Tyson Foods, Inc. v. Bouaphakeo questions statistical averaging in FLSA collective and class actions
The Tyson Foods, Inc. v. Bouaphakeo case, which a lot of watchers believe is a test of statistical averaging used by Courts to calculate damages in class actions, is particularly interesting now given the upheaval in the Supreme Court caused by Justice Scalia’s death. A number of SCOTUS watchers now believe that the Court will split 4-4 and trial court’s Tyson Foods’ decision concerning the statistical averaging trial by formula approach will be upheld by default.
The Tyson Foods, Inc. v. Bouaphakeo case questions the use of statistics and the statistical averaging trial by formula approach to determine damages in a class-action. Specifically, the th
ree issues in Tyson Foods, Inc. are
(1) Whether differences among individual class members may be ignored and a class action certified under Federal Rule of Civil Procedure 23(b)(3), or a collective action certified under the Fair Labor Standards Act, where liability and damages will be determined with statistical techniques that presume all class members are identical to the average observed in a sample; and
(2) whether a class action may be certified or maintained under Rule 23(b)
(3), or a collective action certified or maintained under the Fair Labor Standards Act, when the class contains hundreds of members who were not injured and have no legal right to any damages.
As for background, the class of plaintiffs in the Tyson Foods case were workers at a Tyson Foods facility and they allege that they were not paid for the time that they spent putting on and taking off their equipment and required uniforms. The plaintiffs engaged to experts Dr. Ken Mericle and Dr. Leisel Fox. Dr. Mericle is a time and motion expert, who performed a study of the time that it took for the workers to don and doff their equipment. Dr. Fox took the statistical averages generated from Dr. Mericile’s time and motion study extrapolated those results to the Tyson Foods class as a whole.
At trial, plaintiffs’ economic expert estimated the damages for the Tyson Foods classwide damages using what some refer to as the statistical averaging trial by formula approach. According to Dr. Fox, the Tyson Foods Plaintiffs damages were $6,6 million for the Rule 23 class and 1.6 million for the FLSA collective class, based on her assumption that every single member worked the purported “average” donning and doffing times. Dr. Fox conceded, that over 200 Tyson Foods class members never worked more than 40 hours in a given work week, and therefore were not entitled to any relief. Tyson Food’s pretrial motions for decertification and judgment as a matter of law were denied, and the jury eventually rendered a verdict for the plaintiffs or nearly $3 million to compensate them for donning and doffing time.
Tyson Foods appealed to the Eighth Circuit, which in a 2-1 ruling affirmed. After its petition for rehearing en bane was denied by a 6-5 vote, Tyson Foods filed a petition for a writ of certiorari with The United States Supreme Court on March 19, 2015.
Prior to Justice Scolia’s death, most SCOTUS watchers believe the Court was leaning toward some type of narrow Tyson Foods ruling that would had had an impact on how statistical average and trial by formula could be used in FLSA class actions.
However now, with the upheaveal in the Supreme Court, the outcome is less decided. According to most watchers, the trial court’s decision would likely be affirmed by default due to a likely 4-4 split in the Tyson Foods case. This in essence would be a win for the plaintiffs, since the trial decision would stay intact.
The amicus briefs filed in the case are particularly enlightening concerning the arguments made by Tyson Foods and the class of plantiffs lead by Bouaphakeo.
The briefs supporting Tyson Foods tend to take two approaches. On the one hand Tyson Foods supporters argue that the use of the statistical averaging trial by formula approach violates their rights and the rule of law. For example, the brief filed by the Cato institute discusses the Mt. Clemens (1946) case and the inappropriateness of the ‘trial by formula’ approach in the Tyson Foods case.
Other briefs supporting Tyson Foods, such as one filed by DRI, argue that the use of statistical averaging to calculate damages in these types of cases is inherently a bad idea.
The DRI brief argues that the trial by formula and statistical averaging improperly allows uninjured Plaintiffs to recover and allows Plaintiffs to manipulate the amount of damages.
That is, the DRI brief argues, the use of averages can distort the overall damages picture for the plaintiffs and award damages to individuals that are in fact not injured.
Even worse, according to the DRI brief, formulas and extrapolations like those respondents relied on in the Tyson Foods v. Bouaphakeo trial by formula case tend to be opaque, inaccessible, and (what is most problematic) easily misunderstood. Because few lawyers and judges (and even fewer jurors) are trained statisticians, the use of flawed statistics in trial by formula cases leads them to draw inferences that the data cannot reasonably support the DRI brief argues.
Further, the DRI brief argues that extrapolating from an average for a sample can obscure class members’ dissimilarities, such as lack of injury in the Tyson Foods v. Bouaphakeo case.
