McNeil & Ors v Revenue And Customs

The Claimants in the proceedings which give rise to this appeal, who are the appellants before us, are a group of female employees of the Commissioners for Her Majesty’s Revenue and Customs (“HMRC”), who are the respondents. Jobs in HMRC below Senior Civil Service level are allocated to one of seven grades, each of which attracts its own “pay band” – that is, a range within which the salary of a person in that grade will fall. The Claimants are (or were) employees in grades 6 and 7 (6 being the higher) whose salaries have, over the periods to which their claims relate, been at the lower end of the pay band for their grade. They claim equal pay with male colleagues whose salaries are at or near the top of the relevant band. I shall have to return in due course to how their claim is formulated, but the essence is that statistics show that at the material times women were disproportionately represented at the lower end of the bands and men at the higher – what they describe as “clustering”. It is common ground that 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 that they are entitled to be paid the same as them unless HMRC can rely on the “material factor defence” under section 69 of the Act, which I set out at para. 9 below.

The Employment Tribunal directed a preliminary hearing on two agreed issues, as follows:

“(i)

What is or are the 'factors' within s. 69 (1) of the Equality Act 2010 causing the difference in basic pay between any Claimant and comparator who has a higher basic pay?

(This issue will include resolution of the parties' respective positions as to the correct definition of the relevant 'factors' and whether, or to what extent, the precise definition has a material bearing on the correct overall analysis under s. 69)

(ii)

Whether in light of the proper definition of the factor or factors, that factor or factors put the Claimants and women at a particular disadvantage when compared with men in Grades 6 and/or 7

(respectively) for the purposes of s. 69 (2) of the Equality Act 2010?

(This issue will include resolution of the parties' respective positions as to the appropriate pool(s) of comparison and whether the statistical and/or other evidence demonstrates a relevant particular disadvantage within the appropriate pool(s)).”

By a Judgment with Reasons promulgated on 17 June 2016 the ET (Employment Judge Snelson, sitting alone) held that the factor causing the pay differentials complained of was “length of service” and that it did not put the Claimants, or women generally, at a particular disadvantage. That meant that HMRC’s material factor defence was proved, and the claims were dismissed accordingly.

The Claimants appealed to the Employment Appeal Tribunal. By a judgment handed down on 27 February 2018 Simler P dismissed the appeal.

This is an appeal against that decision. The Claimants have been represented by Mr Ben Cooper QC and Ms Betsan Criddle and HMRC by Mr Thomas Linden QC, leading Mr Robert Moretto. All counsel save Ms Criddle appeared also in both the ET and the EAT. Their submissions have been of extremely high quality.

The rights formerly conferred by the Equal Pay Act 1970 are now to be found, albeit differently framed, in Chapter 3 of Part 5 of the 2010 Act, under the heading “Equality of Terms”. Chapter 3 contains a distinct regime applying to discrimination in the terms of a contract of employment: it is thus not concerned only with discrimination in pay, but in what follows I will for convenience refer to the “equal pay regime”. The intention is plainly that the underlying principles of the general law of discrimination should apply to the equal pay regime, but its provisions are different in their details from those of Chapter 1, which covers other cases of discrimination in the employment field, and indeed from those of the rest of the Act. There was presumably a good reason for the choice made to maintain in the 2010 Act a distinct regime for equal pay, adopting the essential framework of the 1970 Act, rather than assimilating it to the regime applying generally to discrimination at work; but it has left the law unsatisfactorily complicated.

The structure of Chapter 3 is that it confers rights on an employee (“A”) who is employed on “equal work” with a comparator of the opposite sex (“B”). One of the circumstances in which employees are to be treated as doing equal work is where they are doing work “rated as equivalent”, as the Claimants and their comparators are here. B must not only be doing equal work with A but must also satisfy the criteria for being a comparator in section 79 of the Act: I will refer to someone who satisfies those criteria as “properly comparable”. I will in the following paragraphs assume that the complainant is a woman, and I will for ease of distinction refer to the putative discriminator as “he”.

The principal operative provision of Chapter 3 is section 66, which provides, so far as relevant:

“(1)

If the terms of A's work do not (by whatever means) include a sex equality clause, they are to be treated as including one.

(2)

A sex equality clause is a provision that has the following effect—

(a)

if a term of A's is less favourable to A than a corresponding term of B's is to B, A's term is modified so as not to be less favourable;

(b)

if A does not have a term which corresponds to a term of B's that benefits B, A's terms are modified so as to include such a term.

(3)-(4) …”

Section 69 provides for the material factor defence. It reads, so far as relevant:

“(1)

The sex equality clause in A's terms has no effect in relation to a difference between A's terms and B's terms if the responsible person shows that the difference is because of a material factor reliance on which –

(a)

does not involve treating A less favourably because of A's sex than the responsible person treats B, and

(b)

if the factor is within subsection (2), is a proportionate means of achieving a legitimate aim.

(2)

A factor is within this subsection if A shows that, as a result of the factor, A and persons of the same sex doing work equal to A's are put at a particular disadvantage when compared with persons of the opposite sex doing work equal to A's.

(3)-(6) …”

The effect is that the defence is available if but only if the differential complained of can be shown to be caused by a “factor” which is neither directly (sub-section (1) (a)) nor indirectly (sub-sections (1) (b) and (2)) discriminatory. In the present case we are concerned only with indirect discrimination.

Sub-sections (1) (b) and (2), taken together, closely mirror the terms of section 19 of the Act, which is headed “indirect discrimination” and which applies to discrimination otherwise than under the equal pay regime. Section 19 reads (so far as material):

“(1)

A person (A) discriminates against another (B) if A applies to B a provision, criterion or practice which is discriminatory in relation to a relevant protected characteristic of B's.

(2)

For the purposes of subsection (1), a provision, criterion or practice is discriminatory in relation to a relevant protected characteristic of B's if—

(a)

A applies, or would apply, it to persons with whom B does not share the characteristic,

(b)

it puts, or would put, persons with whom B shares the characteristic at a particular disadvantage when compared with persons with whom B does not share it,

(c)

it puts, or would put, B at that disadvantage, and

(d)

A cannot show it to be a proportionate means of achieving a legitimate aim.

(3)

…”

The phrase “provision, criterion or practice” in section 19 is usually shortened to “PCP”. I should also note that the term “indirect discrimination” is sometimes inaccurately used to refer to a case satisfying only requirements (a)-(c) (or, in the case of section 69, sub-section (2)), i.e. to a case which will constitute indirect discrimination unless justified: I will refer to such a case as one of “prima facie indirect discrimination”.

The definition in section 19 derives from EU legislation. The Directive currently applying to equal pay between men and women is the Equal Treatment Directive 2006/54/EC, article 2.1 (b) of which defines indirect discrimination as occurring

“… where an apparently neutral provision, criterion or practice would put persons of one sex at a particular disadvantage compared with persons of the other sex, unless that provision, criterion or practice is objectively justified by a legitimate aim, and the means of achieving that aim are appropriate and necessary”.

