1. We should start with how the Council formulated its defence under section 1 (3). Mr. Cavanagh and Mr. Breen pleaded an elaborate and careful document headed "The Respondent's Genuine Material Factor Defences". So far as material for present purposes this can be summarised as follows:

(1) Paras. 20-26 plead a "brief summary of the GMF defence"12. Para. 20 reads as follows:

"Bonusability

The main GMF defence that will be relied upon by the respondent in relation to the period up to 31st March 2007 is that, where there is a pay differential between the claimant and comparators who, at the material time, were engaged on work rated as equivalent or work of equal value, the differential is explained by the fact that the comparators received a performance related bonus and the claimant jobs were not suitable for a performance related bonus. This will be described as the "bonusability defence."

The remaining paragraphs of the summary are immaterial.

(2) At paras. 42-53 the pleading states "the nature of the Respondent's defence". It is said that it is not understood that direct sex discrimination is alleged, but that if it is it is denied (para. 44). "PCP-type" indirect discrimination is denied at para. 46. It is also denied that the statistics raise a case of "Enderby-type" discrimination (paras. 47-51). At para. 52 the conclusion is pleaded that:

"Accordingly, the respondent will say that it is sufficient in all cases, in order to succeed with its GMF defence, that the respondent can prove that there is a reason other than the difference in sex, for the pay differential. The respondent will further say that in all cases it is able to prove such a reason other than the difference in sex for the pay differential."

However, at para. 53 it is pleaded in the alternative that objective justification can be proved if necessary.

(3) There follows a detailed exposition of "the GMF defences". We are concerned with "bonusability". As to that, the pleading deals first with why the comparators receive bonus. Para. 54 reads:

"The bonuses that were received by the holders of the comparator jobs were genuine performance related bonuses (PRBs). If they need to be justified, they are justified on the basis of efficiency, effectiveness and productivity. The bonuses incentivised the employees concerned to work more productively and/or efficiently and/or at a faster pace, thus enabling the respondent to make efficiency savings. They were wholly or largely self-financing, and, in particular, enabled the work to be done by fewer employees than would otherwise have been the case and/or the saving of overtime. The PRBs enabled the respondent to require the employees concerned to work at a pace and performance level that was consistently above standard level."

The point is made that the schemes were self-financing (para. 56) and that at first performance was carefully measured (para. 57). It is acknowledged (para. 58) that:

"As the years passed, the bonuses stabilised, in the sense that it became unnecessary to measure the performance of the employees in the way that it had previously been done. Employees were known to be working to the higher productivity and so it was no longer cost-effective to expend money and resources on measurement, evaluation and monitoring. Moreover it was known that the workforce was working to the level of performance which merited the full bonus, because the tasks given to the employees for a particular day could only be accomplished if the employees worked to the level of performance which qualified for a full bonus."

After some further detail, and justification, of the "stabilisation" process, para. 63 pleads:

"The bonuses continued to be genuine performance related bonuses, notwithstanding the fact that they were not measured in the same way as before. The fact that employees usually or always received the full bonus resulted from the fact that they were always working to the full bonus performance level. During the period when bonuses were paid, the standard and productivity of the work done in the departments which paid bonuses was at a high level, and the respondent was much more successful than most local authorities in retaining work that was required to go to tender in-house."

As to why the Claimants do not receive bonus, the Council pleads, at para. 65:

"The reasons for providing a PRB to the holders of the comparator jobs did not apply to the claimants. Their jobs were not suitable for PRBs and/or efficiency savings could not be made through the use of PRBs. No bonus could be designed for the claimant jobs that would have been beneficial for service delivery or which would have been cost-effective."

That proposition is developed in the following paragraphs.

The case as so pleaded was maintained in the Council's opening and closing submissions before the Tribunal.

1. The initial case of both groups of Claimants was that Bury's bonus schemes were, as it was put in Mr. Napier's opening skeleton before the Tribunal, "introduced as a means of remedying the problem of low pay for male manual workers in the face of government-imposed pay restraint [our emphasis]": it followed that the explanations by reference to the need to improve productivity and the "non-bonusability" of the Claimants' jobs were not genuine. But in the course of the hearing before the Tribunal Mr. Napier abandoned that position. He formally conceded in his closing submissions that "there is no sufficient evidential basis for the contention that the original bonus schemes ... were a 'sham' when they were introduced" and acknowledged the good faith of the two witnesses from whom it had heard who had at least partial responsibility for the schemes, Messrs. Rutter and Hale13. Ms. Bourke made a similar concession. Their case thereafter was that, whatever the position initially, by at least 2001 (being the earliest relevant date) it had changed. As Mr. Napier put it at paras. 20 and 21 of his closing submissions:

"20. It is certainly possible that a scheme which was sex-tainted at its inception could, because of changed circumstances, cease to be so at a later date. Equally a bonus scheme that was not sex-tainted when first introduced might become so in the light of changed circumstances. There were very significant changes in the bonus schemes and how they operated in the course of the years between their first introduction and 2001. In particular, well before 2001, the link with measured work had disappeared and different reasons for the payment of bonus at particular levels had emerged. The transition was associated with the dates when bonus became consolidated as part of basic pay, and with the onset of CCT. The original linkage which had, to an extent, existed between measured productivity and pay ceased to exist when there was a move to what is conveniently described as "tender-led" values. In other words, the earnings of employees came to be fixed not by how much work they did or how fast they did it, but by what could be paid to them if their employer was to operate in a competitive market. The reach of the schemes themselves, however, did not change with this fundamental move away from measured work – there was never any serious attempt to bring in female-dominated occupations; indeed there would appear to have been no thought given to this even as a possibility, and no awareness that within Bury that the pay inequality arising from the existence of bonus schemes might give rise to equality questions. That was not appreciated by all accounts until the arrival of Mr. Berry as Director of Personnel in 2001.

21. What did happen over time, however, is that bonus became automatic, so that it was never withheld. It became part of pay, notwithstanding the continuation of formal documentation by the administrative systems of the Respondent which suggested the contrary. The Tribunal is invited to look at the reality of the situation, and the reality is that bonus payments were, at the very latest by the mid-1990s, to all intents and purposes, part of pay, and seen as such by almost all of those who received them. After measurement of work ceased to be a determining principle in the fixing of pay, it was no longer possible to link payment to any defined level of performance (i.e. level 75, 100, 104 etc.). Neither was there any interest on the part of the Respondent in so doing."

The Claimants' case was that throughout the relevant period the payment of bonus to the comparator groups was indirectly discriminatory and could not be justified.

1. The structure of the Tribunal's Reasons is as follows:

(1) After an introductory section it makes, at paras. 22-70, its findings of fact as to the history of local authority bonus schemes in the relevant period and the Council's schemes in particular. In the latter section it makes some findings of fact which are in practice dispositive of some of the legal issues which it comes formally to consider at a later stage.

(2) At paras. 70-251 it examines in turn the bonus arrangements for each of the comparator groups.

(3) At paras. 252-296 it considers each of the Claimant groups, and in particular the extent to which their work lent itself to bonusability.

(4) At paras. 297-303 it sets out what it takes to be the applicable law. This includes a reference to the decision of this Tribunal in Dolphin and to Judge Clark's definition of "sham" in King's College London v Clark (EAT/1049/02), namely that:

"A sham or non-genuine explanation is a false one, designed to disguise the true reason for the difference in pay, itself tainted by sex … ."

The Tribunal at para. 306 describes the phrase "itself tainted by sex" as "otiose", since:

"If the explanation is a sham, then it matters not whether the true reason is tainted by sex."

It continues at para. 307:

"The respondent therefore has, in short, to prove: an explanation which is genuine and not a sham; that the more favourable treatment is due to that reason and is significant and relevant; and that the reason is not the difference of sex. Should the respondent fail to prove any of those matters then the genuine material factor defence fails and there is no room for a justification defence. On the other hand, if the respondent succeeds, the burden reverts to the claimant to show prima facie indirect discrimination."

(5) The Tribunal's reasoning and conclusion in relation to the issue appear in a final section entitled "Application of the law to the facts as found" (paras. 324-368), which we analyse in the following paragraphs.

1. The Tribunal's decision in favour of the Claimants was based entirely on the first of the questions which it identified at para. 307 (see para. 21 (4) above), namely whether the Council had proved an explanation for the non-payment of bonus to the Claimants "which is genuine and not a sham". Its conclusion is summarised at para. 351 in the following terms:

"In conclusion accordingly, upon this issue, we consider that while not a sham when first introduced, the bonus schemes became a sham and were not genuinely linked to productivity. There being no satisfactory evidence of a link between the bonus schemes and productivity we find … that the genuine material factor [defence] fails."

Having rejected the section 1 (3) defence on that basis, the Tribunal did not proceed to consider whether the explanation was "tainted by sex" or the question of objective justification: on its analysis, those issues simply did not arise – cf. its observation at para. 306.

