Asda Stores Ltd v Brierley & Ors

When the 1970 Act first came into force, which was on 29 December 1975, the right to equal pay was accorded by section 1 (2) in two cases – (a) where a woman was employed on “like work” with a man “in the same employment” and (b) where she was employed on work “rated as equivalent” with that of a man in the same employment. With effect from 1 January 1984, as a result of a decision of the ECJ in Commission v UK (Case 61/81), [1982] ICR 578, a third case was added – (c) where a woman was employed on work “of equal value” with that of a man in the same employment.

The phrase “in the same employment”, which is common to all three cases, was defined in section 1 (6) of the 1970 Act as follows:

“… [F]or purposes of this section … men shall be treated as in the same employment with a woman if they are men employed by her employer or any associated employer at the same establishment or at establishments in Great Britain which include that one and at which common terms and conditions of employment are observed either generally or for employees of the relevant classes.”

Section 1 (3) of the 1970 Act afforded an employer a defence where the differential which is the subject of the claim is shown to be “due to a material factor which is not the difference of sex”.

The 2010 Act reproduces essentially the same scheme as the 1970 Act but a different drafting technique is adopted. Section 65 (1) provides for the same three grounds of comparison as under section 1 (2), but the comparator is defined simply as “B” rather than “a man in the same employment”. The scope of permitted comparisons is prescribed, so far as relevant for our purposes, by section 79 (2)-(4), which reads:

“(2)

If A is employed, B is a comparator if subsection (3) or (4) applies.

(3)

This subsection applies if —

(a)

B is employed by A's employer or by an associate of A's employer, and

(b)

A and B work at the same establishment.

(4)

This subsection applies if —

(a)

B is employed by A's employer or an associate of A's employer,

(b)

B works at an establishment other than the one at which A works, and

(c)

common terms apply at the establishments (either generally or as between A and B).”

Section 80 (2) (a) provides (so far as material for our purposes) that “[t]he terms of a person's work are … the terms of the person's employment that are in the person's contract of employment”. The equivalent to section 1 (3) of the 1970 Act, albeit differently worded, is at section 69.

It will be observed that although the elements of the relevant parts of section 1 (6) of the 1970 Act are mostly reproduced in substantially identical terms in section 79 (2)(4) of the 2010 Act, the parenthesis in sub-section (4) (c) uses the phrase “as between

Article 119 of the Treaty of Rome began:

“Each member state shall during the first stage ensure and subsequently maintain the application of the principle that men and women should receive equal pay for equal work.”

There are two further short paragraphs by way of amplification but they are not material for our purposes.

In 1975 the Council of Ministers adopted Council Directive 75/117/EEC. This defined “the principle of equal pay for men and women outlined in Article 119 of the Treaty” as (by article 1):

“for the same work or for work to which equal value is attributed, the elimination of all discrimination on grounds of sex with regard to all aspects and conditions of remuneration”.

Article 2 required member states to introduce domestic legislation whereby employees could enforce the principle of equal pay. In Worringham v Lloyds Bank Ltd (case 69/80) [1981] ICR 558 the ECJ said, at para. 21 of its judgment (p. 589E):

“Although article 1 of the directive explains that the concept of ‘same work’ contained in the first paragraph of article 119 of the Treaty includes cases of ‘work to which equal value is attributed’, it in no way affects the concept of ‘pay’ contained in the second paragraph of article 119 but refers by implication to that concept.”

The necessary implication of that is that the phrase “equal pay for equal work” in article 119 covers “work to which equal value is attributed”.

In Defrenne v Sabena (case no. 43/74) [1976] ICR 547 the ECJ held that article 119 gave employees a directly enforceable right to equal pay in some circumstances notwithstanding a failure by the relevant member state to implement the Directive. The Court said, at paras. 18-24 of its judgment (pp. 566-7):

“18.

For the purposes of the implementation of these provisions a distinction must be drawn within the whole area of application of article 119 between, first, direct and overt discrimination which may be identified solely with the aid of the criteria based on equal work and equal pay referred to by the article in question and, secondly, indirect and disguised discrimination which can only be identified by reference to more explicit implementing provisions of a community or national character.

19.

It is impossible not to recognise that the complete implementation of the aim pursued by article 119, by means of the elimination of all discrimination, direct or indirect, between men and women workers, not only as regards individual undertakings but also entire branches of industry and even of the economic system as a whole, may in certain cases involve the elaboration of criteria whose implementation necessitates the taking of appropriate measures at community and national level.

20.

This view is all the more essential in the light of the fact that the community measures on this question, to which references will be made in answer to the second question, implement article 119 from the point of view of extending the narrow criterion of ‘equal work’, in accordance in particular with the provisions of Convention No. 100 on equal pay concluded by the International Labour Organisation in 1951, article 2 of which establishes the principle of equal pay for work ‘of equal value’.

21.

Among the forms of direct discrimination which may be identified solely by reference to the criteria laid down by article 119 must be included in particular those which have their origin in legislative provisions or in collective labour agreements and which may be detected on the basis of a purely legal analysis of the situation.

22.

This applies even more in cases where men and women receive unequal pay for equal work carried out in the same establishment or service, whether public or private.

23.

As is shown by the very findings of the judgment making the reference, in such a situation the court is in a position to establish all the facts which enable it to decide whether a woman worker is receiving lower pay than a male worker performing the same tasks.

24.

In such situation, at least, article 119 is directly applicable and may thus give rise to individual rights which the courts must protect.”

The Treaty of Rome was amended by the Nice Treaty in 2003, at which point article 119 became article 141 and was somewhat reformulated. Paragraph (1) read:

“Each Member State shall ensure that the principle of equal pay for male and female workers for equal work or work of equal value is applied.”

The new drafting thus spells out what the Court had in Worringham held was implicit in the old.

Article 157 (1) of the TFEU is in identical terms to article 141.

We are at this stage, as I have said, concerned only with whether the Claimants are entitled to compare themselves with employees in the distribution operation at all. If the decision is that they are, it remains to be established whether they do work of equal value with their comparators and, if so, the extent of any differential and whether Asda has “material factor” defence. As to that issue, they put their case on two bases, which I take in turn.

As a matter of domestic law, the Claimants acknowledge that they do not work in the same establishments as any of their comparators, because Asda’s stores and its depots are entirely separate; but they claim that common terms of employment apply at both, either generally or as between themselves and their comparators, so that they can rely

Alternatively, the Claimants rely on the direct effect of EU law, as established by Defrenne. They contend, in reliance on the decision of the CJEU in Lawrence v Regent Office Care Ltd(C-320/00), [2003] ICR 1092, that comparison is possible in any case where there is a “single source” for the terms of employment of the claimant and the comparator, and they say that that is plainly so in their case since they are both employed by the same employer. As to this, Asda does not accept that the effect of Lawrence is that as a matter of EU law an employee can compare herself with any other employee of the same employer. But it contends that, even if it is, EU law does not have direct effect in a claim under section 65 (1) (c) (i.e., equal value) because a claim of that kind – unlike a claim under heads (a) or (b) (i.e. like work or work rated as equivalent) – falls into the category recognised in para. 18 of the judgment in Defrenne where more specific implementing legislation is required.

In the ET and the EAT the primary focus was on the direct effect case, and they both dealt with the “single source” issue in that context. I prefer to take the domestic law case first, though I will consider issues of EU law (and specifically single source) in that context to the extent that they inform the interpretation of the domestic legislation.

DOMESTIC LAW

There are two decisions of the House of Lords and one of the Supreme Court addressing the meaning and effect of section 1 (6) of the 1970 Act as regards crossestablishment claims. Although we are now primarily concerned with section 79 (4) of the 2010 Act those authorities remain the essential basis for any analysis, and I take them in turn.

In Leverton v Clwyd County Council [1989] AC 706 the applicant was employed by the council as a nursery nurse at an infant school. She brought a claim under the then new equal value provisions of the 1970 Act seeking equal pay with a number of male clerical workers in various council offices: none of them worked at her school. Both she and they were classified by the council as working in its Administrative Professional Technical and Clerical (“APT&C”) Services; and the terms and conditions for both nursery nurses and clerical workers were set out in the same national collective agreement (“the Purple Book”), determined by the National Joint Council for Local Authorities’ Administrative Professional Technical and Clerical Services. The Agreement set out a number of salary points on a single spinal column and assigned different kinds of employee to different ranges of the column. The applicant sought to be paid in accordance with the same range as her comparators.

