ct. 2541 ; see also Comcast , 569 U.S. at 35, 133 S.Ct. 1426 (requiring a "rigorous analysis" to confirm that the damages model is "consistent with its liability case").

We first stated that the "amount of damages is invariably an individual question and does not defeat class action treatment" in Blackie v. Barrack , 524 F.2d 891, 905 (9th Cir. 1975). That was a securities fraud class action, and we recognized that "computing individual damages will be virtually a mechanical task " because "the amount of price inflation during the period can be charted." Id. (emphasis added). Put another way, damages can be easily calculated because it is a plug-and-play exercise: Look at the number of shares bought by each shareholder and the price of the share that day, and compare it to the price inflation caused by the misrepresentation. While each class member may have individualized damages, the damages can be easily calculated for the entire class in "one stroke." See Wal-mart , 564 U.S. at 350, 131 S.Ct. 2541.

Since Barrack , we have applied that concept mostly in employment and wage-and-hour cases. See, e.g., Vaquero , 824 F.3d at 1152 (suing for payment for unpaid hours on non-sales work); Leyva v. Medline Indus. Inc. , 716 F.3d 510, 514 (9th Cir. 2013) (class action based on wage and hour claims in which defendant's "computerized payroll and time-keeping database would enable the court to accurately calculate damages"). Wage-and-hour cases present another mechanical application scenario: a class administrator can easily look at the employer's payroll records and calculate the number of hours or wages that each employee was underpaid. At times, however, we have quoted that language without determining whether damages could be calculated mechanically or if the court would have to engage in individualized mini-trials for damages. See, e.g., Yokoyama v. Midland Nat. Life Ins. Co. , 594 F.3d 1087, 1094 (9th Cir. 2010) (stating that individualized damages do not defeat class certification in case involving misleading statements in annuities promotional materials). But here, it will not be a "mechanical task" to calculate the damages for each class member. Blackie , 524 F.2d at 905. The district court will need to conduct individualized mini-trials to determine whether each class member suffered an injury, and if so, what the damages are for each member. That would upend Rule 23's commonality requirement. The majority opinion notes that commonality may still be met, even if a defendant "might attempt to pick off the occasional class member here or there." Maj. Op. 682, n. 31 (citing Halliburton Co. v. Erica P. John Fund, Inc. , 573 U.S. 258, 276, 134 S.Ct. 2398, 189 L.Ed.2d 339 (2014) ). But our case does not involve a "pick off" of a few uninjured class members, but rather a massive grab bag of class members—perhaps almost a third of the class—who may not have suffered any harm. The district court thus will have to engage in individualized mini-trials to figure out who suffered an injury.

Finally, the majority suggests that an oversized class with unharmed class members does not pose a practical problem if a method can separate the uninjured from the injured at trial. No harm, no foul, the majority implies. But that cannot be so if a large number of class members (certainly, a third) suffered no injuries. Suppose that 80% of the putative class members suffered no harm. Could a district court still certify a class just because it could later winnow out the 80% who were uninjured? Would Rule 23(b)'s predominance of common issues be met even if only 20% of the putative members belong in the class? By definition, a class with 80% uninjured members cannot present a predominance of common issues because they have nothing in common with the remaining sliver of injured members.

If we allow a court to certify a class in which a large number of putative class members have suffered no injury, we will allow plaintiffs to weaponize Rule 23 to impose an in terrorem effect on defendants. The "[c]ertification of the class is often, if not usually, the prelude to a substantial settlement by the defendant because the costs and risks of litigating further are so high." Amgen Inc. v. Conn. Ret. Plans & Tr. Funds , 568 U.S. 455, 485, 133 S.Ct. 1184, 185 L.Ed.2d 308 (2013) (Scalia, J., dissenting). Indeed, "when 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." AT&T Mobility LLC v. Concepcion , 563 U.S. 333, 350, 131 S.Ct. 1740, 179 L.Ed.2d 742 (2011).