1. We need not set out the full detail of the reasoning leading to that conclusion, but we should note the following points:

(1) At para. 325 the Tribunal refers to Mr. Napier's concession set out at para. 20 above and observes that it would in any event have unequivocally rejected any contention that the schemes "were a sham from the start". It continues:

"We found Mr. Rutter and Mr. Hale both to be honest witnesses. Genuine schemes were introduced which increased productivity, which were monitored and were self-financing. Up to the advent of CCT, the Tribunal is of the view there is more than ample evidence that the bonus schemes introduced by Mr. Rutter and/or maintained and monitored by him and his work study team were genuine. It was due to that factor that the male bonus earners were earning more than the female job groups. It was accordingly, a relevant material factor."

(2) At the end of para. 326 it holds that "the non-introduction of bonus schemes into the female work groups was not caused by the difference of sex" and gives its reasons for that conclusion. In essence it accepts the argument that the Claimants' jobs were "not apt or amenable to a bonus scheme of this type".

(3) Thus far, therefore, the Tribunal's findings favoured the Council. However, it went on to accept Mr. Napier's case that the position as at the introduction of the schemes later changed fundamentally. At paras. 329-330 it says:

"329. We are, of course, concerned principally with the period between 2001 and 2007. The evidence demonstrates that there were very significant changes in the bonus schemes and how they operated over the 20-year period (or thereabouts) to 2001. Mr. Napier submitted that well before 2001, the link with measured work disappeared and different reasons for the payment of bonus at particular levels had emerged. The transition was associated with the dates when bonus schemes became consolidated as part of basic pay and with the onset of CCT. He goes on to say (in paragraph 20 of his submission) that the original link between measured productivity and pay ceased to exist and there was a move to "tender-led" values. The earnings of employees came to be fixed not by how much work they did or how fast they did it, but by what could be paid to them if their employer was able to operate in a competitive market. The payment of bonus became automatic.

330. We consider that Mr. Napier and Miss Bourke (who made similar submissions) are correct in their assertion that the link between productivity and pay had long since ceased prior to 2001. Accordingly, we conclude that although not a sham from the start, the bonuses became a sham well before 2001."

(4) The Tribunal then goes on to make good that general conclusion in relation to each of the particular schemes in issue. It summarises the results at paras. 349-350 as follows:

"349. The Tribunal considers that, looking at the work undertaken in the comparator work departments, the reality was that employees were simply deemed to be working at the required performance level. There was no evidence of any measurement to check that this was so. With the advent of CCT, there was often a need for employees to work at a rate which was competitive. There was no evidence that the respondents had used work study for planning bonus schemes for more than 10 years. No proper costings of schemes were undertaken. Costings were often the result of negotiation rather than based on work study and measured performance and productivity. There was, in the comparator departments, no satisfactory evidence of a link between bonus and productivity from well before 2001.

350. The position is, we think, more than adequately summarised in Mr. Berry's letters sent to the comparator groups of December 2007 (an example of which appears at C2-766). He says there that during the consultation meetings, "concerns were raised that the bonus schemes were outdated and the links between payment of bonus, attendance and productivity may have been lost". That, in our judgment, is an accurate reflection of the position as the respondent itself appears to have recognised. Mr. Berry, indeed, expressed concerns which the respondent had about its vulnerability to equal pay claims (see in particular D-327 and 328 where Mr. Berry raises issues of pay protection and also the need to negotiate with relevant unions in order to protect the respondent against unfair dismissal claims from the bonus earners)."

1. The foundation of Mr. Cavanagh's argument on this appeal is that it was not open to the Tribunal to find, as it did at para. 351 (para. 22 above), that the explanation for the differential between the Claimants' pay and their comparators' between 2001 and 2007 was a sham. It had been conceded that the schemes were at their inception genuinely intended to reward greater productivity, and although on the Tribunal's findings they subsequently ceased to do so there was no evidence that that change was the result of, or involved, any deliberate decision to conceal the true position: indeed the Claimants had not so asserted. The result of the Tribunal taking this route was that it had failed to consider the real issues in the case – namely whether the explanation was "tainted by sex" and, in so far as any tainting took the form of indirect discrimination, whether objective justification had been shown. If it had done so, on its own findings of fact there was no sex taint and the defence under section 1 (3) was accordingly established.

1. We accept the first step in that argument. The Claimants' (eventual) answer to the Council's proffered explanation was that the factor relied on was by the start of the relevant period no longer operative. That is not necessarily, and was not in the present case, the same as saying that the explanation was not "genuine" in the sense in which that term has been glossed in the authorities, i.e. that it was a "sham": see para. 18 above. The Claimants were not saying – and we agree with Mr. Cavanagh that there was no basis on which they could say - that the Council's maintenance of what were described as productivity bonus schemes in the period 2001-7 was designed to conceal a different (and presumably discriminatory) explanation for the non-payment of bonus in the jobs which they did. Rather, they were saying that the disappearance of the link between productivity and receipt of bonus by itself, and irrespective of any dishonest motivation, undermined the Council's defence. We shall have to consider below whether that argument is good or bad, but it is wrong to analyse it in terms of whether the explanation was genuine or a sham. As explained above, the apparent attention given to the concept of genuineness in the authorities – and in particular in Dolphin, to which the Tribunal evidently had particular regard - is misleading if it gives the impression that whether an explanation "works" for the purpose of section 1 (3) depends on whether it is a sham in the understood sense of that word.

1. But the fact that the Tribunal may have used the wrong language, or indeed the wrong analytical tools, does not necessarily mean that its conclusion was wrong, or even that the case requires to be remitted for reconsideration on the correct basis. The Tribunal states explicitly in para. 351 that the reason for its finding of "sham" was that it accepted the Claimants' case that by the relevant dates the original link between productivity and the receipt of bonus had been lost. That finding stands, even if its characterisation as a finding of "sham" is wrong. It is a finding of fact and is indeed unchallenged by the Council as far as it goes. It is the Claimants' contention that that finding is by itself sufficient to defeat the section 1 (3) defence. The central issue on this appeal is whether that contention is correct.

1. Mr. Cavanagh contended that the Tribunal's finding that the link between productivity and bonus had by the relevant dates been lost was immaterial. What mattered was the position at the inception of the schemes. As to that, the Tribunal had found at paras. 325-6 that the provision of productivity schemes for the comparators but not for the Claimants was originally not discriminatory: that finding was conclusive. The statutory question was to what factor the differential was due; in other words the issue was one of causation. The cause of the differential was the original introduction of the schemes for the comparators' jobs (but not for the Claimants'); and that remained the cause even if the schemes had failed to achieve their intended purpose. Since that cause had been held to have nothing to do with sex no question of justification arose: that was what Armstrong 1 had decided. Not only should the appeal be allowed but the claims should be dismissed on the basis that the Council had, on the Tribunal's findings, proved its defence under section 1 (3).

1. Mr. Napier's primary submission, supported by Mr. Allen, was that the Tribunal's finding that the link between productivity and bonus had been lost necessarily meant that the Council failed at the first stage of the "structured approach" because it had failed to prove an explanation for the differential complained of. The essence of the Council's explanation was that the bonuses were intended to reward productivity, so that once it was found – as it was - that they were no longer even understood to do so the "factor" to which the differential was "due" was not proved. It followed that the Tribunal had been right not to consider the subsequent question of whether the explanation was tainted by sex (including, in so far as indirect discrimination was relied on, the question of justification): since no explanation had been proved, that question simply did not arise. Thus the reasoning of the Tribunal should be upheld in its entirety (subject to its inappropriate use of the language of "sham", which did not however undermine its substantive reasoning) and the appeal should be dismissed.

1. We do not accept either submission. We will state our approach and conclusion in summary and then elaborate our reasoning. Contrary to Mr. Napier's submission, we believe that the Council did prove an explanation for the difference in treatment between the Claimants and their comparators, namely that the comparators' jobs were – in the terms of the Council's pleading - "bonusable"14 and the Claimants' were not. However, contrary to Mr. Cavanagh's submission, that was an explanation which (subject to one qualification – see para. 33 below) involved prima facie indirect sex discrimination of the Enderby type and which thus required objective justification. On the Tribunal's findings the difference in treatment cannot be justified. The Council's defence under section 1 (3) thus (subject to the qualification referred to) fails. We proceed to address the essential elements in that reasoning in more detail.

1. The reason why we reject Mr. Napier's primary submission is that in our view it misunderstands what is required by an "explanation" for the differential complained of - or, in the statutory language, what may constitute a "factor" to which that differential is "due". It is not necessary that the explanation should carry with it its own justification. It is essentially descriptive. Although formally the burden of proving the explanation is on the employer, that burden is discharged simply by showing at a factual level how the state of affairs complained of comes about. That being so, in cases such as this defences under section 1 (3) will rarely fail at the first stage. The real battleground will be at the second stage, where the tribunal has to consider whether there has in fact been sex discrimination – which will in turn, where the necessary gender disproportion is shown, turn on the issue of objective justification. That is as it should be: as has been repeatedly emphasised since Wallace, an entitlement to equal pay ought to depend on whether there has been discrimination. In fact, the argument advanced by Mr. Napier is very close, if not identical, to that rejected by the House of Lords in Marshall. In that case the applicants argued that since the employer had done no more than show that they and their comparators were on different pay scales, without showing any specific reason for that difference, it had in truth proved no explanation at all: it was simply saying "we pay less because we pay less". But at p. 204 D-G Lord Nicholls, while describing that argument as "beguiling", explicitly rejected it. If Mr. Napier's approach were right the Council would be required to prove not only how the differential came about but that that explanation was "good", before any question of potential sex discrimination arose. That is contrary to the whole message of Wallace and Marshall. It would also produce undesirable over-complication: it would be necessary for the tribunal to carry out an evaluative exercise not only at the stage of assessing justification but also at the prior stage of establishing the explanation.

