The Supreme Court issues a narrow decision on the use of statistical evidence in a class action



March 25, 2016

Employment Law Alert

Author(s): Stephen J. Jones, Todd R. Shinaman, Christopher M. Mason

This alert discusses what businesses need to know about the Supreme Court’s most recent ruling on class certification.

While refusing to establish a blanket rule on the use of statistical evidence in class certification decisions, the Supreme Court held this week that a representative statistical sample could be used to support certification in certain circumstances. Justice Kennedy, writing for a 6–2 majority in Tyson Foods, Inc. v. Bouaphakeo, No. 14-1146, 2016 U.S. LEXIS 2134 (U.S. Mar. 22, 2016), held that where an employer does not keep detailed records regarding employees’ overtime pay, the employees may use statistical sampling as evidence to help prove hours worked, so long as the experiences of the sample group are “probative as to the experiences of all” the plaintiffs. Id. at *25.

Factual and procedural background

The winners in Tyson Foods are hourly workers at a pork-processing facility who alleged that the company failed to compensate them for time spent “donning and doffing” personal protective equipment and walking to and from their work stations. The District Court certified the classes sought by the employees under both the Fair Labor Standards Act (“FLSA”), 29 U.S.C. § 216(b) (which has an “opt-in” mechanism) and state overtime law pursuant to Fed. R. Civ. P. 23(b)(3) (which has an “opt-out” mechanism), finding that common questions, such as whether the unpaid activities were compensable “work,” predominated over individual questions, such as the differences in the amount of time that individual employees actually spent on these activities (including many employees who spent no time on them). See 2016 U.S. LEXIS 2134, at *11-12. At trial, the plaintiffs used statistical evidence of the average donning, doffing and walking times for employees to prove liability and damages. The jury returned a verdict in favor of the employees, and the district court entered a $5.8 million judgment against Tyson Foods.

On appeal, the company contended that class certification was improper because employees’ individual routines varied, and the litigation therefore could not generate common answers apt to drive resolution as required under Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541 (U.S. 2011). Tyson Foods also argued that the trial court incorrectly permitted the jury to infer liability and damages from statistical evidence in violation of the Supreme Court’s prohibition on “trials by formula” in Dukes and the use of damages models that ignore the basis of defendant’s alleged liability as set forth in Comcast v. Behrend, 133 S. Ct. 1426 (2013). Finally, the company argued that class certification had been inappropriate because some employees did not work any overtime and were thus not entitled to any damages.

A divided panel (2 to 1) of the Eighth Circuit rejected these arguments. The majority recognized that individual plaintiffs varied in their donning and doffing routines, but held that class certification was appropriate because Tyson Foods had a compensation policy for such activities, and the employee class members all worked at the same plant using similar equipment. After the Eighth Circuit refused to rehear this decision en banc, the company sought certiorari from the Supreme Court.

The Supreme Court’s decision

The Supreme Court granted certiorari to decide two questions about certification of a class under Fed. R. Civ. P. 23 or a collective action under Section 216(b) of the FLSA. First, the Court agreed to decide whether such certification would be proper where liability and damages will be determined with statistical techniques that presume all class members are identical to the average observed in a sample. Second, the Court agreed to decide whether a class may be certified “if it contains ‘members who were not injured and have no legal right to any damages.’” 2016 U.S. LEXIS 2134, at *27.

Tyson Foods argued to the Court that an approach that permits individuals to recover without proving any personal injury would deprive employers of the ability to litigate any defenses to individual claims. Relying in large part on Wal-Mart Stores, Inc. v. Dukes, the company urged the Court to adopt a broad rule forbidding representative evidence of the nature used against it at trial. In particular, it argued that the use of statistical evidence, such as the average number of minutes an employee took to don and doff protective equipment, amounted to the “trial by formula” specifically disavowed in the Wal-Mart Stores, Inc. v. Dukes decision.

