Heskett v Secretary of State for Justice (Rev 1)

This ground is founded on the proposition that, as Mr Menzies put it in his skeleton argument, “cost alone cannot justify indirectly discriminatory treatment” and that the financial burden of acting in a manner which does not discriminate can only be relied on as an element in a justification which involves other factors. This is referred to in employment lawyers’ jargon, not in my view very aptly, as the “cost plus” principle. The Tribunal, as we have seen, recognised that there was such a principle but held – Mr Menzies says wrongly – that it did not apply in the present case. By his Respondent’s Notice the Secretary of State puts the existence of the principle in issue. His “additional ground 2” reads:

“The saving or avoidance of costs can, without more, amount to the achieving of a legitimate aim under EU and/or UK law. There is no ‘bright line’ preventing reliance upon cost-based justification in the absence of a ‘plus factor’.”

It is accordingly necessary that I consider whether there is indeed a “cost plus” principle and if so what it means. This will, I fear, mean a rather lengthy excursion through the authorities.

The early cases are all decisions of the ECJ or the CJEU. Most of them are concerned with the disadvantageous treatment of part-time workers. I should therefore note by way of preliminary that it has long been established in EU law that since part-time workers are disproportionately female such treatment will be regarded as indirect sex discrimination and will be unlawful unless objectively justified.

The relevant case-law starts with De Weerd v Bestuur van de Bedrifsvereniging voor de Gezondheid (case no C-343/92, [1994] ECR 1571. (The claimant’s unmarried name was Roks, and the case is sometimes referred to by that name.) This was not a decision in the employment field but it is referred to in the authorities which follow. It concerned a Dutch social security measure which had a disproportionate impact on women. One of the questions referred to the ECJ was whether the measure in question could be justified “on budgetary grounds”. As to that the Court said:

“35.

[A]lthough budgetary considerations may influence a Member State’s choice of social policy and affect the nature or scope of the social protection measures it wishes to adopt, they cannot themselves constitute the aim pursued by that policy and cannot, therefore, justify discrimination against one of the sexes.

36.

Moreover, to concede that budgetary considerations may justify a difference in treatment as between men and women which would otherwise constitute indirect discrimination on grounds of sex, which is prohibited by Article 4(1) of Directive 79/7, would be to accept that the application and scope of as fundamental a rule of Community law as that of equal treatment between men and women might vary in time and place according to the state of the public finances of the Member States.

37.

Finally, as the Court has stated in connection with the second question, Community law does not prevent Member States from taking budgetary constraints into account when making the continuance of entitlement to a social security benefit dependent on conditions the effect of which is to withdraw the benefit thereof from certain categories of persons, provided that when they do so they do not infringe the rule of equal treatment as between men and women laid down in Article 4(1) of Directive 79/7.”

The first case which addresses the issue of the admissibility of “cost” as a justification in an employment context is Hill and Stapleton v Revenue Commissioners and Department of Finance (Case no C-243/95) [1999] ICR 48. The claimants were Irish female civil servants who worked for the Revenue Commissioners. Pay for civil servants was determined partly by reference to length of service. Both claimants had been working on a job-sharing basis, under which they worked one week on and one week off; but later moved to full-time employment. For the purpose of computing their full-time pay they were treated as only having had half their actual service in respect of the period for which they were job-sharing. They complained that this constituted indirect sex discrimination: almost all job-sharers were women. One aspect of the justification advanced by the employer was that the practice of pro-rating years of service worked “ensures that the incremental cost of job-sharing staff is the same as that of full-time staff” (see n. 20 in the Opinion of the Advocate General). As to that the Court said, at para. 40 of its judgment:

“So far as the justification based on economic grounds is concerned, it should be noted that an employer cannot justify discrimination arising from a job-sharing scheme solely on the ground that avoidance of such discrimination would involve increased costs.”

That statement in Hill and Stapleton is the first appearance of the phrase “solely on the ground that avoidance of such discrimination would involve increased costs” – for short, “solely to avoid increased costs”. The Court refers generally to whether the employer can “justify” discrimination: it does not employ the dichotomy of “aim” and “means”. As noted at para. 17 above, the distinction between the two is not always clear-cut. I think, however, that the Court must be understood to have been referring to the legitimacy of the employer’s aim rather than to the proportionality of its means; and that is how Rimer LJ understood it in his judgment in Woodcock to which I refer below.

In Jørgensen v Foreiningen af Speciallaeger and Sygesikringens Forhandlingsudvalg (Case no C-226/98) [2000] IRLR 726 the issue was (in very summary form) whether the Danish health service could pay lower fees to a female doctor on the basis that she had previously worked part-time. The referring court asked whether considerations relating to budgetary stringency, savings or medical practice planning might be regarded as objective considerations justifying a measure which adversely affected a larger number of women than men. In answering the question, the Court repeated paras.

In Kutz-Bauer v Freie und Hansestadt Hamburg (Case no C-187/00) [2003] IRLR 368, the claimant was a female employee of the City of Hamburg. A scheme was provided for under German legislation by which employees approaching pensionable age were encouraged to take part-time employment at an enhanced rate of remuneration, with the difference being reimbursed to the employer by the German government; but the benefit ceased to be funded if they worked beyond pensionable age. Since pensionable age was 60 for women and 65 for men the scheme had an indirectly discriminatory effect. The city ceased to pay the claimant at the enhanced rate after she reached 60. She brought proceedings against both the government and the city, no doubt on the basis that both were in different ways responsible for the detriment of which she complained. One of the justifications advanced by the government was that it would be expensive to fund the benefit for women aged over 60. As to that the Court said, at paras. 59-61 of its judgment:

“59.

