Last week, the Supreme Court issued its opinion in Tyson Foods v. Bouaphakeo, No. 14-1146 (March 22, 2016), a closely watched case out of the 8th Circuit that presented two meaty issues relevant to class action practitioners: 1) whether a plaintiff can use statistical sampling to establish class-wide liability, aka “Trial-by-Formula”; and 2) whether a certified class can include uninjured claimants.  In its opinion affirming both class certification and the trial verdict, the Court did not make any broad pronouncements on the use of statistical evidence in classwide proceedings, but instead took a more measured approach on when such evidence may be used.  Our thoughts on the opinion are below.

  • On the use of statistical evidence, Justice Kennedy, writing for the majority observed that “[a] categorical exclusion of that sort, however, would make little sense. A representative or statistical sample, like all evidence, is a means to establish or defend against liability.”  Instead, whether 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.”
    • The Court explained, that in a collective action brought under the FLSA, “when employers violate their statutory duty to keep proper records, and employees thereby have no way to establish the time spent doing uncompensated work,” a plaintiff could “introduce a representative sample to fill the evidentiary gap created by the employer’s failure to keep adequate records.”
    • Thus, the issue in Tyson Foods was whether the statistical evidence proffered by plaintiffs (a time study performed by their expert showing the average time spent donning and doffing equipment needed to perform their jobs) would be permissible as a way of proving liability because Tyson had not kept adequate records of the time spent by employees in that activity.
  • The Court distinguished the admonition against Trial-by-Formula in Wal-Mart v. Dukes based on two reasons: 1) the proposed class in Dukes failed to meet the commonality requirement under Rule 23(a) because there was no evidence of a common policy of discrimination that affected each employee; 2) the proposed statistical evidence could not have been used in individual suits brought by different class members.  Specifically, because the class in Tyson Foods was limited to single facility (and even more narrowly, two departments in that facility), the Court said that even if the employees had proceeded with 3,000+ individual suits, each employee could have used the expert’s time study to prove the hours he or she worked, because of the similar nature of their work, and all were subjected to the same policy that did not compensate them for their donning and doffing activity.  By extension, if the employees [in Dukes] had brought 1 1∕2 million individual suits, there would be little or no role for representative evidence because “the experiences of the employees in Wal-Mart bore little relationship to one another.”
  • As an aside, the Court chided the defendant for failing to raise a Daubert challenge and stated there was “no basis in the record” to review whether the admission of the proffered statistical evidence was “legal error.”  This is a bit surprising, as the Supreme Court (and a majority of Circuit courts) has never affirmed the applicability of Daubert during class certification proceedings, so it’s a helpful reminder to always file such a motion if you have a basis to do so.
  • Chief Justice Roberts’s concurring opinion identifies the practical problem with relying on statistical evidence: if the jury’s verdict does not exactly conform to the statistical expert’s conclusions, or does not specify which (or how many) class members were injured, how does a court later figure out who is entitled to damages without running afoul of the principle that uninjured class members are not entitled to any form of compensation.  Chief Justice Roberts observed that logically, the jury must have determined that “at least one” of the statistical expert’s two averages to be “too high” but because the jury never made a finding (or it wasn’t on the verdict form) it would be nearly impossible for a court to later determine which subset of employees was not properly credited for their donning and doffing time.

As always, SCOTUS Blog has full coverage of the case, including all briefs and an analysis of the oral argument and opinion.  Also, in the interests of full disclosure, our firm submitted an amici brief on behalf of the National Association of Manufacturers, American Tort Reform Association, and other trade organizations.