T he recent wave of food and beverage class actions

Product Safety & Liability Reporter Reproduced with permission from Product Safety & Liability Reporter, 42 PSLR 1125, 10/06/2014. Copyright 2014 by The Bureau of National Affairs, Inc. (800-372-1033) http://www.bna.com FOOD TRENDS Recent class certification decisions in food and beverage advertising cases reveal common fact patterns that consistently weigh for or against certification, and confirm the importance of developing a robust factual record when opposing a motion for class certification, attorneys Scott A. Elder and Jenny A. Mendelsohn say in this BNA Insight. The authors also identify and discuss several issues, including ascertainability, that appear ripe for potential circuit splits and possible additional guidance from the Supreme Court. All You Can Eat: Food and Beverage Class Actions in 2014 BY SCOTT A. ELDER AND JENNY A. MENDELSOHN Scott A. Elder is a co-leader of Alston & Bird s products liability group, and represents corporations in complex litigation, including consumer class actions, personal injury cases and commercial matters. Elder is available at scott.elder@alston.com. Jenny A. Mendelsohn is a senior associate in Alston & Bird s products liability group, where she focuses on products liability, class and mass actions, and other complex litigation. Mendelsohn can be contacted at jenny.mendelsohn@alston.com T he recent wave of food and beverage class actions has moved into the class certification stage, and courts issued many class certification opinions in the first half of 2014. These opinions demonstrate the continued difficulty in certifying food and beverage class actions, as only four 1 of the 11 2 opinions analyzed certified a class. In all cases, the Third Circuit s ascertainability analysis in Carrera, the Supreme Court s predominance analysis of damages models in Comcast, and the Ninth Circuit s choice of law decision in Mazza are emerging as the precedents having the most impact on the certification decision. The recent class certification decisions also reveal several common fact patterns that consistently weigh 1 Brazil v. Dole Packaged Foods, LLC, N.D. Cal., No. 12- cv -01831, 5/30/14; Ebin v. Kangadis Food, Inc., 297 F.R.D. 561 (S.D.N.Y. 2014); Lanovaz v. Twinings N. Am., Inc., N.D. Cal., No. C-12-02646, 4/24/14. 2 Astiana v. Ben & Jerry s Homemade, Inc., N.D. Cal., No. C 10-4387, 1/7/14; Brazil; Bruton v. Gerber Prods. Co., N.D. Cal., No. 12-CV-02412, 6/23/14; Caldera v. J.M. Smucker Co., C.D. Cal., No. CV 12-4936, 4/15/14; Ebin, 297 F.R.D. 561; In re POM Wonderful, C.D. Cal., No. ML 10-02199, 3/25/14; Jones v. ConAgra Foods, Inc., N.D. Cal., No. C 12-01633, 6/13/14; Lanovaz; Sethavanish v. ZonePerfect Nutrition Co., N.D. Cal., No. 12-2907, 2/13/14; Stewart v. Beam Global Spirits & Wine, Inc., D. N.J., No. 11-5149, 6/27/14; Werdebaugh v. Blue Diamond Growers, N.D. Cal., No. 12-cv-2724, 5/23/14. COPYRIGHT 2014 BY THE BUREAU OF NATIONAL AFFAIRS, INC. ISSN 0092-7732

2 for or against certification and confirm the importance of developing a robust factual record when opposing a motion for class certification. Carrera, Mazza and Comcast The Carrera, Comcast, and Mazza decisions continue to shape the class certification landscape, particularly in food and beverage cases, so a brief review of those decisions is in order. In Carrera v. Bayer Corp, 3 a putative class alleged that Bayer falsely advertised its One-A-Day WeightSmart vitamins. The District of New Jersey certified the class despite Bayer s argument that the class was not ascertainable. The court likened ascertainability to manageability, and stated that speculative problems with case management are insufficient to prevent certification. 4 On appeal, the Third Circuit vacated the decision, finding that the two proposed methods of identifying the class, retailer records and customer affidavits, did not satisfy Rule 23 s implicit ascertainability requirement. First, the court held that there [wa]s no evidence that a single purchase of WeightSmart could be identified through retailer records. Second, the court held that affidavits would not be sufficiently reliable and would not afford Bayer its due process right to raise individual challenges and defenses to class membership. 5 Regarding the affidavits, the Court accepted Bayer s argument that plaintiff confused [the product at issue] with WeightSmart Advanced and other generic or similar products (none of which are part of this litigation), indicating that the reliability of class members recall should not be presumed. 6 In Comcast v. Behrend, 7 television subscribers brought a putative antitrust class action against cable service providers. Defendants argued unsuccessfully in the district court and in the Third Circuit that the case was certified improperly because plaintiffs proposed damages model did not link the alleged damages to the only theory of injury that remained in the case at class certification. 