Rad Kohanzad provides a summary on McNeil v HMRC  EWCA Civ 112
McNeil v HMRC  EWCA Civ 112 answers the question of whether a tribunal should approach “particular disadvantage” in equal pay and indirect discrimination cases by reference to averages or differential distribution? Differential what?!
Owing to a disproportionate recruitment of women by the HMRC into the grades 6 and 7 in recent years and the fact that pay within a pay grade is determined, in part, by length of service, that meant that women were disproportionately represented at the lower end of the pay bands and disproportionately under-represented at the higher end (“differential distribution”). The Claimants' case was that the Respondent's use of length of service as a determinant of pay placed women at a particular disadvantage compared with men because those grades were historically male-dominated and women have only more recently begun to be recruited or promoted into those grades in greater numbers. The result is that women tend to be disproportionately over-represented at the lower end of the pay scale for each grade and disproportionately under-represented at the upper end of the pay scale for each grade.
The Claimants brought equal pay claims on the basis of their over and under representation at the lower and higher ends of the pay bands. The Claimants and their comparators do work rated as equivalent, within the meaning of section 65 (1) (b) of the Equality Act 2010, and accordingly they are entitled to be paid the same as them unless the HMRC could show a “material factor defence” under section 69 of the Act.
It was not in dispute that the statistics showed the differential distribution (also referred to as “clustering”) contended for. The HMRC’s case was, however, that the correct approach to compare the differences in pay between men and women was to compare the averages within the pay grade rather than by looking at the lower and upper quartiles (i.e. by splitting the pay grade into four quarters) of the pay grades and comparing the percentage of men and women in the lower and upper quartiles. The average difference in pay was about 2%.
The ET, EAT and Court of Appeal all favoured the approach of looking at the averages of the pay difference between men and women across the pay band.
Reiterating the Supreme Court’s decision in Essop v Home Office  UKSC 27, Underhill LJ reminded us that the reason why a particular pay system produced a disparity between men and women was not material to whether prima facie indirect discrimination had been shown, per Lady Hale.
Turning to the question of whether it was right to compare pay through a differential distribution, the Court noted that group disadvantage relied upon in the present case was not of the usual kind. Ordinarily, when analysing whether a group is put at a particular disadvantage by a PCP or factor the question is binary. What proportion, for example, of men and women receive a particular benefit? Here, however, the disadvantage complained of was not binary because there are a wide range of individual salaries within the pay bands. It is what the HMRC’s expert called “a semi-continuous variable”. There is, therefore, no fixed point on the pay band to draw a comparison to try and determine whether women are disadvantaged as a group.
The Court rejected the argument that disadvantage could be measured by reference to the proportion of men and women towards the bottom and top of the pay bands because that tells you nothing of what is happening in the middle two quartiles. It may well be the case that whilst women are over-represented at the bottom of the pay scale and underrepresented at the top that in the middle quartiles they do better than men. That is likely to be the case where there is the clustering described but yet nevertheless there was no, or little, difference in the average pay. Underhill LJ illustrated the point at paragraph 64 in the following way:
Take the case of a pay-band of between £50,000 and £80,000 p.a., with 100 men and 100 women in it. Assume that there are only three pay-points, as follows:
£50,000: 55 women; 45 men
£70,000: 55 men
£80,000: 45 women.
On [the Claimant’s] approach the women in the pool are at a particular disadvantage because they are, in terms of numbers, disproportionately at the bottom “end” of the scale (and men disproportionately at the top). However, that leaves out of account the fact the men in the top half are only just over the mid-point, whereas the women are all at the maximum, so that a large minority of the women earn much more than any man – and the average pay of the women in the pool is greater than that of the men.
So whilst women may disproportionately be in the bottom quartile, that does not mean that women as a whole are disadvantaged. Because pay is a semi-continuous variable, the “only way to ensure that the distribution analysis is entirely reliable and accurate is to create a segment for each distinct pay-point: otherwise you are, as Dr Brown put it in his oral evidence, ‘not using all the data available’. But, as he said, if you do in fact use all the data available you are simply calculating an average”. A distribution analysis which shows clustering was, therefore, held to be an “inherently inferior surrogate for the more reliable and accurate analysis provided by the comparison of averages”.
What lessons can employment lawyers learn from this case? Firstly, its application may be wider than equal pay alone because the concept of “particular disadvantage” is one that appears in indirect discrimination claims. Although such cases tend to present binary alternatives, this case provides authority for how to approach cases of a semi-continuous variable. Secondly, paragraphs 13 to 20 of Underhill’s LJ Judgment provide a very useful summary of the case law on indirect discrimination claims. Thirdly, you should instruct an expert in statistics if you can. Here, the HMRC’s expert outgunned the Claimants’ expert in terms of qualifications and knowledge. Finally, statistics are hard; but you probably already knew that. In any event, at least you know what a semi-continuous variable and differential distributions are.