As regards the German Government's argument concerning the additional burden associated with allowing female workers to take advantage of the scheme at issue in the main proceedings even where they have acquired entitlement to a retirement pension at the full rate, the Court observes that although budgetary considerations may underlie a Member State's choice of social policy and influence the nature or scope of the social protection measures which it wishes to adopt, they do not in themselves constitute an aim pursued by that policy and cannot therefore justify discrimination against one of the sexes (Case C343/92 De Weerd and Others [1994] ECR I-571, paragraph 35).

60.

Moreover, to concede that budgetary considerations may justify a difference in treatment between men and women which would otherwise constitute indirect discrimination on grounds of sex would mean that the application and scope of a rule of Community law as fundamental as that of equal treatment between men and women might vary in time and place according to the state of the public finances of Member States (De Weerd and Others, cited above, paragraph 36, and Jørgensen, cited above, paragraph 39).

61.

Nor can the City of Hamburg, whether as a public authority or as an employer, justify discrimination arising from a scheme of part-time work for older employees solely because avoidance of such discrimination would involve increased costs (see, to that effect, Hill and Stapleton, paragraph 40).”

It will be noted that in those paragraphs the Court referred separately to the authorities relating to the entitlement of the government to rely on “budgetary considerations”, i.e. De Weerd and Jørgensen (paras. 59-60), and to Hill and Stapleton as regards the entitlement of an employer to rely on the “increased costs” of avoiding discrimination (para. 61). The distinction is evidently deliberate and reflects the fact that both the government and the city were responsible for the discrimination.

In Steinicke v Bundesanstalt für Arbeit (Case no C-77/02) [2003] IRLR 892 German legislation imposed pre-conditions on the rights of civil servants to continue to work part-time after age 55 which discriminated against employees who had previously worked part-time. The claimant, who worked for the Federal Labour Office, complained of its application of those conditions in her case. The Federal Labour Office sought to justify them by reference, among other things, to “considerations of cost-neutrality”. At paras. 66-68 of its judgment the Court repeated in (mutatis mutandis) identical terms paras. 59-61 of its judgment in Kutz-Bauer.

In Schönheit v Stadt Frankfurt-am-Main (Case no C-4/02) and Becker v Land Hessen (Case no C-5/02) [2004] IRLR 983 the claimants were female civil servants employed by, respectively, the city of Frankfurt and the state government of Hesse. The terms of German legislation relating to civil service pensions were disadvantageous to part-time employees. The German government sought to justify the terms in question on the basis of seeking to limit public expenditure. The Court said:

“84.

It must be observed at the outset that the aim of restricting public expenditure, which, according to the national court, was invoked by the State when the pension abatement first became part of national law, cannot be relied upon for the purpose of justifying a difference in treatment on grounds of sex.

85.

The Court has already held that budgetary considerations cannot justify discrimination against one of the sexes. To concede that such considerations may justify a difference in treatment between men and women which would otherwise constitute indirect direct discrimination on grounds of sex would mean that the application and scope of a rule of Community law as fundamental as that of equal treatment between men and women might vary in time and place according to the state of the public finances of Member States (Roks, paragraphs 35 and 36 … and Kutz-Bauer paragraphs 59 and 60).”

It will be noted that the Court does not refer to para. 61 of Kutz-Bauer or, therefore, to Hill and Stapleton. That presumably reflects the fact that the discrimination in question derived wholly from a government measure.

Ms Darwin also referred to two decisions of the CJEU about the compatibility of legislation imposing mandatory retirement ages with Directive 2000/78 – R (Age Concern England) v Secretary of State for Business, Enterprise and Regulatory Reform (Case no C-388/07) [2009] ICR 1080, and Fuchs v Land Hessen (Case no C-159/10) [2012] ICR 93 – but the issues were not the same as those before us, and I do not think that considering them here adds anything useful.

The CJEU case-law was first considered by a domestic tribunal in Cross v British Airways plc [2005] UKEAT 0572/04, [2005] IRLR 423. In that case members of BA’s female cabin crew were complaining of the effect on their pension rights of the differential retirement age which applied as between male and female employees. The ET found that the indirectly discriminatory effect of that differential was justifiable. So far as relevant, it directed itself that “economic (which includes cost) grounds can properly be a factor justifying indirect discrimination, if combined with other reasons”; and it held that in the particular circumstances of the case it should take into account, albeit not as the only factor, the cost to the employer of altering its retirement age, which would have been very substantial.

The EAT upheld that decision. The judgment of Burton P is very fully reasoned, but the essential points can be summarised as follow.

First, at para. 62 he referred to a number of decisions in which “economic considerations, either the same as or analogous to costs, have been permitted or envisaged as justification”, and in particular to four cases – Jenkins v Kingsgate (Clothing Productions) Ltd [1981] ICR 592; Bilka-Kaufhaus v Weber von Hartz [1987] ICR 110; Rainey v Greater Glasgow Health Board [1987] AC 224; and Allonby v Accrington & Rossendale College [2001] EWCA Civ 529, [2001] ICR 1189. He continued, at para. 63:

“It seems to us, as a matter of obvious common sense (and in accordance with the principle of the concept of proportionality), … that, albeit that, in the weighing exercise, costs justifications may often be valued less, particularly if the discrimination is substantial, obvious and even deliberate, economic justification such as the saving, or the nonexpenditure, of costs (which must, for example, include the avoidance of loss) must be considered. It would, in our judgment, need clear reasoning and binding authority to prevent that occurring.”

He then conducted a thorough analysis of the CJEU authorities, noting among other things the distinction to which I have referred at para. 53 above. He concluded, at para. 72:

“We conclude that the European Court has laid down a perfectly comprehensible structure. A national state cannot rely on budgetary considerations to justify a discriminatory social policy. An employer seeking to justify a discriminatory PCP cannot rely solely on considerations of cost. He can however put cost into the balance, together with other justifications if there are any.”