So if a court certifies a class with many uninjured class members, it dramatically expands the potential exposure and artificially jacks up the stakes. It matters little that the uninjured class members can be separated at trial because with "the stakes so large ... settlement becomes almost inevitable—and at a price that reflects the risk of a catastrophic judgment as much as, if not more than, the actual merit of the claims." In re Bridgestone/Firestone, Inc. , 288 F.3d 1012, 1016 (7th Cir. 2002). The opportunity at trial to jettison uninjured members from the certified class is a phantom solution because defendants will have little choice but to settle before then.

II. The majority's rejection of a de minimis rule creates a circuit split.

I believe the majority also errs in rejecting a de minimis rule. To be sure, a plaintiff need not show that every single putative class member has suffered an injury. But the number of uninjured class members should be de minimis —based on Rule 23's language, common sense, and precedent from other circuits. First, as noted above, the words "common" and "predominate" in Rule 23(b)(3) suggest that the class should include only (or mostly only) people who have suffered an injury. If one-third—or half or two-thirds—of the class members suffered no injury, it follows that "common" issues would not "predominate," as required under the text of Rule 23, because those uninjured class members have little in common with those who have been harmed. In short, Rule 23 allows a de minimis number of uninjured members but no more.

Second, allowing more than a de minimis number of uninjured class members tilts the playing field in favor of plaintiffs. By expressly rejecting a de minimis rule, the majority's opinion will invite plaintiffs to concoct oversized classes stuffed with uninjured class members—with little fear of having their class certification bids being denied for lack of "predominance" or "commonality." And in creating these grossly oversized classes, plaintiffs will inflate the potential liability (and ratchet up the attorney's fees based in part on that amount) to extract a settlement, even if the merits of their claims are questionable.

Finally, the majority opinion needlessly creates a split with other circuits that have endorsed a de minimis rule. The D.C. Circuit, for example, suggested that "5% to 6% constitutes the outer limits of a de minimis number." In re Rail Freight Fuel Surcharge Antitrust Litig. , 934 F.3d 619, 624–25 (D.C. Cir. 2019) (cleaned up). The district court had found that the class of 16,065 members (12.7% of whom were uninjured) failed to meet the predominance requirement because more than a "de minimis " number were uninjured. Id. at 623–24. The D.C. Circuit on appeal affirmed, ruling that the plaintiffs' model "even if sufficiently reliable, does not prove classwide injury." Id. at 623. Put another way, "even assuming the model can reliably show injury and causation for 87.3 percent of the class, that still leaves the plaintiffs with no common proof of those essential elements of liability for the remaining 12.7 percent." Id. at 623–24

Likewise, the First Circuit suggested that "around 10%" of uninjured class members marks the de minimis border. See In re Asacol , 907 F.3d at 47, 51–58. The First Circuit was perhaps willing to look past "a very small absolute number of class members" who have suffered no injury because they "might be picked off in a manageable, individualized process at or before trial." Id. at 53. But if "there are apparently thousands who in fact suffered no injury ... [t]he need to identify those individuals will predominate." Id. at 53–54.

* * * * *

While this case centers on the narrow issue of price-fixing of canned tuna, its implications extend beyond to a wide sea of class action cases. I fear that today's decision will unleash a tidal wave of monstrously oversized classes designed to pressure and extract settlements.

I respectfully dissent.

Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC

Summary

Opinion

AFFIRMED.

I. The district court did not "rigorously" scrutinize the dueling experts' opinions about uninjured class members.

II. The majority's rejection of a de minimis rule creates a circuit split.

Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC

Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC

Case Details

Citations

Citing Cases

Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC

Summary

Opinion

AFFIRMED.

I. The district court did not "rigorously" scrutinize the dueling experts' opinions about uninjured class members.

II. The majority's rejection of a de minimis rule creates a circuit split.

Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC

Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC

Case Details

CitationsCopy Citation

Citing Cases

Citations