Notes prepared by Thomas Linden , Matrix Chambers, 15-17 September 2006

Introduction

1.      The speech of Lord Browne-Wilkinson in Strathclyde Regional Council v Wallace [1] is a very useful starting point to understanding the potentially difficult area of equal pay. He identified the mischief at which the relevant provisions are directed as follows:

“The purpose of section 1 of the Equal Pay Act 1970 is to eliminate sex discrimination in pay not to achieve fair wages.”

2.      Lord Browne-Wilkinson also said [2] that:

“The cases establish that the Equal Pay Act 1970 has to be construed so far as possible to work harmoniously both with the Sex Discrimination Act 1975 and Article [141].  All three sources of law are part of a code dealing with unlawful sex discrimination …” (emphasis added).

3.      These passages seem to provide a guiding light through the thicket of technicality and case law in this area.  They suggest that as long as all concerned keep well in mind the point that the equal pay provisions are concerned the elimination of sex discrimination in relation to pay they will get to the right answer. Unfortunately, however, Lord Browne-Wilkinson did not go as far as to suggest that the concept of sex discrimination has the same scope whether or not the claimant’s complaint relates to unequal treatment or unequal pay. The case law also demonstrates that the boundaries of this concept in the context of equal pay are both wider and less certain. 

4.      These Notes attempt to map out some of those boundaries and tentatively suggest that the decisions of the House of Lords in the Wallace case and the subsequent case of Glasgow City Council v Marshall [3] should, indeed, be regarded as no more that a useful starting point [4] .

Where do the equal pay provisions fit in?

5.      Although the Equal Pay Act 1970 received the Royal Assent on 29 May 1970, it did not come into effect at that point.  Ultimately the 1970 Act was amended and enacted as Schedule 1 to the Sex Discrimination Act 1975, and the equal treatment and equal pay regimes both came into effect on 29 December 1975.  Parliament could, of course, have simply included discrimination in relation to pay amongst the categories of prohibited discrimination in the employment field in Part II of the 1975 Act.  The fact that it chose not to do so suggested at the outset that it was contemplated that distinct regimes would apply to complaints about unequal treatment and unequal pay, and that the concept of discrimination under each regime would not necessarily be the same.

6.      The precise demarcation between complaints which are properly brought under the Equal Pay Act 1970 and complaints which should be brought under the Sex Discrimination Act 1975 is potentially complex and not central to the present discussion. [5]    Suffice it to say that the preamble to the 1970 Act states that it is “an Act to prevent discrimination, as regards terms and conditions of employment between men and women”.  Thus, where the claimant’s complaint relates to terms and conditions and/or she seeks equal terms to a male comparator on the basis that she is doing equal work, her claim should ordinarily be brought under the 1970 Act.  Where, however, her complaint relates to other forms of treatment by the employer which fall within section 6 of the Sex Discrimination Act 1975 it should generally be brought under the 1975 Act. 

7.      Furthermore, although the 1970 Act obviously applies to claims for equality in terms and conditions relating to pay, the concept of “pay” is a broad one.  In  Defrenne v Belgium [6] the ECJ held that the concept of pay comprises any consideration, whether in cash or in kind, whether immediate or future, provided that the worker receives it, albeit indirectly, in respect of his employment from his employer.  Indeed, under Article 141 EC Treaty the concept of “pay” is not restricted to contractual entitlements; it may include indirect benefits and non-contractual bonuses [7] .  Thus, Article 141 has, for example, been held to cover complaints about voluntary bonuses, travel facilities [8] , occupational pensions [9] , severance payments [10] and compensation for unfair dismissal [11] .

8.      It is also important to bear in mind the relationship between the domestic legislation ie the Equal Pay Act 1970 and the Sex Discrimination Act 1975 on the one hand and Article 141 EC Treaty and the Equal Pay and Equal Treatment Directives [12] on the other.  Employment Tribunals do not have jurisdiction to hear “freestanding” claims brought under the EU legislation. The Claimant therefore has to bring her claim under the appropriate domestic anti discrimination legislation which confers jurisdiction on the Tribunal. Thus, for example, the 1970 Act confers jurisdiction on Employment Tribunals to hear equal pay complaints, and any such complaint therefore has to be brought under this Act.  However, since the Equal Pay Directive is directly effective against emanations of the State and Article 141 is directly effective regardless of the nature of the respondent, in so far as any provision of the 1970 Act (or indeed the Sex Discrimination Act) is incompatible with Community law the claimant is entitled to rely directly on the EU provisions and case law and to argue that the incompatible domestic provision should be disapplied. [13]

9.      This point may be significant where the claimant complains about discretionary bonus decisions. As noted above, such complaints may fall within Article 141. However, arguably such complaints to not relate to terms and conditions and therefore do not fall within the 1970 Act [14] . If this is right, discretionary bonus complaints should be brought under the Sex Discriminations Act 1975. This has advantages and disadvantages for the claimant. An advantage is that the procedural strictures of the 1970 Act in relation to comparators etc may not apply. Arguably, the claimant can rely on an actual or a hypothetical comparator and need not prove equal work etc. Arguably, the limits on who may be a comparator in s1(6) of the 1970 Act (as to which, see below) do not apply either. The major disadvantage for the claimant is that the limitation regime under the 1970 Act (ie, broadly, within 6 years of breach during employment and up to 6 months after its termination) is much more relaxed than under the 1975 Act, where a complaint must be brought within 3 months of the act complained of etc. Since, in a given case, it may be difficult for an employer to prove the reason for a discretionary bonus decision taken some time earlier, this point may be important. 

