McNeil & Ors v Revenue And Customs

Where alleged group disadvantage concerns a continuum of pay rather than a binary benefit, particular disadvantage under the Equality Act 2010 must be assessed by reference to actual pay across the relevant pool. Where average pay figures are available, comparison of averages is the reliable measure. Selective distribution or clustering analysis cannot replace it, although it may provide secondary evidence.

Basic pay is not divided into a notional length-of-service element where no contractual or practical distinction exists. The 3% and 5% figures in equal-pay guidance are investigative indicators, not legal thresholds. An alternative case requiring factual investigation will ordinarily not be permitted for the first time on appeal.

Female HMRC employees in grades 6 and 7 claimed equal pay with male colleagues paid more within the same pay bands. It was common ground that the work was rated as equivalent under the Equality Act 2010. HMRC relied on the material factor defence, identifying length of service as the relevant factor.

The Employment Tribunal held that length of service caused the pay differences but did not put the claimants or women generally at a particular disadvantage. It dismissed the claims. The Employment Appeal Tribunal dismissed the appeal: [2018] UKEAT 0183/17. The Court of Appeal considered the correct method for assessing particular disadvantage where pay varied across a band, and whether alternative statistical cases could be advanced on appeal.

The appeal was dismissed unanimously. Underhill LJ gave the leading judgment, with Ryder LJ and Holroyde LJ agreeing.

1. Assessment of particular disadvantage. The alleged disadvantage was a continuum of salary levels, with no fixed point identifying who was disadvantaged. In that situation, the court must consider the actual sums paid to all properly comparable employees in the relevant pool. Where average pay figures are available, comparison of average pay is the reliable and definitive measure. A distribution analysis based on selected quartiles or deciles cannot displace it. It may be useful as secondary evidence where average female pay is lower, but it is an inferior surrogate.
2. Basic pay. The claimants could not isolate a notional variable element above the pay-band minimum. Basic pay was indivisible in the absence of any contractual or practical distinction between base salary and a length-of-service element. The term relied on was basic pay as a whole.
3. Statistical thresholds. The 3% and 5% figures in the Equality and Human Rights Commission guidance were indicators for investigation, not legal thresholds. A small differential could in principle establish particular disadvantage if caused by the relevant factor, although that issue had not been properly advanced on the evidence.
4. Alternative case on appeal. The claimants had expressly disavowed reliance on average-pay figures below. It would be wrong to permit them to advance a new alternative case on appeal where factual issues, including causation and possible alternative explanations, would require investigation and remittal. The exceptional power to allow a new point of pure law did not apply.

• Court of Appeal (Civil Division) — appeal dismissed.
• Employment Appeal Tribunal — appeal from the Employment Tribunal dismissed: [2018] UKEAT 0183/17; [2018] ICR 1529.
• Employment Tribunal — held that length of service was the relevant factor but did not create particular disadvantage; claims dismissed on 17 June 2016.