Duran v. U.S. Bank (A125557 & A126827) was decided by the First Appellate District on February 6, 2012.  It ordered the Alameda California trial court to decertify the class action case and had strong criticism of the statistical evidence the trial court used to find U.S. Bank liable at trial.

Judge Robert Freedman certified a class of 260 sales representatives of the bank.  The court order defined the class as: "all current and former California-based salaried employees with the title -small-business banker (SBBs) and/or -business banking officers (BBOs) employed by defendant any time between December 26, 1997 and April 28, 2005."

U.S. Bank had designated the class members as outside sales representatives.  Accordingly, the bank classified them exempt and did not pay them overtime.  Duran claimed that the sales representatives worked less than 50% of the time outside the office and were, thus, not exempt under the outside sales exemption.

Before certification, the bank offered deposition testimony and 75 declarations of class members who said that they worked more than 50% of the time outside the office.  Plaintiff offered 34 declarations from class members who said that they performed sales activities outside the office less than 50% of the time.  The trial court certified the class.

At trial, the court limited the trial testimony to 20 class members who were randomly selected.  Those class members who had opted out were not allowed to testify.  Additionally, class members so selected were given the opportunity to opt-out after they had been selected to testify.  U.S. Bank argued that those procedures skewed the evidence because opt-outs did not believe U.S. Bank to be liable.

The appellate court said that the trial court's reliance on 20 class members to find liability for 260 class members was not statistically justified.  In addition, not allowing opt-outs to testify skewed the statistical results even further.  Accordingly, the court found that the statistical sample of evidence to prove liability violated U.S. Bank's due process rights.  Finally, the appellate court ordered the trial court to decertify the case.

Contradictory facts concerning liability proffered by the defendants do not necessarily deny class certification.  (See, Jaimez v. Daiohs, (2010) 181 Cal.App.4th 1286.)  However, maybe a significant volume of contradictory facts can be used to deny class certification.  At the very least, they may make it nearly impossible to try a class action case based on statistical sample of evidence because the defendant appears to have a due process right to put on large volumes of evidence that contradict the evidence plaintiffs proffer.

Plaintiff's counsel may still find ways to avoid the problem that counsel in Duran faced:

1) A different class definition might have saved the case.  Maybe the following definition would have worked: "all current and former California-based salaried employees with the title -small-business banker (SBBs) and/or -business banking officers (BBOs) employed by defendant any time between December 26, 1997 and April 28, 2005, who worked less than 50% of the time outside the office."  Then, all those who worked more than 50% of the time outside of the office would not be in the class.

2) File the case under the FLSA and California law and only represent those who elect to opt-in under the early opt-in rule of the FLSA.

3) Get the names of the class in discovery and file a mass action with those class members who sign individual contracts.

4) Allow the defendant to put on a statistically significant amount of contradictory evidence on liability.

The bottom line appears to be that the use of a statistical sample of evidence in class actions on the issue of liability may violate a defendant's due process rights.  Using statistics strategically and fairly may avoid the result in Duran.