1. Turning to Mr. Cavanagh's submission, the reason why we reject it is that we do not accept that showing that the schemes were not discriminatory at their inception necessarily means that they were not discriminatory ten or twenty years later. The explanation, or cause, of a state of affairs is not definitively established simply by showing its historical origins. In the case of direct discrimination it may be pertinent to consider not only why the differential in question first arose but why it has been maintained, particularly if the relevant circumstances may have changed. In the case – more pertinently for this appeal – of indirect discrimination, gender proportions as between the advantaged and disadvantaged groups may have changed, or there may be reasons why a justification which was once good no longer remains so.

1. We accept that the route which the Tribunal took meant that it did not in terms consider either of the issues on which on our analysis the claims turned – that is, whether there was a sufficient disproportion between the gender make-ups of the Claimant and the comparator groups to raise a prima facie case of Enderby-type *indirect discrimination or, if so, whether the differential could be justified – and that it did not therefore make any explicit findings on those issues. *But, subject to the qualification already referred to, that does not mean that the appeal must be allowed. We take the two issues in turn.

1. As regards gender disproportion, it was common ground before the Tribunal that most of the Claimants belonged to groups which were predominantly female and which attracted the operation of Enderby. But the Council contended that there were some exceptions, specifically in the cases of the caretakers; bar staff; two groups known as the "driver/carers" and "CPU drivers"; and an individual Claimant described as "the vending assistant"15. In their cases it was said that although the lead Claimants were women the groups to which they belonged were either predominantly male or were too small for any statistically significant conclusions to be drawn about them. The Tribunal made no findings on that question. That was consistent with the basis of its reasoning, but on the analysis which we believe to be correct it was wrong not to do so. Indeed, if the Council is right on the facts, the position as regards these Claimants neatly illustrates the vice inherent in the Tribunal's approach, since a group of employees will have succeeded in a claim for equal pay in circumstances where there is no reason to suppose that any gender discrimination was operating. As regards these Claimants the case will have to be remitted to the Tribunal to consider (to the extent that there is any issue) what groups they should be regarded as belonging to for Enderby purposes – a question in relation to which we heard no argument - and what the relevant gender proportions in the relevant group or groups are.

1. As regards justification, although the Tribunal did not consider this as such, in our judgment the finding that throughout the relevant period the bonus schemes enjoyed by the comparators no longer reflected any link between receipt of bonus and actual productivity necessarily means that the non-payment of equivalent sums to the Claimants could not be justified. The object of rewarding productivity, in jobs which lent themselves to measured assessment of work, was the only possible justification for the different treatment of the Claimants and their comparators.

1. We would therefore dismiss the Council's appeal, save in the case of the Claimants identified in para. 33 above, where we order remittal on the basis indicated.

1. Sunderland's section 1 (3) defence was essentially to the same effect as Bury's. Paras. 20-24 of its "Generic Response Form" read as follows:

"20. Where particular Claimants are able to rely on the provisions of Section 1 (2) (b) of the Act, the First Respondent will contend that at all material times variations in pay were genuinely due to material factors which were not the difference of sex and which amounted to a material difference between the Claimants' cases and the relevant Comparators.

21. In relation to the Comparators who were formerly employed under the provisions of the White Book in the jobs listed in Paragraph 10.5 above, differences in pay between them and other workers rated as equivalent were explained by cost-neutral bonus payments arising out of productivity/efficiency incentive schemes which was a material difference which is not the difference of sex. The use of such a scheme to meet the objectives of the First Respondent is objectively justifiable if the same needs to be justified which is denied.

22. Although it is denied that the First Respondent is required to prove that a bonus scheme should, or could have been made available to the Claimants, the nature of the work of the Claimants is/was such that no similar scheme, or term could in any event have been transposed in their case because the nature of the work was quite different and it was not possible to provide bonuses based on enhanced productivity. The work done by the Comparators involved the provision of different types of services. It is denied that a bonus scheme, cost-neutral to the First Respondent could have been developed for the Claimants that would be beneficial for service delivery.

23. The local negotiations and agreements entered into with the trade unions were free of discrimination and reflected an objectively assessed need on the part of the First Respondent for such bonus schemes.

24. There could be no advantage to the First Respondent, to service users, or to ratepayers in introducing any bonus scheme to the Claimants in that the Claimants' salaries already compared favourably with that of the private sector. Further, bonus schemes were introduced on a cost neutral basis and, where appropriate, were introduced alongside a reduction in the numbers of the relevant workforce. The nature of the work performed by the Comparators was such that such an outcome was achievable. The nature of the Claimants' work was such that a reduction in the workforce associated with an increase in output would be detrimental to service delivery."

(The "market forces" arguments adumbrated at para. 24 was not in the end persisted in.)

1. The Claimants did not have to plead an answer to the Council's response and we have not seen their written submissions before the Tribunal. However, the nature of their case appears sufficiently from the Tribunal's Reasons as summarised below. In particular, it is clear that they took the point that, at least as regards the relevant period, the Council's proffered explanation for the differential complained of was a sham.

1. The structure of the Reasons is as follows:

(1) After an introductory section the Tribunal at paras 9-18 gives, as section I and II, a summary of the relevant law.

(2) Section III is headed "The Issues and the Parties' Submissions". The parties had agreed an elaborate list of issues, which the Tribunal sets out but which we need not reproduce here. We note however that issue 10 was:

"Whether the bonuses paid to the comparators by the first respondent were paid by the first respondent genuinely to encourage and reward productivity."

The Claimants had sub-divided that into several sub-issues, though the accuracy of the division was disputed. One of the sub-issues was "were the bonuses genuine or a sham?"; and Mr. Engelman is recorded as having submitted that that was the most important question. Mr. Reade is recorded as having responded at length on the question of whether the bonus schemes in question were a sham, which he submitted was equivalent to whether they were entered into and maintained in bad faith.

(3) Section IV sets out the Tribunal's findings of fact over some 54 pages. We have already sufficiently summarised the broad background. At paras. 20.12-41 the Tribunal examines with great care each of the three schemes under which the comparators had been entitled to bonus – the gardeners', the refuse collectors' and road sweepers' schemes – examining the extent to which productivity was in fact measured and whether employees in practice ever received less than the maximum bonus figure. In the case of the gardeners' scheme it concludes, at para. 20.24:

"In our judgment, by 1998 the purported bonus payments had long since ceased to have anything to do with productivity – even on the basis of maintaining rather than increasing standards – and were, as they were seen to be by workforce and management alike, a fixed part of salary."

Essentially the same conclusion is reached in the case of the other two schemes. At para. 20.34 the Tribunal finds in relation to the refuse collectors' scheme that:

"… by 1998 the payments had … long since ceased to have anything to do with productivity … and were perceived as being and were a fixed part of salary."

In relation to the road sweepers it finds, at para 20.41:

"As with the other schemes, it may very well be that the amount of work actually being carried out by each employee … was important to the respondent in calculating its tenders under CCT but the bonus had long since ceased to be, or to be seen as being, linked to productivity. In our judgment, neither the workforce nor management would have thought otherwise."

At para. 20.44 the Tribunal considers whether the work done by the Claimants would have lent itself to the application of a true productivity bonus scheme, an issue which it describes as "bonusability"16. In the case of all categories save cleaners it holds that it would not.

(4) The Tribunal's conclusions appear at section V of the Reasons, comprising paras. 21-24, which we analyse more fully in the following paragraphs.

1. The Tribunal's final conclusion is stated in para. 24 as follows:

" … for the reasons we have given, we came to the conclusion that the bonus schemes were not genuine and thus that they were sex-tainted in relation to all claimants other than the leisure centre attendants … Accordingly, with regard to all [except the leisure centre attendants], the purported defence under section 1 (3) fails."

The route by which that conclusion is reached is not as easy to follow as it might be because it is presented in three very long paragraphs without any headings or breaks to act as signposts. But we can summarise it as follows.

1. Para. 21 addresses "the genuineness of [the Council's] explanation, pointing out that Mr. Reade had described that as "the first question" and Mr. Engelman as "the most important question". The Tribunal refers to its earlier finding that for many years, and certainly throughout the period covered by the claims, "the bonus payments were seen by the workforce and management alike to be a fixed part of salary". It refers to a passage from the judgment of this Tribunal in Dolphin (above), at para. 30 (p. 174), where Judge McMullen said that the question for the tribunal was not whether the scheme relied on by the employers was good or bad but

"… simply … whether it is genuine. Were the schemes intended to and do they in fact achieve productivity improvements?"

It continues:

"Whatever these three bonus schemes were originally intended to do and actually did in the 1970s, we had to consider that question in the context of a period commencing in the late 1990s. The position as it was more than 20 years earlier was, of course, of relevance and evidential value. However, in the intervening period a great deal had changed. In our judgment, with regard to none of these schemes was there the maintenance of the achievement of productivity improvements. Within the meaning of the test laid down in Marshall as explained in Dolphin, each of these three bonus schemes was a sham and a pretence and therefore could not be a explanation for the variations in pay."