The applicant’s claim was dismissed by a majority in the industrial tribunal (“the IT”), on the basis both (a) that she was not in the same employment as her comparators because none of them was employed at the same establishment as her and (b) that, even if she was, the difference between her terms and theirs was due to a material factor other than the difference of sex. The EAT dismissed her appeal, though only on the basis of the same employment point. The Court of Appeal, by a majority, upheld the decision of the EAT. The House of Lords dismissed her appeal, but only on the material factor ground. On the same employment point, which is what is relevant for our purposes, it upheld the view of the minority members in both the IT and the Court of Appeal that the applicant and her comparators were in the same employment.

I need not set out the reasoning of the IT or the EAT: the reasoning of the minority member in the IT which the House of Lords approved is quoted in the passage from the speech of Lord Bridge which I set out below. I should, however, set out the relevant passage from the dissenting judgment of May LJ in the Court of Appeal. At p. 717 D-E he says that the essential question is the correct construction of the phrase “common terms and conditions”. He continues, at pp. 717-9 (I add paragraph numbers for ease of reference):

“[1] First, I respectfully cannot accept the industrial tribunal's construction of this phrase. There is in my opinion no warrant for construing ‘common’ as ‘broadly common’. The only two legitimate constructions of the phrase are either that ‘common’ means ‘the same’, or alternatively that the terms and conditions of all the relevant employees belong equally to more than one, are generally known, or are in general use, to employ some of the definitions in the Shorter Oxford English Dictionary. It may well be, and is indeed the position in the present case, that they are contained in a book such as the ‘Purple Book’. Further, as the words themselves contemplated, such general terms and conditions may be common to all the employees of the particular employer, or common to the relevant classes.

[2]

I reject the first of the above meanings of the phrase for the reasons given by the industrial tribunal … . If it were the correct construction then the consequent required identity of the terms and conditions of employment of the applicant and comparators would defeat the whole purpose of the legislation.

[3]

Thus I am driven to adopt the alternative construction of the phrase ‘common terms and conditions’ to which I have just referred. Though so driven, however, it seems to me that by this route one does arrive at the purposive destination to which the draftsman was directing one. In support of this construction it is to be observed that the phrase is just ‘at which common terms and conditions … are observed’, not ‘terms and conditions common to the applicant and her comparators are observed’, or anything similar. Thus one is led to a concept of terms and conditions in a general sense, applicable across the board, or across the particular boards of the relevant classes of employee.

[4]

In other words, the object of section 1(2)(c) and 1(6) is to ensure that a woman doing work of the same value ‘in the same employment’ enjoys equal terms and conditions. If the woman and the man are employed by the same or an associated employer at the same establishment then they are, well understandably, to be treated as in the same employment under the first two lines of the last phrase of section 1(6). ...

[5]

It may be, however, that the man to whose work the woman alleges her work for the same or any associated employer is of equal value may be employed at a different establishment of the employer or associated employer. Then she is to be treated as in the same employment as the man if her establishment and his establishment are in the same concern and if at those establishments (this follows from the use of the words ‘at which’ in the subsection) common terms and conditions are observed either generally or for employees of the relevant classes, that is to say, the class of employee of which the woman is a member and the class of which the man is a member. Before the woman can have an equality claim there must, either throughout the employer's business, or at least in relation to the classes of employee to which each belong, uniformity of employment. A woman working in an establishment A, ex hypothesi doing work of the same value as the man, cannot have an equality claim in respect of that man working in establishment B, cannot to that end claim to be in the same employment as the man, unless at least she and her fellow employees doing the same work in establishment A and the man and his fellow employees doing the same work in establishment B each are subject to common terms and conditions. Otherwise either the woman or the man or both might be a ‘rogue’ enjoying uncommon terms and conditions of employment, possibly because of the particular establishment in which they work.”

In the House of Lords Lord Bridge delivered the only substantial speech on the same employment point. After setting out the history, he summarised the rival contentions as follows, at pp. 744-5:

“On the question whether the appellant was in the same employment as the comparators working at different establishments, the view which prevailed with the majority of the industrial tribunal, the Employment Appeal Tribunal, and the majority of the Court of Appeal was that the comparison called for by section 1(6) was between the terms and conditions of employment of the appellant on the one hand and of the comparators on the other and that it was only if this comparison showed their terms and conditions of employment to be ‘broadly similar’ that the test applied by the phrase ‘common terms and conditions of employment’ in section 1(6) was satisfied. The majority of the industrial tribunal, affirmed by the Employment Appeal Tribunal and the majority of the Court of Appeal, held that the difference in this case in working hours and holidays was a radical difference in the ‘core terms’ of the respective contracts of employment which prevented the comparison from satisfying the statutory test. The contrary view embraced by the dissenting member of the industrial tribunal and by May L.J. in the Court of Appeal was that the comparison called for was much broader, viz. a comparison between the terms and conditions of employment observed at two or more establishments, embracing both the establishment at which the woman is employed and the establishment at which the men are employed, and applicable either generally, i.e. to all the employees at the relevant establishments, or to a particular class or classes of employees to which both the woman and the men belong. Basing himself implicitly on this view, the dissenting member of the industrial tribunal expressed his conclusion in the matter tersely. Having referred to the Purple Book, he said:

‘3. Within that agreement there are nine sections and numerous clauses. They do not apply, with few exceptions, to any particular grade. It is clearly a general agreement and not specific to any particular group or class of employee.

4.

It is, in my opinion, beyond doubt that the applicant and the comparators are employed on common terms and conditions, i.e. the A.P.T. & C. agreement, and clearly it is within the provisions of section 1(6).’”

At p. 745 D-E Lord Bridge says that he prefers the minority view in the courts below.

In the two following paragraphs Lord Bridge goes on to explain why what he regards as the clear meaning of the statutory language also accords with “the manifest purpose of the legislation”. He says, at pp. 745-6:

“That purpose is to enable a woman to eliminate discriminatory differences between the terms of her contract and those of any male fellow employee doing like work, work rated as equivalent or work of equal value, whether he works in the same establishment as her or in another establishment where terms and conditions of employment common to both establishments are observed. With all respect to the majority view which prevailed below, it cannot, in my opinion, possibly have been the intention of Parliament to require a woman claiming equality with a man in another establishment to prove an undefined substratum of similarity between the particular terms of her contract and his as the basis of her entitlement to eliminate any discriminatory differences between those terms.

On the construction of section 1(6) which I would adopt there is a sensible and rational explanation for the limitation of equality claims as between men and women employed at different establishments to establishments at which common terms and conditions of employment are observed. There may be perfectly good geographical or historical reasons why a single employer should operate essentially different employment regimes at different establishments. In such cases the limitation imposed by section 1(6) will operate to defeat claims under section 1 as between men and women at the different establishments. I take two examples by way of illustration. A single employer has two establishments, one in London and one in Newcastle. The rates of pay earned by persons of both sexes for the same work are substantially higher in London than in Newcastle. Looking at either the London establishment or the Newcastle establishment in isolation there is no sex discrimination. If the women in Newcastle could invoke section 1 of the Act of 1970 to achieve equality with the men in London this would eliminate a differential in earnings which is due not to sex but to geography. Section 1(6) prevents them from doing so. An employer operates factory A where he has a long standing collective agreement with the ABC union. The same employer takes over a company operating factory X and becomes an ‘associated employer’ of the persons working there. The previous owner of factory X had a long standing collective agreement with the XYZ union which the new employer continues to operate. The two collective agreements have produced quite different structures governing pay and other terms and conditions of employment at the two factories. Here again section 1(6) will operate to prevent women in factory A claiming equality with men in factory X and vice versa. These examples are not, of course, intended to be exhaustive. So long as industrial tribunals direct themselves correctly in law to make the appropriate broad comparison, it will always be a question of fact for them, in any particular case, to decide whether, as between two different establishments, ‘common terms and conditions of employment are observed either generally or for employees of the relevant classes’. Here the majority of the industrial tribunal misdirected themselves in law and their conclusion on this point cannot be supported.”

Mr Jeans submitted that the two examples given in the second paragraph of that passage showed the essence of Lord Bridge’s ratio, namely that terms were only common between two establishments when the “regimes” at both had a common genesis, in terms of either history or geography. I do not accept that. As I have noted, Lord Bridge had already stated his fundamental ratio, based on what he took to be the plain meaning of the statutory language. This passage simply gives illustrations of circumstances in which “non-common” terms and conditions might be observed as between different establishments of the same employer. (Footnote: ¹) The word “regime” is perfectly appropriate for the point that he was making, but it is not useful as an allpurpose alternative to the statutory language.

