U.S. Supreme Court Affirms Class Certification Based on “Representative Evidence” of Liability and Damages
On March 22, 2016, the Supreme Court of the United States issued a 6-2 opinion in Tyson Foods, Inc. v. Bouaphakeo,1 affirming the certification of a class based on the “representative evidence” of a statistical sample used to establish liability and damages. The Court, however, declined to adopt “general rules” regarding the use of statistical evidence in class action cases, limiting its decision to the circumstances before it and stating that “[w]hether and when statistical evidence can be used to establish classwide liability will depend on the purpose for which the evidence is being introduced and on the elements of the underlying cause of action.”2 Accordingly, class action plaintiffs who seek to invoke Tyson in different circumstances, including the data security context, may well face skepticism from the courts.
Plaintiffs in Tyson had alleged that Tyson’s failure to fully compensate employees for time spent donning and doffing protective gear before and after working at a pork processing plant resulted in unpaid overtime in violation of the Fair Labor Standards Act (FLSA) and Iowa Wage Payment Collection Law.3 Because Tyson kept no records of the actual time spent donning and doffing by class members, Plaintiffs sought to establish damages through a study that measured the time spent donning and doffing for a sample of 53 employees. Plaintiffs then assumed that individual employees spent the average amount of time arrived at through the study and combined this average time with employee time sheets to arrive at an estimate of overtime pay wrongly withheld. The jury found for Plaintiffs, but was apparently only partially convinced by Plaintiffs’ experts’ calculations of $6.7 million in damages, awarding only $2.9 million with no explanation as to how they arrived at the lower figure.4 The Eighth Circuit affirmed.
Tyson’s primary argument before the Supreme Court was that a class5 cannot be properly certified where liability and damages are determined using an average obtained through a sample of the proposed class. Tyson relied on Wal-Mart v. Dukes,6 in which the Supreme Court found that a Title VII class was improperly certified where the employer had no common policy of sex discrimination and plaintiffs attempted to infer discrimination toward any given class member through a sampling of employees which revealed an estimated “percentage of claims determined to be valid.”7 The Wal-Mart Court found this “Trial by Formula” impermissible under the Rules Enabling Act because it enlarged the class’s substantive rights, allowing the class to recover where individual plaintiffs could not.8
The Tyson Court rejected Tyson’s argument, citing an early FLSA case in which evidence of a statistical sample was allowed “to fill an evidentiary gap created by the employer’s failure to keep adequate records.”9 The Court reconciled its holding with Wal-Mart, explaining that inference from sampling was improper in Wal-Mart because, without a common policy of sex discrimination, the class members were not “similarly situated.”10 Class members in Tyson, on the other hand, were similarly situated because each member “worked in the same facility, did similar work, and was paid under the same policy,” and thus in Tyson “the experiences of a subset of employees can be probative as to the experiences of all of them.”11 The underlying question both in Wal-Mart and in Tyson was “whether the sample at issue could have been used to establish liability in an individual action.”12 Given the statistical evidence in Tyson was properly admitted13 and sufficiently persuasive, then, the jury was entitled to rely on statistical evidence to establish damages just as it would if such evidence were presented in individual suits.
The Court declined to address a further argument made by Tyson – that a class cannot be properly certified where plaintiffs have not established a means of insuring that uninjured members will not share in the damages award.14 While the Court acknowledged that “the question whether uninjured class members may recover is one of great importance,” it found that the question was not yet presented because damages had not yet been disbursed.15 Chief Justice Roberts wrote a concurring opinion to argue that, given the jury’s unexplained reduction in the damage award, he saw no way to infer which class members the jury determined had unpaid overtime, and that it therefore “remains to be seen whether the jury verdict can stand.”16
Whether or not the jury verdict stands after the district court attempts to properly disburse damages, we can expect class-action plaintiffs’ lawyers to argue that Tyson’s approval of statistical evidence should apply not only in circumstances like those presented in Tyson, but also in other contexts where such lawyers have propounded statistical averages to show injury and damages, including the data security context. But rather than announcing a broad rule with regard to statistical evidence, the Tyson holding was expressly limited to its facts, and defendants will often have strong arguments that Tyson does not counsel in favor of class certification. In data breach cases, for instance, class members are not nearly so “similarly situated” as the class members in Tyson. Consumers whose information is alleged to have been exposed in a security breach have widely differing post-breach experiences relevant to the issues of injury, causation, damages and other elements of their claims, including in regard to (1) whether and to what extent they experienced fraudulent charges, and (2) whether, to what extent and for what reason they purchased credit monitoring or took other preventative measures to mitigate the risk of fraud. Proposed classes of banks that issue payment cards allegedly exposed in a breach are also varied – not only does the existence and amount of fraud experienced vary, so too does the existence, extent and appropriateness of mitigation measures such as reissuance of payment cards and implementation of fraud monitoring.17
As a result of the marked differences in the impact of data security breaches across class members, data breach plaintiffs have typically been unable to obtain class certification,18 and will likely continue to face such difficulties in the wake of Tyson.19 Proposed classes of banks and consumers are arguably more like the proposed class in Wal-Mart, where no common discriminatory policy bound them together – individual banks and consumers each have their own idiosyncratic reactions to a data breach and have nothing akin to the common workplace, work tasks, and payment policy that bound the class together in Tyson.20 Ultimately the question is whether statistical, representative evidence would suffice to establish liability and damages if a data breach class action were brought as individual actions. Given the variability within proposed classes in data breach suits, it seems likely that such evidence would not suffice.