The decision in Cross – and specifically the last two sentences of para. 72 – is the origin of the “cost plus” approach in the domestic case-law, although, to be fair to him, Burton P did not use that phrase.

In Redcar & Cleveland Borough Council v Bainbridge [2006] UKEAT 0135/06, [2008] ICR 249, the council had terminated a pay structure under which its employees in predominantly male jobs received bonus payments which were not available to employees in predominantly female jobs of equal value, and a new structure common to both kinds of job was substituted under which no such bonuses were available. In order to cushion the blow to employees in the predominantly male jobs, the council for a limited period made “pay protection” payments. The female employees in the latter group contended that it was discriminatory not to make equivalent payments to them. The ET held that the difference in treatment was justified. One of the matters on which it relied was that the employer was “impecunious”.

The EAT upheld the finding of justification on other grounds, but it did address the “impecuniosity” aspect on an obiter basis. Elias P at para. 90 of his judgment characterised the argument of counsel for the employees as being that “the only proper construction of [the] decisions of the European Court is that costs can never be taken into consideration when considering the question of objective justification”. He pointed out that that argument had been rejected in Cross and that it would be wrong to depart from that decision unless it was plainly wrong. He continued:

“… We do not think that such a broad statement is consistent with the case law. We accept that the cases show that it is not legitimate to discriminate where the aim or purpose is to save costs; see De Weerd, nee Roks v Bestuur van de Bedrifsvereniging voor de Gezondheid (‘Roks’) [1994] ECR 1571 where the ECJ held that a state cannot rely upon budgetary considerations to justify a discriminatory social policy. But we do not think that the case law supports the conclusion that the question of cost should always be irrelevant.”

He continued:

“91.

We would add this. It is not in our view helpful simply to talk about costs in an abstract way. Almost every decision taken by an employer is going to have regard to costs. Given an unlimited purse there need be no losers at all. We wholly accept that where a benefit is introduced and where costs determine the scope and size of that benefit, as they inevitably will, then it would be unlawful to allocate the benefit on a discriminatory basis. In that sense it would not be open for an employer to say that the restriction on cost prevented him from conferring the benefit on the disadvantaged group.

92.

If there are cost constraints, they must be allocated in a way which limits any discriminatory impact as much as possible: see for a recent example [Secretary of State for Defence v Elias [2006] EWCA Civ 1293, [2006] 1 WLR 3213]. This in our view is the explanation of the Schönheit case. Usually, however, the issue of costs may become material when an employer is being asked to put right some alleged continuing discrimination. Cross suggests that an employer cannot defeat the right to equality by pointing to financial burdens alone, but he can pray the financial burdens in aid as some support for a decision which is objectively justified on other grounds. Pay protection arrangements provide a good example. Transitional arrangements of such a kind will sometimes be appropriate (and often unavoidable in practice) to cushion the pay of those moving to lower pay. It would theoretically be possible to confer the benefit of the higher pay on everyone, but the cost may reinforce the justification limiting the benefit.

93.

In our view, that is the position here. The council has identified a significant material factor defence which explains the difference in pay. The cost of bringing about equality is in that context merely a supportive reason, but it was never relied upon as the principal basis for the objective justification.”

In Woodcock, to which I have already referred, the claimant was a senior employee in an NHS Trust. He was liable to be dismissed for redundancy but there was a long delay, essentially for his benefit, in giving him the twelve months’ notice of dismissal that was required: the latter part of the delay was as a result of a proposed consultation meeting having to be postponed. Notice was finally given, at a time when formal consultation procedures had not been completed, in order that the notice should expire before his fiftieth birthday, at which point he would have become entitled to early retirement and enhanced pension rights, at very considerable cost to the employer. He claimed (direct) age discrimination. It was common ground that the decision to make the claimant redundant had nothing to do with his age: the challenge was to the timing of its implementation. As to that, the ET found that in the particular circumstances of the case the enhanced benefits that would have accrued if the claimant had remained employed when he reached fifty would have constituted a windfall, in the sense that

The EAT – [2010] UKEAT 489/09, [2011] ICR 143 – upheld the ET’s decision on the basis of the “cost plus” approach. I need not summarise that aspect of the reasoning. However, I did at para. 32 of my judgment express our reservations about whether the distinction between “cost alone” and “cost plus” correctly reflected the applicable principles. I need not quote the whole passage, but the following passage may be material:

“The ‘cost plus’ approach propounded in Cross represents the current orthodoxy. … But [counsel for the employer] submitted, as one alternative basis of his case, that the cost plus approach was wrong, and we have to say that we do not find it convincing. For reasons which will appear, we need not reach a concluded view, but we will briefly indicate our thinking in case the matter falls for decision elsewhere. We respectfully agree with Burton P’s observation … that, as a matter both of principle and of common sense, considerations of cost must be admissible in considering whether a provision criterion or practice which has a discriminatory impact may nevertheless be justified ... . But we find it hard to see the principled basis for a rule that such considerations can never by themselves constitute sufficient justification or why they need the admixture of some other element in order to be legitimised. The adoption of such a rule, it seems to us, tends to involve parties and tribunals in artificial game-playing – ‘find the other factor’ – of a kind which is likely to produce arbitrary and complicated reasoning: deciding where ‘cost’ stops and other factors start is not straightforward (cf. the observations of Elias P. in Bainbridge, at para. 91 …) ...”

I went on to say that if the matter were free from authority it would be our view that an employer should be entitled to seek to justify a PCP producing a discriminatory impact on the basis that the cost of avoiding that impact, or rectifying it, would be disproportionately high.

