Statistical Sampling in Wage-and-Hour Defense: Building a Duran-Compliant Framework

In PAGA and class action matters involving large employee populations, the battle over statistical methodology often determines the outcome. Duran v. U.S. Bank National Association sets the constitutional floor.

Duran v. U.S. Bank National Assn. (2014) 59 Cal.4th 1 held that trial by statistical sampling must satisfy due process — the sample must be representative, the methodology must be sound, and the defendant must have the opportunity to challenge individual claims. Applied to PAGA, this means that a plaintiff cannot simply extrapolate a violation rate from a handful of employees to the entire aggrieved population without a defensible statistical framework.

The defense opportunity is in controlling the sampling methodology: defining the sample universe (which job classifications, which locations, which time periods), the sample size (typically 25-30% for statistical significance), the selection method (random, stratified, systematic), and the analysis framework (what constitutes a 'violation' in the sample data). Each of these choices shapes the resulting violation rate.

In practice, I coordinate with forensic economists to define expert tasks: pull a random sample of employee records, analyze time punches against policy requirements, calculate violation rates per category, and extrapolate to the population with confidence intervals. The defense benefit is that actual data almost always shows lower violation rates than plaintiff's blanket allegations assume.