In a landmark decision, the California Court of Appeal has ruled that a trial court erred when it decided that a class of 260 employees with state wage and hour law claims could present their case of class-wide liability based on the evidence relating to two named plaintiffs and 19 randomly-chosen so-called Representative Witness Group members ("RWGs").  Based on this approach, and in disregard of sworn statements from 78 class members declaring that they had not been misclassified as exempt by the employer, the trial court had awarded $15 million in damages and $18 million in attorney fees.  At the time of trial, it was not uncommon in California for class-wide liability in class actions to be proven with evidence relating to a so-called "statistical sampling" of individuals.  Duran v. U.S. Bank Nat'l Ass'n.*  The Court of Appeal held that, not only was liability not proven, but the class should not have been certified due to variations in the facts for the assorted class members.  *  Duran v. U.S. Bank Nat'l Ass'n, Case No. A125557 (Cal. Ct. App. Feb. 6, 2012).

The 260 class members in Duran were so-called Business Banking Officers and Small Business Banking Officers.  The state wage and hour claims were based on California overtime pay and meal break laws.  The trial court denied the employer's legal motions that the employees in these positions were exempt under the California version of the "administrative" and "retail sales" exemptions.  Trial of the case, which extended over many months, related only to whether the class members fit in the "outside sales" exemption.  The trial judge ruled that, in Phase I of the case, relating to liability, the evidence would be restricted to that pertaining to two plaintiffs and the 19 RWGs.  He refused to receive the sworn declarations of 78 class members who said they had not been misclassified as exempt.  Based on the class-wide liability extrapolated from the evidence regarding the two plaintiffs and 19 RWGs, the judge - in Phase II of the case - made his damages findings based on the "statistical sampling."  Throughout the trial court proceedings, U.S. Bank opposed certification of the proposed class. 

The Court of Appeal based its decision on numerous grounds.  For example, it found six flaws in the plaintiffs' "statistical sampling" technique.  Moreover, because of the trial court's refusal to consider the 78 sworn declarations, and because the facts differed for each of the class members, the due process rights of the employer had been impaired.  Further, because each individual class member's circumstances and facts were different, the trial court's certification of the class was inappropriate.  The Court of Appeal also stated that its decision was bolstered by the U.S. Supreme Court's decision in Wal-Mart Stores Inc. v. Dukes,** analyzed in our July 2011 newsletter, which reversed the certification of the largest class ever approved in an employment discrimination case.

In reaching this conclusion, the Court of Appeal did not express opposition to the use of "statistical sampling" when calculating damages or relief for a prevailing class.

Duran is the latest example of many class actions disapproved of by the courts since the Supreme Court's Dukes decision.  But it also reverses California's previously accepted approach on the use of the "statistical sampling" technique for liability.  If Duran is accepted widely, it could eradicate the model of proof widely used by the EEOC in so-called "pattern and practice" discrimination cases.