In a unanimous decision, the California Supreme Court on May 29 reversed a class action verdict for a class which was based on a flawed statistical model to determine liability and damages. Duran v. U.S. Bank National Association.
In Duran, plaintiffs brought a class action to challenge U.S. Bank’s (“USB”) classification of its business banking officers (“BBO’s”) as exempt under the “outside salesperson” exemption under California Labor Code section 1171.
At trial, USB sought to introduce evidence demonstrating that a substantial amount of the class members performed more than 50% of their work engaged in outside sales and thus fell within the “outside salesperson” exemption. The trial court barred this evidence and instead relied on testimony from its sample size of 21 class members to find USB liable to the entire class for misclassification. With respect to damages, the trial court accepted the estimate of plaintiffs’ expert (with an admitted 43.3% margin of error) that each class member on average worked approximately 11.87 hours of overtime per week.
In reversing the verdict, the Supreme Court gave a scathing review of the class action trial plan at issue, specifically, its sole reliance on evidence from a flawed statistical sample of the class and its refusal to permit litigation of relevant affirmative defenses outside of the sample.
While recognizing that statistical models may be appropriate in class actions, and that trial courts enjoy great latitude in structuring trials, the Court simply could not ignore the flawed model used by the trial court in this case.
The Court held the statistical model used to determine liability and damages failed because:
- the sample size was too small (i.e. no statistical evidence supporting a sample size of 21 for a class of 260),
- the sample was not random (i.e. small sample was plagued with selection bias); and
- the margin of error was intolerably large (i.e. 43.3% margin of error).
The Court concluded:
With no sensitivity to variability in the class, the court forced the case through trial with a flawed statistical plan that did not manage but instead ignored individual issues… ¶ This result cannot stand.”
Significantly, the Court advised that a “class action trial management plan may not foreclose the litigation of relevant affirmative defenses, even when these defenses turn on individual questions.” The Court was critical of the trial court’s reliance on a sample size of 21, while ignoring the 75 class member declarations submitted by USB showing more than half of the time of class members was dedicated to outside sales.
The court’s decision to extrapolate classwide liability from a small sample, and its refusal to permit any inquiries or evidence about the work habits of BBOs outside the sample group, deprived USB of the ability to litigate its exemption defense . . . . Instead, extrapolating findings from its small sample size and ignoring all evidence proffered to impeach these findings, the court found that the entire class was misclassified. The injustice of this result is manifest.” (Boldface emphasis added.)
In the end, the Court did not rule whether statistical sampling could ever be used in a misclassification action to prove liability to absent class members.
For remand and in future cases, the Court noted: “Assuming that sampling may be an appropriate means of proving liability or damages in a wage and hour class action, the sample relied upon must be representative and the results obtained must be sufficiently reliable to satisfy concerns of fundamental fairness.” Specifically, “a class action trial management plan may not foreclose the litigation of relevant affirmative defenses, even when these defenses turn on individual questions.”