Recent U.S. Supreme Court decisions have opened the door to use of statistical evidence in proving age discrimination. Most of these cases have focused on corporate reductions in force that allegedly affect older employees in numbers beyond those predicted by a random sampling of employees. A January case from the Third Circuit Court of Appeals demonstrates how failure to review the impact of the RIF on older employees resulted in an expensive age discrimination verdict.

In Marcus v. PQ Corp., after acquiring the company, the new owners laid off thirty employees, including eight in its R&D department. Every employee in that department over the age of 55 was terminated. No employee under 55 was affected. Several of the terminated employees sued, claiming age discrimination. The jury agreed, awarding over $5 million in damages before that award was reduced by the District Court. The employer appealed, claiming that statistics regarding the age of employees affected by the RIF did not overcome its business reasons for the selections.

The Third Circuit disagreed, affirming the jury verdict. The court noted that the plaintiffs introduced expert testimony stating that the age distribution of the terminated employees was statistically unlikely to have resulted, even given those business reasons. When combined with circumstantial evidence indicating an intent to get rid of older employees, this was sufficient for the jury to find that age was the motivating factor for the selections.

Employers engaging in reductions in force should review initial selection criteria for possible bias. Preliminary selections for layoff should be reviewed to determine if employees in protected classifications are affected by the choices in statistically significant numbers. If so, the employer should document and be prepared to demonstrate the legitimate, non-discriminatory reasons for the selections. If such reasons are not persuasive, the employer should revisit the selection process to ensure that bias has not entered into the picture.