For more information regarding the Supreme Court’s Tyson decision, or to discuss data security practices generally, please feel free to contact David McIntosh, Mark Szpak, or another member of Ropes & Gray’s leading privacy & data security team.
1 577 U.S.__, 2016 WL 1092414 (2016).
2 Id. at *8 (internal quotations omitted).
3 While Tyson paid at least some employees for the time estimated necessary to don and doff protective gear, Tyson did not compensate employees for the actual time spent donning and doffing and did not record the amount of time spent. Id. at *4.
4 Id. at *14 (Roberts, C.J. concurring).
5 Certification under the FLSA is of “collective actions” under 29 U.S.C. § 216(b) rather than classes under Rule 23, but the Court assumed, without deciding, that standards for certification of a collective action are no more stringent than they would be under Rule 23 and based its holding on an analysis of standards under Rule 23. Tyson, 2016 WL 1092414, at *7.
6 564 U.S. 338 (2011).
7 Tyson, 2016 WL 1092414, at *10.
9 Id. at *9 (citing Anderson v. Mt. Clemens Pottery Co., 328 U.S. 680 (1946)).
10 Id. at *10.
11 Id. at *11.
12 Id. at *10.
13 The Court noted that Tyson had failed to object to the admissibility of Plaintiffs’ study under Daubert, and thus “there is no basis in the record to conclude it was legal error to admit that evidence.” Id. at *11.
14 Tyson originally presented the question of whether a class may be certified if it contains “members who were not injured and have no legal right to any damages,” but later revised its argument, and the Court did not address the original question. Id. at *12.
16 Id. at *13 (Roberts, C.J., concurring).
17 See In re TJX Cos. Retail Sec. Breach Litig., 246 F.R.D. 389, 398-99 (D. Mass. 2007).
18 See id. at 398-99 (denying certification of class of banks that issued payment cards where identification of damages allegedly stemming from a data breach would necessarily be an individualized issue in the absence of “an acceptable method for determining damages in the aggregate”); Stollenwerk v. TriWest Healthcare Alliance, No. CV-03-0185-PHX-SRB, Slip Op. 7-8 (D. Ariz. June 10, 2008) (denying class certification in consumer data breach action where individual issues predominated as to causation, affirmative defenses, and damages); In re Hannaford Bros. Co. Customer Data Sec. Breach Litig., 293 F.R.D. 21, 35 (D. Me. 2013) (denying class certification because no expert testimony was presented that could demonstrate damages by statistical methods); but see Target Corp. Customer Data Sec. Breach Litig., 308 F.R.D. 482, 489-90 (D. Minn. 2015) (certifying class of banks that issued payment cards and allegedly incurred costs as a result of data breach, but reserving right to decertify class should classwide damages prove unworkable).
19 Adding to the difficulty of obtaining class certification in data security suits is the fact that many commonly alleged harms are not legally cognizable – for example: (1) increased risk of future harm, (2) the cost of mitigation measures, (3) diminished value of personal information, (4) invasion of privacy, and (5) lost benefit of the bargain. See, e.g., In re SuperValu, Inc. Customer Data Sec. Breach Litig., No. 14-MD-2586 ADM/TNL, 2016 WL 81792, at *4-8 (D. Minn. Jan. 7, 2016); In re Zappos.com, Inc. Customer Data. Sec. Litig., 108 F. Supp. 3d 949, 962 (D. Nev. 2015).
20 Moreover, in contrast with Tyson’s “failure” to keep records of employee overtime, evidence of damages in data breach suits is generally in the hands of the individual class members, such that any deficiency in that evidence cannot be the fault of the breached company.