How to Run a Pay Equity Regression with Explanatory Factors

A raw pay gap between two groups is a starting point, not an answer. Some of that gap may come from legitimate factors like experience, education, tenure, or performance. Regression analysis separates the explained part of the gap from the unexplained part, and the unexplained part is where your risk is.

In this walkthrough, you’ll see how to:

  • Choose a regression model in EquityPath’s Build the Model panel
  • Select a measure of pay, such as total compensation, and group results by state, job category or country
  • Set the protected and advantaged classes you’re comparing
  • Add explanatory variables like experience, education and performance
  • Read the results table, including the estimated gap, p-value and gap as a percentage of pay
  • Use the waterfall chart to see how much of the gap each factor explains, and how much remains
  • Spot high-risk groups at a glance with the treemap and risk meter

Why it matters

Regulators, attorneys and the EU Pay Transparency Directive all ask the same question: is the gap explained by something other than gender, race or another protected characteristic? Regression with explanatory factors answers that question in a way that holds up to review. EquityPath also offers Glass Ceiling and Oaxaca-Blinder regression, plus non-parametric tests like Mann-Whitney, so you can match the model to the question you need answered.

EquityPath lets you do more and do it better

Fast drill-down insight, flexible adjustment tools, and reporting crafted for leadership and legal scrutiny

Categories

Share this article

Related articles

Schedule your personalized demo

Fast drill-down insight, flexible adjustment tools, and reporting crafted for leadership and legal scrutiny

This field is for validation purposes and should be left unchanged.