After discussing the case of one individual comparator in more detail it concludes para. 21 by saying:

"Accordingly, we rejected the genuineness of the first respondent's proffered explanation."

1. At the start of para. 22 the Tribunal says that its rejection of the Council's explanation in the previous paragraph "was not entirely the end of the matter": in principle it was still necessary to consider whether the factor relied on was discriminatory. (It is convenient to note at this point that Mr. Allen submitted that that approach did not make sense: if the Tribunal had rejected the Council's proffered explanation for the differential then the section 1 (3) defence failed at the first hurdle and it was not necessary to proceed any further. We see some force in that criticism but we need not pursue it.) The Tribunal proceeded to consider the relative gender balances in the groups who received bonus and in those who did not. As regards the former it found that the employees receiving bonus were all, or almost all, men. As regards the latter, it thought it right to do the exercise separately for each of the six groups identified at para. 6 above. Having done so, it concluded that each of the groups save the leisure centre attendants was overwhelmingly female. Thus in relation to those groups "the … material factor defence failed and Mr. Reade did not suggest that it could do otherwise". That conclusion appears, with respect, rather to jump the gun. All that the Tribunal had at that point established was a sufficient gender disproportion between the claimant and the comparator groups to raise a prima facie case of Enderby-type indirect discrimination, but that would not mean that "the material factor defence failed" unless the Council were unable to prove objective justification: as to this, however, see below.

1. At para. 23 the Tribunal turns to consider whether, if, contrary to its finding, the schemes had indeed been genuine, the defence under section 1 (3) would have succeeded. It starts by concluding that they were not directly discriminatory. It then addresses the question of indirect discrimination, making it clear that the case advanced was of the Enderby type. It points out that its conclusion in para. 22 meant that the statistics established (except in the case of the leisure centre attendants) a prima facie case of sex discrimination. As for objective justification it proceeds to "[look] at this question as if we had found that at the material times the male bonus schemes had been genuine". On that basis it holds that the real question is whether it would have been justifiable to maintain such schemes without introducing similar schemes for the groups to which the Claimants belonged. Consistently with what it had held at para. 20.44 (see para. 38 (3) above), it finds that the distinction in treatment would have been justifiable except in the case of the cleaners, because the Claimants' jobs in question were genuinely not "bonusable" – that is, they did not lend themselves to the measurement of productivity.

1. Mr. Reade submitted that the Tribunal had plainly misunderstood what was involved in a finding that the Council's explanation for the payment of bonus to the comparators but not to the Claimants constituted a sham. In particular, it appeared to have been misled by Judge McMullen's statement in Dolphin, quoted at para. 21 of the Reasons (see para. 40 above), to the effect that the test of genuineness was whether the schemes were intended to and did in fact achieve productivity improvements. It was clear that the Tribunal's reasoning was not based on any consideration of the honesty of the Council's witnesses or decision-takers and that it had not found any intention deliberately to conceal the true reasons for the payment/non-payment of bonus. Rather, it found the explanation to be a sham simply because the relevant schemes no longer in fact rewarded productivity. That did not justify a finding of sham. To the extent that it was held in Dolphin that it was sufficient to show that the schemes no longer in fact achieved productivity improvements, that was wrong: there were many perfectly genuine reasons why a scheme might fail to achieve its objectives whether initially or as a result of subsequent developments.

1. It will be apparent from what we say at para. 18 above and from our conclusions on the Bury appeal that we regard those submissions as well-founded. "Sham" connotes an explanation which has been deliberately fabricated in order to present things otherwise than as they are. We agree that if in Dolphin this Tribunal intended to give it a wider application it was, with respect, wrong. It is clear from the Tribunal's reasoning that it did not intend to make any such finding.

1. But it will also be apparent from our earlier reasoning that we do not regard the focus on the question of whether the Council's explanation was genuine (in the sense identified) or a sham as helpful. The Tribunal's approach may be understandable because it seems to have reflected the way in which the case was put by the parties, but it nevertheless misses the real points, namely (a) whether the Council had in fact advanced a factual explanation, good or bad, for the different treatment of the Claimants and the comparators and (b) whether that explanation was tainted by sex.

1. As to those points, there is considerable overlap with the questions already considered in Bury, but the arguments are not identical and we should take it step-by-step.

1. The starting-point is that the Council did in our view prove an explanation for the differential complained of, namely that the comparators enjoyed the benefit of a productivity bonus scheme whereas the Claimants did not. Mr. Allen advanced a similar argument to Mr. Napier's, to the effect that the Tribunal's finding that the schemes no longer in fact rewarded productivity, and indeed were no longer even understood to do so, deprived the explanation of any content. For the reasons given at para. 30 above we reject that submission.

1. The next question is whether that explanation was "tainted by sex". As to this, the Claimants' case depends on Enderby-type indirect discrimination. The issue of disproportion in the gender balance as between the Claimant and comparator groups is decided in the Claimants' favour (save in the case of the leisure centre attendants, in respect of whom, as noted above, there is no challenge to the Tribunal's finding) by the findings at para. 22 of the Reasons (see para. 41 above).

1. The remaining question is whether the difference of treatment between the two groups was objectively justified. If the Tribunal was entitled to find that at the material times "the purported bonus payments had long since ceased to have anything to do with productivity … … [and had become] … a fixed part of salary" (see para. 38 (3) above), then in our view it follows that their extension to the comparators but not to the Claimants could not be justified, for the reasons given at para. 34 above. (In this connection it should be noted that para. 23 of the Reasons is not concerned with the question of objective justification in this context. Rather, it is addressing the question whether, if the schemes did genuinely reward productivity, it was justifiable not to put in place similar schemes for the Claimant groups. But the premise of that question is negatived by its earlier findings.)

1. Unlike in Bury, however, the Council challenges the Tribunal's finding that by the material dates the schemes had ceased to have any connection with productivity. In order to explain the nature of the challenge it is necessary to set out, in bare outline, what the Tribunal found about each of the three schemes with which it was concerned:

Gardeners*. The original bonus scheme for gardeners assigned numerical values to levels of productivity in accordance with the then common science, or pseudo-science, of "work study". The amount of work which an employee might be expected to carry out while working at normal pace was described as "75". Under the scheme if an employee worked at "100" he would receive a bonus of 33?% of basic pay, and if he worked at "122" he would receive 50%. Systems were in place to measure and record an employee's achievement of the relevant levels of productivity. But the scheme changed in 1990 as a result of the changes associated with the introduction of CCT. Employees were assigned tasks for a given week which it was calculated (though not by way of any formal re-measurement) that they could only achieve by working at "122 performance", and on the basis that they would indeed perform those tasks they were to be paid what was described in the relevant agreement as "a fixed, pre-determined bonus" of 50%. If the tasks were finished within the week they could go home early on Friday (i.e. a variant of "task and finish"). If they failed to do the work assigned they would in theory have to stay on in their own time to complete it. However, the evidence was doubtful as to whether that ever occurred, and if it did it was not only highly exceptional but, on the evidence, highly contentious. The Tribunal found that in practice the bonus was always paid and constituted simply part of what employees were paid for working at the rate expected of (and by) them. Since bonus was now earned automatically there were no arrangements to monitor productivity: though employees still submitted paperwork the timings shown in the returns were entirely notional.

Refuse collectors*. As with the gardeners, the original scheme provided for bonus to be paid by reference to achievement of specified targets set using elaborate work study methodology. Bonus was initially variable, depending on achievement of particular targets, but from a date in the late 1980s was fixed at 50%. A new agreement was introduced in 1994, in response at least in part to the introduction of new ways of working (and in particular the introduction of wheelie-bins capable of being emptied mechanically into the refuse truck), which resulted in a reduction in the number of rounds and in the size of the workforce. Rounds were redesigned, without any re-measurement using work study techniques, and thenceforward employees received a 50% bonus simply for doing the rounds assigned: when the work was done they could go home. There was no monitoring of productivity: given the scheme as now operated it would have had no purpose. The rounds were comfortably achievable within the times assigned.

Road sweepers.* The evidence and findings with regard to the road sweepers' scheme were rather less full but they were to substantially the same effect. Bonus was paid simply for completing the assigned tasks for a week and was never withheld.

1. In its Notice of Appeal the Council pleaded what appears to be a very detailed attack on the Tribunal's factual findings in relation to all three schemes. The recurrent theme is that the Tribunal had only heard evidence about the schemes from the Council's witnesses, who had consistently maintained that the schemes retained a link with productivity: that evidence was essentially factual and had not been significantly shaken in cross-examination, and the Tribunal had accordingly not been entitled to reject it. The number of particular points made was somewhat reduced in Mr. Reade's skeleton argument, though he told us that that was simply because he was focusing on his best examples. However, in the course of oral argument it became clear that the underlying issue did not depend on any question of detail. Mr. Reade's essential point, echoing that made by the Council's witnesses, was that because the weekly tasks for the performance of which the employees received their pay, including bonus, were allocated on a basis which reflected what had originally been measured as achievable only by optimal productivity – 122 in the case of the gardeners' scheme – the performance of those tasks necessarily involved a high degree of productivity for which the employees were entitled to reward. We do not accept that argument. It is important not to lose sight of the purpose for which the question is being asked. The comparators were being paid more than the Claimants for doing work which had been rated as being of the same value. That could only (so far as relevant for present purposes) be justified if the additional amount represented a reward for doing more than the basic work expected of them. The effect of the Tribunal's findings was that that was not the case. The fact that, as a matter of history, the level of work expected reflected what had once been regarded as exceptional does not mean that performance of the relevant tasks was exceptional during the relevant period or that that was what the bonus was being paid for.

