Distinctly Average: statistics, rate of pay and shifting burden of proof in equal pay claims

The Court of Appeal in McNeil v Commissioners for her Majesty’s Revenue and Customs has recently held that when determining particular disadvantage in non-binary situations, Tribunals should look at male and female average pay, not the distribution of male and female pay within a band.  Daphne Romney QC, an equal pay specialist, explains Underhill LJ’s judgment.

Importance of McNeil

When Simler J gave judgment in the EAT in McNeil, the primary focus was on her confirmation that the principle in Newcastle NHS Hospitals Trust v Armstrong,[1] was finally pronounced dead. The employer could no longer avoid objective justification where the claimants had proved particular disadvantage by showing that even though there was an indirectly discriminatory effect on women, the reason was not tainted with sex.  Simler J applied the judgment of Baroness Hale in Essop v Home Office/Naeem v Department of Justice,[2] who had held that it was unnecessary to show why there was particular disadvantage; it sufficed to prove that particular disadvantage existed[3].  However, McNeil is also important in its analysis of the proper way to determine particular disadvantage when looking at comparative pay figures for men and women in claims of equal pay and indirect discrimination. Simler J’s decision on particular disadvantage was appealed to the Court of Appeal – the Armstrong point was not.Underhill LJ’s analysis therefore focuses our attention on the vexed issue of statistics and how to present and interpret them.[4] As he noted, there was no authority explicitly providing guidance on a situation where claimants argued particular disadvantage in a case of a ‘continuum’ case, namely where the pay constantly changes through a number of variables.

Proving particular disadvantage

It is well established that it for the Tribunal to decide whether the statistics advanced by a claimant are sufficient to prove indirect discrimination – see Mummery LJ in Haq v Audit Commission.[5] Equally, the Tribunal must select the suitable pool for comparison appropriate to the facts of the case. There are no hard and fast rules about this and certainly there is no statutory prescription or formula, save that the pool should consist of those affected by the question as summarised by Sedley LJ in Grundy v British Airways[6] andby Baroness Hale in Naeem.

Usually, indirect discrimination arises either through a PCP or ‘barrier-type’ claim like Seymour-Smith[7] or through an Enderby[8]-type situation, where men and women work in different occupations which is effectively “female work” and “male work”.

Particular disadvantage in “continuum” cases

In McNeil, as the claimants themselves argued, their case was neither of those situations. The claimants were female HMRC employees in grades 6 and 7. The grades attracted minimum and maximum pay, but no fixed spinal points in between. New entrants usually started at the bottom of the band (unless they were assimilated at their pre-existing salary) and then progressed towards the maximum through pay rises, which were normally made annually, although there had been a freeze for a few years, which preserved the status quo. Underhill LJ described the position as follows:

‘..the amount of salary payable to any particular employee in the two grades will depend on a number of variables – the point at which they entered the grade, how many years they have been in it, the amount of any increase in the minimum and maximum for the band and the amount of any pay awards, and their performance, in those years. The result is that in any year there is in each of the pay bands a very wide range of salaries – typically well over a hundred in each band – reflecting different pay histories in the individual cases.[9]

The claimants’ case was the length of service criterion operated to their particular disadvantage because the grades 6 and 7 had been historically male-dominated, and more women had only recently started to be recruited or promoted into them; as a result, women were disproportionately clustered at the bottom of those grades and men were disproportionately clustered at the top of those grades. It was argued that matters were exacerbated by various other factors, including the failure to reduce the length of the grades, the length of time it took to progress to the top of the grades and the pay freeze. The claimants also contended that the usual method for determining pay differentials, namely either a mean or a median average, was ‘not a meaningful, logical, or accurate method’  of testing the ‘core allegation’ in cases such as this, being neither a Seymour-Smith nor an Enderby case. As Underhill LJ pointed out in the Court of Appeal, length of service in this case did not equate to moving up the ladder in specified spinal points, because that was not the way that the grades were structured – rather, length of service allowed longer-serving employees to get the benefit of more annual pay increases, thereby moving quicker towards the top of the grade in question than someone who had been on that grade for a shorter period.