This Court likewise upheld the decision of the ET. The only substantial judgment was given by Rimer LJ, with whom Arden LJ and Ryder J agreed. At paras. 55-63 of his judgment he carried out a full review of the authorities, mainly through the prism of Cross. At para. 65 he summarised the submission of counsel for the employee as being that:

“… the saving or avoidance of costs alone cannot be a legitimate aim. It can only be so if it is linked to a non-cost factor. Thus a consideration that, by itself, is inadmissible as justification becomes admissible if so linked; and, in such a case, it can play a part in the proportionality assessment.”

He continued:

“66.

There is, it seems to me, some degree of artificiality about such an approach to the question of justification. As Elias J observed in the Redcar & Cleveland case, [2007] IRLR 91, at paragraph 91, ‘Almost every decision taken by an employer is going to have regard to costs.’ [The relevant provision of the regulations then governing age discrimination], however, says nothing of the extent to which considerations of cost may feature in the justification exercise. It provides merely that what would otherwise be discriminatory treatment may be justified if it was ‘a proportionate means of achieving a legitimate aim’. The relevant question must therefore be whether the treatment complained of was such a means. Accepting, as I make clear I do, that the guidance of the Court of Justice is that an employer cannot justify discriminatory treatment ‘solely’ because the elimination of such treatment would involve increased costs, that guidance cannot mean more than that the saving or avoidance of costs will not, without more, amount to the achieving of a ‘legitimate aim’. That is entirely unsurprising. To adopt a simple example given by [counsel for the employer], it is hardly open to an employer to claim to be entitled to justify the discriminatory payment to A of less than B simply because it would cost more to pay A the same as B. Such treatment of A could not, without more, be a 'legitimate aim'.

If the Trust's treatment of Mr Woodcock is correctly characterised as no more than treatment aimed at saving or avoiding costs, I would accept that it was not a means of achieving a ‘legitimate aim’ and that it was therefore incapable of justification. It would fall foul of the limitations upon justification explained in cases such as Hill and Stapleton. On the unusual facts of this case, I would not, however, regard that as a correct characterisation … .”

That reasoning could be described as a version of “cost plus”, but I note that Rimer LJ does not use that phrase.

The most authoritative recent consideration of the position is in the judgment of the Supreme Court, given by Lord Hope and Lady Hale in O'Brien v Ministry of Justice [2013] UKSC 6, [2013] ICR 499. The claim arose out of the denial of pension rights to Recorders (and other part-time judges), which was said to be contrary to the PartTime Workers (Prevention of Less Favourable Treatment) Regulations 2000, which implements the EU Part-Time Workers Directive. The primary issue before the Court was whether Recorders were “workers” within the meaning of the Regulations (and the Directive). Following a reference to the CJEU, it concluded that they were. It followed that the denial of pension rights constituted direct discrimination. Under the Regulations direct discrimination may be justified. The issue accordingly arose of whether the issue of justification needed to be remitted. The Court held that it did not because no arguable basis of justification had been advanced.

One aspect of the justification advanced by the Minister was the high cost of providing pensions to part-time judges. The Court considered that issue at paras. 63-70 of its judgment, under the heading “cost”. It started by observing, at para. 63:

“The Ministry accept that cost alone cannot justify discriminating against part-time workers. But they argue that ‘cost plus’ other factors may do so. This is a subtle point which is not without difficulty.”

I would respectfully echo the last sentence.

The Court then proceeded at para. 64 to quote paras. 35 and 36 from the judgment in De Weerd. It glossed para. 35 as follows:

“In other words, richer states may have more generous benefits systems than do poorer states. Cost may inform how much the state will spend upon its benefits system, but the choices made within that system must pursue policy aims other than saving cost.”

As to para. 36, it said:

“It is one thing to set benefits at a particular level for budgetary reasons. It is another thing to pay women less than men because it is cheaper so to do. Sex discrimination is wrong whether the state (or the employer) is rich or poor.”

At paras. 65-68 the Court considered the effect of the CJEU’s observations in Jørgensen about the legitimacy of “reasons relating to the need to ensure sound management of public expenditure”, observing at para. 67 that:

“Sound management of the public finances may be a legitimate aim, but that is very different from deliberately discriminating against part-time workers in order to save money.”

It drew support from the decision of the Court in a non-employment context, in European Commission v The Netherlands C-542/09. It concluded, at para. 69:

“Hence the European cases clearly establish that a Member State may decide for itself how much it will spend upon its benefits system, or presumably upon its justice system, or indeed upon any other area of social policy. But within that system, the choices it makes must be consistent with the principles of equal treatment and non-

discrimination. A discriminatory rule or practice can only be justified by reference to a legitimate aim other than the simple saving of cost.”

Finally, at para. 70 the Court noted that its attention had been drawn to Woodcock, but it declined to express any view about whether it was rightly decided.

It will be noted that in those passages the Court treats the governing principle as being that stated in De Weerd and refers to decisions taken by “states” (e.g. “a Member State may decide for itself how much it will spend upon its … justice system”), subject to one parenthetical reference to employers (in para. 64). That is unexceptionable because the exclusion of part-time judges from the judicial pension scheme was enshrined in primary legislation (the Judicial Pensions and Retirement Act 1993). It does not in fact seem that the Court was referred to Hill and Stapleton or Kutz-Bauer – or to Cross, in which the EAT carefully noted the difference in the formulations applicable to governments and to employers.

When it came to its conclusion on the Ministry’s justification case as a whole the Court found, having dismissed the other factors relied on, that it came down to the proposition that “if recorders get a pension, then the pensions payable to circuit judges will have to be reduced”. It continued, at para. 74:

“That is a pure budgetary consideration. It depends upon the assumption that the present sums available for judicial pensions are fixed for all time. Of course there is not a bottomless fund of public money available. Of course we are currently living in very difficult times. But the fundamental principles of equal treatment cannot depend upon how much money happens to be available in the public coffers at any one particular time or upon how the State chooses to allocate the funds available between the various responsibilities it undertakes. That argument would not avail a private employer and it should not avail the State in its capacity as an employer. Even supposing that direct sex discrimination were justifiable, it would not be legitimate to pay women judges less than men judges on the basis that this would cost less, that more money would then be available to attract the best male candidates, or even on the basis that most women need less than most men.”

