How to do a gender pay gap analysis.

What the analysis has to answer

Article 9(1) of the Directive lists the information employers must provide. Your analysis is finished when it answers all seven points from one consistent dataset.

1. The gender pay gap.
2. The gender pay gap in complementary or variable components.
3. The median gender pay gap.
4. The median gender pay gap in complementary or variable components.
5. The proportion of female and male workers receiving complementary or variable components.
6. The proportion of female and male workers in each quartile pay band.
7. The gender pay gap between workers by categories of workers, broken down by ordinary basic wage or salary and complementary or variable components.

The first six describe the organisation as a whole. The seventh is the demanding one: it needs categories of workers, which means grouping people who do the same work or work of equal value. Most of the effort in a first analysis goes there.

Step 1: Fix the scope and the reference period

Each report covers the previous calendar year. The size of the employer sets the deadline and the cycle:

Two questions to settle before you touch any data. First, which legal entity reports: the Directive speaks of the employer, so a group with several companies usually runs one analysis per employing entity. Second, who counts as a worker: national law decides, and the national law that transposes the Directive may set rules for headcount dates, temporary staff and people on leave. Write both decisions down. You will be asked how you made them.

Step 2: Collect complete pay data

The Directive defines pay broadly. Article 3(1)(a) covers the ordinary basic or minimum wage or salary and any other consideration, in cash or in kind, that a worker receives from the employer. An analysis built on base salary alone is incomplete.

• Basic pay for the reference year, and the contractual hours behind it, so you can compare full-time and part-time workers on the same basis.
• Complementary or variable components: bonuses, commissions, overtime premiums, allowances, long-term incentives.
• Benefits in kind with a monetary value, such as a company car or housing.
• Sex, as recorded for each worker.
• Role or job, department, location and the other attributes you will use to form categories of workers.

Article 3(1)(b) defines pay level as gross annual pay and the corresponding gross hourly pay. Check which one your national law asks you to report. Hourly pay keeps part-time workers comparable with full-time workers, which matters because part-time work is unevenly split between women and men in most workforces.

Most HR systems can export this as a spreadsheet. Before calculating anything, check the export: missing sex values, duplicate workers, people who left mid-year, and variable pay recorded in a different system. Errors here flow into every indicator. Collect only what the analysis needs; this is personal data under the GDPR.

Step 3: Build categories of workers

Article 3(1)(h) defines a category of workers as workers performing the same work or work of equal value, grouped in a non-arbitrary manner on the gender-neutral criteria of Article 4(4). Those criteria include skills, effort, responsibility and working conditions.

Job titles and departments are not enough. Two different titles can describe work of equal value, and one title can cover very different work. The usual route is job evaluation: score every role on the same factors, then group roles with similar scores into grades. Each grade becomes a category of workers. Our guide to work of equal value explains how to do this without building old biases back in.

Where workers’ representatives exist, the Directive expects the criteria to be agreed with them. Involve them at this step, not after the numbers are out.

Step 4: Calculate the seven indicators

The Directive defines the gap the same way for every indicator: the difference between men’s and women’s pay, expressed as a percentage of men’s pay. A positive figure means men are paid more.

• Mean gap: (men’s average pay − women’s average pay) ÷ men’s average pay × 100.
• Median gap: the same formula with the median, the pay level at which half the group earns more and half earns less.
• Variable pay gaps: the same two formulas, applied only to complementary or variable components.
• Share receiving variable pay: the percentage of women, and separately of men, who received any complementary or variable component.
• Quartile pay bands: rank everyone by pay level, split the list into four groups of equal size, and give the share of women and men in each.
• Gap by category of workers: the mean gap inside each category, shown separately for basic pay and for complementary or variable components.

Report the mean and the median side by side. A few very high salaries move the mean far more than the median, so the two often tell different stories. Our pay gap analysis page has a worked example with ten fictional salaries.

Step 5: Look behind the numbers

The indicators describe the gap; they do not explain it. For every category with a material gap, find out why. Some differences have objective, gender-neutral reasons that you can document. Others do not, and those are the ones the Directive wants removed.

This is where the adjusted gap helps: it shows how much of the raw gap remains once documented, gender-neutral factors are taken into account. It is a diagnostic, not something you report instead of the indicators. Our guide to the adjusted and unadjusted pay gap explains the difference with numbers.

Step 6: Confirm, report and share

Article 9(6) requires management to confirm the accuracy of the information after consulting workers’ representatives, and gives the representatives access to the methodology you used. Keep a record of how each figure was produced: which data, which version of your method, which decisions.

• Indicators (a) to (g) go to the national authority that compiles them (Article 9(7)).
• You may publish indicators (a) to (f) on your website (Article 9(7)).
• Indicator (g), the gap by category of workers, goes to all your workers and their representatives (Article 9(9)).
• Workers, their representatives, the labour inspectorate and the equality body may ask for clarifications, and you must give a substantiated reply within a reasonable time (Article 9(10)).

Step 7: Deal with any gap of 5% or more

If the average pay of women and men in any category of workers differs by 5% or more, and you can neither justify the difference on objective, gender-neutral criteria nor remedy it within six months of submitting the report, Article 10 requires a joint pay assessment with your workers’ representatives. Start the justification work before you report, not after. Our guide to the joint pay assessment covers what it involves.

Common mistakes

• Using base salary only, and leaving out variable pay and benefits in kind.
• Using job titles as categories of workers.
• Comparing annual pay of part-time and full-time workers without converting to a common basis.
• Treating the analysis as a spreadsheet exercise with no record of how each number was produced.
• Showing workers’ representatives the results without the method behind them.
• Starting in May of the reporting year.

Frequently asked questions