At the Tribunal, the HMRC’s expert maintained that the claimants’ approach was flawed and that the correct methodology was to compare the average pay for men and women across each grade; those figures, which the claimants accepted, showed that the difference in average pay was minimal. Over a seven- year period, they were between 2.3% and 1.2% in grade 7, and 1.9% to 1.5% in grade 6. The Employment Judge agreed. Clustering was not the same as pay and ‘whilst it may provide a partial picture of apparent actual advantage, it says nothing about actual pay difference’. The law called for a comparison of the difference in terms, in this case, pay, between the claimant and her comparator. The figures showed that there was no material difference between the average pay for men and women, and the claimants’ expert’s methodology was inappropriate. Basic pay was a term of the contract and could not be divided into sub-elements, namely average pay and variable pay. The Employment Judge further noted that the claimants’ methodology meant, ‘it would never be possible to guard against a complaint that, on this or that selective statistical analysis of a portion of the relevant population, one gender group appeared to enjoy an advantage over the other’. As a result, an employer could never form a view as to whether its pay structure was indirectly discriminatory and could be forced to objectively justify differences in pay which it would not otherwise have to do.

On appeal, Simler J endorsed the Tribunal’s view and emphasised that basic pay was an indivisible term, on the principle of Hayward v Cammell Laird,[10]and Brownbill v St Helens Hospitals NHS Trust.[11] Just as basic pay cannot be aggregated with other elements of pay, such as bonus or overtime, so basic pay cannot be subdivided into separate elements, particularly with as many variables as here.

The claimants’ primary appeal to the Court of Appeal was that both the Tribunal and the EAT had measured ‘the wrong thing’, namely measuring the average differences in pay relevant to total payinsteadof ‘measuring the disparate impact on women compared with men caused by the factor of length of service by analysing the relative distribution of men and women along the pay range in each of the relevant grades’. Underhill LJ pointed out that this was not a typical case of particular disadvantage, where the non-receipt of a defined benefit affected one sex more than the other; this could be measured through the use of a pool and a comparison of the proportion of advantaged and disadvantaged groups. Here the disadvantage was not binary; rather base pay was ‘a semi-continuous variable’ and it was not possible to identify ‘a particular touchstone of disadvantage’. He continued:

The application of the concept of particular disadvantage in a “continuum” case of this kind has not been explicitly discussed in any of the authorities, and it has required some careful thinking. But I have come to the clear conclusion that the ET and the EAT were right to reject the approach argued for by [the claimants]. I do not believe that what the statute requires to be measured is the “incidence” of disadvantage. As Simler P points out at paras. 45-46 of her judgment, the distribution of men and women within a pay band only means anything in so far as it represents what sums they are in fact paid. It seems to me necessarily to follow that it is essential to take into account the actual amounts paid to each person in the group; and the only way in which that can be done definitively is by taking an average.’[12]

He echoed the point made by the Employment Judge – where distribution analysis is used to show indirect discrimination by reference to whatever part of the pool the claimant may choose to focus on, an employer can never be sure that its system was not potentially unlawful.

The essence of the judgment, therefore, is that in cases such as these, only the average pay for each group will give a reliable indication of whether pay discrimination exists. That is not to say that clustering is always irrelevant – in a case where there is a material difference between male and female pay, it is likely that there would be disproportionate clustering of men at the top end and women at the bottom end. But to start from the distribution angle where there is the more reliable test of average pay is the wrong approach.

The issue was not under appeal, but it is interesting that Underhill J set out the test for the burden of proof in paragraph 18 of the judgment, which did not mention the Armstrong stage, therefore giving tacit approval to Simler J’s view expressed in the EAT.

[1] [2006] IRLR 124
[2] [2017] IRLR 558
[3] Paras. 71-73
[4] [2019] EWCA Civ 1112
[5] [2013] IRLR 206 at paras. 45-51
[6] [2008] IRLR 817 at para. 27
[7] [1999] 1 AC 554
[8] Enderby v Frenchay Health Authority [1994] ICR 112
[9] Para. 31
[10] [1988] 1 AC 894
[11] R v Secretary of State ex p Seymour-Smith [1999] 2 AC 554
[12] Court of Appeal at para 63.