Faced with what defense lawyers hoped would be another arrow in their quiver to fight class certification, the Supreme Court refused to slam the door on the use of “representative evidence” in proving predominance under Federal Rule of Civil Procedure 23(b)(3). Instead it held that, in certain cases, “representative evidence” – evidence representative of a large group but not necessarily of any member of the group – can be used to demonstrate class-wide liability. Tyson Foods, Inc. v. Bouaphakeo, No. 14-1146. The decision, though, should not make the plaintiffs’ bar cheer too much, or the defense bar overly weepy.

The facts

Plaintiffs are employees of Tyson Foods, Inc. (Tyson), who work in the kill, cut and retrim departments of a pork processing plant in Iowa. Plaintiffs’ work requires them to wear certain protective gear, which varied depending on the department in which the person worked on a given day. A previous federal court injunction required Tyson to pay all its employees an additional four minutes a day, referred to internally as “K-code time,” to compensate them for the donning and doffing of their protective gear. In 2007 Tyson stopped uniformly paying K-code time to its employees and instead compensated some employees—but not others. Importantly, at no point did Tyson record the time that each employee spent donning and doffing their protective gear, notwithstanding its statutory obligation to do so.

Arguing that the donning and doffing of their protective gear were activities “integral and indispensable” to their regular work, and therefore compensable under the Fair Labor Standards Act of 1938 (FLSA), plaintiffs bought suit as a class action. Under FLSA and a similar state law, workers must be compensated “time-and-a-half” for all time worked in excess of 40 hours per week. But, because Tyson did not maintain records of the actual time it took employees to don and doff their protective gear, plaintiffs could not demonstrate precisely how much overtime pay a given employee was entitled to, or whether adding the donning/doffing time would mean that a given employee worked more than 40 hours in a given week.

To overcome this hurdle, plaintiffs presented “representative evidence” to establish how much time an average employee spent donning and doffing. A plaintiffs’ expert conducted 744 videotaped observations of various donning and doffing activities and then averaged the observations (hence, the representative nature of the evidence) to arrive at a time estimate specific to each department. The estimates included the time it took employees to don and doff at the beginning and end of the day, as well as before and after meal breaks. Notably, the expert’s analysis demonstrated that individual times varied greatly – as much as ten minutes. Another expert used these average time estimates to calculate the total time worked by each class member. That expert added the donning and doffing time estimates to the time each employee worked on the floor minus any K-code time paid to the employee to determine which employees, if any, worked more than 40 hours without overtime compensation. Plaintiffs’ experts concluded that Tyson failed to pay overtime to thousands of its employees, totaling approximately $6.7 million. Tyson did not challenge the reliability of either of plaintiffs’ experts’ methodology under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), nor did Tyson offer a rebuttal expert witness.

After trial the jury found Tyson liable, but awarded the class only $2.9 million in unpaid wages. According to the special verdict, the jury found that only the time spent donning and doffing at the beginning and end of the work day, as opposed to the time spent donning and doffing at breaks during the day, was compensable. No other information regarding how the jury arrived at the final damages figure was provided. Tyson moved to set aside the jury verdict, arguing, among other things, that the class certification was improper because the variation in time that it takes each employee to don and doff defeats the predominance requirement of Fed. R. Civ. P. 23(b)(3). Tyson further argued that because the jury adjusted the damages figure, it obviously did not accept the averages presented and that even a slight modification to the averages eliminated hundreds of putative class members. Thus, uninjured class members will be compensated, denying Tyson the ability to challenge individual awards. The District Court denied Tyson’s motion and both the 8th Circuit and Supreme Court affirmed the ruling.

The Court’s analysis

The Supreme Court began by noting that whether the representative evidence presented was inaccurate or unrepresentative is a question common to the class, and therefore does not defeat predominance. The central question, rather, was whether it is permissible to assume that each employee donned and doffed for the same average time calculated by plaintiffs’ experts. The Court reasoned that because such evidence could be relied upon by each class member in an individual action to establish liability, then it must be a permissible method by which to prove class-wide liability, pursuant to the Rules Enabling Act.

In deciding that representative evidence was a permissible means of establishing liability, the Court relied on its prior decision, Anderson v. Mt. Clemens Pottery Co., 328 U.S. 680 (1946). In Mt. Clemens, the Court held that representative evidence, similar to the evidence presented by plaintiffs, is permissible to fill in evidentiary gaps “when employers violate their statutory duty to keep proper records.” The Court explained that Mt. Clemens created a burden shifting when the employer failed to maintain the required employment records. The initial burden falls on the employee to prove that he has performed work for which he was not compensated. The burden then shifts to the employer to produce evidence of the precise amount of work performed or to show that the inferences drawn from the employee’s evidence is not just and reasonable. The Court found that plaintiffs met their burden and Tyson, offering no rebuttal expert witness, failed to rebut it.


The Court made clear that it was not adopting any bright line rule regarding the use of representative and statistical evidence. Ultimately, “[w]hether a representative sample may be used to establish class-wide liability will depend on the purpose for which the sample is being introduced and on the underlying cause of action.” Throughout the opinion, the Court strongly signaled that its decision was based in large part on the remedial nature of the FLSA and Tyson’s failure to keep the required time records. Accordingly, the case may not have as far-reaching consequences as some might have hoped (or feared). Class action plaintiffs will cheer the fact that the Court did not seize the opportunity to issue some broad pronouncement against using evidence that produces averages in order to prove liability and predominance. This is particularly crucial in antitrust and consumer protection class actions, which often use statistical analyses that produce averages. On the other hand, the Court’s heavy reliance on the fact that the plaintiffs here could not prove individual estimates any other way because of Tyson’s failure to adhere to its statutory obligations leaves open the question whether the use of averages or other representative evidence will be sufficient where the data is arguably more available. Similarly, the defense bar remains hopeful that a more appropriate case might serve as a vehicle to limit the use of such evidence. Nevertheless, the Court’s refusal to categorically prohibit the use of representative evidence and its vague guidance that the appropriateness of such evidence “will depend on the purpose for which the sample is being introduced and on the underlying cause of action” prevents defense lawyers from taking any real comfort as well. As always- stay tuned.