8 Ultimately, the Supreme Court reversed and held that because plaintiffs damages model did not apply across the entire class, that class failed to meet Rule 23(b)(3) s predominance requirement and could not be certified. In other words, [q]uestions of individual damage will inevitably overwhelm questions common to the class, absent a common damages model. 9 Thus, Comcast establishes that [c]alculations need not be exact, but at the class-certification stage (as at trial), any model supporting a plaintiff s damages case must be consistent with its liability case. 10 In Mazza v. American Honda Motor Co., 11 plaintiffs filed a putative nationwide class action against Honda, alleging that Honda misrepresented the benefits of an optional brake system on its Acura RL vehicles. The class representatives were non-california residents who purchased their vehicles from dealerships located in Florida and Maryland. In the district court, plaintiffs argued that California law should apply to the consumer protection claims, because Honda s corporate headquarters were located in California and the alleged misrepresentations were developed in and emanated from California. The lower court certified a nationwide class. The Ninth Circuit reversed, concluding that the district court abused its discretion in certifying a class under California law that contained class members who purchased or leased their cars in different jurisdictions with materially different consumer protection laws. 12 While the court held that California had significant contact to the claims of each class member, application of California s choice of law test dictated that the other states interests in applying their own distinct consumer protection law outweighed California s interest. Ascertainability In the wake of Carrera, many of 2014 s food and beverage class certification decisions turned on whether the proposed class was ascertainable. While there is no explicit ascertainability requirement in Rule 23, courts have held that the class must be objectively defined and clearly ascertainable before a class action may proceed. 13 Said differently, the class definition must be sufficiently definite so that it is administratively feasible to determine whether a particular person is a class member. 14 For example, in Brazil v. Dole Packaged Foods, LLC, 15 the Northern District of California certified a class of all California residents who purchased 10 Dole products that were labeled All Natural, but contained allegedly synthetic ingredients. 16 Relying on Carrera, Dole argued that the class was not ascertainable because it lacked company records to identify relevant purchasers. 17 The court disagreed, holding that [w]hile [Carrera] may now be the law in the Third Circuit, it is not currently the law in the Ninth Circuit. 18 The court continued, [i]n this Circuit, it is enough that the class definition describes a set of common characteristics sufficient to allow a prospective plaintiff to identify himself or herself as having a right to recover based on the description. 19 Importantly, the court distinguished Carrera because here all of the purchasers of the identified products are included in the class definition and all identified Dole products bore the same alleged misstatements, 20 so plaintiffs could likely proffer reliable affidavits demonstrating class membership. Thus, a consumer would 3 Carrera v. Bayer Corp., 727 F.3d 300, (3d Cir. 2013). 4 Id. at 304. 5 Id. at 307-312. 6 Id. at 304. 7 Comcast v. Behrend, 133 S. Ct. 1426 (2013). 8 Id. at 1431. 9 Id. at 1433. 10 Brazil v. Dole Packaged Foods, LLC, N.D. Cal., No. 12- cv -01831, 5/30/14. 11 Mazza v. American Honda Motor Co., 666 F.3d 581, (9th Cir. 2012). 12 Id. at 590. 13 Astiana (citing Xavier v. Philip Morris USA Inc., 787 F. Supp. 2d 1075, 189 (N.D. Cal. 2011). 14 Id. 15 No. 12-cv-01831 (N.D. Cal. May 30, 2014). 16 Id. 17 Id. 18 Id. (quoting McCrary v. The Elations Co., LLC, C.D. Cal., No. 13-242, 1/13/14. 19 Id. 20 Id. 10-6-14 COPYRIGHT 2014 BY THE BUREAU OF NATIONAL AFFAIRS, INC. PSLR ISSN 0092-7732

3 21 Werdebaugh v. Blue Diamond Growers, N.D. Cal., No. 12-cv-2724, 5/23/14. 22 Id. 23 No. C-12-02646 N.D. Cal. Apr. 24, 2014). 24 Id. 25 Id. 26 McCrary v The Elations Co., C.D. Cal., 1/13/14. 27 Lanovaz v. Twinings N. Am., Inc., N.D. Cal., No. C-12-02646, 4/24/14. 28 297 F.R.D. 561 (S.D.N.Y. 2014). 29 Id. at 567. 30 In re Visa Check/Mastermoney Antitrust Litigation, 280 F.3d 124, 140 (2d Cir. 2001). only have to remember that she bought the product as opposed to identifying a version with a particular label. Notably, the decision does not mention due process but focuses instead only on the reliability of the affidavits, apparently implying that a sufficiently reliable affidavit provides the due process to which a defendant is entitled. The same judge issued a similar order in Werdebaugh v. Blue Diamond Growers, 21 holding that here all of the purchasers of Blue Diamond s almond milk products are included in the class definition, and all cartons of the challenged almond milk products bore the alleged mislabeling, 22 so plaintiffs could likely proffer reliable affidavits confirming class membership. The court again focused on the affidavits reliability and did not discuss due process explicitly. In Lanovaz v. Twinings North America, Inc., 23 where the court only certified an injunctive relief class, plaintiffs alleged that Twinings had improperly labeled 51 varieties of teas as a Natural Source of Antioxidants. 