1. Accordingly, we dismiss the Council's appeal.

1. The basic facts giving rise to the pay protection issue can be very briefly stated. As already noted, the formal implementation of a single status agreement introducing Green Book terms meant that White Book employees previously in receipt of bonus ceased to be entitled to it. In Bury this occurred with effect from 1st April 2007.17 However, this would mean an immediate and substantial decrease in weekly earnings for many employees and was understandably extremely unpalatable. Accordingly the unions insisted on terms under which employees who lost out by the removal of bonus would continue to receive additional payments over, potentially, a three-year period, so as to cushion the blow. Under the scheme employees who would suffer a loss of earnings of more than 40% received payments equivalent to the entirety of their loss in the first year, to three-quarters of that loss in the second year, and to half in the third year. Those suffering a loss of between 20% and 40% would receive protection only for the first two years and those suffering a loss of under 20% only for the first.

1. As, again, already noted, prima facie any female employee doing comparable work – i.e. work rated as equivalent or work of equal value - with a male employee in receipt of pay protection is entitled to payment of an equivalent amount. At first sight the Council has a straightforward defence to that claim under section 1 (3): the denial of pay protection to the Claimants is due to a material factor other than the difference of gender, namely that the comparators have previously been in receipt of a benefit (i.e. bonus) which is now being withdrawn whereas the Claimants have not. No doubt there is a gender disproportion between the two groups, but the difference in treatment is justified because the former need a cushion and the latter do not. But it is not as simple as that. The reason that the Claimants have not been in receipt of bonus is – on the findings already made - sex-tainted, and the result of the Tribunal's decision on the primary issue is that they will indeed now receive it, albeit in arrears. Viewed in this way, "pay protection" is in fact simply a continuation of the previous discrimination, albeit on a diminishing basis, and it is hard to see why the Claimants should not be entitled to whatever their comparators receive.

1. Those issues, in the context of local authority equal pay claims, were authoritatively considered by the Court of Appeal in Redcar and Cleveland Borough Council v Bainbridge [2009] ICR 133. The judgment covers not only the appeal in Bainbridge itself (and another appeal from a different decision between the same parties) but also an appeal from the decision of this Tribunal in Middlesbrough Borough Council v Surtees [2007] ICR 1644 which raised similar issues. When we need to refer to the two cases separately we will do so as "Redcar" and "Middlesbrough". In both cases the local authority employers had entered into single status agreements incorporating pay protection arrangements under which, for a limited period, employees were allowed to retain benefits enjoyed under the old pay system. The female claimants claimed the benefit of those pay protection arrangements under the 1970 Act. In Redcar the council had as at the date that it concluded the agreement already conceded that the benefits in question constituted indirect sex discrimination (of the Enderby type); and the claimants said that it followed that the pay protection arrangements were also discriminatory. In Middlesbrough no such concession had been made. In both cases the council contended that, whatever the justifiability of the benefits under the old system, their temporary continuance by way of pay protection could be justified on the basis that they represented legitimate transitional arrangements to cushion the impact of their withdrawal on existing employees. It is necessary to set out separately how that defence fared in the two cases:

* In Redcar the employment tribunal rejected the defence. Its reasoning is analysed by the Court of Appeal at paras. 122-3 (pp. 173-4) (see also para. 136 (p. 177)). The tribunal accepted that affording pay protection to the (male) employees who would lose by the abolition of bonus was legitimate. But the council neither considered at the time nor advanced any arguments subsequently as to why equivalent payments should not be made to female employees who should have been receiving the same bonus: it was aware that its non-payment of bonus to at least some groups of female employees was discriminatory. No evidence was given as to why those women should not have been included in the scheme. The tribunal would have been prepared to consider evidence about cost but it was given none. On appeal this Tribunal (Elias P presiding) found that the employment tribunal had misdirected itself, but it considered the justification issue for itself and came to the same conclusion. As summarised by the Court of Appeal (para. 138 (p. 177 G-H)) its essential reasoning was that

"where … the employer knew that he had been discriminating against the women and failed to consider whether [he] could exclude that discrimination after change over, it was impossible … to justify his arrangements."

The Court referred in particular to a comment by Elias P that:

"It would frustrate the principle of equality to deny [the claimants] benefits which, as everyone accepted, they were legally entitled to receive [our emphasis]. Had the employers corrected the inequality when they ought to have done, then there can be no doubt that these women would have been in exactly the same position as their comparators. They would have had the benefit of the pay protection."

* In Middlesbrough the employment tribunal also rejected the section 1 (3) defence and thus upheld the claimants' claims to pay protection. It found that, although at the time that it entered into the single status agreement the council had not accepted that its previous pay arrangements were discriminatory, it should have appreciated that there was a real risk that that would be held to be the case: see the Court of Appeal's summary of the reasoning at para. 148 (p. 180 D-E). Against that background it held that:

"[a]lthough … a period of protection of previously discriminatory pay, limited in time in order to phase it out, would be a legitimate aim, it had not been shown that this scheme was an appropriate and proportionate means of achieving that objective when the size of the disadvantaged group and the cost of including or excluding that group had not been calculated"

(see, again, the Court of Appeal's summary of the reasoning at para. 149 (p. 180G)). In that connection it attached particular weight to the fact that the pay protection arrangements had not been costed. That decision was overturned by this Tribunal, again presided over by Elias P, which held that the only possible conclusion on the facts as found was that the council had established its justification defence.

1. The Court of Appeal held that in both cases the decision of the employment tribunal was unimpeachable, and in the case of Redcar it added that the decision was plainly correct. It also endorsed the reasoning of this Tribunal in Redcar on the question of justification (albeit that it held that that reasoning had been unnecessary because the employment tribunal had not in fact misdirected itself). The analysis is very lengthy, but it can be sufficiently summarised as follows18:

(1) In any case where the existence of past direct discrimination has been "recognised", the continuation for the future of such discrimination in the form of transitional or phasing-out arrangements cannot be justified: see para. 129 of the judgment (pp. 175-6), applying the decision of the European Court of Justice in Smith v Avdel Systems Ltd [1995] ICR 596. The removal of such discrimination must be "immediate and full".

(2) The position as regards transitional arrangements which continue past indirect discrimination is more subtle. The key passage in the judgment is at paras. 131-5 (pp. 176-7). There appear to be some differences of emphasis in different parts of that passage, but the essential message is at para. 134 (p. 177 A-B), where the Court holds that in cases of indirect discrimination a "more flexible principle" should apply, under which the continuation of past indirect discrimination will not be unlawful if it can be justified. The tribunal will need to apply the proportionality test, but in doing so it will be relevant to consider "the employer's state of knowledge about the discriminatory effect of his provisions and the extent to which he tries to minimise that effect". That observation is picked up and amplified at para. 156, where the Court says:

"As we have said, we do not think that it was correct to say that, merely because the employer knew that it had been discriminating in the past, it became impossible to justify the continuation of some discrimination after the reorganisation. We think it would still be possible (although more difficult) for an employer to justify where he was aware of the past discrimination. He could do so by demonstrating that he had done all he could to minimise the effect of the continuing discrimination but he was unable to eliminate it immediately."

(3) The employer in Redcar, having made the concession that it did (see para. 54 above), knew or ought to have known at the time that it agreed the pay protection arrangements that its previous pay system was discriminatory in the relevant respects. That being so, both the employment tribunal and this Tribunal were right to hold that the transitional arrangements could not be justified, and the claimants were entitled to the benefit of the pay protection provisions: see paras. 136-140 (pp. 177-8).

(4) In Middlesbrough the employer had at the material time made no such concession. The Court of Appeal reversed the decision of this Tribunal that the employment tribunal's reasoning, as summarised at para. 54 above, was flawed and held that its conclusion was legally open to it: see paras. 169-171 (p. 186).

(5) Although that was enough to dispose of the appeal, the Court of Appeal made some observations on the reasoning of this Tribunal in deciding that Middlesbrough's pay protection arrangements could be justified. At paras. 172-3 it commented on a particular passage which we have to consider below (see para. 66) and which we will accordingly not set out here. It then proceeded to criticise an observation to the effect that "all schemes of this nature" were likely to be justifiable. It said:

"174. We are concerned about this … statement which appears, in this kind of case, to take away from tribunals the task of deciding whether the employer's arrangements can be objectively justified notwithstanding the fact that they are prima facie indirectly discriminatory. The EAT seems to think that the same answer will apply in every case where, following pay reorganisation, pay protection is given to the employees who have been the advantaged group under the old arrangements and is denied to the gender group who have been discriminated against in the past. We do not think it can be right to take that essential evaluation away from tribunals. Moreover, we cannot accept that the answer will always be as the EAT thinks.