Again, in that passage the Court addresses itself specifically to the position of states, although it does say in terms that the position would be the same for a private employer. Unison

In R (Unison) v Lord Chancellor one of the challenges advanced to the lawfulness of the Employment Tribunal and the Employment Appeal Tribunal Fees Order 2013 was that it was indirectly discriminatory against (among other groups) women, and that that discrimination could not be justified since the only aim was budgetary, or to reduce

“cost”. The Court of Appeal – [2015] EWCA Civ 935, [2016] ICR 1 – rejected that argument. At para. 91 of my judgment, with which Moore-Bick and Davis LJJ agreed, I identified the impugned provision – what in an employment context would be the PCP – as being the charging of a higher level of fees for what were perceived to be more complex cases. I continued:

“The underlying rationale is that there ought to be a relationship between the level of the fee and the degree of the demand on the Tribunals’ resources. I do not regard that as invoking ‘cost’ or ‘budgetary considerations’ in the same sense as the Ministry of Justice was said to be doing [sc. in O’Brien] when it argued that it was unduly expensive to accord pension rights to part-time judges. Given that it is legitimate to charge tribunal fees in the first place, a graduated level of fees which reflects the extent to which the resources of the tribunal are engaged is no more than an application of ordinary principles of economic efficiency, which has regularly been accepted as relevant to justification: it is not necessary to cite more than the celebrated decision

of the ECJ in Bilka-Kaufhaus GmbH v Weber von Hartz (C170/84), [1987] ICR 110 – see para. 36 (p. 126 E-G). To avoid any possible misunderstanding, I am not to be taken as saying that economic efficiency will always be a sufficient justification for a PCP having a discriminatory impact, but [the claimant’s] challenge was to the admissibility of this consideration at all.”

At para. 92 I cited, to the same effect, an observation of Elias LJ at para. 89 of his judgment in the Divisional Court, where he said

“… I would not in fact describe the first objective as costs saving. This is not a case of government refusing to correct discrimination because it would be too expensive. Rather it is more accurately characterised as requiring a contribution towards the cost of running the Tribunal Service, charging equal amounts from all who bring claims within class B.”

Although the case proceeded to the Supreme Court, the claim based on indirect discrimination was not pursued there.

I start with two preliminary points.

The first is the point which I have noted at para. 53 above, and which I have picked up in relation to O’Brien at para. 73. In expressing the applicable principle the CJEU authorities distinguish between cases where the discrimination is the result of a measure taken by central government, where they use the De Weerd “budgetary considerations” formulation, and cases where it results from the decision of the employer, where they use the Hill and Stapleton “solely [to avoid] increased costs” formulation. Although Burton P attached importance to the distinction in Cross, I have noted that the Supreme Court assumed that the principles applying in both kinds of case were the same. I do not in fact believe that the distinction is significant, at least in this appeal, but the CJEU’s use of different language was clearly deliberate, and in what follows I will use the language of Hill and Stapleton.

The second point is that in the CJEU cases the employer, where proceeded against, was in each case a public authority (though not always an emanation of central government).

I turn to the fundamental question, which is what is meant by the phrase “solely [to avoid] increased costs”, and more particularly what is the effect of the word “solely”. On this, it seems to me that we are bound by the guidance given by Rimer LJ at paras. 66-67 of his judgment in Woodcock; but even if we were not, I would respectfully agree with it. He says in para. 66 that the CJEU’s language “cannot mean more than that the

That might seem too trite to need saying – “unsurprising”, as Rimer LJ puts it – but it is not difficult to understand why the CJEU thought it important to spell it out. It is the same obvious but important point that the Supreme Court makes at several points in O’Brien: see para. 67 of its judgment (“very different from deliberately discriminating against part-time workers in order to save money”), para. 69 (“a legitimate aim other than the simple saving of cost”) and the example given at the end of para. 74 (“it would not be legitimate to pay women judges less than men judges on the basis that this would cost less”).

It follows that the essential question is whether the employer’s aim in acting in the way that gives rise to the discriminatory impact can fairly be described as no more than a wish to save costs. If so, the defence of justification cannot succeed. But, if not, it will be necessary to arrive at a fair characterisation of the employer’s aim taken as a whole and decide whether that aim is legitimate. The distinction involved may sometimes be subtle (to adopt the Supreme Court’s language in O’Brien) but it is real. The nature of the distinction can be illustrated by reference to some of the cases.

I start with Woodcock itself. In one sense the Trust’s aim could indeed be said to be to avoid the cost of having to pay the claimant the enhanced benefits to which he would become entitled if he were still employed when he reached fifty. But that was not the whole story because it omitted the fact that those were benefits which he had no legitimate entitlement to expect. It was for that reason that Rimer LJ held at para. 67 that it would not be correct to characterise the Trust’s aim as being “no more than treatment aimed at saving or avoiding costs”.

In Bainbridge likewise it might be said that the council’s aim in not making the same temporary pay protection payments to employees in the predominantly female jobs as were being made to employees in the predominantly male jobs was to save the cost of having to do so. But, again, that was not the whole story, because the reason why payments were being made to (for short) the men was to provide a temporary cushion against a loss of earnings which the women were not suffering. It is legitimate for a public authority (or indeed any employer) to seek not to make payments to employees to whom the rationale for those payments does not apply (and for whom it would indeed, in the language of Woodcock, be a windfall). (Footnote: ⁵)

Similarly in Unison both this Court and the Divisional Court did not accept that it was fair to characterise the Secretary of State’s aim as being “budgetary” or concerned solely with “cost” in the relevant sense. Rather, he had the plainly legitimate aim of

By contrast, in the CJEU cases the justifications advanced by the employer in the domestic proceedings appear to have amounted to no more than the fact that it would cost more to treat job-sharers/part-timers in the same way as full-time employees. (Footnote: ⁶) The same is true of O’Brien (ignoring for the present the subtlety that the Court was using the De Weerd formulation): the Secretary of State’s case was, as the Court analysed it, simply that it would be expensive to give pension rights to part-time judges.

