Step 5

Unadjusted and adjusted pay gaps

Unadjusted gaps of more than 5% need to be explained by adjusting for factors influencing pay.

The core distinction

Unadjusted gap

The actual difference in average or median pay between women and men before controls. This is the governmental reporting view: it shows what the pay system has produced.

Adjusted gap

A modelled comparison after adjusting for factors that influence pay levels. It asks whether gender is still associated with pay among otherwise comparable employees.

An adjusted-gap example

An Illustrative split between the observed gap, the part associated with objective factors, and the residual gap.
Unadjusted gap 10%

The observed average difference of actual payouts before controlling for pay-relevant factors.

Explained by observed factors 8%

The share explainable with pay-relevant factors such as grade, role, location, tenure, or other gender-neutral defensible variables.

Adjusted or residual gap 2%

The remaining modelled gap after the selected factors have been taken into account.

How to use adjusted analysis

Adjusted analysis is a diagnostic step, that partially separates the observed gap into parts that may be associated with objective, gender-neutral factors and a residual part that needs closer review.

  • Start with the reported pay gap.
  • Add objective, gender-neutral pay drivers.
  • Estimate how much of the gap is associated with those drivers.
  • Review the remaining residual gap and the quality of the model.

Typical factors are HR master data or job and people data, such as grade, employee group, job family, location, tenure, qualifications, or performance where the data is robust and the factor is defensible.

Cautions

The adjusted gap is only as credible as the model, data, and assumptions behind it. A small residual gap does not prove the pay system is fair, and a large residual gap does not by itself prove discrimination.