175. It is apparent that the EAT in the Middlesbrough case was sympathetic to the position of an employer who, on finding that he has been or might have been discriminating against women seeks to reorganise his pay structure so as to avoid discrimination in the future. The EAT recognises that employers (or at least public employers such as these local authorities) will have a limited budget and many calls upon their resources. The argument accepted by the EAT in the Middlesbrough case is that in such circumstances, the employer will always be entitled to say that it must continue to discriminate against the women for another three or four years (albeit to a reducing extent) because it cannot afford to bring them into line with the men at the time of reorganisation. We find that a very surprising and undesirable general conclusion. We accept that a large public employer might be able to demonstrate that the constraints on its finances were so pressing that it could not do other than it did and that it was justified in putting the need to cushion the men's pay reduction ahead of the need to bring the women up to parity with the men. But we do not accept that that result should be a foregone conclusion. The employer must be put to proof that what he had done was objectively justified in the individual case.

176. A significant part of Mr. Jeans' submission to this Court was concerned with the difficulties faced by local authorities with large workforces. His submissions began from the premise that the authorities were not in any way to blame where the pay of women employees had fallen behind the men's in a discriminatory way. It was unrealistic to expect the authorities to be able to prevent such inequalities arising. So, in cases of this kind, he submitted, that was where they were starting from. "We are where we are". Authorities such as Middlesbrough were having a very difficult time because large numbers of equal pay claims were cascading down upon them. At times, in the submissions of both Mr. Cavanagh and Mr. Jeans, it appeared that the authorities' stance was that they were under a duty to get rid of sex discrimination in pay only when claims were made and the inequality of pay had been demonstrated to them. In short, they only had to ensure equal pay in response to a successful claim. The EAT appears to have been sympathetic to those submissions. But, as Mr. Allen pointed out, equal pay legislation came into force in 1975 and, at that time, employers had been given 5 years in which to reorganise their pay structures so as to get rid of sex discrimination in pay. Mr. Allen submitted, and we accept, that an employer's duty since then has been to ensure that there was no sliding back into discrimination. Employers have been and are under a continuing duty to avoid sex discrimination in pay, regardless of whether their female employees seek to assert their rights by litigation.

177. If the general rule suggested by the EAT were to apply, employers would be able to allow their pay structures to fall out of compliance with the law and then, when forced to do something about it as the result of claims being brought, would be able to assert that they could legitimately take a further three to four years to bring their pay structures into compliance. We do not think that such a situation is consistent with the provisions of the 1970 Act which have now been in force for over 30 years. We consider that it will be possible for an employer to justify the continuance of indirect sex discrimination through the discriminatory application of a pay protection scheme but not as a matter of course and only where the employer satisfies the test of justification as set out in Barry/Cadman."

(The reference in the final paragraph is to the summaries of the requirement of objective justification by Lord Nicholls in Barry v Midland Bank plc [1999] ICR 859, at p. 870, and in the judgment of the ECJ in Cadman v Health and Safety Executive [2006] ICR 1623, at para. 32: these are set out at paras. 51-52 of the Court of Appeal's judgment (p. 154 D-F).)

1. It appears from Bainbridge that there is no universal answer to the question whether pay protection arrangements of the kind put in place by Bury (which appear to be broadly typical of the arrangements negotiated by local authorities on the implementation of single status) can be justified, nor even any single touchstone by which those arrangements that can be justified can be distinguished from those that cannot. The question is one for the assessment of the tribunal in each case applying the established test of proportionality. But it is fair to observe that the Court of Appeal's approval of the decisions of both the Redcar and the Middlesbrough tribunals upholding the women's claim to pay protection (and indeed the comment that the former was plainly right) suggests that, to put it no higher, employers will not be able to justify withholding pay protection from claimants without advancing cogent and specific reasons for what is in effect a continuation (albeit limited) of past discrimination.

1. We should note for completeness that since Bainbridge this Tribunal has had to consider the question of pay protection in two further cases – Coventry City Council v Nicholls [2009] IRLR 345 and Pulham v London Borough of Barking and Dagenham [2010] ICR 333. Both decisions seek to apply Bainbridge (though – see n. 18 above - Pulham may not have summarised its effect entirely accurately), and there is no further development of the relevant principles.19

1. The Council's case on pay protection is pleaded at length in the document referred to at para. 19 above. In summary it is:

(a) that it was essential, both as a matter of fairness, and to prevent industrial relations unrest, to offer a cushion to those employees who would lose out financially by the withdrawal of bonus;

(b) that it would not have been practicable as at 1 April 2007 to make any payments to the Claimants since until the primary claims had been decided it could not be known what entitlements fell to be protected;

(c) "that … it would have been too expensive to extend pay protection in that manner" (para. 110 (4);

(d) "that if pay protection was offered to those who had not actually suffered a drop in pay it would be a payment for a different reason from the reason why the payments were being made to the comparators" (para. 110 (5)).

1. As to (a), this was not challenged either before the Tribunal or before us: the Claimants' case was not that the payment of pay protection to the comparators was not justified but that it should have been paid to them as well. As to (d), it is clear from Bainbridge that this argument is bad: as already noted at para. 54 above, the only reason why the Claimants have not suffered a drop in pay is that the payments in question had been unlawfully withheld.

1. Accordingly the real issues were (b) and (c). As to those, the Council's evidence was given by Mr. Berry, its Director of Personnel, at para. 18 of his witness statement. He said:

"It would not have been practicable for the Council to offer pay protection to the claimants who had not actually suffered a drop in pay, but who might have suffered a drop in pay had they been paid the same as the bonus earners, because it would have been impossible to be certain who had a good claim, until the litigation had run its course. And at the time when pay protection was under consideration for the bonus earners, no equal pay claims had been lodged against the Council. Even if it were possible to determine who had a good claim in advance, it would not have been possible to calculate their "loss" and therefore what they would have been entitled to receive by way of pay protection, because the amount would have varied depending upon which bonus earner was selected for that exercise. It would therefore have been grossly irresponsible and contrary to the Council's fiduciary duty to offer pay protection to the claimants on this basis and in any event it would have imposed a financial liability on the Council which it could ill-afford. And had the Council offered pay protection to those who had not actually suffered a drop in pay, this would have been a payment made for different reasons from the reasons why payments were being made to the comparators."

No attempt was made, in the statement or the documents referred to (nor, to anticipate, in his oral evidence), to put a figure on the proposition that extending pay protection to the Claimants was something which the Council could "ill afford".

1. It is necessary to set out the entirety of the Tribunal's reasoning on the pay protection issue, which reads as follows:

"353. … Mr. Cavanagh accepts that, in the circumstances of our findings, there will be prima facie discrimination in relation to pay protection regardless of the disparate impact issue … . This is because the Court of Appeal held in Bainbridge that the existence of historic pay discrimination which benefited the people who are now receiving pay protection means that there is prima facie discrimination as regards pay protection. It follows, therefore, that historically, the respondent in this case having chosen to benefit a predominantly male group in circumstances where the female group was disadvantaged, the application of pay protection post 1 April 2007 is, inevitably, tainted by indirect sex discrimination.

354. At paragraph 156 of Bainbridge, the Court of Appeal said this:

"We do not think it was correct to say that, merely because the employer knew that it had been discriminating in the past, it became impossible to justify the continuation of some discrimination after the reorganisation. We think it would still be possible (although more difficult) for an employer to justify where he was aware of the past discrimination. He could do so by demonstrating that he had done all he could to minimise the effect of the continuing discrimination but he was unable to eliminate it immediately."

355. In the context of pay protection, the Employment Appeal Tribunal in Surtees said the following at paragraph 108:

"We consider given that the purpose of the scheme was to cushion employees from the potentially disastrous effects of a sudden drop in pay, the Council was entitled to take the view that it should limit the benefit to those actually in that group and to exclude all others even if some of them ought to have been in the group. Unless the pay was actually being received, there was nothing to protect. We think that is itself sufficient justification, but it is reinforced by the fact that the need to reach a protected pay arrangement, with the agreement of the unions, was crucial to the making of a job evaluation scheme. Any assessment of future costing would inevitably be highly speculative and would undermine the ability to obtain agreement for the scheme."

356. Mr. Cavanagh submitted that only the comparators were going to suffer a drop in pay and that is the reason why the comparators were given pay protection and the claimants were not.

357. The Tribunal does not accept that the granting of pay protection to the comparators was not tainted by sex. It clearly was. On any view, the GMF having failed, there arises the statutory presumption of the difference of salary between the claimants and their comparators as being due to the difference of sex. That presumption may be displaced by the respondent establishing a genuine material factor defence under Section 1(3). This respondent cannot so do. The reason why the comparators were in receipt of pay protection (and therefore a higher salary) than the claimants was tainted by sex discrimination. The Equal Pay Act provides for that statutory presumption of sex discrimination.

358. Nevertheless, the Tribunal considers that the respondent has made out its case that it can justify the indirect discrimination resulting from the pay protection policy. The measures adopted by the respondent correspond to a real need upon the part of the respondent to implement single status and abolish the bonuses. There was recognition by the respondent that the bonuses were potentially discriminatory (as indeed we find they were). The co-operation of the unions was essential to the implementation of single status and, as a prelude thereto, of undergoing a job evaluation. We agree with Mr. Cavanagh's submissions, accordingly, that pay protection for the previous bonus earners was essential in industrial relations terms. Mr. Berry experienced a vivid illustration of the strength of the unions when industrial action was threatened by the refuse collectors in March 2008.