The upshot of all this is that there is certainly an established principle that, to take Rimer LJ’s formulation in Woodcock, “the saving or avoidance of costs will not, without more, amount to the achieving of a legitimate aim” for the purpose of the defence of justification in a discrimination claim; but that that principle needs to be understood in the way that I have sought to explain it in the preceding paragraphs. It only bites where the aim is, as the CJEU put it in Hill and Stapleton, “solely” to avoid costs.

That being so, the “cost plus” label (and its cognates such as “cost alone” and “the plus factor”) cannot be said to be incorrect, and it is sometimes too convenient a shorthand to eschew. However, that language is not in fact used either by Burton P in Cross or by Rimer LJ in Woodcock, and I would prefer to avoid it so far as possible. In my experience it can lead parties, and sometimes tribunals, to adopt an inappropriately mechanistic approach (see my observations in Woodcock quoted above). It is better, in any case where the issue arises, to consider how the employer’s aim can most fairly be characterised, looking at the total picture. It is only if the fair characterisation is indeed that the aim was solely to avoid increased costs that it has to be treated as illegitimate.

Despite the length of that exposition, it is no more than the foundation for the particular issue raised by this appeal, which is (in short) whether the “plus factor” can consist of the fact that the employer is subject to financial constraints which oblige it to reduce its costs: can that be said to be that different from “cost alone”? I turn to that issue below.

Mr Menzies’ case as pleaded and as advanced in his skeleton argument can be summarised as follows. His submission to the ET had been that the Respondent’s justification amounted to no more than that it was too expensive to maintain the previous rate of pay progression and accordingly constituted reliance on “cost alone”. The Tribunal had not rejected that submission on the basis that this was a “cost plus” case. Instead it had drawn a distinction between cases where the aim was “cost-cutting” and cases where an “absence of means” forces the employer to do the act complained of: see para. 57 of the Reasons. That is a distinction without a difference. In either case the employer does the discriminatory act with the sole aim of reducing the financial burden on it: that is just what the authorities say should not be done. What the Tribunal

I should start by saying that I do not believe that that submission accurately reflects the Tribunal’s reasoning in para. 57. It was not evading the need to identify a “plus factor”. On the contrary it found it in the fact that NOMS’s budget for paying its employees had been frozen, so that it was required to reduce the rate of pay progression in order to “live within its means”; that was an important part of the picture and it held that it meant that NOMS’s aims “cannot simply be described as cost cutting”.

However, that only answers the form of Mr Menzies’ submission. It is not an answer to the underlying substantive point, which is that “absence of means” cannot be treated as a distinct factor from “cost”; and that is the real point on which the argument before us focused. To put it another way, the issue is whether a justification can fairly be characterised as “cost alone” where the employer argues not simply that it would be more expensive for it to avoid the discriminatory impact of the measure in question but that the cost of doing so is positively unaffordable.

As to that, Ms Darwin relied on three cases which she said established that it was legitimate for an employer to take steps to ensure that it “lived within its means”. I take them in turn.

The first was the decision of the EAT (myself presiding) in HM Land Registry v Benson, to which I have already referred. The Land Registry is a public body financed by the fees which it charges. It is required to break even year-on-year. At the time with which the case was concerned it had reserves but it could only draw on them with Treasury approval. A slump in its revenues meant that in order to break even it needed to reduce its workforce. It received Treasury approval to spend £50m from its reserves to finance voluntary redundancy or early retirement schemes, of which £12m was allocated to a particular “merging offices” scheme. The terms under which employees were selected for redundancy under that scheme had an age-discriminatory effect. There was some debate about how the relevant aim should be characterised for the purpose of the justification issue, given that it could be said to include several elements. At para. 34 of my judgment I said:

“The uncertainty about how to characterise [the relevant aim or aims] … does not, fortunately, matter since in our view all the various potential elements are plainly legitimate. It is (to put it no higher) legitimate for a body such as the Appellant, like any business, to seek to break even year-on-year … It is likewise legitimate to offer voluntary redundancy/early retirement schemes of the kind with which we are here concerned: the Tribunal found in terms that the Appellant had a

‘real need’ to implement the Merging Offices Scheme in 2008/9 …. And, most pertinently, it was in our view legitimate for the Appellant to impose a budget on the amount to be spent on such schemes in 2008/9, even if that might mean that selection had to be made between applicants. Like any business, it was entitled to make decisions about the allocation of its resources.”

It is fair to say that the employees did not in the EAT seek to argue that the employers’ justification constituted reliance on “cost alone” and was thus inadmissible: see para. 23 of the judgment (Footnote: ⁷).

The second case was the decision in Edie v HCL Insurance BPO Services Ltd [2015] UKEAT 0152/14, [2015] ICR 713. In that case the employer was faced with severe financial difficulties: the details are summarised at paras. 13-19 of the judgment of Lewis J in the EAT, and I need not summarise them here, save to note that its staff costs alone amounted to 115% of its revenue. In an attempt to reduce staff costs it introduced new terms and conditions which resulted in loss of some benefits which were particularly advantageous to a group of older employees. They brought age discrimination claims. At para. 57 of his judgment Lewis J quoted the ET’s characterisation of the employers’ aim as being “to reduce staff costs to ensure its future viability and to have in place market competitive, non-discriminatory terms and conditions”. He went on, at para. 58, to cite Benson as authority for the proposition that “[i]t is, as a minimum, a legitimate aim for a business to seek to break even year-onyear … [and] … to make decisions about the allocation of its resources”.