In a narrow holding, the Court rejected these arguments, deciding 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.’” 2016 U.S. LEXIS 2134, at *20 (quoting Erica P. John Fund, Inc. v. Halliburton Co., 563 U. S. 804, 809 (2011)). In Tyson Foods, the plaintiffs’ statistical analysis was permissible, according to the Court, in large part because it was introduced by the employees “to fill an evidentiary gap” created by the company’s failure to keep adequate time records. 2016 U.S. LEXIS 2134, at *22. If the individual class members had proceeded with separate lawsuits, the Court reasoned, each employee could have offered the statistical study to try to prove the hours he or she worked; therefore, it was permissible for them to do so collectively in a class action as well. Id. at *21-22. As Justice Kennedy wrote for the majority, “[b]ecause a representative sample may be the only feasible way to establish liability, it cannot be deemed improper merely because the claim is brought on behalf of a class.” Id. at *20-21. For much the same reason, the majority also rejected Tyson Food’s argument that the decision below conflicted with Wal-Mart Stores, Inc. v. Dukes. See id. at *23 (that case “does not stand for the broad proposition that a representative sample is an impermissible means of establishing classwide liability.”). Justice Kennedy cautioned, however, that the Court’s decision against Tyson Foods also was not one establishing a general rule upholding the use of representative and statistical evidence in class actions. Instead, the admissibility of such statistical evidence “will depend on the purpose for which the sample is being introduced and on the underlying cause of action.” Id. at *27. While “[i]n FLSA actions, inferring the hours an employee has worked from a study . . . has been permitted by the court so long as the study is otherwise admissible,” and while Justice Kennedy obviously considered the state overtime wage claims in the Tyson Foods case sufficiently similar to an “FLSA action” to apply the same rule, “[t]he fairness and utility of statistical methods in contexts other than those presented here will depend on facts and circumstances particular to those cases.” Id. at *27.

As to the second question on which it granted certiorari, “whether a class may be certified if it contains ‘members who were not injured and have no legal right to any damages,’” id., Tyson Foods, in effect, dropped it. The Court therefore did not address it. See id. at *27-28. But in response to a slightly different question Tyson Foods did try to ask instead, whether plaintiffs must “demonstrate . . . that there is some mechanism to identify . . . uninjured class members prior to judgment and ensure that” they do not inflate the damages awarded and cannot recover any for themselves, id., at *28, Justice Kennedy concluded that the question was “not . . . yet fairly presented by this case, because the damages award has not yet been disbursed, nor does the record indicate how it will be disbursed.” Id. at *28-29. Tyson Foods could decide to challenge any proposed method of allocation when the case returns to the trial court.

What’s next?

Plaintiffs’ class action attorneys are certain to make much ado of the Supreme Court’s approval of statistical averaging of employee donning and doffing time in this case. Significantly, however, the majority declined to establish any broad, categorical rules governing when representative and statistical evidence may be received to establish classwide issues. The Court cautioned instead that the admissibility of such evidence depends on the elements of the underlying claim, the purpose for the evidence’s introduction and its reliability. Thus, while use of statistics from a sample of employees was upheld in this limited instance, the decision should not have broad precedential implications in most employment class actions, including wage-hour cases, much less in most class action cases generally.

For example, there are strong arguments that in wage-hour litigation, the decision should be limited to certain donning and doffing-type cases in which time comparisons to other employees performing the same tasks may be probative to the amount of time the individual spent on those tasks. In other types of wage-hour cases, such as off-the-clock and misclassification cases, the differences in primary job duties and time spent on preliminary, postliminary and meal-break activities will often vary too much for the use of representative testimony to have any probative value. In non-wage-hour class action litigation, there will probably be even stronger arguments against expansion of Justice Kennedy’s opinion, given his heavy reliance on FLSA precedent. Finally, in all kinds of cases, defendants opposing class certification or liability based on statistical evidence will undoubtedly (and correctly) note that the majority also emphasized that Tyson Foods had not challenged the expert methodology used to develop the statistical model offered against it (and offered no rebuttal expert, either), meaning there was no Daubert error committed by the district court in admitting the evidence. See id. at *14 (citing Daubert v. Merrell Dow Pharms., Inc., 509 U. S. 579 (1993)). The Tyson Foods case should encourage such defendants to give additional early consideration to using Daubert challenges to any purported statistical expert testimony and reports.

The foregoing has been prepared for the general information of clients and friends of the firm. It is not meant to provide legal advice with respect to any specific matter and should not be acted upon without professional counsel. If you have any questions or require any further information regarding these or other related matters, please contact your regular Nixon Peabody LLP representative. This material may be considered advertising under certain rules of professional conduct.

Back to top