It is relevant to note that the definition of indirect discrimination in the original sex discrimination legislation was differently expressed. Section 1 (1) (b) of the Sex Discrimination Act 1975, as originally enacted, provided that a person discriminated against a woman where:

“… he applies to her a requirement or condition which he applies or would apply equally to a man but—

(i)

which is such that the proportion of women who can comply with it is considerably smaller than the proportion of men who can comply with it, and

(ii)

which he cannot show to be justifiable irrespective of the sex of the person to whom it is applied, and

(iii)

which is to her detriment because she cannot comply with it.”

Although the change to the current formulation (which preceded the 2010 Act) was intended to avoid some undesirable complexities of analysis which the original language was thought to encourage, the essential character of the requisite comparison has not changed. That is apparent in particular from the judgments of Lady Hale in Chief Constable of West Yorkshire Police v Homer[2012] UKSC 15, [2012] ICR 704, (esp. at para. 14 (pp. 709-710)) and Essop v Home Office [2017] UKSC 27, [2017] 1 WLR 1343: in the latter she emphasises the “salient features” common to all the various iterations of the indirect discrimination concept (see paras. 23-29 (pp. 13534)).

As background to the issues on this appeal, I need to make a number of particular points about the effect of those provisions; but what follows is not intended as any kind of overall analysis of this difficult area. Although some of the points are of general application, I will refer more specifically to sex discrimination in the employment field; and I will assume that the claimant is a woman.

(1) “Factor” and “PCP”. One consequence of retaining a distinct regime for equal pay is that section 69, in so far as it covers indirect discrimination, does not employ the familiar terminology of a “provision, criterion or practice”. But in order to preserve coherence in this area the term “factor” in section 69 (2) must be understood to be to the same effect.

(2) Collective disadvantage. The definitions of indirect discrimination in both section 19 and section 69 (2) are concerned with group disadvantage. The essential comparison is between the positions of the men within the relevant pool, viewed as a group, and the women in the pool, likewise viewed as a group. The requirement that the claimant herself should suffer the same disadvantage is additional, and merely ensures that she should not be entitled to complain of a disadvantage to which she is not herself subject. The relevant pool for the purpose of the comparison is all (properly comparable) employees doing equal work to whom the factor in question applies. It follows that it is not necessary that the position of every woman must be worse than that of every man. At para. 27 of her judgment in Essop (p. 1354 B-C) Lady Hale says:

“A fourth salient feature is that there is no requirement that the PCP in question put every member of the group sharing the particular protected characteristic at a disadvantage. The later definitions cannot have restricted the original definitions, which referred to the proportion who could, or could not, meet the requirement. Obviously, some women are taller or stronger than some men and can meet a height or strength requirement that many women could not. Some women can work full time without difficulty whereas others cannot. Yet these are paradigm examples of a PCP which may be indirectly discriminatory.”

(3) “Particular” disadvantage. In CHEZ Razpredelenie Bulgaria v Komisia za zashtita ot diskriminatsia, case C-83/14, [2016] 1 CMLR 14, the CJEU confirmed that the word “particular” in the phrase “particular disadvantage” was not intended to connote a disadvantage which was “serious, obvious and particularly significant” but simply to make clear that it was persons with the relevant protected characteristic who were disadvantaged: see paras. 99-100 of its judgment (p. 546). (That being so, the word would appear to be, strictly speaking, to be redundant, since that would be the effect of the statutory language even without it.)

(4) “Barrier” and “Enderby-type” discrimination. Although the concept of a PCP, or a “factor”, “putting” women at a disadvantage might seem most naturally to connote something identifiably done by the employer which creates the disadvantage – sometimes called “barrier-type” indirect discrimination – that is not the only situation in which indirect discrimination can be found. In Enderby v Frenchay

(5) Burden of proof. The position about the burden of proof is a little complex. Before we get to the material factor defence a claimant establishes a prima facie case under section 66 simply by identifying a properly comparable man with whom she does work of equal value but who is paid more than her. However, in a case of the present kind that is essentially a formality (Footnote: ²), and the real issue turns on the material factor defence. As to that, the initial burden is on the employer to prove that the differential is the result of an identified factor and that that factor is not directly discriminatory. It is then for A to prove that any such factor puts her, and other women doing equal work with her, at a particular disadvantage compared with men; but if she does so the burden of proving that the factor is a proportionate means of achieving a legitimate aim – for short, that it is justified – is on the employer.

(6) Statistics and “significance”. It will often be the case that claimants will need, or seek, to establish particular disadvantage by a statistical comparison of the extent of the disadvantage (in this case, differential pay) as between men and women – though, as Lady Hale makes clear in Homer, it will not be necessary to do so in every case.

(7) Degree of differential and degree of impact. As already noted, the CJEU in CHEZ confirmed that a disadvantage does not need to be “serious” or “particularly significant” (Footnote: ³) in order to give rise to prima facie indirect discrimination. But that is distinct from the question of the size of the differential between the proportions of men and women who suffer the disadvantage. As to that, it is clear from the case-law that the differential must be “considerable”, or, to put it another way, that “far” more people in the protected group must suffer the disadvantage: see, for example Seymour-Smith at paras. 59-64 (pp. 598-9), and CHEZ, at para. 101. (That is distinct from the point made in para. 19 above about the “significance” of any statistics relied on; but the points may overlap forensically, since one reason why statistics which show only a small differential may not be significant is that small differentials are more likely to arise by chance or be the result of confounding factors.)

Section 14 (1) of the Equality Act 2006 (as amended) empowers the Equality and

The Commission has issued a Code of Practice under section 14 (1) of the 2006 Act. Part 2, headed “Good Equal Pay Practice”, recommends employers to conduct regular equal pay audits in order to identify whether there is in their organisations a “gender pay gap” and suggests a five-step model for doing so. Step 3 is headed Collect and compare pay data to identify any significant pay inequalities between roles of equal value. Two paragraphs are relevant for our purposes:

“177.

Once employers have determined which male and female employees are doing equal work, they need to collate and compare pay information to identify any significant inequalities by:

•

calculating average basic pay and total earnings, and

•

comparing access to and amounts received of each element of the pay package.

…

179.

Employers then need to review the pay comparisons to identify any gender pay inequalities and decide if any are significant enough to warrant further investigation. It is advisable to record all the significant or patterned pay inequalities that have been identified. The Commission's toolkit gives detailed advice and guidance on collecting and comparing pay information and when pay gaps may be regarded as significant.”

The “toolkit” referred to in paragraph 179 is published separately. The following passage appeared in the version of the guidance on “step 3” current at the time of the hearings below:

“As a general guide, any differences of 5% or more, or 3% or more if there is a pattern of gaps favouring one sex, will require exploration and explanation.”

