Gender Pay Gap KPI

The Gender Pay Gap KPI measures the disparity in earnings between genders, serving as a critical performance indicator for organizational equity.

Addressing this gap influences employee satisfaction, retention rates, and overall company reputation.

Companies that actively track and improve this metric often see enhanced operational efficiency and better financial health.

By fostering a more equitable workplace, organizations can also improve their strategic alignment with broader societal expectations.

This KPI is essential for data-driven decision-making and can significantly impact recruitment and talent management strategies.

How Gender Pay Gap Connects to Your Strategy

Gender Pay Gap belongs to five KPI groups in the KPI Depot library, and the five read it in very different ways.

Its strongest placement is Diversity, Equity, and Inclusion (DEI), where it ranks ninth. That group opens with Employee Diversity Ratio and Leadership Diversity Ratio, then Diversity in Candidate Interview Selection, Diversity Hiring Goal Achievement, Minority Talent Acquisition Rate, and Diversity Talent Pipeline Strength, with Employee Retention Rate and Promotion Rate Disparity rounding out the headline set. In that company Gender Pay Gap is the settlement metric. Everything ahead of it measures access: who applies, who gets interviewed, who gets hired, who is ready for the next level. Gender Pay Gap measures what those access decisions eventually pay out, which is why it lands where it does in the ordering rather than at the top.

In Compensation and Benefits it is a supporting metric at twenty-fourth, well behind a financial-perspective front end of Total Compensation Cost, Compensation and Benefits as Percentage of Revenue, and Benefits Cost As a Percentage of Payroll. The neighbour that matters most there is Pay Equity Ratio at sixth, and the two are not the same measurement. Pay Equity Ratio compares people doing comparable work; Gender Pay Gap as defined by this KPI's formula compares averages across the whole population regardless of role. An organization can hold a clean Pay Equity Ratio and still carry a wide Gender Pay Gap, because the gap is mostly a statement about who occupies which grade. Customers who report only the first number and call the issue closed usually get corrected in public.

The remaining three placements put it further down. In Environmental, Social, Governance (ESG) it sits thirty-first, behind a group whose headline metrics are environmental: Carbon Footprint Reduction, Renewable Energy Consumption, Energy Intensity Reduction, Water Usage Intensity, Waste Diversion Rate. It earns its place there as one of the few social-pillar figures that external raters can verify against filed disclosures. In ISO 26000 (IEC 26000) it ranks forty-eighth among metrics led by Employee Satisfaction Index, Diversity and Inclusion Index, Occupational Health and Safety Incidents, and Employee Turnover Rate, where it supports the labour-practice and human-rights sections rather than driving them. In Workforce Planning it is sixty-seventh, far behind Headcount, Turnover Rate, Vacancy Rate, Time to Fill, and Cost per Hire. That distance is honest: workforce planning runs on supply and cost, and the pay gap is a constraint on how you staff, not a dial you turn during a hiring cycle.

The canonical placement is the learning and growth perspective, and the placement is right. This is a lagging figure. It reports the accumulated result of years of hiring grade, promotion timing, and starting-offer decisions, and a single reporting period rarely contains enough movement to shift it.

The sharpest tension in the graph runs against Employee Diversity Ratio, the first-priority metric in the DEI group. Hire women in volume into entry and early-career grades and the diversity ratio improves immediately, while an unadjusted pay gap gets worse, because the new headcount sits at the bottom of the salary distribution. Both metrics are behaving correctly and a reader watching only one of them will draw the wrong conclusion. A second pull comes from Compensation and Benefits: Compensation Ratio (Compa-Ratio) and Market Competitiveness Ratio both reward paying accurately to the market band for a grade, which leaves a grade-composition gap completely untouched, and any corrective uplift lands directly on Total Compensation Cost, the group's first-priority metric. The metric that reconciles all of this is Promotion Rate Disparity, eighth in the DEI group, since it shows whether the grade distribution is actually changing or whether the pay gap is being managed with one-off adjustments.

Measuring Gender Pay Gap in Practice

The formula divides the difference between average male and average female pay by average male pay, and the word doing the work is average. Every hard decision in this metric is about what goes into that average, who is in the population when it is taken, and at what moment.

The data sits across at least four systems, and joining them honestly is most of the job. The payroll ledger holds actual amounts paid per pay period. The HRIS holds effective-dated job, grade, location, and contract records. Time and attendance holds contracted and worked hours. Equity administration holds grants, vesting, and exercises, usually in a separate system with its own identifiers. Payroll is transactional and periodic while job data is effective-dated, so a naive join on employee identifier returns whatever job record happens to be current rather than the one in force during the pay period. Pick a snapshot date, resolve every job attribute as of that date, and derive a pay rate rather than summing paid amounts.