24 Finding that the class was ascertainable despite the absence of records identifying purchasers, the district court noted that courts in the Ninth Circuit had found many similar classes to be ascertainable and that ruling otherwise would be the death of consumer class actions. 25 Like the courts in Brazil and Werdebaugh, the Lanovaz court cited McCrary v. The Elations Co. LLC, 26 holding that [i]n this Circuit, it is enough that the class definition describes a set of common characteristics sufficient to allow a prospective plaintiff to identify himself or herself as having a right to recover based on the description. 27 The court again did not discuss defendant s due process rights. Ebin v. Kangadis Food, Inc., 28 is one of two class certifications decisions in the first half of 2014 that comes from a non-california court. In Ebin, the Southern District of New York certified a class consisting of all persons who purchased Capatriti 100% Pure Olive Oil, which plaintiffs alleged was improperly labeled as 100% Pure. Interestingly, the Ebin court questioned the ascertainability of the class but declined to deny certification on this ground. The court opined that plaintiffs d[id] not point to any records that can objectively determine membership in the proposed class. Nor is it likely that consumers consistently maintain[ed] receipts of their purchase or the actual tins or bottles. Indeed, plaintiffs here have neither the Capatriti they purchased nor any receipts or documentation proving their purchases. 29 But, the Ebin court, quoting In re Visa Check/Mastermoney Antitrust Litigation, 30 reasoned that the Second Circuit has instructed that failure to certify an action under Rule 23(b)(3) on the sole ground that it would be unmanageable is disfavored and should be the exception rather than the rule. 31 The court continued, the class action device, at its very core, is designed for cases like this where a large number of consumers have been defrauded but no one consumer has suffered an injury sufficiently large as to justify bringing an individual lawsuit. Against this background, the ascertainability difficulties, while formidable, should not be made into a device for defeating the action. 32 Once again, the court failed to address defendant s due process right to challenge class membership, instead equating ascertainability only with manageability. Despite California s seemingly permissive approach to ascertainability, the Central District of California decertified a class in In re POM Wonderful, LLC, 33 on ascertainability grounds. It reasoned that because of limited sales records, and because the allegedly misleading 100% Juice statement was made only in Pom s advertising, but not on the product label, it would be too difficult to identify those purchasers who saw the alleged misleading statement. 34 Several other 2014 decisions refused to certify classes on ascertainability grounds for similar reasons. For example, in Astiana v. Ben & Jerry s Homemade, Inc., the court held that it was impossible to determine which consumers ice cream contained synthetic alkalized cocoa (the alleged unnatural ingredient) because only one of 15 cocoa suppliers used the allegedly synthetic ingredient. 35 Similarly, in Sethavanish v. ZonePerfect Nutrition Co., another All Natural case involving alkalized cocoa powder, the court held that the proposed nationwide class was unascertainable because it [wa]s unclear how Plaintiff intends to determine who purchased ZonePerfect bars during the proposed class period, or how many ZonePerfect bars each of these putative class members purchased. 36 Finally, in Bruton v. Gerber Products Co., 37 the same judge who certified the classes in Brazil and Werdebaugh declined to certify a class of California purchasers who bought Gerber baby foods that were allegedly mislabeled Excellent Source of, Good Source of, Healthy, or No Added Sugar. Gerber argued that the class was unascertainable because it did not keep records of who purchased its products, but the court rejected that argument in light of its rulings in Werdebaugh and Brazil. 38 The Bruton court, however, ultimately held that the class was unascertainable because of [t]he number of products at issue in this case, the varieties included and not included in the class definition, the changes in product labeling throughout the class period, the varied and uncertain length of time it takes for products with new labels to appear on store shelves, and the fact that the same products were sold with and without the chal- 31 297 F.R.D. at 567. 32 Id. 33 In re POM Wonderful (C.D. Cal. Mar. 25, 2014). 34 Id. 35 Astiana v. Ben & Jerry s Homemade, Inc., N.D. Cal., No. C 10-4387, 1/7/14. 36 Sethavanish v. ZonePerfect Nutrition Co., N.D. Cal., No. 12-2907, 2/13/14. 37 Bruton v. Gerber Prods. Co. 12-CV-02412 (N.D. Cal., June 23, 2014). 38 Id. PRODUCT SAFETY & LIABILITY REPORTER ISSN 0092-7732 BNA 10-6-14