10.  Finally, it is worth noting that the effect of a successful equal pay claim is that an equality clause is implied into the claimant’s contract of employment. In so far as she has not been employed on equal terms and conditions to those of her comparators, the employer is in breach of contract and she is entitled to recover damages for a period of up to 6 years prior to proceedings being issued, as in the County or High Court.

The stages in the equal pay argument

11.  In Glasgow City Council v Marshall [15]   Lord Nicholls said this:

“The scheme of the Act is that a rebuttable presumption of sex discrimination arises once the gender based comparison shows that a woman, doing like work or work rated as equivalent or work of equal value to that of a man, is being paid or treated less favourably than a man. …  The burden passes to the employer to show that the explanation for the variation is not tainted with sex.”

12.  He went on to say that

“If there is any evidence of sex discrimination, such as evidence that the difference in pay has a disparately adverse impact on women, the employer will be called upon to satisfy the tribunal that the difference in pay is objectively justifiable” [16] .

13.  Thus, subject to what is said below about the decision of the Court of Appeal in Armstrong v. The Newcastle upon Tyne NHS Hospital Trust [17] , the arguments in an equal pay claim may go through the following stages:

13.1.                    First, the claimant (a) identifies a male comparator in the same employment as her and (b) proves that he is employed on like work, work rated as equivalent or work of equal value (“equal work”) to her but that (c) one or more of the terms of his contract is more generous. If this is established, a rebuttable presumption of sex discrimination arises.

13.2.                    Second, the burden then shifts to the employer to explain the difference in terms by showing that the variation is genuinely due to a material factor which is not the difference of sex between the claimant and her comparator.  This may be achieved by establishing a gender neutral explanation for the difference.

13.3.                    Third, however, the claimant may argue that although the explanation for the difference in terms is apparently gender neutral it operates in a discriminatory way or has a discriminatory impact.

13.4.                    Fourth, if this is established, the employer is required objectively to justify the difference in terms.  If the employer fails to do this, liability is established.

13.5.                    Fifth, if she is successful, the claimant is entitled to damages for breach of contract for any period during which (in broad terms), in the 6 years prior to proceedings, she was paid less than her comparator.

14.  Each of these stages will be considered below.

Stage 1: establishing the presumption of discrimination

Introduction

15.  As the case law currently stands, the claimant is required to identify an actual comparator [18] of the opposite sex in all cases other than challenges to national legislation. [19] The claimant and her comparator must be in common employment as defined by s1(6) of the 1970 Act or, at least, their pay must be “attributable to a single source” and they must be employed on like work or work rated as equivalent or work of equal value. Four points are worth a mention.

Does the Comparator Need to be Contemporaneously employed?

16.  First, it is important to bear in mind that because the provisions are concerned with eliminating discrimination Article 141 is flexible in terms of who the comparator may be. The question at the heart of the case is whether the claimant has been paid less than her comparator because she is a woman (or, at least, for reasons linked to gender). It therefore does not necessarily matter that the male comparator is not employed contemporaneously with the claimant provided he did equal work.  Thus, the claimant may compare herself with a predecessor, [20] a contemporary or a successor [21] . 

17.  This, combined with the fact that the claimant is not limited to alleging like work, may mean that in a given case she is able to name a wide range of comparators employed in different jobs at different times.  In the short term, this may cause difficulties for the respondent and for the Tribunal in terms of case management, but ultimately the weakness in such an approach by the claimant may be that any differences in terms which are established are readily explained by the difference in the circumstances when the relevant decisions were taken.  Thus, as will be seen below, a non-sex discriminatory explanation for a difference in terms might be that relevant decisions were taken by different people at different times and/or related to different jobs and/or were taken without reference to each other.

Does the Comparator have to be employed by the same employer?

18.  The second point is that there have been greater difficulties in the case law in relation to the degree of connection that must exist between the claimant and her comparator.  This question is relevant to the scope of the concept of discrimination in this context because it requires consideration of whether there must be a discriminator in the sense of a person or persons who determine the pay of both the claimant and her comparator on an illegitimate basis.  The 1970 Act appears to contemplate that there must be a close connection in that the claimant and her comparator are required to be “in the same employment” and this is defined in section 1(6) as follows (so far as material):

“Men shall be treated as in the same employment with a woman if they are men employed by her employer or an 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”.

19.  In other words, the claimant and her comparator must be employed by the same employer or an associated employer and either they must be employed at the same establishment or they must be employed at establishments at which common terms and conditions of employment are observed [22] .  Furthermore, the establishments in question must be in Great Britain.

20.  Article 141 contemplates a looser connection, but how much looser remains to be seen.  In Lawrence and Others v Regent Office Care Ltd [23] the ECJ was asked to consider whether school catering and cleaning staff who had been transferred by a local authority into the private sector pursuant to TUPE could compare their pay with male comparators still employed by the local authority whose work had been rated as of equal value in a job evaluation study.  The ECJ held that they could not:

“although there was nothing in the wording of Article 141 (1) EC to suggest that the applicability of that provision was limited to situations in which men and women worked for the same employer, where, as in the present case, the differences in pay of workers of different sex performing equal work or work of equal value could not be attributed to a single source, such as a collective agreement, there was no one body responsible for the inequality which could restore equal treatment; … such a  situation did not come within the scope of Article 141(1) EC, and the work and pay of the different sets of workers could not be compared on the basis of that provision”.