I do not think that there is any doubt about the ratio of Leverton, but because the history of this case suggests that there is still some uncertainty it may be useful if I state it concisely here. The essential concept conveyed by the reference to “common terms and conditions … [being] observed” at two establishments is that no distinction is made between the establishments as regards what terms and conditions apply (at least for the relevant classes). That is what May LJ, whose reasoning was approved by the House of Lords, means by referring to terms and conditions applying “across the board” (para. [3] in the passage quoted). And that is why Lord Bridge says that the existence of a single collective agreement like the APT&C agreement is the paradigm case of common terms and conditions being observed at different establishments, because such an agreement (typically) prescribes terms without reference to a particular workplace. In their submissions in the House of Lords counsel for the applicant – Mr Anthony Lester QC and Mr David Pannick – described the relevant test as being whether the terms and conditions “differ according to the establishment at which that person is employed” and continued: “in other words, are the terms and conditions applicable to the relevant jobs irrespective of the establishment at which the employees work?” (p. 731 C-D). That is in my view the essence of the test adopted by the House in Leverton.

That ratio involves the explicit rejection of the approach taken by the majority in the IT and this Court, which had involved considering the degree of similarity or otherwise between the terms of the applicant and the comparator(s): see, most explicitly Lord Bridge’s statement that Parliament did not intend that a woman “should have to prove an undefined substratum of similarity between the particular terms of her contract and [her comparator’s]”. Although this approach should therefore be dead, the history of the present case shows that it is not willing to lie down.

It also follows from the correct approach that it is not necessary that any employees in the comparator’s class should actually be employed in the claimant’s establishment. If, to take Lord Bridge’s paradigm case, the employer has negotiated a collective agreement providing for the terms and conditions of its employees (or at least those in the relevant classes) at all its establishments common terms and conditions for those employees will, necessarily, apply at those establishments irrespective of who actually works there. One way of making the same point is to say that if someone in the comparator’s class was employed at the claimant’s establishment he would enjoy the terms in question. Although that hypothesis is at the centre of the analysis in the later cases considered below, it is not explicitly deployed in Lord Bridge’s speech in Leverton: however, the substantive point is illustrated by the facts of the case, since there were no clerical workers employed at the applicant’s school. (Footnote: ²)

In British Coal Corporation v Smith [1996] ICR 515 the applicants were female canteen-workers and cleaners employed at British Coal establishments (Footnote: ³) seeking the same terms and conditions as surface mineworkers at a number of pits. Some of the applicants worked at the same pits as their comparators, but others did not and needed to make cross-establishment claims. Terms and conditions of employment for canteen-workers and cleaners were governed by a single national collective agreement which applied at every establishment, but the position as regards surface mineworkers was not so straightforward. Most of their terms and conditions were also governed by a single national collective agreement, but – crucially to the issues on the appeal – in two particular respects that was not the case: entitlement to concessionary fuel was determined on an area-by-area basis and incentive bonus payments (which might represent as much as 15% of earnings and were accordingly certainly not de minimis) were the subject of pit-by-pit negotiation.

It was British Coal’s case that the differences as regards concessionary coal and incentive bonus meant that mineworkers could not be said to be employed on common terms and conditions: it was necessary for the purpose of section 1 (6) that all terms and conditions be common (subject possibly to a de minimis exception). That issue had not arisen in Leverton because it was not contended that the APT&C agreement was anything other than comprehensive as regards the range of terms covered.

The IT rejected that argument. Its essential reasoning is set out at pp. 528-9 in the speech of Lord Slynn in the House of Lords. I need not reproduce it in full here. In short, it held that the terms and conditions of mineworkers could still be regarded as common across all British Coal’s establishments notwithstanding the limited degree of local variation as regards incentive bonus and concessionary coal. All mineworkers were entitled to incentive bonus structured in essentially the same way, albeit that there were differences in the detailed implementation; and, as it put it, “local variations do not destroy the centralised, industry-wide nature of the entitlement” (p. 529E). Likewise the fact that in some areas the entitlement to concessionary coal was higher than in others did not mean that common terms and conditions were not observed between the areas in question.

That decision was upheld by the EAT but reversed in this Court. It was, however, restored by the House of Lords. Lord Slynn delivered the only speech. He described the IT as having taken “a broad common sense approach … in accordance with the speech of Lord Bridge [in Leverton]” (p. 528 G-H).

In my view the ratio of Smith is, and is only, that in considering whether common terms and conditions are observed as between employees in the comparator’s class (Footnote: ⁴) at two establishments – that is, the claimant’s establishment and the comparator’s – it is not necessary that all the terms are common: “broad commonality” is enough (Footnote: ⁵). Nothing else was in issue. However I should refer to three other points.

First, Lord Slynn recognised, though the point was not in dispute, that it was not necessary for the purpose of a cross-establishment claim that anyone in the comparator’s class should actually be employed at the claimant’s workplace. He said, at p. 526 F-G:

“If there are no such men at the claimant's place of work then it has to be shown that like terms and conditions would apply if men were employed there in the particular jobs concerned.”

That is of course the hypothesis to which I have already referred in the context of Leverton: see para. 36 above.

Secondly, it was common ground in Smith that what was required was a comparison between the terms applied to employees in the comparator class employed (a) at the comparator’s establishment and (b) (though possibly only hypothetically) at the claimant’s establishment – i.e. between the terms enjoyed by mineworkers at different establishments. The comparison is between the terms applicable at the twoestablishments, not between the terms applicable to claimant and comparator. Exactly that point was of course decided in Leverton: see para. 35 above.

Thirdly, Lord Slynn characterised the applicants’ case, which he accepted, as being that it was sufficient that “there be a broad similarity of terms” (p. 526H) and said that it was right for the IT to carry out a “broad comparison” (pp. 528 F-G and 530 C-D). That is unexceptionable, but the history of the present case shows that the phrase is capable of being (mis)understood as referring to a similarity between the applicants’ terms and their comparators’. As I have just said, that is not the relevant comparison. The “broad similarity” which Lord Slynn said was required was between the terms of the surface mineworkers at different pits: if there was no such similarity it would be impossible to say that mineworkers enjoyed the same terms wherever they worked and thus that a mineworker employed at an office in, say, London where a claimant worked would be paid the same as her comparator at a pit in Yorkshire.

In North v Dumfries and Galloway Council [2013] UKSC 45, [2013] ICR 993, the claimants were classroom assistants and nursery nurses employed at schools operated by the respondent council who sought to compare their pay with that of manual workers employed by the council at various depots. The terms of both the claimants and their comparators were governed by (different) national collective agreements – “the Blue Book” in the case of the claimants and “the Green Book” in the case of the manual workers. The ET considered the hypothesis identified at para. 42 above and found that if the manual workers had been employed at the claimants’ schools they would have been employed on Green Book terms: as a collective agreement it applied to all the council’s employees wherever they worked. On that basis it held that common terms and conditions applied (for the relevant classes) at the schools and the depots.

The core facts in North are very close to those in Leverton, and the ET’s conclusion might have appeared inevitable. The only apparent difference is that there were separate collective agreements for the claimants’ class and the comparators’, whereas in Leverton a single agreement covered both. But, as Lady Hale, who delivered the only judgment, pointed out at para. 10 (p. 997 D-E), that can make no difference: what matters is that the terms under both agreements apply wherever employees of the relevant classes are employed. However, the council’s case was that it did not make sense to describe terms as being “observed” at an establishment if an employee of the relevant class would never in practice be employed there, and that in such a case the hypothesis applied by the ET was illegitimate. It had to be at least “feasible” that a manual worker would be employed at one of the council’s schools, and the evidence in the ET had been that it was not, or in any event that if he were his terms and conditions would in practice have to be varied: see para. 32 of Lady Hale’s judgment (p. 1003 G-H).

Lady Hale rejected the council’s case. She expressly accepted the argument of Ms Dinah Rose QC for the claimants, which she summarised as follows, at para. 30 (p. 1003 B-D):

“Not surprisingly, Ms Rose … argues that the tribunal should not speculate about the adjustments to the comparators' present terms and conditions which might be made in the unlikely event that they were transferred to the comparators' workplace. The hypothesis is that the comparators are transferred to do their present jobs in a different location. The question is whether in that event, however unlikely, they would remain employed on the same or broadly similar terms and conditions to those applicable in their current place of work. As Lord Slynn had recognised in [Smith], the object of the legislation was to allow comparisons to be made between workers who did not and never would work in the same workplace. An example might be a manufacturing company, where the (female) clerical workers worked in an office block, whereas the (male) manufacturers worked in a factory.”