The forks to settle before you calculate anything:

• Mean or median. Decide, and publish both. The mean is what a handful of executive salaries do to your figure; the median is what your middle grades do. Reporting only the one that reads better is transparently self-serving, and readers of your disclosure notice.
• Hourly rate or annual pay. Hourly is the only basis that survives part-time employees, and the tracked national sources that state a formula use an hourly basis. Converting salary to hourly requires contracted hours, not worked hours, and contracted hours are frequently stale in the HRIS for anyone who changed schedule without a new contract record.
• Which pay elements. Base only, base plus allowances, base plus bonus, or total remuneration including equity. Bonus and equity are far more skewed than base salary and are concentrated in senior and revenue-carrying roles, so including them usually widens the figure. Report a bonus component separately rather than blending it into a single headline figure.
• Population. Employees only, or employees plus contractors, agency workers, and fixed-term staff. Consolidated group, or per legal entity. The external sources here split on exactly this axis, with one restricted to the private sector and another to full-time workers.
• Adjusted or unadjusted. Both are legitimate, and they answer different questions. Never publish one under the other's label.

Snapshot censoring is the trap that catches the most organizations. Anyone on unpaid or reduced-rate leave on the snapshot date has a distorted pay figure for that period, and most reporting regimes handle this by excluding them from the population. The people excluded are disproportionately those on parental leave, which means the population definition quietly removes a group whose treatment is central to why the gap exists. If your rule excludes them, say so, and track the excluded count by sex as a diagnostic in its own right.

Joiners and leavers behave the same way. A mid-period hire has partial pay in the ledger and will look underpaid if you sum amounts instead of using a rate. Leavers carry final-pay items, notice pay, and accrued leave payouts that are not salary. Strip those or exclude the population, and stay consistent year over year, because a change in that rule alone will move the figure.

Grade mapping is where multi-entity organizations lose comparability. If you have grown by acquisition, the same grade label frequently means different things in different entities, and a global grade band that was mapped once and never revisited will place the same job at different levels. Salary structures also differ by country, so a cross-border figure converted at a single rate mixes genuine pay differences with labour market and currency effects. Run the calculation per country and per legal entity first, then decide whether a consolidated figure means anything.

Segmentation that pays for itself: by grade band, which separates a within-grade problem from a distribution problem and is the most diagnostic cut available; by function, since a gap concentrated in one function is a hiring or promotion issue in that function rather than a company-wide pay issue; by full-time and part-time status; by tenure band and by hire cohort, which reveals whether recent starting offers still reproduce the gap or whether the figure is a legacy position working its way out. Cell sizes get small fast in these cuts, so set a minimum cell size before you look, since one senior hire can swing a small cell dramatically and that swing is noise.

Two further instrumentation notes. Statutory reporting regimes apply only to employers above a headcount threshold, so any external distribution you compare against describes large employers and nothing else. And the adjusted version carries its own risk: a regression that controls for current grade, or for starting salary, absorbs the very inequity being measured and will return a reassuring result no matter what is happening underneath. Controls should describe the work, not prior pay decisions.

Common Pitfalls

Ignoring the gender pay gap can lead to reputational damage and hinder talent acquisition efforts.

• Failing to conduct regular pay audits creates blind spots in compensation practices. Without this analysis, organizations may overlook systemic biases that perpetuate disparities.
• Neglecting to communicate pay structures can breed distrust among employees. Transparency is crucial for fostering a culture of fairness and accountability.
• Overlooking intersectionality may lead to incomplete analyses. Failing to consider factors like race and age alongside gender can mask deeper inequities.
• Relying solely on averages can distort the true picture. Averages may hide significant disparities within departments or roles, leading to misguided conclusions.

Improvement Levers

Addressing the gender pay gap requires a multifaceted approach that prioritizes transparency and accountability.

• Implement regular pay audits to identify disparities. These audits should be comprehensive, covering all roles and departments to ensure equitable compensation practices.
• Enhance transparency around pay structures to build trust. Clear communication regarding how salaries are determined can help mitigate feelings of inequity among employees.
• Establish mentorship and sponsorship programs to support underrepresented groups. By fostering professional development, organizations can help bridge the gap in advancement opportunities.
• Set clear targets for reducing the gender pay gap and hold leadership accountable. Regular reporting on progress can drive focus and commitment across the organization.

KPI Depot is trusted by consulting, strategy, finance, and analytics teams at leading organizations worldwide, including those listed below.

Gender Pay Gap Benchmarks

We have 6 relevant benchmarks in our benchmarks database.

Browse the Top Benchmarked KPIs in Diversity, Equity, and Inclusion (DEI)

Reading the Benchmarks for Gender Pay Gap

The benchmark records tracked for this KPI come from Stats NZ, the Workplace Gender Equality Agency, the Office for National Statistics, Eurostat, and the OECD. Every one of them publishes a figure called a gender pay gap or a gender wage gap. No two of them measure quite the same thing, and none of them measures what a single employer measures on its own payroll.

Start with central tendency, because it is the divergence that moves reported figures the most. The OECD defines its gender wage gap on median earnings. Eurostat and the Workplace Gender Equality Agency work from averages. Mean and median answer different questions about a salary distribution: the mean is pulled by a small number of very high earners, who in most organizations and most economies are disproportionately men, while the median tracks the middle of the distribution and barely notices the top tail. A source switching from one to the other changes the published figure without anything changing in the underlying pay.

