How to Calculate Gender Pay Gap Accurately

A gender pay gap calculation can be technically correct and still tell an incomplete story. A single percentage may reflect occupational segregation, a recent acquisition, commission timing, part-time working patterns or a small senior population. For employers, the task is not simply to calculate gender pay gap figures for publication. It is to establish whether the result points to a structural reward issue, understand its commercial and employee implications, and decide what action is proportionate.

For UK employers with 250 or more employees, statutory gender pay gap reporting provides a defined methodology. Used properly, that methodology is a reliable starting point for wider pay equity analysis, workforce planning and reward governance.

Start with the statutory measures

UK gender pay gap reporting compares the pay and bonus outcomes of men and women across an organisation. It does not compare whether two individuals performing the same or comparable work receive equal pay. That is an equal pay question, requiring a different analysis of roles, job value and pay decisions.

The statutory calculation has six published measures: mean and median hourly pay gaps; mean and median bonus gaps; the proportion of women and men receiving a bonus; and the gender distribution across four hourly-pay quartiles. Each measure answers a different question. Reading only the headline hourly-pay gap is a common and costly mistake.

A positive pay gap generally means that men’s average or median hourly pay is higher than women’s. The figures should be presented clearly, including the direction of the gap, rather than assuming all audiences will interpret a percentage consistently.

Mean and median are not interchangeable

The mean is the arithmetic average. Calculate the mean hourly pay for women and the mean hourly pay for men, then apply this formula:

`(mean male hourly pay – mean female hourly pay) / mean male hourly pay × 100`

The median is the middle hourly-pay value when each gender group is arranged from lowest to highest paid. Apply the same percentage formula using median male and female pay.

Mean pay is more sensitive to a relatively small number of highly paid employees. Median pay is often a clearer indicator of the experience of the employee in the middle of each population. A large difference between the two can be revealing. For example, a modest median gap combined with a wider mean gap may indicate that representation in the most highly paid roles is a principal driver.

Neither result is inherently more meaningful. The right interpretation depends on workforce shape, seniority structure and the questions the organisation needs to answer.

How to calculate gender pay gap hourly rates

The quality of the output depends on disciplined payroll preparation. For private, voluntary and most commercial employers, the statutory snapshot date is 5 April. For public-sector bodies listed in the regulations, it is 31 March. The reporting year and snapshot date must be controlled carefully, particularly where payroll periods, acquisitions or payroll-system changes complicate the data extract.

The hourly-pay calculations use full-pay relevant employees. In broad terms, these are relevant employees who are receiving their full basic pay or paid leave on the snapshot date. Employees on reduced or nil pay because of maternity, paternity, adoption, parental or shared parental leave, sick leave, special leave or similar arrangements may be relevant employees but excluded from the hourly-pay calculations if they are not on full pay.

For each included employee, calculate an hourly rate from ordinary pay in the relevant pay period. Ordinary pay normally includes basic pay, paid allowances, shift premiums, piecework pay and certain bonus payments paid in that period. It excludes overtime pay, redundancy or termination payments, pay in lieu of leave, benefits in kind and most salary-sacrifice benefits.

The statutory hourly-rate rules require more care than simply dividing annual salary by contracted hours. The applicable pay period, pay elements, irregular hours and salary sacrifice arrangements can materially affect the result. Employers should document the calculation logic and reconcile it to payroll before producing any reported figure.

Once hourly rates are established, separate the data by gender, calculate the mean and median for each group, and calculate the percentage gaps. The same hourly-rate dataset is then used for the quartile analysis.

Calculate bonus gaps and pay quartiles separately

Bonus reporting covers a 12-month period ending on the snapshot date. The mean and median bonus gap calculations compare bonus payments received by male and female relevant employees during that period. The formula is the same as for hourly pay:

`(mean male bonus – mean female bonus) / mean male bonus × 100`

Do not confuse the bonus gap with bonus participation. A business may have a narrow average bonus gap among recipients but a material difference in the proportion of women and men who receive any bonus. Both measures should be reviewed together, alongside eligibility rules for sales incentives, annual bonus plans and long-term incentive awards.