The DRI brief points out the Plaintiffs time and motion expert found that the “average” uncompensated overtime that supported the damages calculation in this case was drawn from a sample that exhibited what respondents’ expert (Dr. Mericle and Dr. Fox) conceded to be “a lot of variation”: some workers took less than a minute to don, doff, etc., while others took five or even ten minutes to perform the same tasks.
Obviously there would also be some variation as well in the amount of time it took each worker to travel between his or her locker and the production floor, depending on where one’s locker is in relation to one’s work station. And, the trial evidence also showed, not every worker had the same equipment to don and doff, and not every worker went about those tasks the same way: some continued donning their equipment while on the line, others did not. In the end, these myriad variations were set aside entirely the DRI brief argues, and instead respondents (Bouaphakeo) treated every worker as if he or she were entitled to be paid for the “‘all-in’ average” of either 18 or 21.25 minutes (depending on which floor each worked on), and the jury presumably calculated a seven-figure plaintiffs’ award on that basis. According to the DRI Tyson Food’s brief, this result is doubly flawed.
First, the brief argues that the judgment shortchanged any workers who may have been entitled to more than this “average.” And second, the Tyson Foods v. Bouaphakeo judgment gave a windfall to those who had worked less—including more than 700 class members who had not logged enough hours to be entitled to any overtime, or whose overtime pay would have been merely de minimis (and thus nonrecoverable).
DRI argues that had they brought their cases individually, instead of as a class or collective action using the statistical averaging trial by formula approach, they should have taken nothing at all.
Amicus brief filed on behalf of the workers and Bouaphakeo, are equally interesting and enlightening. The respondent briefs take on Tyson Foods’ briefs by showing how commonplace and widely used statistics and averaging is used in all types of court cases. For example, a brief by Braun and Hummels in support of the respondents discusses why the use statistics, and the statistical averaging trial by formula approach, is correct both from a legal standpoint and a practical standpoint.
Braun and Hummels brief argues that employers throughout the country create and retain a wide range of business records to comply with or receive benefits from federal and state tax laws, employment laws, and other laws. Employers routinely rely on their own personnel records and a range of other business records to successfully defend against actions filed by their employees. And while workers commonly rely on their employers’ records to prove their employment-related claims, they usually have a far greater need to discover and proffer employers’ records as evidence, because employers ordinarily have exclusive access to the relevant records, while workers ordinarily have the burden of proving that their employers violated the law. When workers seek to vindicate their rights collectively, Braun and Hummels argue that it is vital that they can rely on their employers’ business records to prove their claims
The brief argues that parties frequently use statistical evidence to support factual findings as well. Of course, the opposing party has the opportunity to challenge these findings by contesting the methodology employed to reach the conclusions, but these challenges are common to the class. When properly compiled and described, such evidence is routinely admitted. In class actions and complex business disputes, such evidence is often the only feasible way for the fact finder to answer certain questions.
Litigants commonly rely on aggregate, statistical evidence in a variety of areas of the law—such as securities, antitrust, and commercial litigation—and the courts, juries, and parties are capable of comprehending and using this evidence in civil litigation. Braun and Hummels argue that precluding the use of this evidence would deny factfinders access to an important category of evidence frequently used in courtrooms across America, as well as in science and business every day. And because the underlying data are admissible, the alternative would be admission of the very same evidence, sliced into individual strands, in an endless series of individual trials, along with the other common evidence. The brief in support of the respondents and Bouaphakeo further argues that both fairness and efficiency mandate the approach taken by trial courts in conducting a single class action relying on the same types of evidence typically used in individual cases and class actions.
The brief further argues that statistics, statistical averaging and trial by formula have been used successfully in myriad class cases, as well as non-class cases including wage and hour, antitrust, consumer cases, employment discrimination, securities, and business and commercial litigation
Additionally a brief filed by economist and statisticians, shows how statistics and economics like those in the Tyson Foods case are a widely used tool in the courts.
Court room shenanigans, statistical averaging and trial by formula in Tyson Foods
While this is an aside, and not all that relevant to the Supreme Court decision, the shenanigans at court trial were particularly interesting in this case.
Basically near the start of the first trial, Tyson Foods moved to essentially exclude their own expert, because they did not want those findings to be presented at court. According to trial transcripts at one point the plaintiffs wanted to call the defendants expert to testify to his estimate of the donning and doffing time incurred by the plaintiffs. According to the Tyson Foods own expert, the plaintiffs had in fact incurred off the clock donning and doffing time that was unpaid by Tyson Foods. The plaintiffs time and motion expert wanted to use the defendants expert analysis and findings to sensually corroborate his findings. Tyson of course did not want that to occur.
At one point it appears according to email correspondence that the plaintiffs were actually engaging the defendants expert as their own.