I should sound a note of caution here. When a tribunal is referred to Part 2 of the Guidance in the context of an equal pay claim it needs to bear in mind that the conduct of an equal pay audit is not the same as the determination of an equal pay claim. The figures of 3% and 5% given in the passage quoted from the toolkit are put forward, avowedly, as useful indicators of whether the employer’s pay system may incorporate gender-discriminatory elements: as it says, they will “require exploration and explanation”. It is plainly not the case that any differential of over 5% (or 3% persisting) between the average pay of men and women in the relevant pool will be unlawfully discriminatory: that will depend on what the cause is found to be and if necessary on justification. Nor, equally importantly, will the fact that the differential is lower than those figures necessarily exclude a case of prima facie indirect discrimination. The toolkit itself makes the latter point, concluding step 3 with the caveat that “any pay difference, even when less than 3% or 5%, may be open to legal challenge”.

In the ET the parties produced a very full and clear statement of agreed facts, which I attach to this judgment as Annex 1; and the Judge made some further findings. For present purposes I need only summarise the essential points.

As already noted, we are concerned with grades 6 and 7: each has in fact two pay bands, depending whether the employee is based in London or outside (“nationally”).

The bands are defined only by reference to their maximum and minimum points. Unlike in many older pay schemes in the public sector, there are no fixed points in between: this is not a ladder with defined rungs. Typically the maximum and minimum figures have been raised on an annual basis with effect from 1 June (though latterly the two figures have not gone up at the same rate – see below); but there is no contractual obligation on HMRC to do so, whether by reference to inflation or otherwise, and there was in fact no increase in the three years from 2011 to 2013.

It has for some time been HMRC policy to shorten the range of the bands – that is, to reduce the difference between the minimum and the maximum points. This has been done by increasing the minimum figure, in the years where there has been any increase at all, by a greater amount than the maximum. In the first year covered by the claims (2009/10) the minima for grades 7 and 6 (for London) respectively were £50,888 and £61,317 and the maxima £60,980 and £73,474 – a range of 20%; but by 2014/15 the corresponding figures were £53,196 and £61,590 (grade 7) and £64,098 and £74,209 (grade 6) – a range of 15.8%. There has been an equivalent shortening for employees outside London, but I need not give the figures.

In principle new entrants to the grade will start at the bottom of the band. But the position is complicated by the fact that the present structure dates back only to 2006, following the merger of the Inland Revenue and the Customs and Excise: that required a process described as “assimilation”, under which existing employees might come into the pay bands associated with the new grades part-way up, in order to preserve their previous level of earnings. And even since assimilation individuals may start at a point above the minimum if for some particular reason their pay level has to be preserved.

The very concept of a pay band implies that employees have the opportunity to move up through the band by increases in their annual salary relative to the minimum until they reach the maximum. There is no contractual entitlement to any such increases, and there was in fact a two-year pay-freeze in 2011 and 2012; but the expectation in HMRC, and the usual practice, has been for annual pay awards to be made with effect from 1 June. Awards are expressed as a percentage increase on the existing salary, with the rate depending on the employee’s performance rating for each year: in the period with which we are concerned, however, only those employees who were rated as “poor” (who were a small minority) did not receive an award. The details of the awards made vary over the period, and there are some refinements which can for present purposes be ignored. The broad picture is that except in the two pay-freeze years most employees in grades 6 and 7 received awards that brought them closer to the maximum for the band, with those at or near the minimum benefiting disproportionately (i.e. to the extent that the minimum went up by more than their individual award), and those already at the maximum benefiting only if and to the extent of any increase in the maximum itself. This is not equivalent to a system of “pay progression” in the traditional public sector sense, where employees move up each year through a pre-determined number of pay points; but it does mean that they have an expectation of a pay increase each year (unless they perform positively badly) and also that that increase will involve an element of progression towards the maximum for the band.

It will be apparent from the foregoing that 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.

There have throughout the period with which we are concerned been substantially more men than women employed in grades 6 and 7. It is common ground that HMRC has over the period covered by this claim sought to redress that imbalance by recruiting more women. This had increased the proportion of women in grade 6 between 2009 and 2015 from 32.6% (out of a total headcount of 862) to 39.2% (out of

It appears from the figures at para. 28 above that a woman at the bottom of either band, whether in London or nationally, would be earning over £10,000 less than a man at the top of the same band, though we were told by Mr Cooper that the differentials complained of in these proceedings were “up to around £5,000”: of course individual comparisons are bound to yield varying figures.

It is important, for the purposes of the arguments before us, to understand the evolution of the Claimants’ case. I therefore start by summarising the pleadings and evidence chronologically.

The earliest ET1s, which date from early 2014, were accompanied by comparatively summary “Details of Complaint”. These, strictly unnecessarily, pleaded a positive case in response to HMRC’s anticipated reliance on the operation of “length of service” as a determinant of pay. It was contended that length of service was a factor which placed women at a particular disadvantage. I need not summarise all the points made in support of that contention, because they were subsequently narrowed, but they included (at para. 7 (b)) an allegation that the operation of length of service resulted in “clustering” of women towards the lower end of the pay bands and of men towards the higher end; and (at para. 7 (c)) an allegation based on the result of an equal pay audit conducted in 2010 which showed that the median pay (Footnote: ⁴) for women was

HMRC’s ET3 attached full Grounds of Resistance pleaded by Mr Moretto and dated 23 May 2014 in terms common to all the claims. Again, I need not summarise all the points made, but it was denied that HMRC’s pay system put women at a particular disadvantage or, more particularly, that there had been statistically significant differences in the pay of men and women. In response to the case based on the difference in the median pay of men and women, HMRC set out at paras. 34 and 35 of the Grounds figures for the differential in the average – that is, mean – pay of men and women. I attach at Annex 2 to this judgment tables setting out the figures in question. (Footnote: ⁵) It will be seen that they show that, although there has throughout the period covered by the claims been a difference in the average pay of men and women in both grades, the differential has never been more than 1.9% for grade 6, taking the London and national figures together, and never more than 2.3% for grade 7 (though there are slightly greater differentials for London if taken on its own); and that it has, viewed overall, been falling over the period.

The Claimants were required to particularise their case on particular disadvantage. They did so by Further Particulars, pleaded by Mr Cooper, dated 1 October 2014. I need to give a fairly full summary of these.

Under the heading “The core allegation of particular disadvantage”, the Particulars plead (at para. 3):

“It is the Claimants' case that the Respondent's use of length of service as a determinant of pay in both Grades 6 and 7 places, and has at all material times placed, women at a particular disadvantage compared with men because those grades are historically male-dominated and women have only more recently begun to be recruited or promoted into those grades in greater numbers, with the result 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. That effect has been exacerbated by the Respondent's failure over the years to reduce the length of the pay scales and the period taken to move from minimum to maximum for the two grades in question and by the imposition of the public sector 'pay freeze' in 2010, which has had the effect of protecting the higher pay of longer-serving employees and precluding any narrowing of the gap for employees with shorter service.”