Lady Hale also endorsed Ms Rose’s submission, summarised at para. 31 (p. 1003 EF), that it was “unnecessary and illegitimate” to hypothesise the existence of a completely new all-purpose handyman, who might be the kind of manual worker who could plausibly be employed at a school.

At paras. 4-14 (pp. 996-8) Lady Hale reviewed the decisions in Leverton and Smith. At para. 12 (p. 998 D-E) she summarises the relevant principles to be derived from them as follows:

“First, the ‘common terms and conditions’ referred to in section 1(6) are not those of, on the one hand, the women applicants and, on the other hand, their claimed comparators. They are, on the one hand, the terms and conditions under which the male comparators are employed at different establishments from the women and, on the other hand, the terms and conditions under which those male comparators are or would be employed if they were employed at the same establishment as the women. Second, by ‘common terms and conditions’ the subsection is not looking for complete correspondence between what those terms are, or would be, in the woman’s place of work. It is enough that they are, or would be, broadly similar.”

The first of those points is the one which I have made at paras. 35 and 43 above, and the second (though slightly compressed) is the point which was the subject of the actual decision in Smith.

Lady Hale continues, at para. 13 (p. 998 F-G), by spelling out Lord Slynn’s point that the commonality of terms between the two establishments may be hypothetical. She says:

“It is also plain from the reasoning of both Lord Bridge in the Levertoncase [1989] ICR 33 and Lord Slynn in the British Coal Corpncase [1996] ICR 515 that it is no answer to say that no such male comparators ever would be employed, on those or any other terms, at the same establishment as the women. Otherwise, it would be far too easy for an employer so to arrange things that only men worked in one place and only women in another. This point is of particular importance, now that women are entitled to claim equality with men who are doing completely different jobs, provided that the women are doing jobs of equal value. Those completely different jobs may well be done in completely different places from the jobs which the women are doing.”

At paras. 33-41 (pp. 1003-6) Lady Hale gave five reasons for preferring the claimants’ case on the “feasibility” issue. The first is that on the reported facts in Smith it might well not have been feasible for a mineworker to be employed at some of the applicants’ workplaces (see n. 3 above). The second is that to require such a test (or the cognate “real possibility” test) was an unwarranted gloss on the language of the statute. The remaining three reasons I should set out in full.

Lady Hale’s third reason is that the adoption of a requirement that an employee in the comparator’s class might realistically, even if in fact they did not, work at the claimant’s establishment would defeat the policy objectives of the equal pay provisions. She says, at para. 34 (p. 1004 A-D):

“In the third place, to adopt such a test would be to defeat the object of the exercise. This is not just a matter of preventing employers from so organising their workplaces that the women work in one place and the men in another. There may be perfectly good reasons for organising the work into different places. But the object of the legislation is to secure equality of treatment, not only for the same work, but also for work rated as equivalent or assessed by the experts to be of equal value. It stands to reason, therefore, that some very different jobs which are not or cannot be carried out in the same workplaces may nevertheless be rated as equivalent or assessed as having equal value. One example is the (female) office worker who needs office equipment in a clean environment and the (male) factory worker who needs machines which create dirt and dust. But another is the (female) factory worker who puts microscopic circuits on silicon chips in one factory and the (male) factory worker who assembles computer parts in another. The fact that of necessity their work has to be carried on in different places is no barrier to equalising the terms on which it is done. It is well known that those jobs which require physical strength have traditionally been better rewarded than those jobs which require dexterity. It is one of the objects of the equality legislation to iron out those traditional inequalities of reward where the work involved is of genuinely equal value.”

The fourth reason is given at para. 35 (p. 1004 D-G). Lady Hale starts by saying:

“… [I]t is not the function of the same employment test to establish comparability between the jobs done. That comparability is established by the like work, work rated as equivalent and work of equal value tests.”

She refers also to the possibility of explaining any differential by reference to section 1 (3) of the 1970 Act. She continues:

“The ‘same employment’ test should not be used as a proxy for those tests or as a way of avoiding the often difficult and complex issues which they raise (tempting though this may be for large employers faced with multiple claims such as these). Its function is to establish the terms and conditions with which the comparison is to be made. The object is simply to weed out those cases in which geography plays a significant part in determining what those terms and conditions are.”

In his oral submissions Mr Jeans submitted that that passage showed that Lady Hale only understood comparability to be possible where an employer agrees to collective agreements applying across its workforce and “is not operating separate businesses in separate locations”. That was important because Asda’s retail and distribution operations were in substance different businesses, with wholly different origins and locations: see para. 6 above. I do not believe that Lady Hale meant any such thing. She was doing no more than acknowledging the role of the “same employment” test as a filter, while emphasising its limited purpose. The passage is not directed at defining the circumstances in which common terms and conditions apply across establishments. It is in fact clear from the passages quoted at paras. 49 and 57 above that she envisaged cross-establishment comparisons being possible between very different kinds of operation of the same employer.

Lady Hale’s fifth reason is that the claimants’ construction of section 1 (6) “is more consistent with the requirements of European Union law”: see para. 36 (p. 1004 G-H). She develops that point at paras. 37-41 by reference specifically to the decision of the CJEU inLawrence(seepara. 24above).It is necessary to consider what she says with some care.

In Lawrence local authority staff had been transferred under TUPE to the employment of a private contractor. They sought to bring equal pay claims by reference to the terms of employees who had remained in the council’s employment. On a reference the Court held that such comparison was not legitimate. At paras. 17-18 of its judgment (pp. 1108-9) it said:

“17.

There is, in this connection, nothing in the wording of article 141(1) EC to suggest that the applicability of that provision is limited to situations in which men and women work for the same employer. The court has held that the principle established by that article may be invoked before national courts in particular in cases of discrimination arising directly from legislative provisions or collective labour agreements, as well as in cases in which work is carried out in the same establishment or service, whether private or public: see, inter alia, Defrenne[1976] ICR 547, 568, para 40; Macarthys Ltd v Smith (Case129/79)[1980] ICR 672, 690, para 10, and Jenkins v Kingsgate(Clothing Productions) Ltd (Case 96/80)[1981] ICR 592, 613–614, para 17.

18.

However, where, as in the main proceedings here, the differences identified in the pay conditions of workers performing equal work or work of equal value cannot be attributed to a single source, there is no body which is responsible for the inequality and which could restore equal treatment. Such a situation does not come within the scope of article 141(1) EC. The work and the pay of those workers cannot therefore be compared on the basis of that provision.” (Footnote: ⁶)

At paras. 36-39 of her judgment Lady Hale refers to that passage and sets out corresponding passages from the opinion of the Advocate-General. At the beginning of para. 40 (p. 1006 A-B) she says:

“The position is thus that, for the principle of equal pay to have direct effect, the difference in treatment must be attributable to a single source which is capable of putting it right.”

She goes on to observe that there appears to be “no case in the Court of Justice in which the principle of equal pay has not been applied between men and women who work for the same employer”. She reviews the decision of this Court in Robertson v Department for Environment, Food and Rural Affairs [2005] EWCA Civ 138, [2005] ICR 750, in which a comparison was not permitted between the terms of civil servants in different departments: although all civil servants are formally employed by the Crown, responsibility for setting terms and conditions had been delegated by legislation to the different Ministers.

At para. 41 Lady Hale concludes that it is unnecessary to decide whether Robertson was correctly decided. (I note in passing that that means that lower courts, including this Court, remain bound by it.) She continues (p. 1006 D-F):

“In this case it is quite clear that the difference in treatment between the claimants and their comparators is attributable to a single source, namely the local authority which employs them and which is in a position to put right the discrepancy if required to do so. If section 1(6) were to operate as a barrier to a comparison which was required by EU law in order to give effect to the fundamental principle of equal treatment, it would be our duty to disapply it. However, for the reasons given earlier, it sets a low threshold which does not operate as a barrier to the comparison proposed in this case.”

