The Pay Equity Audit at Halcyon Software
Halcyon Software, a 1,100-employee SaaS company, commissioned a pay equity audit after a media inquiry. The analysis found an 8% unadjusted median pay gap between men and women. After controlling for job level, function, and tenure, the adjusted gap falls to 2.5%, which is not statistically significant at the 95% level. However, the audit found that women are concentrated in Customer Success and Support (68% female) while men dominate Engineering and Sales (74% male), and that Engineering and Sales carry higher pay bands and larger bonus targets. Two female directors in Customer Success have flagged that their roles were re-benchmarked downward 18 months ago. The CEO wants to publish the adjusted 2.5% figure as evidence of fairness. Legal counsel is cautious. The board's ESG committee wants a public commitment.
1. Explain what the adjusted and unadjusted figures each measure, and evaluate the CEO's plan to publish only the adjusted figure.
Read the worked answer
The two figures answer different questions and neither alone tells the full story.
Unadjusted 8% median gap: Measures the raw difference in what men and women earn across the whole workforce. It captures composition - who holds which jobs - as well as any pay differences within jobs. It is the number most stakeholders and regulators care about because it reflects lived outcomes.
Adjusted 2.5% gap: Measures whether men and women in comparable roles (same level, function, tenure) are paid differently. A non-significant result is genuinely good news: it suggests no systematic within-role discrimination in base pay.
Why publishing only the adjusted figure is risky:
- It is technically defensible but rhetorically evasive. Observers will ask why the headline number is 8% if the company is fair, and the answer - occupational segregation - is itself a fairness problem the company has not addressed.
- It invites accusations of cherry-picking and may damage trust more than transparency would.
- The re-benchmarked Customer Success roles are a live within-role issue that the adjusted model may have masked if the re-benchmarking was applied consistently to both genders - consistency does not make a downward re-benchmark of a female-dominated role equitable.
Recommendation: Publish both figures with a clear explanation, and pair them with a concrete action plan on segregation and the re-benchmarking review. Transparency plus remedy is more credible than a single favorable statistic.
2. The two Customer Success directors allege their roles were unfairly downgraded. How should HR investigate, and what legal and reward considerations apply?
Read the worked answer
This is the highest-risk element because it alleges a specific, recent, within-role action rather than a statistical pattern.
Investigation steps:
- Document the re-benchmarking decision: Who initiated it, what job-evaluation methodology was used, what comparator roles were chosen, and what evidence supported the new band.
- Test consistency: Were comparable male-dominated roles re-benchmarked in the same period? If only Customer Success was downgraded, the burden shifts to the company to explain why.
- Review the job evaluation itself: Check whether the evaluation captured the full scope of the role (e.g., technical complexity, revenue influence) or undervalued relational and client-facing work - a common source of systemic undervaluation.
- Interview the directors and any others affected, and preserve records.
Legal considerations: Depending on jurisdiction, equal-pay or pay-transparency laws may apply, and a downgrade concentrated in a female-dominated function can support an inference of indirect discrimination even absent intent. Counsel should be involved from the outset, and the investigation should be privileged where possible.
Reward considerations: If the downgrade is unsupported, remediate through band restoration and back pay rather than one-off bonuses, which do not fix the structural inequity. If it is supported, document the rationale thoroughly and communicate it to affected staff to avoid a grievance escalating. Either way, the outcome should feed the audit's action plan and be reported to the board committee.