In short, therefore, the factor which the Claimants said put women in the two grades at a disadvantage was “[HMRC’s] use of length of service as a determinant of pay” (for short, “length of service”); and the reason why it was said to do so was that disproportionate recruitment of women into the grades in recent years (for short, “differential recruitment”) meant that they were disproportionately represented at the lower end of the pay band and disproportionately under-represented at the higher end (“differential distribution”).

The Particulars continue with the heading “The appropriate analytical method(s) for testing the core allegation”. Paras. 5 and 6 contend that the “core allegation” does not neatly correspond to either barrier discrimination (of which Seymour-Smith is referred to as an exemplar) or to Enderby-type discrimination and that the methods applied to prove discrimination in those kinds of case cannot be directly transplanted. Para. 7 reads:

“The most appropriate method for testing the core allegation set out at paragraph 3 above is to analyse whether the proportion of women within the lower part (e.g. the bottom quartile or decile) of the pay scale for each grade is significantly greater than the proportion of men in the same part of the pay scale and, conversely, whether the proportion of women within the upper part (e.g. the top quartile or decile) of the pay scale for each grade is significantly lower than the proportion of men in the same part of the pay scale.”

That proposition is then developed in the following paragraphs.

Paras. 10 of the Particulars sets out in some detail the reasons why “the method outlined at paragraph 7 above is more appropriate as a means of logically testing the

Paras. 12-14 of the Particulars, under the heading “Statistical analysis”, summarise the results of an analysis following the method advanced at para. 7 conducted by Dr Alison Hall, whose report was served with the Particulars. Dr Hall is an HR professional specialising in equal pay analysis, though she is not a qualified statistician. In sections 5 and 6 of her report Dr Hall exhibited tables which showed, for each of the two grades (separately for London-based and “national” employees), and for each of the six years with which we are concerned (2009-2014), the proportion of male and female employees in each quartile and decile of the relevant band (Footnote: ⁶). The tables are at Annexes 3 and 4 to this judgment. What the tables in Annex 3 show is that for both grades (both in London and nationally) and for all six years there were substantially more women than men in the lowest quartile and substantially more men than women in the highest quartile. The precise figures vary, but, for example, in 2014 in London 44% of the women in grade 6 were in the lowest quartile, as against 26.5% of the men; the corresponding figures for grade 7 were 53.8% of the women as against 42.4% of men. The disproportions are both smaller and less consistent for the two middle quartiles (though these account for well under half the headcount in both grades). The tables in Annex 4 show an essentially consistent picture when the analysis is done by deciles. Dr Hall said that the figures were statistically significant by reference to the so-called χ² (“chi-squared”) (Footnote: ⁷) test.

Para. 15 of the Particulars reads:

“So far as comparison of average pay is concerned, in accordance with section 4 of the appended report, the Claimants accept the Respondent’s calculations set out in paragraphs 34 and 35 of the Grounds of Resistance ... The Claimants do not therefore allege that the differences in average pay are or have been consistently greater than 3% and do not pursue the allegation in paragraph 7 (c) of the Details of Complaint to that effect. However, for reasons set out above, comparison of crude averages does not constitute a meaningful, logical or appropriate methodology for assessing particular disadvantage in this case.”

There is a slight elision there. The “Respondent’s calculations” which are referred to relate to the mean rather than the median (see para. 36 above), and accepting them would not require the Claimants to abandon their case based on the median as originally pleaded at para. 7 (c) of the Details of Claim (see para. 35): that is of course consistent with the stance adopted in para. 11 of the Particulars. (The Claimants took that stance notwithstanding that Dr Hall herself had expressed the view, at section 4 of her report, that the differentials in both the median and the mean figures, despite the latter being under 3%, tended to suggest that women were at a disadvantage compared to men.)

HMRC served a Response to the Claimants’ Particulars, settled by Mr Linden and Mr Moretto, on 1 December 2014. I need not summarise these because the various respects in which they take issue with the Claimants’ case will sufficiently appear below.

HMRC’s Response exhibited a report dated 27 November 2014 from Dr Gary Brown, the Principal Methodologist at the Office for National Statistics, which challenged Dr Hall’s use of a χ² analysis and her reliance on the value of differentials in median pay (though this had in fact been abandoned by counsel – see para. 42 above). Each expert produced two further reports, partly responding to the other’s views and partly also addressing further data supplied by HMRC. They had a telephone discussion on 17 April 2015, of which an agreed note was prepared. I need not give a full summary of the differences between them. However, I should note that the essential reason why Dr Brown did not accept that it was appropriate to use a χ² analysis in the context of the present case was that it turned “base pay, which is a semi-continuous variable, into a categorical value”.

On 13 May 2015 the Claimants served Further Particulars of Indirect Discrimination. As to these, I need only note that they continued to put their case squarely on “a substantial differential distribution of women compared with men within each pay scale, with a substantially greater proportion of women in each grade than men towards the bottom of each pay scale and a substantially higher proportion of men in each grade than women towards the top of each pay scale” (see para. 26.1).

I wish to say by way of preliminary that the Employment Judge’s reasons are exemplarily thorough and thoughtful.

In his Introduction the Judge identifies the way in which the Claimants’ case has evolved prior to the hearing. He quotes in full paras. 3 and 7 of the October 2014 Particulars, which he describes as “admirably clear”. He notes a number of points which either had been previously pleaded but abandoned or had been specifically disavowed. In particular:

I need say nothing about the Judge’s findings of fact beyond what I have set out above, save that I should perhaps note that at para. 25 of the Reasons he said:

“I have seen no evidence of the gender balance of those seeking appointment to grade 7 or grade 6 positions as against those appointed, no doubt because direct sex discrimination was never alleged and it was not suggested that the female numbers have been depressed by any form of indirect discrimination in the recruitment process [emphasis in original]”

I turn to his conclusions on the two preliminary issues.

The question raised by this issue is what is or are the factor or factors, for the purpose of section 69 of the Act, causing the difference in salary between any claimant and her comparator. As set out above, it was the Claimants’ case that the factor in question was “length of service”. Mr Linden argued that that was not accurate. The salary of any particular employee in any given year was the result of a number of variables, of which length of service is only one: see para. 31 above.

The Judge accepted Mr Linden’s proposition as far as it went, but he held that the Claimants were entitled to argue that length of service was the only one of those factors that had a discriminatory effect and to focus their claim on it accordingly. As he put it at para. 33 of the Reasons:

“Of course, read literally, [the issue] could be answered as Mr Linden proposes: pay differences undoubtedly arise from all of the factors which he identifies. But in the context of this case, the sole criterion that matters is length of service. The Claimants rely on none other. They complain of breach of the equality clause in respect of ‘a [my emphasis] difference”’ in pay (s69 (1)) in so far as that difference is referable to length of service, but not otherwise. In these circumstances, I am satisfied that the ‘material [my emphasis] factor’ within s69 (1) is, as Mr Cooper submits, length of service.”