4 lenged label statements simultaneously ma[d]e Plaintiff s proposed class identification method administratively unfeasible. 39 The court in Jones v. ConAgra Foods, Inc., 40 similarly held that subclasses of purchasers of Hunt s tomato products, PAM cooking spray, and Swiss Miss hot cocoa were not ascertainable because of the potential unreliability of purchaser s memories given the number of products at issue and label changes that occurred throughout the class period. Predominance 39 Id. 40 No. C 12-10633 N.D. Cal. June 13, 2014). 41 The price premium model compares the price of the identified products to the price of allegedly comparable products (often produced by other manufacturers) that do not have the allegedly misleading statement. 42 Brazil v. Dole Packaged Foods, LLC, N.D. Cal., No. 12- cv -01831, 5/30/14. 43 Id. See also Werdebaugh v. Blue Diamond Growers, N.D. Cal., No. 12-cv-2724, 5/23/14. 44 See also Astiana; Caldera v. J.M. Smucker Co.; In re POM Wonderful (decertifying class on this ground). A. Comcast Issues Whether a proposed class could satisfy Comcast s requirement for a classwide damages model was frequently litigated in the first half of 2014. In general, the Comcast analysis hinged on the viability of plaintiffs proposed regression analysis, as simpler theories such as a purchase price model generally failed to support class certification. In Brazil, for example, plaintiffs offered three damages models in their effort to satisfy Comcast. The court concluded that neither a full refund model, nor a price premium 41 damages model satisfied Comcast because neither properly measured the recoverable damages. A full refund model ignored that the plaintiffs received benefits from the product, and a price premium model failed to account for the fact that a premium may be placed on the product for reasons other than an All Natural label. 42 However, the court held that the regression model sufficiently tie[d] damages to Dole s alleged liability under Comcast because it allowed the plaintiff to compare Dole s profits for the same product with and without the allegedly misleading label. 43 As in Brazil, other courts found that damages models did not satisfy Comcast where plaintiffs proffered only a full refund or price premium model and failed to introduce a reliable regression analysis. 44 The decisions also make clear that not just any proffered regression analysis will support class certification. In Lanovaz, for example, plaintiffs could not rely on the initially proffered regression analysis because the Natural Source of Antioxidants statement had been on the product labels over the entire class period, meaning there was no way to establish the price premium attributable to the allegedly misleading label alone. Id. Similarly, in Jones, the court held that plaintiff s expert had not sufficiently described his proffered regression analysis, stating that the expert d[id] not provide a clearly defined list of variables, he ha[d] not determined whether the data related to any or all of his proposed control variables exists, and he ha[d] not determined, or shown how he would determine, which competing and complementary products he would use. 45 B. Choice of Law In the first half of 2014, no court certified a nationwide 23(b)(3) class action in a food or beverage case; instead, courts certified two California classes, one class comprised of purchasers from several Northeastern states, and one 23(b)(2) injunctive relief class. These cases suggest that the Ninth Circuit s Mazza decision continues to have a significant impact on the class certification analysis and that many future classes will be either statewide or limited to smaller groups of states. In Ebin, for example, the olive oil manufacturer defendant tried to defeat certification by arguing that there were multiple questions of law at issue, including different states laws regarding fraud and misrepresentation. 46 However, unlike in other cases, this product was only sold throughout the Northeast, not nationwide. 47 Accordingly, the court held that a survey of potentially applicable state laws [including the laws of New York, Connecticut, Pennsylvania and Vermont], reveal[ed] no material difference that would affect the merits of the class s common law claims at trial. 48 In Brazil and Werdebaugh the court held that it was only proper to certify a class of California plaintiffs. 49 In both cases, the court walked through California s threestep government interest choice-of-law test, noting first that there were material differences between California s consumer protection regime and that of other states... including: (1) injury requirements, (2) deception requirements, (3) scienter, (4) reliance, (5) prefiling notice requirements, (6) statutes of limitation, (7) restrictions on consumer protection class actions, and (8) remedies. 