359. The dilemma facing Mr. Berry was well illustrated by the documents to which we have just referred at D327 to 328. The respondent was clearly caught in the dilemma of facing potential equal pay claims from the claimants and potential unfair dismissal claims from the comparators.

360. Many witnesses said that morale dropped when the issue was raised by Mr. Berry (by his letters of December 2006) that bonus was to be withdrawn. Pay protection was also recommended in the Green Book (see M-8).

361. We also accept Mr. Berry's evidence that there was no way of working out with any degree of precision to whom to make pay protection payments to (other than the former bonus earners). It would be an extremely difficult exercise to calculate how much pay protection payment to make to each individual employee not in receipt of bonus. For example, at A-282, the Stefan Cross claimant cites 94 comparators for her position of Night Care Assistant.

362. We agree with Mr. Cavanagh that this was not a case where the respondent was "bang to rights". We attach particular significance to the fact that the bonuses were not a sham, as we have found, from the outset. There appears to be no challenge from the claimants to the length of the pay protection period. It is limited to a three-year period on a tapering basis. This was a significant departure from the former scheme of protection to which Mr. Berry refers in his witness statement of red circling in effect for four years (with no tapering).

363. Objectively balancing the discriminatory effects of the pay protection on the one hand and the reasonable needs of the respondent on the other, the Tribunal is driven to the view that objectively, pay protection in this case can be justified. Mr. Morton in cross-examination indicated that he was in favour, in his capacity as a trade union official, of the principle of pay protection. The Trade Unions did not ask for pay protection for the female workers. However, the Trade Unions did wish to have pay protection for the former bonus earners.

364. Mr. Berry gave evidence, which we accept, that to extend pay protection to the claimant groups as well as the comparator groups would be prohibitively expensive. There was no actual costing but we can readily accept that extending pay protection in this way would be unaffordable.

365. We therefore conclude that the respondent has failed to establish a genuine material factor defence in relation to the bonus issue but has succeeded in establishing a genuine material factor defence in relation to the pay protection issue."

1. Many of the points made in that passage are directed at a question which was not really in issue, namely the justifiability of the payments made to the comparators. The points relied on by the Tribunal which go to the issue of why the same payments were not made to the Claimants appear to be threefold:

(a) that, while it was recognised that the non-payment of bonus to the Claimants was "potentially" discriminatory (see para. 358), it was far from certain that that was the case - this is what is being referred to at para. 362 by the phrase "bang to rights";

(b) that even if it should have been assumed, in some general sense, that all the Claimants were entitled to be paid bonus, it would have been impossible to calculate the amounts payable, and thus the amount of any pay protection, because all or most Claimants were making multiple comparisons which would have produced very different amounts by way of bonus - see para. 361;

(c) that it was clear, even in the absence (no doubt for the reason summarised at (b)) of any precise calculation, that to extend pay protection to the Claimants would be unaffordable: see para. 364.

1. The arguments advanced by both groups of Claimants in their Notices of Appeal were to some extent refined and developed in the skeleton arguments and in oral submissions. In their final form they can be addressed as follows.

1. Both Mr. Napier and Mr. Allen took the point that it was wrong of the Tribunal to refer to the various factors going to the justification for affording pay protection to the comparators – e.g. need to maintain morale, union support etc. – when that was not the real issue. We agree that it would have been better if the Tribunal had distinguished more clearly between such factors and those which went to the withholding of pay protection from the Claimants, as identified at para. 63 above; but we do not believe that this lack of focus in the drafting can amount to an error of law. The factors in question were not positively irrelevant to the question of substance: they were merely of lesser significance to the argument because they were not in issue.

1. At para. 353 of its Reasons the Tribunal quotes a passage from the judgment of Elias P in Surtees (i.e. the Middlesbrough case) to the effect that (a) it was legitimate for the council "to take the view that it should limit the benefit to those actually in that group and to exclude all others even if some of them ought to have been in the group"; and (b) "any assessment of future costing [sc. of the inclusion of female claimants] would inevitably be highly speculative and would undermine the ability to obtain agreement for the scheme". *That passage was the subject of criticism by the Court of Appeal in Bainbridge*. At para. 173 (p. 187B) it said:

"We would accept that, if the appeal tribunal had been the tribunal of first instance, it would have been entitled to decide the issue of justification in that way. However, it was not and that passage must be treated with reserve. We note that the reasoning in that passage would have been equally applicable to the Redcar case and it would be unsatisfactory if the appeal tribunal were to promulgate two quite different decisions in two cases which were factually very similar without providing a rational basis for distinguishing them."

Mr. Napier and Mr. Allen submitted that in the light of that comment it was a misdirection for the Tribunal to take the passage from Surtees into account in its reasoning, as it must be taken to have done.

1. One answer to that submission would be that the Court of Appeal did not say that Elias P's observations in Surtees were positively wrong, merely that it was unsatisfactory that he should have said something so apparently inconsistent with his earlier judgment in Redcar: indeed it said in terms that "if the appeal tribunal had been the tribunal of first instance, it would have been entitled to decide the issue of justification in that way". But that may be too glib. It seems to us that there is at the very least an inconsistency of emphasis between what we have referred to as "point (a)" in the Surtees passage and the Court of Appeal's observations at paras. 173-7 in Bainbridge (see para. 56 (5) above); and if the Tribunal was going to refer to the passage at all it would have been better if it had referred also to the Court of Appeal's expressed reservations about it and explained precisely what proposition it nevertheless drew from it.

1. A better answer is that there is no sign that the passage in question actually affected the Tribunal's reasoning. It is quoted in what appears to be an introductory section, before the Tribunal gets to its actual reasons on the objective justification issue, which start at para. 358. None of the three points identified at para. 63 above depends on anything said in Surtees.

1. It is apparent from Bainbridge that an important factor in the proportionality exercise will always be the extent to which it must, or should, have been clear to the employer that its previous non-payment of bonus to the claimants was unlawfully discriminatory: see para. 56 (2) above. Mr. Allen drew our attention to a passage in Mr. Napier's cross-examination of Mr. Berry in which he accepted that in late 2006 and early January 2007, when the decision as to the scope of pay protection was being made, he anticipated – on the basis of what was happening in other authorities - that employees in predominantly-female White Book jobs would in due course be bringing equal pay claims against the Council and – on the same basis - that such claims would be likely to succeed. Mr. Allen submitted that that was important evidence on the issue of Bury's knowledge that its arrangements were discriminatory, and that it was not good enough for the Tribunal to deal with it in a single sentence agreeing with Mr. Cavanagh that this was not a case where the Council was "bang to rights".

1. We accept that it would have been better if the Tribunal had addressed this point explicitly, since it is clear that the Claimants placed some weight on it. But we are not prepared to say that its failure to do so was a breach of its obligation to give adequate reasons. Precisely what Mr. Berry knew – or, more accurately, believed - about the merits of the Claimant's claims was not on the facts of the present case a matter of central importance. The question of "knowledge" was important in Redcar because at the time that the pay protection arrangements were implemented the council had already received multiple claims and had made a concession that the non-payment of bonus to at least some claimants was unlawfully discriminatory. In the present case there had at the material time been no claims (or indeed premonitory grievances) and no requirement for Mr. Berry or the Council to form any definitive view as to whether if such claims were brought they were well-founded. On his own account Mr. Berry's views were personal, provisional and formed without the benefit of legal advice: when, later, claims began to come in and the Council's case began to be prepared he became less pessimistic. We do not read the Court of Appeal in Bainbridge as requiring tribunals to attach weight to what may be arbitrary and evanescent changes in the thinking of the decision-takers (who may not indeed all have a common view) about the merits of potential claims; and the necessary evidence would be both expensive and delicate to investigate. What the Court was concerned with was whether, taking a broad view, the employer knew or should have known that his previous pay arrangements were discriminatory. That is a question on which a tribunal which has heard the primary claim will be well-placed to form a view, as this Tribunal did; and the reasons for its answer will not normally require much elaboration.

1. There was no direct challenge to Mr. Berry's evidence that it was not possible for the Council, in advance of the determination of the Claimants' primary claims, to calculate what sums should be paid to them by way of pay protection; and that evidence seems to us plainly correct. It is not simply that it was not known whether the claims would succeed at all (and indeed they may yet not do so in respect of the employees identified at para. 33 above). There was the further point that even if they did succeed to some extent there were differences in the level of bonus paid to different comparators – 33?%, 40% and 50% - and unless and until it were known in respect of which comparators the claim had succeeded it could not be known on what figures to perform the pay protection calculations. The Claimants' cross-examination of Mr. Berry on this aspect was directed simply to establishing that "ball-park" assessments or calculations could be made. But although that might be relevant to the question of affordability it does not address the issues of what sums would have to be paid, pay-day by pay-day, if pay protection for the Claimants were to be implemented from 1 April 2007: for that purpose approximations are no use.