HMRC has not sought to challenge the Judge’s conclusion on issue (1), and we must accordingly proceed on the basis that “length of service” is a factor contributing to the determination of pay in the grades in question, and address the Claimants’ case that it accounts for (at least part of) the differentials between them and their comparators. But it is worth noting what that factor connotes in the context of this case. HMRC’s system is not, as explained above, one where employees move up a ladder with predetermined rungs. They benefit from length of service only in the sense that the more years they have been in the grade the more opportunities they have had to benefit from annual pay awards (when made). In this sense, as Mr Linden pointed out, “length of service” is a factor in any system where an employer tends to increase pay for employees over a period of time, and he suggested that the present case was closer to the Enderby type than to the barrier type. That is a nice point, but it is unnecessary to pursue it for our purposes.

I pause here to note one issue which does not arise. The reason why the use of length of service as a determinant of pay in the present case disadvantages women, as pleaded by the Claimants, is the increase in recent years in the proportion of women recruited, which it is said must tend to operate to depress the pay of women as a group. Differential recruitment of that kind is not in itself discriminatory against women: on the contrary, it might be thought to reflect a proper appreciation of equality issues. However it is clear as a matter of authority that that is irrelevant at

The Judge identified three issues arising under this head, but two are no longer live before us. He dealt with the issue which is relevant for our purposes at paras. 40-50 of the Reasons, under the heading “Statistics: distribution or mean?”. He began, at para. 40, by pointing out that the parties’ statistical evidence did not meet head on. As he said:

“The Claimants’ statistics are said to demonstrate ‘bunching’ or ‘clustering’ of men in the upper quartiles or deciles of the pay ranges and women in the lower quartiles or deciles. The Respondents cite figures which, they say, show a narrow and steadily diminishing gap in average pay across both grades. There was no real dispute about the figures themselves: the contest was as to what they prove.”

He then went on at paras. 41-42 to summarise the parties’ submissions, in terms which I think it is useful to set out in full, although they have to some extent already been foreshadowed in my consideration of the pleadings:

“41.

Mr Cooper submitted as follows.

“(1)

The methodology apt to test the factor relied upon under s.69(2) must be logically fitted to the Claimants' case.

(2)

The Claimants' case is that the system of pay progression based on length of service produces a disparate distribution by gender across the pay scales.

(3)

Comparison of mean averages (as proposed by the Respondents) does not assist in a proper evaluation of the Claimants' case because (a) it brings into account the large part of basic pay which is unaffected by the s.69(2) factor and (b) it masks differential distribution by smoothing it out.

(4)

By contrast, an analysis of the proportions of men and women within each decile or quartile of the pay scales will serve to identify any disparity and demonstrate whether it is significant.

(5)

Dr Hall's Chi-square analysis lends further support to the Claimants' case.

(6)

Application of the methodology referred to in (4) and (5) shows clearly that women are over-represented as the lower end of the pay scales and under-represented at the upper end.

(7)

Accordingly, without prejudice to any objective justification defence, particular disadvantage is made out.

42.

Mr Linden replied as follows.

(1)

The Claimants' statistical approach based on distribution is flawed because it ignores actual pay, replacing it with an approximation.

(2)

The Claimants' figures also ignore distribution both within and outside any selected decile, quartile or other segment of the population.

(3)

The Claimants' case depends on impermissibly breaking down basic pay, which is indivisible, in order to isolate the notional sub-element referable to length of service.

(4)

The Claimants' approach is arbitrary in that the pools for comparison would have to depend (for the purposes, presumably, of the issue under s.69(2) of individual disadvantage, i.e. whether 'A' is put at a particular disadvantage), on the identity of the particular Claimant under consideration.

(5)

The Claimants' approach is also crude, misleading and, in practical terms, unworkable.

(6)

Dr Hall's Chi-squared analysis is unsound and proves nothing.

(7)

The only proper way to test the Claimants' case on group disadvantage is through measuring mean average pay. On that approach, the undisputed figures conclude the particular disadvantage issue in favour of the Respondents.”

The Judge’s reasons for preferring HMRC’s submissions are at paras. 43-50. Again, I think I should set them out in full.

“43.

… I agree with Mr Linden that distribution cannot be equated with, or allowed to supplant, pay. To state the obvious, the law is concerned with ensuring equal pay for equal work. Ascertaining the distribution of men and women within any particular segment of the overall sample may provide a partial picture of apparent advantage, but it says nothing about actual pay difference, within or outside the segment.

44.

The second fundamental problem with the Claimants’ argument is that it ignores the undisputed reality that, as the average figures show, there is no significant long-term difference between the basic pay of men and women in either of the two grades. Given that reality, it necessarily follows that, in so far as selective analysis based on deciles, quartiles or any other slice of the total grade population reveals a ‘clustering’ phenomenon apparently favouring men over women, there must be a (more or less) counter-balancing advantage the other way, within and/or outside the relevant decile, quartile or other slice. Otherwise, the mean figures would not tell the story they do. This being so, the methodology advanced on behalf of the Claimants cannot be regarded as a reliable instrument and certainly does not substantiate the assertion that women are collectively disadvantaged.

45.

Thirdly, I am much more impressed by the evidence of Dr Brown than that of Dr Hall, whose qualifications and experience do not seem to me, with respect, to equip her fully for the task entrusted to her. I accept the entirety of Dr Brown’s evidence, including his remarks about the chi-squared test. It satisfies me that the statistics relied on by the Claimants are statistically insignificant.

46.

Fourthly, I agree with Mr Linden that it is not permissible to divide basic pay into separate elements in order to challenge the length of service criterion. The sex equality clause under s. 66 modifies any ‘term’ shown to be less favourable than the corresponding ‘term’ of any comparator. The term relied on in this case is that which entitles the Claimants and their comparators to basic pay. Basic pay is indivisible, albeit that the figure in any particular case is explained by several factors of which length of service is one. Moreover, for the reasons already given, the Claimants’ statistical evidence does not make good a theory of particular disadvantage attributable to any proposed ‘sub-term’, even if such were permissible.

47.

Fifthly, I agree with Mr Linden that the distribution-based approach favoured by the Claimants would be unworkable in practice and liable to produce most undesirable results. If it was permissible, no employer could be sure of escaping liability under the equal pay provisions, or at least being put at risk in having to make out objective justification, since 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.

48.

Sixthly, the logic of the Claimants’ case does indeed, as Mr Linden pointed out, admit the possibly that distribution statistics could establish the ‘particular disadvantage’ of one gender group in respect of basic pay in a case where overall mean figures showed, by reference to the same term, that it was substantially advantaged. If the law contemplated a finding of particular disadvantage of gender group X as against gender group Y in circumstances where group X was the better paid of the two, Mr Bumble’s celebrated remark would be entirely apposite. I do not, however, accept that view of the law.

49.