The decision of this court in Harrod, to which, again, I have already referred, arose out of a claim by police officers who had been compulsorily retired by the Chief Constables of their respective forces, using a power under regulation A19 of the Police Pension Regulations 1987 which had an age-discriminatory impact. The retirements were required in order to enable police forces to achieve a 20% reduction in their budgets as part of the austerity measures imposed by the coalition government in 2010. The issues in this Court focused more on the proportionality of those decisions than on the legitimacy of the underlying aim, though both Woodcock and O’Brien were referred to in the parties’ submissions. But in the course of his leading judgment Bean LJ, with whom Elias LJ agreed, quoted a long passage from the judgment in Benson which included para. 34, and he said at para. 27 that he agreed with its analysis. In my own concurring judgment I identified the use of regulation A19 as the relevant PCP. I continued, at para. 38:

“It is the choice of that method that has to be justified. The question is whether it was a proportionate means of achieving a legitimate aim. In my view the right way to characterise the forces’ aim is that they wished to achieve the maximum practicable reduction in the numbers of their officers. That is unquestionably a legitimate aim.”

I did not say, because it was not in issue, that the aim of reducing the number of officers was to achieve the required spending reductions; but that was the case.

Neither Benson nor Edie is binding on us, and in neither of them was it argued that the aim identified by the EAT as legitimate fell foul of the “cost alone” principle deriving from Hill and Stapleton. Likewise there was no such contention in Harrod, and I do not think that Bean LJ’s general endorsement of my judgment in Benson nor my own reasoning can be treated as directed specifically to that question. Nevertheless, they do afford some support to the proposition that an employer’s need to reduce its

More importantly, however, I think that that proposition is correct in principle. There is nothing to suggest that the CJEU in Hill and Stapleton, Kutz-Bauer or Steinicke had in mind a case of the present kind, where an employer is having to make choices about how best to allocate a limited budget: the justification advanced in those cases was purely that avoiding discrimination would cost more. I can see no principled basis for ignoring the constraints under which an employer is in fact having to operate. It is never a good thing when tribunals or courts are required to make judgements on an artificial basis. As Burton P in Cross, Elias J in Bainbridge, and myself and Rimer LJ in Woodcock have all observed, almost any decision taken by an employer will inevitably have regard to costs to a greater or lesser extent; and it is unreal to leave that factor out of account. That is particularly so where the action complained of is taken in response to real financial pressures, as was very clearly the case in all three of the authorities relied on by Ms Darwin and as is the case, on the Tribunal’s findings, in the present case. It is also necessary to bear in mind that because age, unlike other protected factors, is not binary it is difficult, to put it no higher, for an employer to make decisions affecting employees that will have a precisely equal impact on every age group, however defined. This makes it particularly important for them to be able to justify such disparate impacts as may occur by reference to the real world financial pressures which they face.

It is important to emphasise at this point that the issue with which we are concerned relates only to the first step in the consideration of the justification defence. If it is permissible for an employer to rely, as a legitimate aim, on a real need to reduce or constrain staffing costs, it still has to show that the measures complained of represent a proportionate means of achieving that aim, having regard to their disparate impact on the group in question. It is in my view entirely appropriate that a proportionality exercise of that kind should be the focus of the justification enquiry. Such an exercise will enable the tribunal to examine carefully the nature and extent of the financial pressures on which the employer relies as well as the possibility that they could have been addressed in a way which did not have the discriminatory effect complained of. That kind of exercise was carried out by the tribunals in Benson and in Edie – and indeed by the Tribunal in the present case.

I recognise that it may sometimes be difficult for a tribunal to draw the line between a case where an employer simply wishes to reduce costs and cases where it is, in effect, compelled to do so. But tribunals often have to make judgements of that kind and there is nothing uniquely difficult about this one. The judgement that would be required if Mr Menzies’ submission were correct – that is, to assess the justification on the basis that any consideration of the employer’s financial position should be excluded – would be at least equally difficult, indeed more so.

Mr Menzies’ principal response on this part of the argument was to rely on para. 74 of the judgment of the Supreme Court in O’Brien and specifically on the statement that “the fundamental principles of equal treatment cannot depend upon how much money happens to be available in the public coffers at any one particular time or upon how the State chooses to allocate the funds available between the various responsibilities it undertakes” and that that argument “would not avail a private employer and … should not avail the State in its capacity as an employer”.

I do not believe that that passage is applicable to the present case, or others of the kind which I have been considering in the previous paragraphs. The Court was concerned with a case of (overt and deliberate) direct discrimination against part-time workers. At the end of para. 74 it expressly equates it with a case of direct pay discrimination against women. As it says, such treatment contravened “the fundamental principles of equal treatment”. In that context it is hardly surprising that it should make it clear that neither a private employer nor the state in its capacity as employer could seek to justify such discrimination (where, untypically, justification is an available defence) on the basis that times were hard and it could not afford to treat part-timers, or women, equally. But the present case does not involve direct discrimination. We are concerned with indirect discrimination – more specifically with a situation where the employer has altered its pay arrangements in a way which has had a disparate impact on employees of different ages (as such changes are very liable to do). I see no sign that the Supreme Court had in mind a case of this kind. What “the fundamental principles of equal treatment” require in such a case is that the disparate impact of the measures in question should be justified. Mr Menzies pointed out that the test of justification was expressed in identical terms in the case of indirect discrimination under section 19 of the 2010 Act and in the case of direct age discrimination: see section 13 (2). That is no doubt correct, but it does not follow that the question of the legitimacy of a particular aim requires the same answer in both contexts.