I believe that the effect of Lady Hale’s reasoning on the single source point can be analysed as follows:

The first three elements in that analysis are plainly ratio, but the final element is obiter. In short, North is in my view binding authority that the fact that claimant and comparator have the same employer will in the ordinary case mean that the terms have a single source and thus that EU law permits comparison between them for equal pay purposes; but it is not binding authority that EU law is in that respect directly effective.

It is convenient at this stage to deal with Asda’s second ground of appeal before us, which is that “‘single source’ is not a sufficient basis for comparison”. Mr Jeans submitted that the parameters for comparison since Defrenne had been that the relevant employees should be “in the same establishment or service”. He emphasised the centrality of the concept of the establishment across many fields of EU employment law and referred us to authority showing that it typically refers to a particular workplace (e.g. the decision of the CJEU in USDAW v WW Realisation 1

I do not believe that this ground of appeal can succeed in the light of North. The Supreme Court decided, as part of its ratio, that on the facts of the case before it identity of employer was as a matter of EU law a sufficient basis for comparison. Those facts are essentially typical; although Lady Hale left open the possibility that there might be cases where the employer was not the “source” of all its employees’ terms that could only be on the basis of exceptional facts of some kind. That being so, I see no advantage in myself conducting any detailed examination of the authorities, including Lawrence. I acknowledge that paras. 36-41 of Lady Hale’s judgment in North are very shortly reasoned and do not address any of the domestic authorities to which we were referred. But the passage is clear, and since it is binding there is nothing more to be said. I would, however, add that even in the absence of authority I think it likely that I would have come to the same conclusion, as Kerr J did in the EAT (see paras. 41-47 of his judgment).

My conclusion on this point overlaps with Asda’s third ground of appeal, which is that even if “single source” is as a matter of EU law a gateway to comparability the fact that it was the employer of both the Claimants and the comparators does not on the facts of the present case mean that their terms had a single source. But it is more convenient to address that point later in this judgment: see paras. 107-114 below.

I should summarise what I believe to be the effect of those authorities as regards cross-establishment comparisons as a matter of domestic law.

The over-arching point is that at least in the ordinary case EU law will permit an employee to compare herself with any employee of the same employer: see para. 58 (1) above. It follows, on Marleasing principles, that unless there is some special feature of the case, as there was in Robertson, a tribunal will be bound to construe the provisions of the 2010 Act, so far as possible, to allow such a comparison. It may be tempting to treat that as a short-cut obviating the need to consider the domestic provisions altogether, but that would be wrong in principle. The domestic provisions must be the starting-point, even if they fall to be creatively construed so as to achieve conformity with EU law. (I am not at this stage considering direct effect.) Accordingly, in the summary that follows I proceed wholly by reference to the domestic provisions.

I make three introductory points:

First, the question whether common terms apply at X and Y depends on whether they apply irrespective of which particular establishment a relevant employee is employed at: see Leverton, as summarised at para. 34 above. The test will be satisfied whether that is so “generally” (i.e. for all employees) or only for employees in the relevant classes – that is, the classes to which A and B respectively belong. Which gateway it makes sense to focus on depends on the circumstances of the particular case (as to this, see further para. 73 below).

Second, that question entails comparing the terms applying as between the two establishments, not as between the claimant and the comparator: that is clear from each of Leverton, Smith and North (paras. 35, 43 and 48 above). At the risk of spelling it out unnecessarily, the tribunal needs to ask either “do common terms of employment apply at X and at Y for all employees ?” (the “generally” alternative) or “do common terms of employment apply for cleaners at X and at Y and for manual workers at X and at Y ?” (the “relevant classes” alternative). In the oral submissions before us the correct comparison was described as “vertical”, and the incorrect as “horizontal”: I am not sure the labels are perfect, but they will do as a shorthand.

Third, common terms apply at X and Y not only where they apply to actual employees in the relevant classes working there but where they would apply, even if a manual worker would never in practice be employed at X or a cleaner at Y. That is, as I have said, implicit in Leverton but it is explicitly confirmed in Smith and North: see paras. 42 and 49 above. This was described in the ET and the EAT as “the North hypothetical”: that is not really accurate, because the point pre-dates North, but I will adopt the label for present purposes. It is important to understand the role of the North hypothetical. The fact that if a manual worker were employed at X he would enjoy the same terms as B is a consequence of the fact (if established) that the same terms apply for manual workers irrespective of where they work: it is not the test as such. Considering the North hypothetical is a potentially useful thought-experiment, but it will often be possible to answer the question whether common terms apply, even if no-one in B’s class is employed at X, without resort to it: it was not considered in Leverton, because it was enough to point to the fact that the Purple Book applied to all the council’s employees wherever they might be employed.

Fourth, even where it is helpful expressly to consider the North hypothetical, doing so does not involve working out a detailed scenario under which someone in the comparator’s class might be employed at the claimant’s establishment, with some adjustment to the facts, and then asking what terms would apply in that scenario. As Lady Hale made clear by accepting Ms Rose’s submissions in North (see para. 47 above), an exercise of creative imagination of that kind is not required. The North hypothetical is, to repeat, simply a way of asking whether the terms for the relevant employees apply irrespective of where they work.

Fifth, it will be straightforward to answer the North hypothetical question in the claimant’s favour if there is a collective agreement governing the terms of the two classes without reference to where they work; and that is so whether it is a single agreement which covers both classes (as in Leverton) or separate agreements for each class (see per Lady Hale in North – para. 46 above). But although that is the paradigm case it is not the only basis on which it may be possible to find that common terms and conditions apply (or, more particularly, that the North hypothetical is satisfied). It would equally be the case, for example, if an employer “imposed” terms on an across-the-board basis; or, simply, if it could be seen that as a matter of fact employees in the relevant classes enjoyed the same terms and conditions wherever they were employed.

Sixth, the requirement that common terms apply as between the establishments does not mean that all the terms of the relevant employees at both must be common. It is enough that terms of cleaners at X and cleaners (actual or hypothetical) at Y and of manual workers at Y and manual workers (actual or hypothetical) at X are “broadly” common, taking a common sense approach: see Smith (and North: paras. 43 and 49 above).

Seventh, it follows from the foregoing – and specifically from para. 67 – that it is irrelevant whether there is any similarity between the actual terms of A and B. The question is whether the terms for cleaners are (or would be) the same (or broadly so) whether they are employed at X or at Y and likewise as regards the terms for manual workers. If that test is satisfied, cleaner and manual worker terms may be identical to each other (save of course in the respect which gives rise to the claim), or wholly different in structure and content, or anywhere in between.

Finally, it should be noted that on a strict analysis the “generally” alternative is redundant. All that a claimant need ever prove is that common terms and conditions apply for the two relevant classes – that is, hers and the comparator’s: that cannot be a more demanding test than proving that they apply generally and will typically be less demanding. To put it another way, the effect of the words is, logically, “generally or at least for employees of the relevant classes”. Proving that broadly similar terms and conditions apply between establishments “generally” – i.e. for all classes of employee – is simply one way of proving that they apply for the classes in question; but it may be a more straightforward route in a particular case.

As already noted, the phrase in section 1 (6) of the 1970 Act “establishments … at which common terms and conditions of employment are observed either generally or for employees of the relevant classes” is replaced in section 79 (4) (c) by “common terms apply at the establishments (either generally or as between A and B)”.

Asda unsuccessfully contended in both the ET and the EAT that the effect of that change is to alter the substantive comparison required, and thus also remove the justification for applying the North hypothetical. It maintains that contention before us as ground 10 in its grounds of appeal. Mr Jeans submitted that the plain effect of the new wording was that the only question was whether the claimant and comparator as individuals enjoyed broadly similar terms with each other, i.e. a “horizontal” comparison, and that the “vertical” comparison between employees of the same class in the two establishments was no longer necessary. He speculated that Parliament might reasonably have regarded the law as stated in Smith as “surprising”.

I agree with the ET and the EAT in rejecting that submission. Since the point is one of pure law I need not set out their reasoning but will simply state my own (though I believe it broadly corresponds to theirs).

My starting-point is that I regard it as very unlikely that Parliament in enacting the 2010 Act intended to make any substantive change for the worse to the rights of equal pay claimants, as it would do if Mr Jeans’ submission were correct. Mr Jeans acknowledged that there was nothing in the White Paper or the Explanatory Notes – or indeed in any other travaux préparatoires – reflecting a concern of the kind which he suggested. In any event, I can see nothing surprising about the effect of section 1 (6) as interpreted in the case-law. On the contrary, it seems to me entirely congruent with the overall scheme. At this preliminary stage of the analysis it makes sense that the focus is only on commonality of terms between establishments: the time for comparing the terms of the claimant and her comparator is at the substantive stage, if a cross-establishment claim is permitted.