Seventhly, although I accept that the categories of indirect discrimination are not closed and the law is constantly developing, it is a material fact that there is no authority to support the Claimants’ case. I do not refer only to decisions of the higher courts. The Equality and Human Rights Commission (‘EHRC’) Code of Practice on Equal Pay (2011) says nothing about distribution-based analysis. It does refer (para. 177) to the need for employers to calculate average basic pay and total earnings. Likewise the guidance in the EHRC ‘Equal Pay Audit Toolkit’ (‘the Toolkit’), which points out that

‘significant’ differences in pay (as referred to, for example, in the Code, para. 179) are to be reckoned in percentages of basic pay and total earnings.

50.

For all of these reasons, I reject the Claimants’ case on the statistics issue. The result is that they fail to establish particular disadvantage by their chosen route of distribution-based analysis. No alternative is advanced and, as I have noted, they do not dispute that average figures do not disclose significant, long-term differences between the basic pay of men and women in either of the two relevant grades. I am in no doubt that they are right to accept that those figures are against them. The Respondents have demonstrated (not that they bore any legal onus) that the differences in basic pay have been consistently marginal, have diminished over time, and will diminish further as time passes.”

The heart of that reasoning is in the first two points, at paras. 43-44. In short, what the Judge holds is that in principle the only reliable way of demonstrating that women in the grades in question were at a particular disadvantage was to show that there was a significant and long-term difference in the average pay of men and women in them; that HMRC’s analysis measured that directly; and that the Claimants’ analysis measured something different, i.e. distribution, which was at best a less accurate surrogate for the average earnings, and whose results logically could not undermine HMRC’s analysis. The remaining points are essentially subsidiary: some may need some elucidation, but I will return to them when I come to address the issues myself.

I mean no disrespect to the careful judgment of Simler P in the EAT if at this point I say no more than that her reasoning is essentially in line with that of the Judge in the ET. The question for us is whether the ET made an error of law, and I do not wish to lengthen an already long judgment by going over the same ground twice: Simler P’s judgment is of course fully reported ([2018] UKEAT 0183/17, [2018] ICR 1529). I will, however, refer to certain passages in the course of my discussion below.

The Claimants plead what is described as their “overriding ground of appeal” as follows:

“The EAT (and ET) erred in adopting a construction of and/or approach to the concept of ‘particular disadvantage’ under Equality Act (‘EqA’) 2010 which measures the wrong thing, in particular by applying a method (comparison of total average basic pay) which measures the size or value of (average) differences in pay relative to total pay, instead of 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.”

They then go on to plead that “that fundamental error underpins each of the following subsidiary grounds of appeal”. Those subsidiary grounds are:

“(1)

The EAT (and ET) erred in construing the concept of ‘particular disadvantage’ in s. 69 (2)EqA 2010 as turning on a measure of the (average) value of the differences between ‘whole’ terms (in this case, average total basic pay), instead of an analysis of whether the material factor (here, length of service) operates in a way that means it is particularly women who suffer disadvantage (i.e. who receive lower pay as a result of being lower down the pay ranges) regardless of the size or value of the differences. The correct approach requires a differential distribution analysis looking at the relative distribution of men and women along the pay range in each of the relevant grades.

(2)

The EAT (and ET) erred in reaching the purported conclusion that the statistics relied on by the Appellants were not sufficiently significant. Both that conclusion and the evidence of the Respondent’s expert, Dr Brown, relied on in support of it, were premised on the fundamental underlying error that it is the (average) amount or value of pay differentials, as opposed to their disparate incidence as between women and men, that has to be measured. In light of the ET’s finding of fact (recorded by the EAT at [§§30 (vii) EAT judgment]) that the statistics confirmed the clustering of women at the bottom of the pay range and men at the top of the pay range, and in any event based on the statistical evidence of differential distribution as a whole, the only proper conclusion was that significant disparate impact (and consequently particular disadvantage) was established.

(3)

Alternatively, if (contrary to the Appellants’ primary case) a differential distribution analysis was not the sole appropriate method for testing particular disadvantage, the EAT (and ET) erred in (a) holding that it was not open to the Appellants to maintain, and in failing to consider, their alternative case that such disadvantage should be analysed by reference to average variable pay together with evidence of differential distribution; and (b) not concluding that analysis of average variable pay established that length of service had put the Appellants and women generally at a particular disadvantage compared with men.”

I will consider those grounds in turn.

The starting-point must be to appreciate that the (collective) disadvantage relied on in the present case is not of the usual kind. In the typical case the disadvantage is the non-receipt of a defined benefit – for example, qualification for a job, or for promotion, or an extra day’s holiday, or a specific level of pay. In such a case particular disadvantage, or disparate impact, is measured by comparing the proportions of men and of women in the pool who enjoy that benefit. Here, however, the disadvantage of which the Claimants complain is not binary. There is a wide range of individual salaries within the pay bands, and it is impossible to say that this or that figure represents the point below which an employee is disadvantaged, or above which they are advantaged. This is what Dr Brown means by saying that “base pay … was a semi-continuous (Footnote: ⁹) variable, [not] a categorical value”: see para. 44 above. There is thus no fixed touchstone of disadvantage.

Mr Cooper submitted that where resort to the traditional approach is not available the touchstone for proving (collective) particular disadvantage must be the incidence of disadvantage – that is, relatively howmany women are disadvantaged (or, to put the same thing another way, relatively how often women are disadvantaged). If you can show that proportionately more women are towards the bottom, and (therefore) proportionately more men towards the top, of the scale in question, that necessarily shows that women are disadvantaged. It is, as he put it in his skeleton argument, logically “the only permissible choice”.

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 Mr Cooper. 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.

The difference between the two approaches can be illustrated by a very stark example. 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:

I do not believe that to find in such a case that the women in the pool, viewed as a group, are at a “particular disadvantage” is consistent with the nature of the statutory exercise, which depends on the impact of the factor in question on pay. In my view that means that it is necessary to bring into the analysis the actual amounts paid to (all) the women in the pool. That is, as I understand it, the point being made by the Employment Judge at para. 43 of his reasons, and developed at para. 46.

Of course in the present case the Claimants have not simply divided the pool into two halves – lower and upper – but have done their analysis by quartiles and deciles. But the conceptual problem is the same. There could in principle be an uneven distribution of salary points, as between men and women, within each quartile or decile, potentially masking the fact that one or other gender is disproportionately represented at the top or bottom of each segment. 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.

Mr Cooper submitted that an approach which, by taking an average, focuses on the position within the whole pool “smooths out” gender differences which occur within it and (as he put it in his skeleton argument) “masks disadvantage experienced by a significant proportion of women found at the bottom of the grade by lumping them together with other women who may not be so disadvantaged …”. That is a tendentious way of putting it, not least because it assumes that the women at the lower end of the band are indeed suffering “disadvantage”, which is assuming what he has to prove. I accept of course that applying my approach to the example at para. 64 means that the fact that there are disproportionately more women at the bottom of the band (55:45) is outweighed by the fact that there are (even) more disproportionately more women at the top (45:0). But I do not think that that can be described as masking some disadvantage suffered by women as a group. Rather, what it means is that the application of the factor has led to some women being disproportionately at the top end of the pay range (however defined) and to others being disproportionately at the bottom end (however defined). In such a case, I do not think it is possible to say that women as a whole are disadvantaged without taking into account what the women in question, and likewise the men in comparison with whom they are said to be disadvantaged, are in fact paid.