21.  The scope of the formulation “attributed to a single source” has not yet been fully explored.  However, at paragraph 49 of his Opinion Advocate-General Geelhoad identified three categories of case where the claimant may be entitled to compare her pay with a male comparator who is not “in the same establishment or service”:

“The first comprises cases in which statutory rules apply to the working and pay conditions in more than one undertaking, establishment or service.  By way of example, one may think of the salaries of the nursing staff working for a service such as the National Health Service. Secondly, there are cases in which several undertakings or establishments are covered by a collective works agreement or regulations governing the terms and conditions of employment.  Finally, the third category concerns those cases in which the terms and conditions of employment are laid down centrally for more than one organisation or business within a holding company or conglomerate.”

22.  The Advocate-General went on to say, at paragraph 51 of his Opinion, that:

“The feature common to the three categories is that regulation of the terms and conditions of employment actually applied is traceable to one source, whether it be the legislature, the parties to a collective works agreement, or the management of a corporate group.”

23.  The reason for this approach was explained as being that the person or persons   could then be held accountable for the difference in pay, which is consistent with the classic notion of discrimination in which a person or persons take discriminatory decisions.  Although the Court was less explicit, it appears to have accepted this rationale in holding, at paragraph 18 of its judgment, that:

“However, where, as in the main proceedings here, the difference is 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..  ……. within the scope of Article 141(1)EC.  The work and pay of those workers cannot therefore be compared on the basis of that provision.” (emphasis added)

24.  This analysis has since been applied in Allonby v Accrington and Rossendale College [24] in holding that a part-time hourly paid lecturer who was provided to Accrington and Rossendale College through an agency known as Education Lecturing Services (“ELS”) could not compare the terms of her contract with ELS with those of lecturers directly employed by the College.  The claimant and her comparator did not have the same employer and their pay was not attributable to the same source.

25.  The Lawrence test was also applied in Department for Environment Food and Rural Affairs v Robertson and Others [25] . Here, the Court of Appeal held that under Article 141 it is neither necessary nor sufficient that a claimant and her comparators have the same employer. Thus, the mere fact that all civil servants are “employed” by the Crown is not sufficient to permit a Civil Servant working in one government department to compare her pay with that of a male comparator in a different department. This is because, although they have the same employer and their pay ultimately comes out of the same coffers, responsibility for determining pay has been delegated to government departments and pay determination is therefore not attributable to a single source.  The fact that the delegated powers could be revoked is immaterial until such time as they have been revoked. At paragraph 13 of his judgment, Mummery LJ said this about the “attributed to a single source” test:

“the focus of this rather imprecise approach is on the location of the body responsible for making decisions on levels of pay rather than on the identification of the relevant legal source of the relevant decision making power”  

26.  He went on to say, at paragraph 41:

“The critical question is: is there a single body responsible for the discriminatory pay differences of which complaint is made? Retention of power by the Crown after delegation to the department means that there is a theoretical legal possibility of the Crown exercising its power at some time in the future but retention of legal power, which has not in fact been exercised……does not make the Crown “the body responsible” for the actual negotiations and decisions on pay by individual departments resulting in pay differences of which complaint is made.”

27.  Robertson is interesting for two main reasons. First, it shows that there may still be a role for s1(6) of the 1970 Act in that, under this provision, employment by the same employer at the same establishment in Great Britain will be sufficient whether or not the pay of the claimant and her comparator is attributable to the same source. Second, it does not answer the question whether “the body responsible” means the body which caused the pay differential or whether it includes a body which, although it did not cause the pay differential, has the power to correct it. In Robertson decisions by the departments were the cause of the differential and they had the immediate power to put the differential right. The point therefore did not strictly arise although, of course, the Court of Appeal ruled out arguments that it is sufficient that there is a legal power to intervene where another body is directly responsible for pay determination [26] .

28.  Robertson also leaves open interesting questions as to how “the body responsible” should be identified given that the mere fact of a common employer is insufficient.  In the typical City financial institution, for example, pay may be determined by different individuals or groups of individuals who may or may not be employed by different but associated companies and may or may not make recommendations to the same or different committees. Their decisions may be in relation to different bonus pools or budgets and employees working in different countries [27] . Terms and conditions in such organizations are rarely governed by common regulations or collective agreements, and it therefore remains to be seen whether tribunals and courts will be prepared to hold that the claimant and her comparator in such cases are not employed “in the same establishment or service” and/or their pay is not attributable to a single source. 

Equal Work

29.  Thirdly, in relation to like work/work rated as equivalent/work of equal value the following points may be worth noting:

29.1.                    First, it is only if like work is not established that any need for consideration of equal value arises.  The like work question is one of fact for the Tribunal and requires consideration of whether the work of the claimant and her comparator “is of the same or a broadly similar nature” and, if so, whether the differences between the things she does and the things the comparator does are “of practical importance in relation to terms and conditions of employment.”  If the answer to the former question is “yes” and the answer to the latter is “no” like work is established. [28] Section 1(4) has not been a source of difficulty. It is the nature of the work that is important [29] and it is important for the Tribunal to concentrate on the work that is actually performed rather than the work that may be performed under the contract [30] . A difference in levels of responsibility may prevent work from being the same or broadly similar [31] ; differences in the time at which jobs are done are unlikely to have the same effect [32] .

29.2.                    As far as work rated as equivalent is concerned, the key authority is Bromley v H&J Quick Ltd [33] where the Court of Appeal held that the definition of work rated as equivalent under section 1(5) of the 1970 Act requires an analytical approach which compares the job of the claimant with the job of her comparator under various headings such as effort, skill, decision etc. [34]

29.3.                    Thirdly, as far as equal value is concerned, it is important to bear in mind that the tribunal is entitled to decide this issue for itself. [35] It is therefore not necessarily the case that this issue will be referred to an independent expert, particularly where the parties have instructed their own experts.

