Step 2

Reporting metrics under the Directive

For each required metric, start with the same questions:
what is being measured, which pay components are in scope, what data is needed, and what can make the result misleading?

The formulas below use the Directive's approach expressing the difference between female and male employees as a percentage of the male pay level.

  • A positive number usually means women are paid less than men on the measure.
  • A negative number means women are paid more on the measure.

a. Mean gender pay gap

The actual average difference between women and men across the company, expressed as a percentage of average male pay.

(average male pay - average female pay) / average male pay x 100

Data needed

  • Gender
  • Pay level
  • Gross annual pay
  • Gross hourly pay
  • Employee population rules

Worked example

If average male gross hourly pay is € 30 and average female gross hourly pay is € 26.40, the gap is (30 - 26.40) / 30 x 100 = 12%.

b. Mean gap in complementary or variable pay

The actual average difference between women and men in pay components beyond ordinary wages or salaries. Various pay components can be used.

(average male variable pay - average female variable pay) / average male variable pay x 100

Data needed

  • Gender
  • Bonus
  • Commission
  • Allowances
  • Benefits in kind where in scope
  • Eligibility population rules

Worked example

If men receiving variable pay average € 7,500 and women average € 5,700, the mean variable pay gap is 24%.

c. Median gender pay gap

The difference between the median-paid woman and median-paid man, expressed as a percentage of median male pay.

(median male pay - median female pay) / median male pay x 100

Data needed

  • Gender
  • Pay level
  • Gross annual pay
  • Gross hourly pay
  • Employee population rules

Worked example

If median male hourly pay is € 28 and median female hourly pay is € 25.20, the median gap is 10%.

d. Median gap in complementary or variable pay

The median-point difference in bonus, commission, allowances, or other variable/complementary components.

(median male variable pay - median female variable pay) / median male variable pay x 100

Data needed

  • Gender
  • Bonus
  • Commission
  • Allowances
  • Benefits in kind where in scope
  • Eligibility rules
  • Zero-payment approach

Worked example

If median male variable pay is € 3,000 and median female variable pay is € 2,500, the median variable pay gap is 16.7%.

e. Proportion receiving complementary or variable pay

The share of women and men who receive variable or complementary pay components.

recipients of a gender / total employees of that gender x 100

Data needed

  • Gender
  • Employee population rules
  • Eligibility criteria of pay components

Worked example

If 80 of 200 women and 150 of 250 men received variable pay, the recipient rates are 40% for women and 60% for men.

f. Proportion in each pay quartile

Rank employees by pay, split into four equal groups, then report the gender composition in each quartile.

65% women 35% men

55% women 45% men

45% women 55% men

30% women 70% men

Illustrative composition: Q1 65 percent women and 35 percent men; Q2 55 percent women and 45 percent men; Q3 45 percent women and 55 percent men; Q4 30 percent women and 70 percent men. 

employees of a gender in quartile / all employees in quartile x 100

Data needed

  • Gender
  • Pay level
  • Quartile allocation rule
  • Employee population rules

Worked example

In a 400-person company, each quartile has 100 employees. If the top quartile has 30 women and 70 men, women make up 30% of the top pay quartile.

g. Pay gap by same or equal-value employee groups

The gender pay gap must be calculated within groups of employees performing the same work or work of equal value, split by ordinary basic wage/salary and complementary or variable components.

(average male group pay - average female group pay) / average male group pay x 100

Data needed

  • Gender
  • Employee group
  • Job evaluation criteria
  • Wage or salary
  • Variable pay
  • Group population
  • Complementary pay components
  • Other explanatory variables

Worked example

In a grade 12 HR role, average male base pay is € 60,000 and average female base pay is € 57,000. The group base pay gap is 5.0%.
For females with below average base-pay the pay gap is even higher.

Common pitfalls and quality checks

Most review points apply across all seven metrics. Treat them as one control layer after the calculations are complete, then add metric-specific checks where the data structure requires it.

Common pitfalls

  • Population drift. Use the same employee population rules.
  • Pay-component mismatch. Do not mix individuals with different eligibility for e.g. variable pay or allowances without neutralizing for them.
  • Missing versus zero values. For variable or complementary pay metrics, separate people with no payment from people with missing data.
  • Variables with high correlation. Variables with high correlation (e.g. age and tenure) blur results, as they could be used as each others proxy.
  • Working-time effects. Annualize pay data to 1 FTE and break down from there.
  • Over-reading the headline number. A low company-level gap is just that, a headline. Dive deeper into grades, job families, locations, or groupings.
  • Weak equal-value grouping. For metric g, job title alone is not enough.

Quality checks

  • Reconcile counts. Tie employee counts back to HR and payroll extracts.
  • High statistical significance. A statistical significance of 90-95% should be achieved in the regression.
  • Document the metric dictionary. Record the pay period, included pay components, exclusions, currency, zero handling, and calculation convention.
  • Inspect distributions. Look for outliers, small groups, tied values, and concentration of women or men in particular grades or quartiles.
  • Test explanations. If pay gaps appear, check whether explanations are objective, gender-neutral, evidenced, and consistently applied.
  • Keep the audit trail. Preserve methodology, employee-representative consultation, and changes made after review.