The truth is that establishing the existence of disadvantage by reference to “distribution” in fact means abandoning any usable reference point: there is no rational basis for selecting one part of the pool rather than another in which to look for disadvantage. What if it were only in the bottom decile that there were disproportionately more women than men and the disproportion were the other way round in the remaining nine ? The logic of Mr Cooper’s position is that “particular disadvantage” could be established by reference to the bottom decile only because it

Mr Cooper referred to the passage which I have quoted at para. 15 above from Lady Hale’s judgment in Essop, which makes the point that a PCP may be indirectly discriminatory even if not all women are disadvantaged by it. But that is to confuse the simple binary case with the case of a (semi-) continuum like the present. In the kinds of cases to which Lady Hale was referring there is no doubt that the women who cannot meet the height requirement, or work full-time, are disadvantaged, and who they are; and the fact that they may be only a minority of the women in the pool is neither here nor there. But the problem is that in our case there is no equivalent touchstone of disadvantage.

There is another problem with Mr Cooper’s approach, though it is ultimately simply an aspect of the underlying error of principle. If a pay system can be found to be (prima facie) indirectly discriminatory on the basis of a distribution analysis (by reference to whatever part of the pool a claimant may choose to focus on), it will be impossible in practice for an employer ever to be sure that its system was not potentially unlawful. Nor, since a distribution analysis affords no way of putting a number on the extent of the disadvantage suffered are there any fixed points for the purpose of examining justification or identifying an appropriate remedy. Only averages give a workable tool for measurement.

Mr Cooper sought to support his proposition by referring to a number of authorities – specifically Seymour-Smith (at paras. 59-60 of the judgment of the Court (p. 597 FH)), Barry v Midland Bank plc[1999] 1 WLR 1465, per Lord Nicholls at p. 1474 EH), Home Office v Bailey (per Peter Gibson LJ at paras. 29-31 (pp.1069-70)) and Essop (per Lady Hale at paras. 23-25 and 27-29 (pp. 1353-4)). However, none of those cases were concerned with the present situation: in each the disadvantage was binary in character, and the passages in question do no more than consider aspects of the conventional approach.

Mr Cooper also submitted that the principle affirmed by the CJEU in CHEZ that particular disadvantage need not be “serious” meant that the “extent” of the differential must be treated as irrelevant. But, again, that is a different point. The point of taking an average of the earnings of the men and women in the pool is to establish whether women are at any disadvantage in the first place: the “CHEZ point” becomes relevant only once it is established that they are.

In fact, such authority as there is goes the other way. In Handels-og Kontorfunktionærernes Forbund i Danmark v Dansk Arbejdsgiverforening (Danfoss A/S) case 109/88,[1991] ICR 74, the ECJ said, at para. 16 (p. 80 C-D):

“where an undertaking applies a system of pay which is totally lacking in transparency, it is for the employer to prove that his practice in the matter of wages is not discriminatory, if a female worker establishes, in relation to a relatively large number of employees, that the average pay for women is less than that for men [emphasis supplied]”.

That passage is repeated in Enderby, at para. 14 of the Court’s judgment (p. 161 D-E). I would not in fact put great weight on it in this context because, again, the Court was not concerned with the particular issue before us; but it is legitimate at least to note that it apparently regarded comparison of average earnings for men and for women as the natural way of identifying discrimination within a particular pool.

I should make it clear that it does not follow that a distribution analysis of the kind done by Dr Hall is of no value. In a case where the average pay of women within a pool is less than that of men, there are likely be disproportionately more men at the higher end of the pay band and disproportionately fewer at the lower end, subject to the important qualification about the evenness of the distribution discussed above. But a distribution analysis which shows that that is the case is an inherently inferior surrogate for the more reliable and accurate analysis provided by the comparison of averages. It cannot in those circumstances be right, where the figures for an analysis of average pay are available, to eschew them in favour of a less accurate analysis. It is important to recall that this part of the argument is put squarely on the ground of principle; how particular disadvantage might be proved in a case where figures for average pay are not available is not something on which we need express a view.

I would therefore reject the Claimants’ case as to the correct approach in principle to the assessment of particular disadvantage in the present case and dismiss the “overriding ground of appeal”. I believe that my reasoning essentially corresponds to that of Simler P at paras. 48-50 of her judgment.

It might be thought to follow that I need say nothing about the three “subsidiary” grounds of appeal, but in fact they do raise some points with which I need to deal.

As pleaded, this ground does not appear to go much beyond the fundamental point of principle which I have already considered. It does, however, incorporate a particular point about how any average differential should be calculated which is also central to the third subsidiary ground.

The premise of the point is that the discrimination complained of in this case affects only a part of total pay: everyone receives at least the minimum rate for the band in question, and length of service can only affect the part of total pay between that figure and the maximum. That being so, Mr Cooper contends, to focus on the average differential in total pay understates any disparate impact which length of service may be shown to have. The correct picture can only be seen by focusing on what he called “the variable element” in pay – that is, the part covered by the pay band. In his skeleton argument (though in fact in the part concerned with his second subsidiary ground) he set out a table showing that the average differentials shown in HMRC’s tables, which were (very broadly) around 2%, would have been (again, broadly) around 20% if calculated only in relation to the pay band range. That distortion may lead to factors which unquestionably have a disparate impact between the sexes, and which may lead to differentials of some thousands of pounds in individual cases, being treated as non-discriminatory.

The same argument was advanced below. Simler P dealt with it at paras 52-56 of her judgment. At para. 52 she says:

“… I do not accept Mr Cooper's contention that where average pay is to be used as a proxy indicator for differential distribution (or relied on together with differential distribution), as a matter of law this can only be by reference to differences in average pay within the variable part of the pay only, in other words in respect of amounts above the pay grade minimum (the so-called variable pay element). This argument amounts to a contention that basic pay can be subdivided into notionally separate elements representing, in effect, base salary and a length of service element, even where there is no evidence that as a matter of contract or practice such a distinction is drawn within basic pay; and where the Employment Tribunal found as a fact that basic pay in this case was indivisible and could be explained by several factors of which length of service was one. Neither the EA 2010, nor authority supports Mr Cooper's approach. Nor does the Code or Commission guidance support it.”

She develops those points in the following paragraphs. At paras. 53-54 she refers to the line of authorities which establish that any comparison under (what is now) section 66 of the Act must be on a term-by-term basis (specifically, Hayward v Cammell Laird Shipbuilders Ltd (no 2)[1988] 1 AC 894 and Brownbill v St Helens Hospitals NHS Trust[2011] EWCA Civ 903, [2012] ICR 68); and points out that the relevant term here is “pay”. At paras. 55-56 she observes that if the analysis were confined to the variable element in pay the percentage differential would be dependent on the relative length of the pay band in any particular case. As she says at para. 55:

“Since it is well established that long pay scales can lead to unequal pay and a means of tackling pay inequality is to shorten the pay scales, it would be a perverse result that the shorter the pay scale the easier it is to establish particular disadvantage.”

