Evaluating the Legal Defensibility of Regression Models Designed to Investigate and Remediate Pay Disparities: Lessons from Rudebusch v. Hughes

By Dan A. Biddle, Ph.D.

“…some employers have been concerned with the potential of “reverse discrimination” or Equal Protection-based lawsuits that may arise after making pay adjustments to a given group. Faced with liability from potential class action suits or audits from “classical” pay disparity issues on one side and Equal Protection cases on the other, some employers view themselves in a “pickle.” Indeed, some employers that act in good faith to conduct (and later act upon) proactive pay equity studies may find themselves open to an Equal Protection suit from the group that received no pay changes.

When the courts evaluate such circumstances, some have extracted three criteria from the U.S. Supreme Court case, Johnson v. Transportation Agency[i]: (1) evaluating whether a “manifest imbalance” exists (which typically includes statistical significance as a minimum threshold); (2) evaluating whether the rights of individuals who were not part of the remedial pay were “unnecessarily trammeled”; and (3) evaluating whether the remedial pay adjustments were along the lines necessary to “attain a balance.”[ii] The Johnson decision also dictates that the burden shifts to the employer to provide a nondiscriminatory rationale for its decision after the plaintiff demonstrates that sex was taken into account in an employer’s employment decision. The employer can establish such a rationale by pointing to an existing affirmative action plan.

When applying the Johnson framework to an Equal Protection pay case, the Ninth Circuit in Rudebusch pointed out that previous Appellate Courts (both the Fourth and Eighth Circuits) have adopted the Johnson approach in their analysis of pay equity claims. While agreeing that the Johnson case provided the proper framework for a pay equity case, they noted that there were “some significant conceptual differences between affirmative action in the promotional context and remedial measures used to cure pay inequity (p. 25)” because the Johnson case involved voluntary affirmative action efforts for promotional decisions—which is quite different than pay equity.
There are two major lessons that can be learned from the Rudebusch case. First, before making pay adjustments to a group, be sure the regression model clearly shows that the gender or race variable is statistically significant after controlling for job qualification factors. Second, make sure that the regression model is sound, accurate, and reliable..."

To view the full article, please click here:

http://www.biddleaap.com/docs/Rudebusch_Lessons.pdf