1. It follows that if the practicability of payment, in that sense, were a material consideration, in our judgment the Council would unquestionably have shown an objective justification for the non-payment of pay protection to the Claimants; and "point (b)" would be unassailable. However in our judgment practicability is not a material consideration. The effect of the equality clause provided for by section 1 of the 1970 Act is, if the claim to pay protection is good, that each Claimant's contract will include a term entitling her to a payment calculated on the same basis as her comparator's. No doubt the content of that obligation must be conceptually knowable as at the moment that it arises, but there is no legal necessity that it be practically ascertainable. There is nothing unheard of in a payment obligation arising whose precise content cannot in practice be ascertained until a court has ruled or some other information necessary for its quantification has been obtained. Indeed even "ordinary" equal pay cases – i.e. ones not concerned with pay protection – may attract problems of this kind: what of a case where it is apparent that some differential between the woman's pay and the man's is justified but the differential is said to be excessive and an apportionment between the legitimate and the discriminatory excess is required ?20 No doubt it is not ideal for an employer to be faced with a unquantifiable contingent liability, but that too is a far from uncommon situation; in the real world it is dealt with by making prudent provisions and if appropriate by setting money aside.

1. We would add that if Mr. Cavanagh's point were good it would apply in pretty well every case of this kind and would make it very difficult for an equal pay claim in relation to pay protection to succeed. That was plainly not the understanding of the Court of Appeal in Bainbridge.

1. It follows that in our view the Tribunal erred in law by relying on what we have referred to as "point (b)", and the Claimants' cross-appeal must be allowed.

1. The starting-point must be what the Court of Appeal said at para. 175 of its judgment in Bainbridge, namely:

"We accept that a large public employer might be able to demonstrate that the constraints on its finances were so pressing that it could not do other than it did and that it was justified in putting the need to cushion the men's pay reduction ahead of the need to bring the women up to parity with the men. But we do not accept that that result should be a foregone conclusion. The employer must be put to proof that what he had done was objectively justified in the individual case."

1. Mr. Napier and Mr. Allen submitted that the Council had put no evidence before the Tribunal to support a case of that kind. As we have already noted, no evidence was put before the Tribunal seeking to quantify the cost to the Council of extending pay protection to the Claimants or, equally importantly, explaining why such a cost was unaffordable. The evidence went no further than Mr. Berry's statement in his witness statement that the cost would be something which the Council could "ill afford". In his oral evidence Mr. Berry accepted that he had done no detailed costings: it was of course his case that such costings were impracticable. Mr. Napier put it to him that in those circumstances he was saying no more that "it would have cost a lot of money". His initial response was "It would have cost more money than the Council had the resources to pay". Mr. Napier pointed out that without costings he could not be sure of that, to which he responded "not precisely, no". The Claimants submitted that mere generalities of that kind were incapable of "demonstrating … that the constraints on [the Council's] finances were so pressing that it could not do other than it did". We were referred to the observation of this Tribunal in Pulham, at para. 46 (p. 356 E-F), that a tribunal faced with such an argument needs to be given "sufficient information … about the alleged costs and the financial background against which the affordability of those costs falls to be judged".

1. Mr. Cavanagh's first point in response was that precise calculation of the costs was impossible, for the reasons canvassed in relation to the practicability argument. We do not believe that this is an answer. It would clearly have been possible, as indeed Mr. Berry accepted in cross-examination, to produce illustrative calculations, on different bases, which would have enabled the Tribunal to assess the scale of the costs in the result.

1. Mr. Cavanagh's other answer was that it was too obvious to require particularised proof that the cost of extending pay protection to the Claimants would be unaffordable, as in effect the Tribunal held at para. 364 of the Reasons. We do not agree. In our judgment, both as a matter of principle and in accordance with the trend of the observations in Bainbridge, a local authority cannot prove unaffordability by mere assertion. A case of justification on this basis can only be proved by adducing sufficiently detailed evidence, both of the costs themselves and of the financial context, to enable the tribunal to reach an informed view. As we observed in Pulham (loc. cit.), that need not involve an exhaustive review of the council's finances, but the tribunal must be put in a position where it can assess the broad picture. That was not done here. The position is indeed hard to distinguish from that in Redcar, where the decisive consideration for the tribunal was that it had been given no evidence on cost (see para. 55 above).21

1. It follows that we accept that there was no evidence before the Tribunal which entitled it to find that the Council had established that extending pay protection to the Claimants was unaffordable.

1. The Claimants' cross-appeal succeeds. It follows that they are in principle entitled to payments by way of pay protection calculated by reference to the bonus payments which they should have received prior to 1 April 2007.

1 There are in both cases a few male claimants, claiming on the basis recently clarified in the judgment of this Tribunal in Hartlepool Borough Council v Llewellyn [2009] ICR 1426; but nothing turns on this for present purposes.

2 In Bury the first-named lead Claimant was a Mrs. Bridges, but she has now settled her claim, and the appeal proceeds under the name of a different lead Claimant, Mrs. Hamilton.

3 The approach in Nagarajan has now been authoritatively re-affirmed by the Supreme Court in R (E) v Governing Body of JFS [2010] IRLR 136: see Martin v Devonshires (UKEAT/0086/10).

4 We note Professor Fredman's reference in a recent article to the Act's "myopic focus on the individual claim": see ILJ vol. 37, at p. 208.

5 This passage is reproduced, with only minor changes, from our judgment in Armstrong 2 (para. 20, at pp. 682F-683C).

6 It is right to record that he scrupulously attributed the credit for this neat expression to Mr. Christopher Jeans QC.

7 For this reason, although the phrase "disparate adverse impact" is sometimes used to describe the element of gender disproportion in an Enderby context, "impact" is not quite the right word because in such a case nothing producing an impact has been shown to have been done. (We made this point in Armstrong 2 (see para. 69), but it was garbled in the ICR report by the word "disparate" rather than the word "impact" being italicised – see p. 704 E-F.)

8 It is right to say that in Bainbridge the approval of Armstrong 1 was somewhat tentative; but in Gibson Smith and Maurice Kay LJJ (loc. cit.) expressed themselves much more firmly (though Pill LJ took the opposite view (see at paras. 49-50 (pp. 724-5)). Permission to appeal in Gibson has been given by the Supreme Court: the appeal is listed for June.

9 The same might be said of the express requirement that the factor be "material". It is an oddity of section 1 (3) that although so few words are used two at least of them are doing very little work.

10 That is a simple matter of English, but if confirmation were needed it can be found in the well-known speech of Diplock LJ in Snook v London & West Riding Investment Ltd [1967] 2 QB 786 (at p. 802). We were also referred to the recent decisions of the Court of Appeal in Consistent Group Ltd v Kalwak [2008] IRLR 505, Protectacoat v Szilyagi [2009] ICR 835 and Autoclenz Ltd v Belcher [2010] IRLR 70, but counsel agreed that they did not in fact advance the argument in the present case.

11 We were tempted to use the phrase "true explanation", but "true" arguably contains the same latent ambiguity as "genuine".

12 "GMF" stands of course for "genuine material factor". Though this piece of jargon is probably now ineradicable, we prefer the description "section 1 (3) defence", since for reasons already suggested we do not think that the phrase "genuine material factor" (which does not precisely track the statutory language) is a very illuminating guide to the exercise required.

13 That concession was qualified in one respect, but the qualification is immaterial for present purposes.

14 NB that in Mr. Cavanagh's usage "bonusable" merely means that a job carries with it, as a matter of fact, a productivity bonus scheme, rather than that its character makes it apt for the application of such a scheme. Although it is part of the Council's case, so far as necessary, that the Claimants' jobs did not so lend themselves, that is not what he means by saying that they were "non-bonusable". We mention this point only because the term appears to have been used somewhat differently in Sunderland.

15 See paras. 13 (9) and 79-83 of the Council's Closing Submissions. We are not entirely clear how these groups fit into the broader groups of Claimants identified at para. 6 above; but nothing turns on this for present purposes.

16 As noted above, this is a different use of the term "bonusable" from that in Bury.

17 In fact we were told that in Bury single status was not implemented, or in any event not fully, until a later date, but that 1st April remained the effective date for the removal of bonus. But nothing turns on this point for present purposes.

18 In this (and the preceding) paragraph we partly reproduce the analysis of Bainbridge at paras. 22-25 of the judgment of this Tribunal in Pulham v London Borough of Barking and Dagenham [2010] ICR 333 (see pp. 346-9). But Mr. Cavanagh pointed out, and we accept, that in one respect – namely whether "recognised" indirect discrimination can be justified - that summary may not have been entirely accurate, and the issues of interest in the present case are not identical to those in Pulham. We have accordingly made some changes.

19 We note, however, that in Nicholls the employment tribunal had held that the withholding of pay protection from the claimants was justified. This Tribunal (Elias P presiding) remitted the issue for rehearing on account of a misdirection. The tribunal reached the same conclusion at the remitted hearing: a further appeal to this Tribunal is pending.

20 The need for an apportionment exercise of this kind in appropriate cases has been recognised at least since the decision of the ECJ in Enderby.

21 We note that such evidence was adduced at the remittal hearing in Nicholls (see n. 19 above) and indeed that it persuaded the tribunal. Without prejudice to any question of the correctness of the decision itself, this illustrates that placing the burden on employers to prove a particularised case is not unrealistic.