In response to that reasoning Mr Cooper submitted that the fact that a term-by-term analysis was required for the purpose of section 66 did not mean that it should be read over into the exercise required by section 69. As for the allegedly distorting effect identified by Simler P, he submitted that HMRC’s approach itself involved the more fundamental distortion to which I have already referred.

I agree with Simler P that Mr Cooper’s argument must be rejected. The essence of the disadvantage alleged is, straightforwardly, that women in the relevant group are paid less than men. To express it in terms of “disadvantage in respect of the top (say) £10,000 of the pay range” is an artificial and unjustified subtlety in circumstances where, as she says at para. 52, no distinction between different categories of basic pay is recognised either by contract or by practice. I think Mr Cooper is probably right to say that the Hayward/Brownbill line of authorities is not directly in point, though they do lend some indirect support to the basic point that it is inappropriate to start dividing up “pay” in this context. However, I do see real force in the point made by Simler P at paras. 55-56 of her judgment, which reinforces the conclusion that I would in any event reach.

Having said all that, I do not believe that the issue has the significance which Mr Cooper attaches to it in this context. I do not believe that calculating the average by reference to total (basic) pay will prevent claimants from establishing particular disadvantage on the basis only that the disadvantage is small in percentage terms. As noted at para. 20 above, there is a distinction in the authorities between the approach to establishing disparate impact/particular disadvantage, for which a “considerable” difference is required in the proportions of men and women affected, and the approach to the size of the disadvantage itself, which need not be “serious”. The latter approach is the relevant one in the present case. In particular, as noted at para. 24 above, the percentage figures mentioned in the Commission’s “toolkit” do not represent any kind of minimum threshold. In response to questions from the Court, Mr Linden accepted that if it were clearly established that the use of length of service as a criterion led to women being paid, on average, less than men, it would not be an answer to say that the differential was less than 3%: he said that that was clear from CHEZ. That being so, there is no need to focus on the variable element in the Claimants’ pay in order to get over the supposed threshold.

Again, as pleaded this ground does not raise any point which I have not already discussed. But it is the vehicle, in Mr Cooper’s skeleton argument, for a challenge to the Employment Judge’s observation, at para. 44 of the Reasons, that if the distribution analysis shows clustering of the kind relied on in any particular segment of the pool there must be “a ... counter-balancing advantage the other way” within or outside that segment – “otherwise, the mean figures would not tell the story they do”. He says that there is no evidence of any such “counter-balancing advantage”. But I think that the Judge was simply making the same point that I have made at para. 67 above: the average necessarily gives the most accurate picture of collective advantage or disadvantage.

This ground on its face goes beyond the point of principle which I have already addressed under the “overriding ground”, because it relies on an alternative case which it is said that the ET and EAT wrongly refused to entertain.

As pleaded, the alternative case relies on the submission that “disadvantage should be analysed by reference to average variable pay together with evidence of differential distribution”, which I have rejected at paras. 79-81 above. In the EAT Mr Linden objected that the alternative case had not been advanced in the ET and that it was too late for it to be advanced for the first time on appeal. Mr Cooper referred to a footnote in the October 2014 Particulars which showed that the point had been made below and said that he had also referred to it in his closing submissions before the Judge – though only, as he candidly acknowledged, as an argument to support the distribution-based analysis rather than as an alternative in its own right. Simler P considered that question carefully at paras. 58-62 of her judgment. She concluded, at para. 62:

“This was a complicated case which evolved over time and the parties' respective cases on particular disadvantage were reduced to writing and refined, and were supported by statistics and expert analysis. In the absence of any written argument to the effect that the Claimants were advancing a positive case in reliance on data containing average variable pay differentials together with or to support their distribution analysis, it does not surprise me that EJ Snelson understood the point advanced in submissions as a point designed only to attack the Respondent's approach rather than to support an alternative positive case. In all the circumstances, if the variable pay argument were sound, I do not consider that it was properly raised before the Tribunal, or that it can fairly be raised now to demonstrate a fatal flaw in the Judgment.”

In the course of his oral submissions before us Mr Cooper made it plain that he would wish, if permitted, to advance a variant of the alternative case referred to in this ground. In particular, in his submissions in reply he referred to Mr Linden’s concession, recorded at para. 82 above, that particular disadvantage could be proved by a differential in average of pay of less than 3%. HMRC’s own figures showed a differential over the entire period of, broadly, 2%; and that was an ample basis for a finding of particular disadvantage.

Mr Linden submitted that it was still less open to Mr Cooper to advance in this Court, for the first time, a case based on the difference in the average total pay of men and women in the two grades than it had been to advance a case in the EAT a case based on differences in the variable element. In response, Mr Cooper referred to the same materials as he had relied on in the EAT.

In my view it would indeed be wrong for this Court to allow the appeal on the basis of the alternative case that Mr Cooper now seeks to advance. I endorse what Simler P said about the similar issue below, but the objection is a fortiori as regards a comparison of average total pay. As appears from the account which I have given of the history of their pleaded case, the Claimants have from at least the October 2014 Particulars onwards clearly disavowed reliance on HMRC’s figures for average pay as a way of proving their case. Para. 15 of those Particulars could not be more explicit. It would be extraordinary to permit the volte-face which Mr Cooper now, albeit only by way of alternative, seeks to make. He said that it was only in this Court that Mr Linden had made the concession that there was no “3% threshold”. But that is not a good answer. It was the Claimants, not HMRC, who had treated the 3% figure mentioned in the toolkit as a threshold (see paragraph 7 (c) of the original Details of Claim); and there was no obligation on HMRC to question whether they needed to do so. The Claimants had HMRC’s figures from the very early stages of the litigation, and they could have based an alternative case on them if they had chosen to do so.

There are, it is true, exceptional cases where a party is permitted to switch horses on appeal, where the issue is one of pure law and this Court could safely determine the new case for itself. But I do not believe that that is the case here; and indeed I understood Mr Cooper to accept in the course of his reply submissions that if he were permitted to rely on this alternative the case would have to be remitted to the ET. It is not enough for the Claimants simply to rely on the differential shown in HMRC’s figures: it is necessary that they should show that it was the result of the factor on

I would add, with all due caution because the issue was not live before us, that even if the Claimants had proved particular disadvantage HMRC might well have had a good case on justification. As I observed in Naeem (see the note referred to at para. 24 above), it would be harsh if an employer who made a deliberate attempt to increase the diversity of its workforce made a rod for its own back because it thereby created a disproportion in the relative lengths of service of men and women in the workforce. I do not suggest that that is a complete answer, since tribunals are entitled to look critically at pay ranges which are too long or where the rate of progress is unreasonably slow; but the Employment Judge’s observations at para. 50 of his

I would dismiss the appeal.