Productivity
Audit pay equity
Where gaps exist after controlling for role and level, and what closing costs.
Fill it in
With role, level, tenure and pay.
Role, level, location and tenure, before concluding anything.
And say what it buys.
Your prompt
Audit pay equity. Control for role, level, location and tenure BEFORE concluding anything. A raw gap is mostly a distribution problem, and reporting it as a pay gap is both wrong and discrediting. The controlled gap is the finding, and where it survives the controls it is much harder to explain away. Report the sample size per comparison. A 9% gap between three people and four is not a finding, and presenting it as one destroys the credibility of the real ones. Look at the distribution across levels too. Equal pay within a level, with one group concentrated at the bottom, is a different problem with the same cause. Cost the remediation and phase it. Anonymise the output.
Use Audit pay equityOpens with everything above already filled in.
Why this works
A raw pay gap is mostly a distribution problem, and reporting it as a pay gap is both wrong and discrediting. This controls for role, level, location and tenure first, reports sample sizes, and separately looks at how groups are distributed across levels.