50 The courts also held that each state had an interest in applying its own law, and finally, that other states interests would be more impaired than California s if those states laws were not applied given those states were the place of the wrong and had an interest in protecting their own consumers. C. Materiality Less than half of the opinions analyzed discussed the materiality of the allegedly misleading statements and only one court held that individual issues regarding materiality predominated, thereby defeating certification. Under California law, a plaintiff can establish materiality by showing that a reasonable man would attach importance to the existence or nonexistence of the information in determining whether to buy the prod- 45 Jones v. ConAgra Foods, Inc., N.D. Cal., No. C 12-01633, 6/13/14. 46 Ebin, 297 F.R.D. at 569-570. 47 Id at 570. 48 Id. 49 Brazil v. Dole Packaged Foods, LLC, N.D. Cal., No. 12-cv -01831, 5/30/14; Werdebaugh v. Blue Diamond Growers, N.D. Cal., No. 12-cv-2724, 5/23/14. 50 Brazil v. Dole Packaged Foods, LLC, N.D. Cal., No. 12- cv -01831, 5/30/14;Werdebaugh v. Blue Diamond Growers, N.D. Cal., No. 12-cv-2724, 5/23/14. 10-6-14 COPYRIGHT 2014 BY THE BUREAU OF NATIONAL AFFAIRS, INC. PSLR ISSN 0092-7732

5 51 Brazil (quoting In re Steroid Hormone Prod. Cases, 181 Cal. App. 4th 145, 157 (2010). 52 Brazil v. Dole Packaged Foods, LLC, N.D. Cal., No. 12- cv -01831, 5/30/14; see also Werdebaugh v. Blue Diamond Growers, N.D. Cal., No. 12-cv-2724, 5/23/14. 53 Lanovaz v. Twinings N. Am., Inc., N.D. Cal., No. C-12-02646, 4/24/14. 54 Jones v. ConAgra Foods, Inc., N.D. Cal., No. C 12-01633, 6/13/14. uct. 51 Only glossing over the materiality requirement, the Brazil and Werdebaugh courts held that the statement must be evaluated from the viewpoint of the reasonable consumer, therefore in Brazil, [w]hether Dole s label statements constitute material misrepresentations does not depend on the subjective motivations of individual purchasers, and the particular mix of motivations that compelled each class member to purchase the products in the first place is irrelevant. 52 In Lanovaz, that court similarly held that no individualized examination of materiality was necessary. Rather, plaintiff need only prove that a reasonable consumer would attach importance to Twinings antioxidant statements, or that Twinings knows or has reason to know that its consumers are likely to regard the label statements as important in making purchasing decisions. 53 On the other hand, the Northern District of California in Jones held that individual inquiries regarding materiality precluded class certification. 54 The Jones plaintiffs alleged that Con Agra s Hunt s tomato products, PAM cooking spray and Swiss Miss hot cocoa were each falsely labeled as natural. Unlike the Brazil, Werdebaugh and Lanovaz courts, the Jones court dug into the materiality inquiry and held that a lack of evidence demonstrating the impact of the challenged label statements defeated predominance. To establish that a reasonable consumer would attach significance to the challenged label statements, plaintiffs submitted an expert declaration that offered nothing more than the expert s unsupported claim that the challenged statements would be material to a reasonable consumer. 55 The expert did not explain how the challenged statements were a factor in any purchasing decisions, she did not survey any customers to assess whether the statements were material to them, and she even conceded that she personally did not pay attention to natural claims. 56 Jones and Brazil, Werdebaugh and Lanovaz appear to treat materiality inconsistently since consumers presumably also have different reasons for purchasing specific fruit products, almond milk and tea. But, even though the materiality question may turn on the court s willingness to analyze the issue, a defendant would be well-served in all instances to generate evidence sufficient to rebut materiality. Though expensive, consumer surveys demonstrating that a given statement is not material to the reasonable consumer can be persuasive and might require plaintiffs to go beyond the pleadings at the class certification stage. Conclusion In the wake of the Supreme Court s recent attention to class actions, many courts are taking a harder look at the class certification question. These 2014 food and beverage cases suggest that defendants should continue to develop a robust factual record at the class certification stage that provides a developed basis for challenging ascertainability and plaintiffs damages model. Contesting materiality has been less successful for defendants, but a successful challenge is possible in the right court, and defendants should continue to push that argument as a basis for defeating certification. Overall, the cases demonstrate that the particular court continues to have a large impact on the certification decision, and inconsistency in the class certification decisions remains the norm. There are several issues, including ascertainability, that appear ripe for potential circuit splits and possible additional guidance from the Supreme Court. 55 Id. 56 Id. PRODUCT SAFETY & LIABILITY REPORTER ISSN 0092-7732 BNA 10-6-14