It is necessary nevertheless to address the actual language of section 79 (4) (c). I accept that as a matter of language both the reference to “A and B” and the phrase “as between” naturally suggest a comparison between the terms of the actual claimant and comparator. But I do not think that that is the only possible meaning, and I am satisfied that it cannot have been the intention here. That is not only for the reason given in the preceding paragraph. More importantly, such a construction simply does not make sense in the specific context of section 79 (4). The condition which head (c) prescribes is that “common terms apply at the establishments”, i.e. the claimant’s establishment and the comparator’s establishment; and the parenthetical words relied on are subordinate to that requirement. It would make no sense for the question whether common terms apply at the two establishments to depend simply on whether A and B themselves have common terms. Such an approach would in practice restore the position which was explicitly rejected in Leverton, which would be an even more drastic change than the rejection of the North hypothetical. In my view it is quite clear that the draftsman has unthinkingly deployed the technique, used throughout the 2010 Act, of referring to claimants and other parties by letters of the alphabet and has failed to appreciate that it could be read as effecting a substantive change. I believe that it is possible to construe the statutory language, having regard to the legislative history and purpose, as continuing to refer to the existence of common terms for employees of A’s class and B’s class at the two establishments. If that were regarded as impossible on any natural use of language, in my view the case meets the requirements for an “amending construction” prescribed by the House of Lords in Inco Europe Ltd v First Choice Distribution [2000] UKHL 15, [2000] 1 WLR 586.

In short, I do not believe that the change in the statutory language effects any change in the substantive law, and there is no basis for treating the North hypothetical as being unavailable under the 2010 Act. The new drafting may perhaps be inept but in context its meaning is clear. There have, regrettably, been several other instances of re-drafting effected by the 2010 Act unintentionally unsettling the previous law. (Footnote: ⁷)

I reach that conclusion on the basis of ordinary domestic construction principles. But I would if necessary rely on a Marleasing approach. For the reasons given at paras. 60-62 above I believe that as a matter of EU law comparison should ordinarily be possible between establishments of the same employer, and a construction of section 79 (4) (c) which inhibits such comparison should be avoided if possible.

It follows from the foregoing that the questions for the ET were whether (broadly) common terms and conditions applied (a) for retail workers irrespective of where they worked and (b) for distribution workers irrespective of where they worked. Since no retail workers were in fact employed at depots, or distribution workers in stores, that question could be framed in terms of the North hypothetical as follows:

I should note one point recorded by the Employment Judge which removed a different potential issue. As noted at para. 9 above, there were some variations between the rates of pay at particular depots. Para. 88 of the Reasons reads:

“It was not suggested that comparators who work in a particular distribution centre had variations between themselves, although the respondents’ case is that terms for pay could vary from depot to depot.”

That is not perhaps crystal clear, but Mr Short explained, without demur from Mr Jeans, that Lord Falconer had been expressly asked in the ET whether Asda was arguing that different regimes applied at different depots and had confirmed that it was not. The concession is not surprising, given that all that is required is “broad” commonality of terms and the existence of minor variations between establishments, as in Smith, is immaterial.

The Employment Judge dealt with the position under domestic law at paras. 192-256 of his Reasons. He says at para. 192:

“Common terms need to be considered in three ways: first, whether there are (and were) common terms generally as between claimants and comparators; secondly, whether there were common terms, in the alternative, ‘for employees of the relevant classes’ under s. 1 (6) EPA

1970; and, thirdly, whether there are common terms ‘as between A and B’ under s. 79 of the EA 2010 …”.

He proceeded to consider those three questions in turn. We are not concerned with the third, because it is disposed of by my conclusion at paras. 74-80 above.

It can be seen that the Judge’s chosen structure proceeds on the basis that the issues involved in deciding whether common terms applied “generally” were substantially different from those involved in deciding whether they applied “for employees of the relevant classes”. That is reflected in his analysis under each: under the first (headed “Common Terms Generally”) he addresses the extent to which the terms for retail and distribution employees were “broadly similar”, whereas under the second (headed

“Common Terms for Employees of the Relevant Classes”) he considers the North hypothetical. That distinction is, with respect, misconceived. As noted at para. 73 above, the “generally” and “relevant classes” gateways are not real alternatives (Footnote: ⁸), and the essential question is the same in either case – that is, whether terms for the relevant classes apply irrespective of where they work.

However, the fact that the Judge structured his approach to the issue wrongly does not matter if he nevertheless addressed the substance of the matter and reached a sustainable conclusion. I take in turn what he says under his two headings, but I substitute headings that reflect their actual contents.

COMPARING THE CLAIMANTS’ AND COMPARATORS’ TERMS

The Employment Judge concluded that common terms were observed generally at Asda’s stores and depots. His reasoning is at paras. 194-217 of the Reasons. For reasons which will appear, I need not analyse it in detail. It is founded on an exercise which he had performed earlier, at paras. 88-112. In those paragraphs he painstakingly compared, with the assistance of tables provided by the parties, the terms and conditions of employment (both contractual and non-contractual) of, on the one hand, Asda’s retail employees and, on the other, its distribution employees, identifying areas of similarity and difference. As regards similarity, at para. 99 he said:

“I find that there is no material difference between Retail and Distribution in the following matters: hourly pay (Footnote: ⁹), admission to the bonus scheme after 6 months service, eligibility to a discount card after 12 weeks in service, matched contributions of 2 or 3 percent to pension, death benefit scheme entitlement, admission to the share save plan, maternity pay, adoption pay, bereavement pay, jury service pay, mobility, dress requirement, deductions from pay by Asda and the right to search.”

As regards differences, he identified at para. 100 a number of respects in which there were what he described as “less significant” differences between retail and distribution terms, and at paras. 102-110 a smaller number of “more substantial” differences: for example, retail employees are not entitled to overtime pay while distribution employees are.

Having performed that exercise, at paras. 212-214 the Judge says (I quote only the points essential to show the nature of the reasoning):

“212.

The analysis I have set out in the findings of fact show that there is a significant correlation or comparison in a broad way between the terms in Retail and Distribution. Although not determinative, the fact that these terms have all been set by staff within the same employer is reflected in the strong similarities in the Handbooks and is for that reason a material factor. The fact that there are similarities in the two classes both of which are hourly paid employees is also capable of supporting the comparison. … .

213.

I reject the respondent’s submission that because these terms were negotiated over periods of time in different ways for the different groups they cannot be common terms. …

214.

I accept that there are some differences in terms but I do not consider that they are so extensive as to undermine the broad comparison which had to be made. …”

He concludes, at para. 217, that it follows from those findings that “common terms apply generally”.

I have to say that I believe that the Employment Judge was in this part of the Reasons conducting wholly the wrong exercise. The question he had posed himself at para.

As I understand it, that point forms part of Asda’s grounds of appeal 6 and 7, but it comes wrapped up with a number of other more particular points about the way in which the Judge conducted the exercise. Asda contends, for example, that he wrongly took into account non-contractual as well as contractual terms, and there is an ancillary debate about which terms are indeed contractual. There were also issues about how similar various different kinds of term, e.g. relating to productivity targets, really were. Mr Jeans submitted that the Judge placed an undue emphasis on “broad” similarities which were in truth inevitable: one sick pay scheme, for example, will inevitably look much like another. Since the exercise was fundamentally misconceived I see no need to consider those questions. I would only say that the fact that the exercise required the lodging of nineteen files (containing over 10,000 documents) and the hearing of days of evidence and argument about the minutiae of retail and distribution terms, without any conceptual framework for measuring their similarity or otherwise, is a further powerful reason for concluding that it is not what the statute requires.

Mr Short submitted that, whatever language the Judge might inadvertently have used, he was in substance performing the right exercise; but it seems to me clear that he was not. In the EAT Kerr J said at para. 90 of his judgment:

“Nor is there any bar to comparing the content of the respective sets of terms. To decide whether terms are common or not, it is useful to know what the terms are. The Judge was right to ascertain them and the parties were right to put them in evidence before him.”