Term by Term Comparison

30.  Finally, it is now well established that a term by term comparison is required when determining whether the equality clause has been breached [36] . It is only where the terms relate to the same subject matter that they can be lumped together, so an employer cannot argue that there has been no breach of an equality clause eg on the basis that although the claimant’s basic pay is lower, her overall remuneration package is of the same value or more generous. Nor can it be said, for example, that the lower pay is compensated for by a longer holiday entitlement.

31.  However, in Degnan v Redcar and Cleveland Borough Council [37] the Court of Appeal confirmed that there are limits to this principle. Here, in broad terms the question was whether the claimants were entitled to a term by term comparison where the elements of the remuneration package to be compared comprised basic hourly pay, a fixed percentage bonus and an attendance allowance. The issue was whether the attendance allowance was such that it could be lumped in with basic pay and bonus, or whether it related to different subject matter, such that the claimants were entitled to chose comparators (whose work had been rated as equivalent) with the highest bonus and comparators with an entitlement to attendance allowance and potentially win against both (the comparators with the highest bonus did not get attendance allowance, so this approach would have a ratcheting up effect). The Court of Appeal said that the claimants were not entitled to take this approach. The Employment Appeal Tribunal had been right to say that all monetary payments for normal working hours should be aggregated and divided by the number of working hours to arrive at an hourly rate. Since the attendance allowance was paid for working normal hours it should be included.   

32.  It is also noteworthy that it has not been authoritatively determined whether an employer may rely on a similar argument to prevent an equality clause from arising in the first place. In other words, it may be open to an employer to say that the reason and/or justification for the difference in basic pay is that the claimant has longer holidays and therefore nothing to do with sex discrimination [38] .

Stage 2: Establishing a genuine material factor (“GMF”) defence

33.  In Glasgow City Council v Marshall [39] Lord Nicholls identified the elements of the GMF defence in the following way:

“In order to discharge this burden the employer must satisfy the tribunal … first, that the proffered explanation, or reason, is genuine and not a sham or pretence.  Second, that the less favourable treatment is due to this reason.  The factor relied upon must be the cause of the disparity.  In this regard, and in this sense, the factor must be a “material” factor, that is a significant and relevant factor.  Third, that the reason is not “the difference of sex”.  This phrase is apt to embrace any form of sex discrimination, whether direct or indirect.  Fourth, that the factor relied upon is or, [an equal value] case … may be a “material” difference, that is, a significant and relevant difference, between the woman’s case and the man’s case”

34.  In truth, however, (and with due respect) this is a somewhat elaborate way of saying that a non-discriminatory reason for the variation in pay must be established by the employer. What follows from this is that the employer’s reason need not be a good one, or one with which the Tribunal agrees, provided it is not tainted by sex discrimination.  Indeed, in Tyldesley v TML Plastics Ltd [40] Mummery J (as he then was) pointed out that:

“even if a differential is explained by careless mistake, which could not possibly be objectively justified, that would amount to a defence under section 1(3) and for the purpose of Article [141] of the EC Treaty, provided that the tribunal is satisfied that the mistake was either the sole reason for it or of sufficient influence to be significant or relevant. If a genuine mistake suffices, so must a genuine perception, whether reasonable or not, about the need to engage an individual with particular experience, commitment and skills”. [41] .

35.  This passage was expressly approved by the House of Lords in the Glasgow City Council v Marshall case. Note, also, that the employer can rely on the same factual feature of the claimant or the comparators’ cases in relation to both the equal work and the s1(3) issues. Thus, in cases where an aspect of the evidence (eg that the fact that the comparator is believed to contribute more to the business) might go to either or both issues there is no need to plead the defence on the basis that it is relevant to one or the other but not both. These sorts of decisions can be left to the Tribunal: Christie v John Haith Ltd [42] .

Stage 3:  Establishing discriminatory impact

Introduction

36.  There was some debate in the Ratcliffe case (discussed below) as to whether it is appropriate to speak of direct and indirect discrimination in the context of equal pay.  However, it is well recognised that in cases where the employer has established a genuine reason for the variation in pay which is apparently gender neutral, the claimant may nevertheless contend that there is a taint of sex discrimination because the employer’s pay practice has an adverse impact on its female employees.  As will be discussed below, there are currently three models for this type of argument but the categories are not necessarily closed.

The Burden of Proof

37.  This follows the discriminatory treatment approach.

37.1.                    The burden of proving the discriminatory impact is on the claimant. [43]   Note, however, that the judgment of the Court of Appeal which reached this conclusion in the Nelson v Carillion Services case emphasised that the claimant may seek the assistance of the employment tribunal in extracting relevant information from the employer, either by way of requests for further and better particulars or written answers to questions or through the new questionnaire procedure.

37.2.                    Once discriminatory impact is established, the burden shifts to the employer to establish justification.

The 1^st model of Discriminatory Impact: Classic Indirect Discrimination

38.  The first model is the familiar one in which the employer applies a requirement or condition which is said to have a disproportionate adverse impact on its female employees.  The classic case is Bilka-Kaufhaus v Weber von Hartz [44] where the employer placed conditions on membership of its occupational pension scheme which made it difficult for part-time employees to join.  The ECJ held that since the majority of the employees excluded from the scheme as a result of these conditions were women, Article 141 would be infringed unless the employer could establish that its policy was based on objectively justified factors unrelated to any discrimination on the ground of sex. 

