Guidance on statistics in class actions

Last week, the state Supreme Court issued its decision in Duran v. U.S. Bank, 2014 DJDAR 6773 (May 29, 2014), which provides additional guidance regarding the use of statistical or representative evidence in employment class actions, the standards applicable thereto, and whether evidence such as declarations or witness testimony outside of a defined sampling group should be considered. In addition, the court clarified that plaintiffs must present viable trial plans at the class certification stage explaining how individualized inquiries into merits and damages issues will be addressed and which allow defendants their due process right to present and try their affirmative defenses.

Duran involved loan officers for U.S. Bank who alleged they were misclassified as overtime-exempt pursuant to the outside sales exception, which applies to employees who spend more than 50 percent of their workday engaged in sales activities outside their home office. U.S. Bank opposed class certification and argued that the adjudication of the plaintiffs' claims and their affirmative defenses required an inherently individualized analysis of how much time any given employee spent (or was reasonably expected to spend) in their home office. The trial proceeded via a supposedly "random" sample of 20 employees chosen by the court, and U.S. Bank was not permitted to introduce any favorable evidence from employees who were not part of the sample. Based on evidence from this 20-employee sample and statistical extrapolations that were applied to the rest of the class, the trial court certified the class. This judgment was reversed by the Court of Appeal. The Supreme Court affirmed the reversal and remanded for a new trial with an instruction that the trial court may entertain a new class certification motion.

Duran clarifies several key issues regarding the certification and trial of class misclassification claims in California. Specifically, the court held that: (1) Before certification can be granted, a plaintiff must provide a viable trial plan that considers the individualized inquiries needed to address merits and damages issues and the defendant's affirmative defenses; (2) that any statistical methodology relied upon by either party must be sound, using an appropriately selected and adequately large sample, and cannot be offered at the exclusion of other types of conflicting evidence; and (3) if representative testimony or random samplings are being relied upon, a preliminary assessment must be conducted by the trial court prior to granting certification to determine how much variability there is in the evidence. Too much variability indicates that class certification should be denied.

Additional guidance can be found in Justice Goodwin Liu's concurring opinion. It states that in determining if employees were properly classified as exempt, trial courts should focus on the "realistic requirements of the job" and whether employees "customarily and regularly" spend more than half of their working time on exempt tasks. Declarations submitted by defendants on these key issues should be considered as relevant evidence on equal ground as statistical evidence, both of which must be assessed by a court for weight and credibility when deciding class certification.Â

Significantly, plaintiffs cannot use statistical evidence as an evidentiary substitute for demonstrating commonality. For example, even though the trial court found certain allegations were common to the class (i.e., that U.S. Bank uniformly classified the loan officers as exempt employees and allegedly failed to train or monitor their compliance with the exemption), these questions did not predominate the litigation because more individualized questions surrounded how each of the 260 class members actually spent their time. Even the evidence presented by plaintiffs' own expert acknowledged that up to 13 percent of the class was properly classified as exempt. The need for individualized inquiries was further pronounced when considering damages issues and the estimated amount of unpaid overtime allegedly worked by the putative class members, for which plaintiffs' expert calculated a more than 43 percent variation rate. The court noted that "[i]f the variability is too great, individual issues are more likely to swamp common ones and render the class action unmanageable."

Citing the U.S. Supreme Court's decision in Wal-Mart Stores Inc. v. Dukes, the court held that a defendant's right to present individual defenses (such as whether employees are properly classified as exempt) was protected by "both class action rules and principles of due process." The court explained that while some cases, such as mass tort claims or securities fraud actions, did not generally involve such individualized defenses, "an employer's liability for misclassification under most Labor Code exemptions will depend on employees' individual circumstances."

The court rejected the plaintiffs' attempt, relying on Bell v. Farmer's Ins. Exchange, to use their trial plan to "conflate liability and damages." The court noted that in Bell, statistical evidence was heard after liability had already been established and was used (on remand) to challenge the amount of damages only. Conversely, in Duran, significant individualized issues related to the determination of liability. Although the court did not establish a strict rule against using statistical evidence to determine liability, it warned that statistical methods may be ultimately incompatible with the nature of the plaintiffs' claims or the defendant's defenses and therefore, the resort to statistical proof may not always be appropriate. Moreover, the court made clear that "[i]f statistical evidence will comprise part of the proof on class action claims, the court should consider at the certification stage whether a trial plan has been developed ... [which] shows how individual issues can be managed at trial" and that "decertification must be ordered whenever a trial plan proves unworkable."

Based on the above, the court found that the attempted statistical sampling performed by the trial court was severely lacking. The court emphasized the importance of using a sufficiently large sample size; of ensuring that any sample is truly random; of avoiding selection bias; and of ensuring that the margin of error is not so large as to cast doubt on the results. In doing so, Duran provides welcome guidance to employment class action litigators and to trial courts, who are now instructed to conduct a more rigorous analysis of the evidence at the class certification stage and to make the necessary credibility and weight determinations at the class certification stage rather than delaying such decision until a later time.

Duran also makes clear that when a statistical model of proof is used to establish either liability or damages, "a defendant accused of misclassification must be given a chance to impeach that model or otherwise show that its liability is reduced because some plaintiffs were properly classified as exempt." To that end, both aggregate methods of proof (such as sampling or representative witness testimony) as well as individualized evidence (such as declarations) "must be considered and weighed together with the full range of evidence bearing on the ultimate issue, including the employer's job descriptions, company policies, industry customs, and testimony of supervisors or managers who monitored, evaluated, or otherwise set expectations for employees in the class."