For pay quartiles, rank all full-pay relevant employees by hourly rate from lowest to highest and split the population into four equal-sized bands, as closely as possible. Report the percentage of women and men in each band. Where the employee count does not divide exactly by four, apply the statutory allocation approach consistently and retain an audit trail.

Quartiles often provide the most accessible explanation of the headline result. A concentration of women in the lower quartiles and men in the upper quartiles points towards questions about recruitment pipelines, progression, job architecture, working patterns and senior-role succession. It does not, by itself, prove discrimination or unequal pay.

Test the data before interpreting the result

Most reporting risk begins before any formula is applied. Payroll data is designed to pay people correctly, not necessarily to answer reward and representation questions. Job titles may be inconsistent, gender records incomplete, contractual hours unreliable, and allowance codes poorly mapped.

A practical quality-assurance process should reconcile employee counts to the relevant payroll population, identify employees excluded from the hourly-pay analysis and explain why, test a sample of hourly-rate calculations, and review outliers. It should also check whether payroll codes have been treated consistently across locations, legal entities and employee groups.

For multinational groups, statutory UK reporting may cover only one part of a wider workforce. A global analysis can be valuable, but it should not be presented as equivalent to the UK statutory result unless the population, definitions and calculation basis are aligned. Local legal requirements, currencies, job markets and data privacy rules need separate consideration.

Governance matters as much as arithmetic. HR, reward, payroll, finance and legal teams should agree ownership of the data, methodology, review process and publication narrative. A board or executive sponsor should understand both the reported figures and the limitations of the analysis before sign-off.

Move from the gap to the drivers

A statutory report describes an outcome. It rarely explains causation. To identify drivers, segment the data by grade, function, location, business unit, contract type, working pattern, length of service and career stage. These cuts should be selected with care: very small populations can produce volatile percentages and may create confidentiality concerns.

The analysis should distinguish between representation effects and pay-within-role effects. If women and men in a grade are paid similarly but women are under-represented in senior grades, the priority may be progression, succession and recruitment into critical roles. If a gap persists within comparable job families or grades, examine starting-salary decisions, performance ratings, allowances, bonus outcomes and pay-review patterns.

Job evaluation and a coherent grading structure are often central to this work. Without a consistent view of role size, it is difficult to assess whether differences reflect legitimate market position, materially different responsibilities or unmanaged inconsistency. Market premiums may be justified, but they require clear criteria, review dates and appropriate approval controls.

There is no universal target or intervention. Freezing pay, applying broad-brush uplifts or changing bonus outcomes without diagnosis can introduce affordability, retention and equal-pay risks. Effective action is targeted, evidence-led and connected to the organisation’s reward strategy and workforce plan.

Build an action plan that can be implemented

A credible action plan identifies the specific decisions that will change the underlying pattern, the accountable owners, the timescale and the measures of progress. It might address recruitment shortlists, promotion criteria, pay-offer governance, bonus-plan eligibility, flexible-working access, career pathways or the design of job architecture.

The strongest plans also model cost and unintended consequences. For example, correcting an identified pay issue may be necessary, but the organisation should understand pension, bonus, payroll-tax and future pay-progression implications. Where representation is the primary driver, leadership should monitor whether hiring and internal movement are changing the talent pipeline rather than expecting an immediate shift in the headline percentage.

Publication should be transparent but measured. Explain what the figures show, acknowledge the main drivers supported by evidence, and state the actions being taken. Avoid unsupported claims that a gap is entirely explained by workforce composition. Employees, regulators and investors increasingly expect organisations to demonstrate that they have tested that explanation.

A well-run gender pay gap exercise creates more than a compliant annual report. It gives leaders a practical view of how reward, representation and career opportunity interact. The useful next step is to retain the calculation model, refresh the diagnostics through the year, and use the evidence when pay, talent and organisation decisions are still being made.