The 2^nd Model of Discriminatory Impact: Danfoss

39.  The second model arises out of the Danfoss case [45] .  In this case, in broad terms, a collective agreement provided that the remuneration of the employees would comprise a basic wage plus various supplements. Eligibility for the first set of supplements would be determined on the basis of the employee’s skill, independence, responsibility and quality of their work.  The second set would be paid on the basis of the employees’ training and length of service. The decision as to whether supplements were awarded was a matter for determination by individual line managers.  The employees’ trade union challenged the operation of the collective agreement. Neither it nor its members could prove that a given decision in relation to a given employee was influenced by gender because the system of assessment of eligibility for the supplements was not transparent.  It could, however, prove that in practice the average pay of the women who were subject to the collective agreement was lower than the average pay of the men.

40.  The central question before the ECJ was whether the Union had adduced sufficient evidence for the burden of proof to shift to the employer to explain or justify its system of pay, or whether it had to be proved by a particular employee that particular decisions in relation to her supplements or the supplements of her comparator were influenced by gender.  The ECJ’s answer was that:

“Where an undertaking applies a system of pay which is totally lacking in transparency, it is for the employer to prove that its 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”. [46]

41.  The scope of the Danfoss principle has not (yet) been fully explored.  Its potential impact is, however, highlighted by the first Employment Appeal Tribunal decision in Villalba v Merryll Lynch [47] .  Here the claimant sought to establish that the respondent’s practice of awarding discretionary annual bonuses resulted in women being paid less, on average, than men.  She therefore sought pay data in respect of a significant number of the respondent’s senior employees with a view to attempting to make good the factual foundation for her case and then going on to argue that the burden was on the respondent to prove that its practice in relation to discretionary bonuses was not discriminatory.  Her application for disclosure of these data was rejected by the Employment Tribunal and this decision was upheld by the Employment Appeal Tribunal.

42.  The importance of the Villalba decision for present purposes is that it addressed two aspects of the Danfoss principle, namely the number of employees required for there to be a statistically significant sample from which inferences might be drawn and, secondly, the criteria on which the relevance of given statistics should be decided:

42.1.                    As regards the first question, the arguments proceeded on the basis that, as indicated in the Danfoss case, a relatively large number of employees is required before any reliable inferences can be drawn. In the context of the Danfoss case this was said to be at least 100. The claimant’s case on appeal was that disclosure of data in respect of a fairly wide pool of employees should therefore be ordered. 

42.2.                    As regards the second question the claimant’s argument was rejected on the basis that, on a true understanding of the Danfoss principle, an inference of discrimination could only be drawn if the statistics related to a comparable group of employees which, for these purposes, meant a group of employees who were employed on like work or work of equal value. If the men and the women in the sample were not employed on equal work, this fact in itself was capable of explaining any differences in pay, and no inference could therefore properly be drawn even if it transpired that the average pay of the women was lower. The Employment Tribunal recognised that the claimant seeking disclosure of information before the hearing of the case will not be able to prove that the sample does indeed comprise employees who are carrying out equal work, but held that a prima facie case that they are so employed needs to be established by the claimant whereas, on the evidence before the tribunal this standard had not been reached.

The 3^rd Model of Discriminatory Impact: Enderby

43.  The third discriminatory impact model was established in Enderby v Frenchay Health Authority [48] in which a group of speech therapists sought to compare their pay with groups of pharmacists and clinical psychologists working in the National Health Service.  Their case was that they were carrying out work of equal value and yet were paid less than their comparators.  The Health Authority sought to meet their claim by explaining that the reason for the difference in pay was that the pay of the three groups was determined by collective bargaining and the collective bargaining unit for the speech therapists was separate to the units for the comparator groups.  Thus, said the Health Authority, there was a gender neutral explanation and no question of direct discrimination arose.  The Authority went on to argue that there was no question of the system of pay being anything other than transparent and  no question of indirect discrimination arose because no requirement or condition was applied to both the speech therapists and their comparator groups.

44.  The ECJ held that on the assumption that the speech therapists were carrying out work of equal value to that of their comparator groups, it was sufficient for them to establish that they were a predominantly female group and that their comparator groups were predominantly male.  In this event, the burden would shift to the Health Authority to provide objective justification for the differences in pay.  Moreover, whilst the fact that there was separate pay bargaining explained the difference in pay it did not necessarily objectively justify it.

45.  The Enderby model obviously represents an important development in that it clearly suggests that the concept of discrimination in the equal pay context is wider than the same concept in the context of equal treatment.  The ECJ did not accept that it was necessary to identify a discriminator or to establish that a requirement or condition has been applied to the claimant group and the comparator groups. Rather, it took the view that the fact that a predominantly female group was being paid less than a predominantly male group carrying out work of equal value called for objective justification.  The ECJ did not define with precision the scope or rationale for the new model, but there are various strands in the judgment and in the Opinion of the Advocate-General. 

45.1.                    Part of the rationale was that which underpinned the Danfoss principle, namely that the law in this area should not place the burden of proof on the claimant where to do so would make it impossible or unreasonably difficult for her to make out her case. 