Next, the earnings base. The Office for National Statistics defines the gap on average hourly earnings excluding overtime, a deliberate exclusion, since overtime hours are unevenly distributed by sex and by occupation. Eurostat works from gross hourly earnings and labels its measure explicitly unadjusted. The OECD also labels its measure unadjusted and builds it from earnings rather than from an hourly rate. The Stats NZ record carries no stated formula at all, so customers should read that release itself before assuming it matches any of the others. Hourly, gross, and total-earnings bases are not interchangeable, and bonus, overtime, and allowance treatment is exactly where employer-level reporting tends to diverge from national statistics.

Population is the third fork, and this KPI's source set contains a clean demonstration of it. Two of the tracked records come from the same Office for National Statistics publication, with the same stated formula, and differ only in population: one covers all employees, the other covers full-time employees only. Same publisher, same definition, two different figures. The difference is almost entirely part-time composition, since part-time work is taken up unevenly by sex and sits in different occupations. The other sources cut population differently again. The Workplace Gender Equality Agency reports on private sector employees, which excludes public administration entirely. The OECD restricts to full-time wage and salary workers, which drops the self-employed along with part-time staff. A customer comparing an internal figure that includes part-time and fixed-term staff against a full-time-only external figure is comparing two different populations and will read the difference as performance.

Geography and coverage compound the problem. The records span New Zealand, Australia, the United Kingdom, the European Union, and the OECD membership. A national figure is an economy-wide composition statistic: it reflects which industries a country has, how its public sector is paid, and how its part-time labour market works. It is not a target for a single employer, and an employer sitting on one side of a national figure has learned very little. The European Union figure is itself an aggregate over member states whose own figures spread widely, so the aggregate describes no particular labour market.

Time period is the last axis, and the tracked sources do not line up on it. One is a single quarter of income statistics. One is an annual employer reporting cycle. The Office for National Statistics records rest on an April reference point, a snapshot that captures whoever happened to be on full pay in that month. One is a reference year whose data was extracted well after the period closed. The OECD record is explicitly the latest year available, which means different countries in the same table are describing different years. Comparisons drawn across these records are comparisons across time as much as across method.

One divergence deserves naming on its own: adjusted versus unadjusted. Eurostat and the OECD state plainly that their measures are unadjusted, meaning no control for occupation, grade, hours, industry, or tenure. Many employers run a regression-adjusted gap internally and report it alongside or instead. An adjusted figure and an unadjusted figure are different quantities and always will be. Setting them side by side is a category error, not a benchmark, and it is the most common way this metric gets misused in public reporting.

The practical conclusion for customers: a figure found on its own, without its central tendency, earnings base, population, geography, reference period, and adjustment status attached, cannot be used for comparison. The benchmark records behind this page carry those dimensions per source, along with the source and its date, which is what makes a comparison defensible instead of merely confident.

OKRs That Use Gender Pay Gap

The Compensation and Benefits group's objective to enhance employee retention by delivering competitive and equitable compensation packages is the natural home for this KPI. That objective already runs on Pay Equity Ratio, Market Competitiveness Ratio, Compensation Ratio (Compa-Ratio), and Turnover Rate Among High Performers, all of which describe whether pay is accurate for the role a person holds. Gender Pay Gap adds the question those four cannot answer, which is whether the distribution of roles is itself fair. Added as a key result it stays directional: narrow the unadjusted gap toward a target the compensation committee sets, while holding Pay Equity Ratio at or above its current level, so that the gap is not closed by letting like-for-like pay drift. The group's own guidance to pair equity improvements with high-performer retention applies here too, since the adjustments that move this metric are visible to everyone who did not receive one.

In the DEI group the objective to establish a leadership team that reflects diverse perspectives and backgrounds is where Gender Pay Gap does its real work, even though the group's worked example lists Leadership Diversity Ratio and Diversity Talent Pipeline Strength as the key results. The group's own best-practice guidance is explicit that Promotion Rate Disparity and Gender Pay Gap should be targeted together, on the reasoning that fixing pay without fixing advancement ignores career trajectory. Carried as a paired key result, that reads as: reduce Promotion Rate Disparity between men and women in the grades feeding senior management, and narrow Gender Pay Gap in those same grade bands, both toward targets the DEI and compensation leads agree on. The pairing also guards against the arithmetic trap in the metric, since strong entry-level diversity hiring pushes an unadjusted gap the wrong way for a while. A key result written on the gap alone, with no promotion measure beside it, quietly penalizes the hiring the same group is asking for.

See OKR Examples for Diversity, Equity, and Inclusion (DEI)

Free Whitepaper

Want to achieve performance excellence in Workforce Planning? Download our in-depth whitepaper: Definitive Guide to Workforce Planning KPIs.

Download the Free Guide

KPI Categories

This KPI is associated with the following categories and industries in our KPI database: