A published gender pay gap is a statistical description of a workforce, not a verdict on equal pay. Treating it as the latter misreads both the law and the number.
Every employer in Britain with 250 or more employees is required, under the Equality Act 2010 (Gender Pay Gap Information) Regulations 2017, to publish an annual gender pay gap figure. What that figure measures is the difference between the median hourly pay of all women employed and the median hourly pay of all men employed, expressed as a percentage of men's median. A mean figure is also required alongside it. Neither calculation compares any woman's pay with any man's pay for the same work. Neither is an equal-pay audit, and the regulations do not make one mandatory.
- Gender pay gap
- the difference between female and male median (and mean) hourly pay across an entire employer; a distribution comparison
- Equal pay
- whether a woman is paid the same as a man for the same, similar or equivalent-value work; a contract-level comparison
- Median gap
- midpoint of ranked pay distributions; resistant to distortion by extreme earners at the top
- Mean gap
- arithmetic average; sensitive to high earners; a large mean–median divergence signals top-of-distribution concentration
- Bonus gap
- the same distribution comparison applied to bonus payments; typically larger than the hourly gap in sectors with grade-linked bonuses
The distinction matters because equal pay — the legal requirement, traceable in British statute to the Equal Pay Act 1970 and consolidated in the Equality Act 2010 — is about like work. It asks whether a woman doing the same or broadly similar job, or work of equivalent value, receives the same remuneration as a man. A pay-gap figure asks something different: across your entire workforce, where does the female pay distribution sit relative to the male one? A company can pass every equal-pay test in employment law and still return a large gap, because the gap reflects who occupies which jobs, not what any individual is paid for hers.
In practice the gap is driven primarily by occupational segregation and grade distribution. If an organisation's senior technical and managerial posts are disproportionately held by men and its administrative, part-time and lower-graded roles are disproportionately held by women, the aggregate distributions diverge — and the reported gap is large — without a single instance of unequal pay for equal work occurring anywhere in the business. In British computing this matters directly: the machine grades that historically channelled women into lower-paid operator and data-entry roles, and the reclassification of programming that attached formal barriers to upward movement, are structural causes of occupational segregation whose effects persist in seniority distributions today.
Mean, median and what each conceals
Requiring both mean and median is deliberate, and reading them together is more informative than either alone. The median — the midpoint of the ranked pay distribution — is relatively resistant to extreme values. A small number of very highly paid executives inflates the mean without moving the median much. Where an organisation has a predominantly male senior leadership and the mean gap is substantially larger than the median gap, the shape of that discrepancy is itself the finding: high-end concentration rather than uniform divergence.
Neither is an equal-pay audit, and the regulations do not make one mandatory.
Read together, the two figures indicate whether the gap is driven by the top of the distribution, the bottom, or both. An organisation with near-equal median pay but a wide mean gap has a concentration problem at the senior level. One with a wide median gap and a modest difference between mean and median has a more evenly distributed structural problem — the whole female distribution sits lower, not just the top.
- Equal Pay Act 1970
- first statutory basis for equal pay for like work in British law
- Equality Act 2010
- consolidated the equal-pay provisions
- Gender Pay Gap Information Regulations 2017
- mandated annual publication for employers of 250 or more; required median hourly, mean hourly and bonus gap figures
Bonus gap figures, also required, compound the picture because bonuses in many sectors are larger and more variable at senior grades. The bonus gap often exceeds the hourly gap substantially, and for the same underlying reason: fewer women in the grades where the largest bonuses are paid.
What the figure cannot be asked to do
Conflating the pay gap with equal-pay liability leads to two errors that run in opposite directions. An organisation with a small published gap can advertise it as evidence of fairness while concealing equal-pay exposures that the aggregate number has averaged away. Conversely, a large published gap does not in itself establish that any law has been broken — it may describe a structurally segregated workforce in which every individual contract is entirely lawful.
The pay-gap figure is diagnostic, not adjudicatory. Its value is precisely that it measures the distribution rather than individual contracts: it makes workforce shape visible. What it cannot do is explain that shape on its own, attribute cause, or determine remedy. For that, the number requires the structural history that sits behind it.