45.2.                    Another strand appears to have been that there was at least a whiff of direct discrimination in the sense that the speech therapists may have been paid less because they were regarded as in a “women’s profession” or doing “women’s work”. [49]  

45.3.                    Other passages in the Opinion of the Advocate-General suggest that the aim of Article 141 is to eliminate disadvantage in relation to pay which is experienced by women because they are women.  Thus, at paragraph 15 of his Opinion the Advocate-General said:

“The purpose of a conceptual scheme is to comprehend methods by which women are placed at a disadvantage in their working lives and not to create additional obstacles to claims being made before the courts in respect of sex related pay discrimination.  For this reason, the formalistic approach should not be adopted when categorising actual instances where women are placed at a disadvantage at work. In accordance with the result orientated line taken by the Court of Justice in the past, a pragmatic approach ought to be pursued. ” (emphasis added) [50]

46.  Before leaving Enderby, however, it is important to point out that the ECJ did not lose sight of the point that there must at least be some form of sex discrimination, and that an essential ingredient for a successful claim is therefore that the comparator group is truly comparable in the sense that the (a) members of the group are carrying out equal work and (b) the statistics in relation to the group are significant.  Thus, at paragraphs 16 and 17, the ECJ said this:

“16.However, if the pay of speech therapists is significantly lower than that of pharmacists and if the former are almost exclusively women where the latter are predominantly men, there is a prima facie case of sex discrimination, at least where the two jobs in question are of equal value and the statistics describing that situation are valid.

17. It is for the national court to assess whether it may take into account those statistics, that is to say whether they cover enough individuals, whether they illustrate purely fortuitous or short-term phenomena, and whether, in general they appear to be significant”.

47.  The ECJ reiterated these points in the Royal Copenhagen case [51] . They are important in that they require the Employment Tribunal to eliminate other possible explanations for the difference in pay, such as that the claimant and her comparator are not employed on equal work and/or that the statistics in question are fortuitous, before placing the potentially onerous burden on the employer to provide objective justification for the difference in pay. Moreover, they at least potentially circumscribe the material which a claimant is entitled to require the employer to produce for the purposes of litigating her claim. 

Stage 4: Justification by the employer

The Test

48.  The key point under this heading is that there is a difference between explaining the differential in pay and justifying it.  Justification may be a significantly more difficult task given that the question arises against the background of equal work being assumed or established and a situation in which the employer’s pay practice adversely impacts on female employees.  In Bilka-Kaufhaus [52] the ECJ said that:

“It is for the national Court, which has sole jurisdiction to make findings of fact, to determine whether and to what extent the grounds put forward by an employer to explain the adoption of a pay practice which applies independently of a worker’s sex but in fact affects more women than men may be regarded as objectively justified on economic grounds.  If the national Court finds that the measures chosen by Bilka correspond to a real need on the part of the undertaking, are appropriate with a view to achieving the objectives pursued and are necessary to that end, the fact that the measures affect a far greater number of women than men is not sufficient to show that they constitute an infringement of Article [141]“. [53] (emphasis added)

49.  This test, with its reference to the need to show that the employer’s approach is “necessary to that end” suggests to the UK lawyer that the test is a fairly strict one. However, the fact that the test is not one of necessity was confirmed by the Court of Appeal in Barry v Midland Bank plc [54] where Peter Gibson LJ said

“it would be wrong to extrapolate from those words written in that context that an employer can never justify indirect discrimination in a redundancy payment scheme unless the form of the scheme is shown to be necessary as the only possible scheme. One must first consider whether the objective of the scheme is legitimate. If so, then one goes on to consider whether the means used are appropriate to achieve that objective and are reasonably necessary to achieve that end.” (emphasis added) [55]

50.  It is interesting to note that although the Guidance Notes to the Part Time Workers (Prevention of Less Favourable Treatment) Regulations 2000 suggest that the Bilka-Kaufhaus test applies to the test of justification in that context, in the Explanatory Notes to the Employment Equality (Sexual Orientation/Religion and Belief) Regulations 2003 and the Employment Equality (Age) Regulations 2006 the DTI points out that subsequent case law suggests that the ECJ in Bilka-Kaufhaus did not intend to establish a test of necessity.  For the avoidance of doubt, however, the 2003 and 2006 Regulations have avoided using this word and refer, instead, to “a proportionate means of achieving a legitimate aim” [56] in defining the justification defence to indirect discrimination. Under this test it is likely that that the approach would be closer to the approach in the context of treatment where the Tribunal is required to balance the discriminatory effect of the condition, practice etc and the reasonable needs of the employer which applies it [57] , although the concept of proportionality also requires consideration of whether the means go beyond what is appropriate and necessary to achieve the relevant end [58] .

When Does the Test Have to be Applied?

51.  A second, important, point is that justification has to be established as at the date of the Tribunal’s decision. This cuts both ways. It means that the employer may have a justification at the time of the introduction of the pay practice under challenge, but it does not follow from this that the practice will be justified forever more. As the House of Lords emphasised in R v Secretary of State for Employment ex parte Seymour Smith (No 2) [59] the employer has to keep its pay practices under review and may be expected to take corrective action where it becomes apparent that they are having a sex discriminatory effect. Conversely, as the ECJ pointed out in Schonheit and Becker [60] , it is not necessary for the employer to have its justification in mind at the time of the introduction of the challenged measure. Since the question of justification is objective, rather than subjective, the employer can also rely on reasons other than those which were put forward at the time.

The Issue in Cadman

52.  The case of Cadman v Health & Safety Executive [61] , which has been referred to the ECJ by the Court of Appeal, usefully illustrates the potential difference between explanation and justification. Here the question is whether length of service is a self justifying criterion. A band 2 principal inspector employed in the HSE was paid  £4000 to £9000 per annum less than her comparators who carried out work rated as equivalent to hers under a job evaluation study. The principal reason for the difference in pay was that, for a number of years, the HSE operated a pay scale based on annual increments. Although this system was altered in 1995, existing pay differentials were preserved, with the result that the men received higher pay by reason of their longer service.

53.  Mrs Cadman was able to show that the length of service criterion had a disproportionate adverse impact on comparable female employees because they tended to have shorter service. However, passages in the judgment in the Danfoss [62]   case suggest that length of service as a criterion does not require justification and/or that it is in any event justified: “since length of service goes hand in hand with experience and since experience generally enables the employee to perform his duties better, the employer is free to reward him without having to establish the importance it has in the specific tasks entrusted to the employee” [63] . The subsequent decisions of the ECJ in  Nimz [64] , Hill [65] and Gerster [66] , however, cast doubt on whether this is the law and the question whether use of length of service requires specific justification has therefore been referred by the Court of Appeal.

54.  The Advocate General’s Opinion in Cadman was delivered on 19 May 2006 [67] . In his view, where an employer uses length of service as a determinant of pay, and this has a disparate impact as between male and female employees, the employer should be required to show that the use of a length of service criterion takes into account the business needs of the undertaking and is applied proportionately so as to minimize its disadvantageous impact on women. Where an employer cannot justify the use of a length of service criterion as a general approach, it should provide specific justification for the difference in pay levels between the claimant and her comparator. Given the uncertainty created by the earlier judgments of the ECJ and the economic effect of the ruling which he advocated, he agreed with the UK and Irish Governments that the temporal effects of the judgment should be limited. His recommendation was that that the ECJ should hold that the judgment cannot be relied upon to support claims of indirect discrimination arising prior to the date of judgment unless the claimant has brought proceedings prior to that date.

The outer limits of discrimination in this context

A Fourth Model?

55.  Various cases suggest that it should by no means be assumed that the three models of discriminatory impact discussed above are the final word on this issue or that the categories of indirect discrimination are rigid or closed. A recent illustration of this point is the decision of the Court of Appeal in Bailey and others v Home Office [68] . Here a very substantial group of administrative staff in the Prison Service compared their work with prison officers and industrial and non industrial support staff who were employed on work rated as equivalent. In the lead case, the comparator groups were predominantly male but the proportion of higher executive officers who were female was only 50.7%. The question was whether a prima facie case of indirect discrimination had been made out.

56.  The Employment Appeal Tribunal held that the facts did not fit within the Enderby model and that it was wholly artificial to analyse the case as being one in which a requirement or condition had been applied. No prima facie case was therefore established and there was therefore no need for objective justification. This was overruled by the Court of Appeal. Peter Gibson LJ held that: 

“it is hard to see why the statistical approach applied in Seymour-Smith should not be capable of being employed in cases where there is no condition or requirement but there is a pay disparity between two occupational groups. Provided that the ET is satisfied as to the validity of the statistics and the appropriateness of their use, in my judgment it is free to use that statistical approach in order to determine whether there has been prima facie discrimination. In particular, I do not see anything in Enderby that precludes such approach” [69] .

57.  In other words, if there was a smaller proportion of women in the advantaged group a prima facie case might be made out whether or not the disadvantaged group was predominantly female and regardless of whether it could be said that a requirement or condition had been applied:

“I can see no justification for the imposition of a high threshold for satisfying the test of prima facie discrimination. Where, as here, there is one group of employees of an employer which contains a significant number, even though not a clear majority, of female workers whose work is evaluated as equal to that of another group of employees of the employer who are predominantly male and who receive greater pay, it would be very surprising if an ET were to be precluded by the presence in the disadvantaged group of a significant number of men from holding that that disparity in favour of men required justification by the employer” [70] .

58.  If this analysis is correct [71] , arguably, it renders the Bilka Kauhaus and the Enderby models redundant and replaces them with a test based simply on a comparison of the proportions of men and women in the advantaged and disadvantaged groups. Providing there is equal work as between claimant and comparator groups, if, on the basis of significant statistics, it is established that there is a considerably smaller proportion of women in the advantaged group, the prima facie case is made out. It is then for the employer to establish objective justification.

Market Forces

59.  A further example of the uncertain scope of indirect discrimination in the context of equal pay is the decision of the House of Lords in North Yorkshire County Council v Ratcliffe [72] which is not easily fitted into recognised concepts of direct discrimination or any of these three models.  Here, the claimants were school dinner ladies employed by the County Council.   Their work was subject to competitive tendering and it was known that one of the competitors to the in-house bid was a commercial company which paid its employees considerably less than Council rates.  In order to keep the work in-house the Council declared the dinner ladies redundant and then re-employed them at rates of pay which were significantly lower than their male comparators who were employed by the Council in similar grades. 

60.  The Employment Tribunal held that the GMF defence had not been established, but the basis on which it did so is less than clear.  The findings noted by the House of Lords [73] suggest that the tribunal rejected the defence because the Council had only been able to cut the pay of the dinner ladies because they were doing jobs which were virtually exclusively carried out by women for reasons which related to their domestic responsibilities.  In other words, the Tribunal appears to have relied on the weak bargaining position of the claimants.  At the same time, however, the Tribunal recognised that the Council and the employees “were over the proverbial barrel” in that they were faced with a stark choice between cutting wages and losing jobs ie that the factors which influenced the approach of the Council were not sex discriminatory in the sense in which the term is generally understood.