That needs some unpacking. There will in fact be cases where it is possible to decide that terms are common between two establishments without knowing what they are. If there is a document headed “Terms and conditions applying to employees [or, to cleaners and manual workers] at X and Y [or, at all establishments]” that will prove commonality by itself, whatever their content (subject only to checking that, despite the title, the document does not provide for different terms at different establishments). In the absence of such a document it may indeed be necessary to go through the actual terms at each establishment for employees in the relevant classes to see if they differ between the establishments; but that does not require comparison between the terms of the claimant and comparator classes.

It seems that the Judge fell into the error that he did because he was a victim of the misunderstanding of the references in Smith to “broad similarity” which I discuss at para. 44 above. I would, however, be slow to criticise him for doing so since the Claimants submitted that such an exercise was required, and it is far from clear that the objections now raised by Asda were developed in the same way below.

THE NORTH HYPOTHETICAL

As already noted, although the Employment Judge’s second heading is “Common Terms for Employees of the Relevant Classes” the actual issue that he considers under it is the North hypothetical. At paras. 220-229 of the Reasons he summarises Asda’s submissions, but I need not set them out because so far as they are relied on in this appeal I address them below. At paras. 230-235 he summarises the Claimants’ submissions, which at para. 236 he says that he accepts. Since that therefore represents his core reasoning I should quote those paragraphs in full:

“230.

The claimants’ submissions were to the effect that Mr Stansfield had accepted in evidence that terms and conditions would be maintained if there was a hypothetical depot at the store. It was accepted by both witnesses that the hourly rate would not change upon a hypothetical relocation since the employees would be paid the rate for the job.

231.

The claimants submit that there is no reason to suspect that other matters of detail would change. This, it is said is not such a case such that all employees who work in Manchester are on one contract and all those who work in London are on another. Mr Short pointed out that Baroness Hale in North accepted a submission that the tribunal ‘should not speculate about the adjustments to the comparators’ present terms and conditions which might be made in the unlikely event that they were transferred to the claimants’ workplace’.

232.

It was submitted there was no factual basis for concluding that there would be significant changes in terms and conditions because there are no Retail workers anywhere employed other than on Retail terms such that for example none receive premiums for hours worked between 2 and 10 pm nor for overtime. Similarly, no Distribution workers are employed other than on Distribution terms. Whilst there are differences to the enhanced rate paid after 2 pm, all receive a premium after that time and all receive overtime.

233.

Where there is great similarity in relation to the terms as between Retail and Distribution they would clearly not change. The claimants point to holiday entitlements after the third year of service and pension arrangements.

234.

Again the claimants submit that the terms do not need to be identical in this hypothetical situation but only broadly similar.

235.

Whilst the claimants accept that the treatment of depot staff temporarily deployed to stores is not directly comparable they suggest it gives a better indication than witnesses’ speculation.”

The Judge begins, as we can see, with what is said to have been a concession by Mr Stansfield. Mr Jeans submitted that what the Judge records does not reflect the totality of Mr Stansfield’s evidence. He referred us to para. 117 of the Reasons, which reads:

“Both Mr Stansfield and Mrs Tatum [Asda’s ‘Executive People Director’] were asked in evidence what would be the position in the event of Distribution employees, however unlikely that might be, performing Distribution work in stores. Both clearly answered that if the Distribution employees were carrying out Distribution work they would be paid the rate for the job they were actually doing. … Both witnesses also maintained their primary position that Retail terms would apply to Distribution employees deployed to work in stores and Distribution terms to Retail employees deployed to work in depots.”

He submitted that the final sentence effectively removes any weight that might otherwise be placed on the second.

I am myself wary about putting too much weight on an answer obtained in crossexamination of a lay witness on an issue which is not one of primary fact, and if the issue turned on what is said to have been the effect of Mr Stansfield’s answer Mr Jeans’ objection might be good. But in truth his evidence is relied on simply as an introduction to the Claimants’ overall case as summarised in paras. 231-232. The summary is not quite as explicit as it might be, but the basic argument is straightforward. The starting-point is that there are common terms for all distribution workers at all depots (that is what the Judge means by saying that terms are the same in London and Manchester) and likewise for retail workers; that is not contentious

In the following paragraphs the Judge addresses various particular points made by Asda in response to that case. I need to quote these because they are challenged in the grounds of appeal:

“238.

I do not attach any weight to the suggestion that a worker moving between depot and depot takes the new depot terms as assisting in this analysis. If I am correct in excluding that then the fact that there are some depots on different terms to others is also not relevant.

239.

Neither am I persuaded that the homogeneity argument is of great weight here. Recognising that this a hypothetical comparison it is a postulation that a depot worker is carrying out his depot work although located at a store. It does not seem to me that that necessarily means that it has to be postulated that he is carrying out that work in the customer facing part of the store. Indeed, recognising the factual hypothesis is inherently unrealistic, it seems to be much more likely that depot workers doing Distribution work would not be in physical proximity to Retail staff and customers. I therefore conclude that homogeneity is unlikely to be a safe basis for concluding that terms would change particularly in view of the evidence of Mrs Tatum and Mr Stansfield that Asda would pay the rate for the job that was being done.

240.

I agree that the temporary redeployment of depot workers into stores, or hypothetically vice versa is not properly comparable. It provides some slight support for the claimants’ case. I do not consider that Mr Short’s attempt to construct a hypothetical depot in a Retail car park is fatal to the claimants’ argument.

241.

In my judgment greater support is derived from the fact that the respondent operates what appear to be more favourable terms for the depot workers and it is inherently unlikely that depot workers would be willing to see those extended to Retail employees if hypothetical relocation of Retail employees occurred in that direction and equally unlikely that depot workers would be willing to give up their terms if there were hypothetical relocation of them into stores.

242.

Furthermore the claimants rightly, in judgment, submit that weight can be placed upon the fact that there is significant similarity of certain parts of the contractual provisions and that these would be maintained in the hypothetical situation even if they are not sufficiently similar to amount to common terms generally.”

Some of the points made in those paragraphs require elucidation, but I will provide that so far as necessary when I consider Asda’s criticisms.

Two grounds of appeal – grounds 8 and 9 – directly address the ET’s reasoning and conclusions on the North hypothetical. I take them in turn.

As developed in Asda’s skeleton argument, this ground comprises two distinct points, as follows:

Both criticisms are in my view flawed for the same reason. They are concerned with an attempt to work out how the necessary hypothesis might be realised in practice, and the consequences of it. That is a wholly irrelevant exercise: see para. 69 above. It might be said that in that case the Judge should not have engaged in it. But it is, again, clear that he was led down this route by the parties; and in any case the fact that

As developed in the skeleton argument, this ground comprises three distinct points, which I take in turn.

The first point reverts to the fact that Asda’s witnesses had given evidence that distribution terms would not apply to those doing depot jobs in a store (see para. 93 above) and contends that the Judge had no basis, and gave no sufficient reasons, for rejecting that evidence. But the question raised by the North hypothetical is not a simple question of primary fact on which the relevant manager’s evidence might be expected to be dispositive. It can only be answered by inference based on how terms for actual workers in the relevant class(es) are applied, and what a lay witness says about that is of limited, if any, value: it is a matter on which the tribunal has to reach its own conclusion.

The second point relies on a finding made earlier in the Reasons to the effect that, if a distribution worker moved from one depot to another where the terms differed, the terms applying at the new depot would apply to him. It is said that the Judge’s statement at para. 238 of the Reasons that that was irrelevant was plainly wrong, because the finding demonstrated that the employee’s terms would certainly change if he went to work at a store, even doing distribution work. I do not accept this. It was common ground that such variations as there are between different depots are not sufficient to prevent there being common “distribution terms” applying at all depots (see para. 82 above). The question raised by the North hypothetical is whether “distribution terms” would apply at all.

The third point depends on what are said to be the ET’s findings “that store terms are specific to stores and depot terms are specific to depots”, and it is said that it necessarily follows that a distribution worker doing distribution work (however unfeasibly) at a store would not enjoy distribution terms. I do not accept that. On one view, saying that “depot terms are specific to depots” is saying no more than that distribution workers only work at depots, which is precisely why the North hypothetical comes into play in the first place; it does not help to answer it. Even, however, if what is meant is that at least some aspects of distribution terms reflect the specific characteristics of distribution work, that is not inconsistent with the proposition that distribution terms apply to distribution workers wherever they are employed: the hypothesis is of course (however unrealistically) that even if working in a store the worker in question will be continuing to do distribution work.

I do not, therefore, believe that either of the pleaded grounds, as developed in the skeleton argument and Mr Jeans’ oral submissions, impugns the conclusion reached by the Employment Judge on this issue.

