Duran v. U.S. Bank National Association Case Analysis

When determining damages in a class action case, the use of statistical sampling is an unreliable means by which to determine liability

A recent decision highlights that with respect to class action cases, requirements related to demonstrating the “commonality” of the class, even with respect to the formation of damages is still a prerequisite to not only class formation, but the determination and assessment of class damages and liability. In the case of Duran v. U.S. Bank Nat. Ass’n, 59 Cal. 4th 1, 172 Cal. Rpt. 3d 371 (2014) prospective class members sought to certify a class action against a national banking association, seeking in relevant part, employment based claims based on purported violations of wage and hour laws, specifically the misclassification or exemption of its sales force for the designed purpose of avoiding having to pay overtime pay.

The trial court letter, making a determination that the prospective class met the various legal prongs for class certification including typicality, adequacy of representation, and commonality, granted Plaintiffs’ motion for class certification and for purposes of assessing damages, effectuated a bifurcated dual phase which included a first phase focused on liability, and then a second phase to assess the issue of restitution. In both phases, the parties relied primarily upon the use of a damages based assessment, evaluation, and quantitative statistical method which involved in pertinent part the use of random statistical sampling, specifically a random sampling of twenty witnesses from the class pool, both on behalf and in opposition to such class claims, in order to establish the liability phase of the damages component post-class certification. Using the same methodology, the parties for the restitution phase, similarly utilized a random sampling from the class pool, by which to proffer a baseline assessment regarding the issue of restitution and the calculating of damages. Using such random sampling and statistical method, the trial court after completion of both phases, rendered a calculation of damages of approximately $15 million in unpaid overtime compensation and prejudgement interest.

On appeal, Defendant took substantive issue with the trial court’s order and judgement granting class certification, and the more significant issue regarding the application of a random sampling statistical method at both the bifurcated damages phases, specifically the liability and restitution phases. In asserting such appeal, the Defendant highlight the statistical unreliability of such random sampling method, including specifically citing to not only substantive flaws in the statistical sampling methodology, but also that it was an inappropriate method by which to determine and otherwise evaluate class wide damages.

The California Court of Appeals found Defendant’s argument persuasive, and not only found the random sampling methodology used at both the liability and restitution damage phases to be unreliable, but also that it was an inappropriate basis by which to evaluate and effectuate classwide damages. Specifically, the Court of Appeals determined that the margin of error from the sample group was significantly too high. The Court of Appeals also found significantly persuasive the fact that there was too many inconsistencies with respect to class formation, including among other issues, the fact that Plaintiffs’ had replaced during different phases of litigation both the class representatives and various employees within the sampling group.

Additionally, the Court of Appeals recognizing that although differences in damages from individual class members may not necessarily itself be a basis by which to decertify a class, that to the extent particular individualized damages issues in the damages phases are apparent to the point the overall class becomes unmanageable, then class certification is inappropriate. In this particular case, the Court of Appeals focused its analysis on the composition of the random sampling of class members in the damage phases, and took issue with Plaintiffs’ changing the composition of the random sampling of employees to the extent that the Court of Appeals found is essence, made the sampling non-random, unfair, non-objective, and otherwise prejudicially weighted. As such, the Court of Appeals determined that based upon the changing composition of the random statistical pool, that the margin of error was too high for such methodology to accurately portray the actual and true liability of the damages incurred pursuant to Plaintiffs’ alleged claims. As a result, the Court of Appeals decertified the class. From a class action perspective, this particular case stands a constant reminder that the Court’s will apply and evaluate the ‘commonality’ of the class through all stages of litigation, including the damages phases, and to the extent that it finds a particular class to be unmanageable, it will and has the complete discretion to decertify a class in this regard.

If you have any employment-related dispute and are considering suing your employer under PAGA, contact the Orange County Employment Lawyers at Nassiri Law Group, practicing in Orange County, Riverside, and Los Angeles. Call (949) 375-4734.