61.  The House of Lords upheld the Tribunal’s decision on the basis that it was a finding of fact.  The difficulty with their Lordships’ decision, however, is that it does not appear to have been suggested by the Tribunal that the gender of dinner ladies operated in any way on the mind of the relevant decision makers on the Council.  Nor does any of the three models of discriminatory impact readily describe what occurred in the Ratcliffe case.  Rather, the result suggests that it was sufficient that the dinner ladies were able to establish that they were at a disadvantage in the labour market owing to the fact that they were carrying out “women’s work” and they needed to carry out this sort of work because of their domestic responsibilities [74] . 

Armstrong v MOD

62.  Such an analysis strikes a chord with the passages in the Enderby case which suggest that the role of the equal pay provisions is to eliminate disadvantage in the labour market experienced by women because they are women.  Indeed, a submission to this effect was accepted by the Employment Appeal Tribunal in Ministry of Defence v Armstrong and Others [75] where Cox J emphasised that the concept of indirect discrimination in relation to pay is broader than the same concept in the context of equal treatment, and that a technical or formulaic approach is not appropriate in the equal pay context.  On this basis, the Employment Appeal Tribunal upheld a decision of the Employment Tribunal that the GMF defence had not been made out because the reason for the difference in pay between the claimants and their comparators was that their pay had been determined on the assumption that they would be in receipt of retirement pay, and was therefore depressed.  In fact, because eligibility for retirement pay depended on establishing 16 years’ pensionable service in the army this assumption was false in the case of the claimants, who were typical in the sense that women serving in the armed forces tend not to have had lengthy careers.  On this basis, the EAT held that the link between the disadvantage experienced by the claimants and the fact that they are women was made out and the reason for the variation in pay was therefore tainted by discrimination. 

Brunnhofer

63.  More radical than decision in the Armstrong v MOD case are the arguments arising out of the decision of the ECJ in Brunnhofer [76] .   Factually, the Brunnhofer case was fairly routine in that it concerned a claimant employed by a bank who was classified in the same salary group as her male comparator and paid the same basic salary.  However, she received a lower monthly supplement than her male colleague and she complained of breach of Article 141. The employer sought to explain/justify the difference in pay by reference to matters of which it was not aware at the time of the impugned decision, such as the claimant’s poor performance. The controversial feature of Brunnhofer is that there are passages in the judgment of the ECJ which arguably appear to suggest that it is not necessary to establish any form of sex discrimination before the employer is required to provide objective justification for the variation in pay or, at least, the claimant only need establish that she is employed on equal work with her comparator but is paid less. 

64.  Clearly, if this is the right reading of the Brunnhofer judgment it represents a substantial change in the position apparently previously adopted by the ECJ and is contrary to the decisions of the House of Lords in Wallace [77] and Glasgow City Council v Marshall where their Lordships emphatically rejected the proposition that objective justification is required in every case where equal work and lower pay are established.  Moreover, in Parliamentary Commissioner for Administration v Fernandez [78] the Employment Appeal Tribunal rejected this analysis of the Brunnhofer decision. Mr Fernanadez was granted permission to appeal but the case then settled.

65.  In Sharp v Caledonia Group Services Ltd [79] , however, a different division of the Employment Appeal Tribunal accepted a submission that Fernanadez was wrongly decided. It therefore fell to the current President of the Employment Appeal Tribunal, Elias J, and his colleagues to seek to introduce certainty (at least at the level of the Employment Appeal Tribunal) in the second appeal in Villalba v Merrill Lynch [80] . They held that Fernandez was right and Sharp was wrong, essentially because the effect of the Sharp analysis was to collapse the distinction between direct and indirect discrimination, which has so painstakingly been developed in the authorities discussed above and was endorsed in successive EC Directives including the Burden of Proof Directive [81] , whereby objective justification is only required where an adverse gender impact is established. The Sharp analysis also fundamentally alters the concept of discrimination in the context of equal pay in that it contemplates that  a claimant may succeed notwithstanding that there is no link of any kind between gender and the pay differential complained of. Moreover this, of itself, would be contrary to the drift at EU level towards common concepts for different types of discrimination, particularly in the context of gender [82] .

66.  The current position is therefore that the conventional approach to indirect discrimination in the context (with the modifications identified above) is the law.

A Spanner in the Works?

67.  In Armstrong v. The Newcastle upon Tyne NHS Hospital Trust [83] the Court of Appeal have created potential difficulties for Employment Tribunals. Rather than adopt the 4 stage analysis of liability arguments discussed above, Buxton LJ said this at paragraph 110 of his judgment [84] :

“if the employer proves the absence of sex discrimination he is not obliged to justify the pay disparity…….Once disparate adverse impact has been established, the burden passes to the employer in respect of two issues. First, that the difference between the man’s and the woman’s contract is not discriminatory, in the sense of being attributable to a difference of gender. Second, if the employer cannot show that the difference in treatment was not attributable to a difference in gender he must then demonstrate that there was nonetheless an objective justification for the difference between the woman’s and the man’s contract.”  

68.  As Elias J pointed out in Villalba [85] this way of putting it is problematic because it appears to suggest the employer has can win by showing that the reason for the difference in pay is gender neutral, whether or not the criterion or practice in question is objectively justified. As emerges clearly from the case law and Directives discussed above, once adverse impact is established the employer has to prove justification and that the criterion, practice etc is not tainted by sex discrimination. Surprisingly, the Petition for leave to appeal in Armstrong was refused.

THOMAS LINDEN
Matrix Chambers
Griffin Building
Gray’s Inn
London WC1R 5LN

tomlinden@matrixlaw.co.uk

15 September 2006