It may be worth stepping back from the specifics of Asda’s challenge and taking a broader view. The essential reason why in my view the Judge’s conclusion was open to him – indeed I believe right – is that for both classes (i.e. retail workers and distribution workers) Asda applied common terms and conditions wherever they work. The effect of the case-law, and of North in particular, is that in such a case

“wherever they work” extends even to a workplace where they would never in practice work because the nature of its operations is so different, as it was in both Leverton and North itself. The contrast is with a situation where there were no common “distribution terms”, so that what terms a distribution worker enjoyed would depend on where they worked. If that had been the case here (as it may have been pre-2003) the outcome would be different, because it would be impossible to say “if a distribution worker worked in a store these are the terms that would apply to him”.

I appreciate that it may seem artificial to say that common terms and conditions apply between depots and stores on the wholly hypothetical basis that if a distribution worker worked (as a distribution worker) in a store distribution terms would apply to him; but the Supreme Court in North confronted that very issue and explained why its conclusion was justified, both on the language of the statute and in policy terms (as to which see in particular para. 34 of Lady Hale’s judgment).

One consequence of this conclusion is that much of the detailed evidence and argument in the ET was in my view beside the point. The preliminary issue could have been decided on the straightforward basis that Asda’s terms for retail workers and for distribution workers both applied wherever they worked. It would in truth be no credit to the law if the kind of elaborate and confusing exercise which the Judge was encouraged to undertake were required in order to establish whether comparison were permitted.

I reach that conclusion, as the Employment Judge did, without seeking any support from EU law. I am, however, prepared to say that I believe that the Claimants’ and the comparators’ terms did in fact have a single source, within the meaning of Lawrence, because they were set by the same employer, which had the power to equalise them. It follows that it would be a breach of EU law if a comparison between them were not permissible; and, like the Supreme Court in North, I regard that as an additional reason for upholding the ET’s finding that the crossestablishment comparisons on which the Claimants rely are permitted by section 79 (4) (c) (and section 1 (6)). Since the point is not central to my conclusion I will give my reasons briefly.

The case advanced by Asda in the ET was that there were different “sources” for retail and distribution terms because they were arrived at by wholly different processes within Asda, reflecting the complete managerial separation between its retail and distribution operations. Retail pay was set by a committee of Asda’s board, and distribution pay by “depot management”, principally (though, as we have seen, not wholly) through collective bargaining. No doubt formally the ultimate responsibility lay with the Asda board and/or Wal-Mart, who inevitably exercised budgetary oversight; but they were not the effective decision-makers as regards actual terms. As regards Distribution in particular Mr Stansfield gave evidence of the high degree of autonomy he enjoyed. Asda’s case was that its effective delegation of responsibility to different internal structures was analogous to the delegation by the Crown of responsibility for pay to the different government departments, which was held in Robertson to preclude inter-departmental comparison.

That case was rejected by the Employment Judge, who held, at para. 190 of his Reasons:

“The Executive Board of Asda and the members or subcommittees of that Board had and exercised budgetary control and oversight over both distribution and retail at all material times. The Executive Board was responsible for the differences in pay and could, or could subject to the overarching control of Wal-Mart, have introduced equality.”

He observed (at para. 184) that the devolution of responsibility in Robertson was wholly different from the delegation of responsibility within a company or corporate structure.

Kerr J in the EAT upheld that reasoning. I quote paras. 56-59 of his judgment in full:

“56.

For all the intensity of Asda’s attack on the Judge’s reasoning and conclusions, far from considering them perverse I find them unassailable. This was an ordinary case of a large organisation delegating the setting of pay to separate internal organs. The Judge avoided the mistake made by the Tribunal in the Glasgow City Council case (see, in particular, paragraph 69 in Lord Brodie’s opinion) of treating [Robertson] as creating a category of dual source cases where power to set pay is delegated and then not interfered with on a regular basis.

57.

In my judgment, [Robertson], if as I assume it is correctly decided, was wholly exceptional and turned on the unique position within the civil service, where the setting of pay and most other terms of civil servants on a departmental not national basis, was enshrined in statutory and other instruments that would have to be expressly revoked if the power was to be reclaimed by the Treasury and exercised centrally.

58.

Here, by contrast, Asda or Wal-Mart could interfere at the stroke of a pen or, more likely, the click of a mouse. The Industrial

Relations Board (“IRB”) included senior employees including the Asda’s CEO, with an interest in pay across both retail and distribution staff. Moreover, the IRB received regular updates on distribution staff pay negotiations, while retail pay was signed off by the executive board.

59.

It would be unjust and little short of absurd if the presence of a single source turned on whether Asda’s or Wal-Mart’s board happened to have overruled Mr Stansfield, or the CEO or the board happened to have overruled Ms Tatum. It was quite enough to justify the Judge’s conclusion that the evidence showed that they regularly scrutinised the work done to set pay in, respectively, retail and distribution, and could overrule their decisions at any time.”

I agree entirely with that passage, save only that I doubt whether it was necessary even to show that the Asda or Wal-Mart boards “regularly scrutinised” the relevant

Mr Jeans advanced before us essentially the same arguments as he did in the EAT, but Kerr J’s reasoning is sufficient to show why I do not accept them.

For the reasons given I would uphold the decisions of both the ET and the EAT that the Claimants are entitled under section 79 (4) (c) of the 2010 Act (and section 1 (6) of the 1970 Act as regards the earlier period) to compare themselves with distribution workers in other establishments. That conclusion is supported by the fact that the result avoids any inconsistency with EU law, but I would reach it in any event.

DIRECT EFFECT

It is of course well-established that article 157 of the TFEU and its predecessors, articles 119 and 141, afford employees a directly enforceable right to equal pay in cases where their comparators are doing the same or similar work or work which has been rated as equivalent: that has been clear since Defrenne. But the authorities do not establish definitively whether the same is true in a case where the claimant’s work is different from the comparator’s and has not been rated as equivalent but where she claims that it is of equal value. There is a question as to the effect of the distinction articulated at para. 18 of the judgment in Defrenne between “discrimination which may be identified solely with the aid of the criteria based on equal work and equal pay referred to by [article 119]”, in respect of which the article has direct effect, and “discrimination which can only be identified by reference to more explicit implementing provisions of a community or national character”, in respect of which it does not: some authorities, both domestic and EU, appear to say that equal value claims fall into the latter category. The authorities, and the arguments both ways, are carefully and lucidly reviewed by Kerr J at paras. 13-35 of his judgment in the EAT. He regarded the issue as difficult and held (at para. 16) that the point was not acte clair, observing (at para. 36) that he would have referred it to the CJEU if the parties had asked him to do so. Since they did not, however, he held (at para. 37) that the better view was that article 157 does have direct effect in an equal value case. That conclusion is in line with Lady Hale’s obiter view in North though Kerr J does not in fact rely on it.

It is not necessary for us to decide this issue in order to determine the appeal, and on balance I prefer not to do so. If the point were in truth determinative my provisional view is that the right course would be to refer it to the CJEU: we are not bound to do so, since we are not the final court, but I agree with Kerr J that it is hard to say that the issue is acte clair, and it would simply cause delay to leave it to the Supreme Court to refer. But it would not be right to make a reference where the point is not determinative; and it will achieve little of value for us to embark on a further elaborate analysis with a view to deciding the point at this level where only a reference will achieve certainty.

I do not think that our declining to pick up this particular baton will cause any practical problems for tribunals. Although it is possible to conceive of cases where the issue will be crucial – essentially, where the requirements of section 79 (4) (c), even as I have construed it, are not satisfied on the facts of the particular case – it remains to be seen whether such cases arise in practice; and if they do the issue of direct effect is determined at EAT level by Kerr J’s decision (and reinforced by Lady Hale’s judgment in North). I am prepared to say, however, that despite the undoubted difficulties presented by some of the authorities, my provisional view at the end of the argument was that Kerr J’s conclusion was right.

I would dismiss Asda’s appeal.

I agree that the appeal should be dismissed for the reasons given by the VicePresident. I should, I think, register that I am more doubtful than he is in para. 116 above that the equal value limb of Article 157 has direct effect. We are in agreement that the point is not acte clair either way.

I also agree that the appeal should be dismissed for the reasons given by the VicePresident. I only add that I note the provisional views expressed by him and by Lord Sales on the subject of direct effect. As the appeal does not turn on this point, which is in any event not acte clair,it is unnecessary for me to say anything about it.