Proving Class Action Cases at Trial Using Statistics

A class, or group, of loan officers sued for overtime, alleging they were unfairly classified as exempt from it as outside salespeople. (This rule applies to people who spent more than half their time selling outside the office.) The trial used a sampling of statistical information and the testimony of 21 employees, chosen randomly, and then extrapolated those results to stand for the hundreds of people involved, rather than having each individual in the class have to testify.

The Defendant, U.S. Bank National Association, was not allowed to introduced evidence about anyone other than the 21 people who were part of the sample. After the court found the entire group had been misclassified – and improperly denied overtime – the Defendant appealed the judgment, which was $15 million dollars.

The Defendant appealed all the way to the California Supreme Court, and argued the results were unfair because the court’s improper use of the sampling prevented them from showing some class members weren’t entitled to anything, presumably because they spent the majority of their time selling. Generally speaking, just because there are differences in damages doesn’t mean a class can’t proceed as a group, but if the court relies on statistical sampling, an expert in that sampling must be used. Also, Defendant has to be allowed a chance to prove the sampling is unfair, or that the damages don’t apply to some members. On May 29, 2014, the Supreme Court ruled that the specific facts of this case meant the decision had to be overturned to give Defendant a fair shot at disputing both the sampling method, and some individual’s entitlement to damages.

There was no precedent for using sampling to establish liability in a case of this type, according to the decision.

This case appears to be fact specific in that the court proceeded with its own plan, instead of those suggested by Plaintiff and Defendant, although Plaintiff’s expert agreed the court’s method was statistically sound. Essentially, the Supreme Court ruled that while the court has discretion in a lot of areas, they cannot exercise that discretion in ways that go against science, such as statistics. The Court’s method in this case appeared to unfairly favor the Plaintiff, and resulted in a high margin of error.   Even the Plaintiff in this case agreed that in the damages phase of the trial, Defendant should have the right to challenge individual damages, a right denied them in this case. Essentially, the estimated damages couldn’t be trusted because the trial itself was determined to be unfair. The issue of statistical sampling and its use in cases of this type is not resolved by this case, which was remanded for a new trial. See Duran v. U.S. Bank National Assn. (SC S200923 5/29/14)

2017-12-13T21:46:45+00:00 July 23rd, 2014|Class Action, General Labor Law|