In the course of Ms Darwin’s submissions there was some exploration, arising out of questions from the Court, of whether it was appropriate to take the constrained budget allocated to NOMS as a given or whether it was necessary to consider the Ministry’s, or the Treasury’s, justification for constraining the budget in the first place. The Court asked for a note from the Respondent explaining the relevant relationships more fully than appeared in the Tribunal’s Reasons. Ms Darwin and Mr Roberts submitted such a note following the hearing and Mr Menzies provided a response. In the event, however, I do not think that this is an area which can be explored in this appeal. No issue of this kind arose in the ET or the EAT or is raised in the grounds of appeal, and Mr Menzies did not in his skeleton argument or opening submissions argue that the ET should have looked beyond the justification advanced by NOMS. In those circumstances I do not think I should express any concluded view. However, leaving aside the kind of case which the Supreme Court was discussing at para. 74 of its judgment in O’Brien, I would take some convincing that it was illegitimate for a government department or agency to seek to keep its pay budget within the limits imposed by the Treasury or a parent department. That was certainly the basis on which Benson was decided.

For the reasons given I do not accept that the Tribunal’s self-direction at para. 57 of the Reasons was wrong in law. It was entitled to treat NOMS’s need to observe the constraints imposed by the pay freeze as a legitimate aim.

It should be appreciated that this ground involves no challenge to the Tribunal’s decision that the reduction in the rate of pay progression which NOMS implemented as a means of achieving that aim were proportionate. One element in its reasoning is the subject of ground 2.

The Tribunal made it clear that a central element in its finding that the reduction in the rate of pay progression was justified was the fact that it had been imposed as a temporary measure and that active consideration was being given to changing the system so as to reduce the age-discriminatory effect of the scheme as it stood. The point is primarily made at paras. 68-69 (see para. 36 above), but its importance is reinforced by what Judge Barklem calls the “shot across the bows” in para. 72.

Mr Menzies submitted that the fact that NOMS was intending to change the system was as a matter of principle irrelevant to the issue of whether it could be justified in the period to which the complaint related. As he put it in his skeleton argument, and reiterated in his oral submissions, “what might happen in the future was not relevant to whether the policy was justified at the present time”.

I do not accept that submission, at least in the context of the Tribunal’s reasoning in this case. An employer may sometimes feel obliged to take urgent measures which have an indirectly discriminatory effect on a group of its employees. Perhaps the disparate impact is something which is or should be appreciated from the start; perhaps it is something which only becomes apparent after a time. In either case I see no reason in principle why it should not be open to the employer to seek to justify those measures on the basis that they represented a proportionate short-term means of responding to the problem in question albeit that they could not be justified in the longer term. It is inherent in such a justification that the employer intends within a reasonable time to discontinue the measures in question, or modify them so that they no longer have the impact complained of; but the essence of the justification does not depend on that intention. Of course in other circumstances it is not open to an employer to defend a discriminatory state of affairs by saying “I know I am discriminating but I will do something about it soon”. But that is not the case which the Tribunal was accepting here. The essential premise was that the reduction in the rate of progression was justifiable on a temporary basis. I note also that an essential part of the justification for the prima facie discriminatory pay protection payments in Bainbridge was that they were being made only for a limited time.

That of course only addresses the question whether it was permissible in principle for the Tribunal to have regard to this factor. There is no challenge to the reasonableness of its conclusion if the temporary nature of the measures taken by NOMS was indeed relevant; and, that being so, I should say nothing about that question. But it is clear from para. 72 that the Tribunal thought that NOMS’s time for addressing the problem was coming very close to running out, and I am not surprised.

It will be seen that both the ET and the EAT referred to the decision of the Supreme Court in Naeem v Secretary of State for Justice [2017] UKSC 27, [2017] ICR 640. That case concerned what Lady Hale, with whom the other members of the Court agreed, called “a system in transition”. At para. 47 of her judgment she said:

“The question was not whether the original pay scheme could be justified but whether the steps being taken to move towards the new system were proportionate. Where part of the aim is to move towards a system which will reduce or even eliminate the disadvantage suffered by a group sharing a protected characteristic, it is necessary to consider whether there were other ways of proceeding which would eliminate or reduce the disadvantage more quickly.”

I am not myself sure that the circumstances in Naeem were truly analogous to those of the present case; but the passage in question does at least confirm that the fact that a system is in the course of change may represent part of the justification for the state of affairs at a particular time.

I would accordingly dismiss ground 2.

As appears from what I have already said, it was an essential element in the ET’s finding of justification that NOMS’s reduction in the rate of pay progression was intended as a short-term reaction to the pay freeze – or, as it put it at the end of para. 69 of the Reasons, that “it has never been viewed as anything other than a stop gap measure”.

Mr Menzies submits that that finding was no more than an inference on the Tribunal’s part, and indeed one which was contrary to what was pleaded in the Secretary of State’s ET3 where the 2011 policy was described as applying in that year “and going forwards”. The Tribunal had heard no evidence about what was in the mind of the decision-makers in NOMS in 2011. Mr Paskin, to whose evidence it attached such weight, had only been in post since 2015. On ordinary principles such an inference should not be drawn except on the basis of appropriate findings of primary fact which need to be properly spelt out: he relies on the well-known statements to that effect in the judgments of Balcombe LJ and Peter Gibson LJ in Chapman v Simon [1994] IRLR 124: see paras. 33 (3) and 43.

I do not see anything in this ground. In the first place, although the Tribunal used the word “infer” at the start of para. 69, I am not convinced that it was drawing an inference in the sense that this Court was concerned with in Chapman v Simon: which was a case where a tribunal had made a finding of unconscious racial motivation against a respondent. The findings in that paragraph are more in the nature of the Tribunal’s overall conclusions. But I accept that that may not go to the heart of the point because it would still be necessary that there be an evidential basis for those conclusions. In my view the short answer to the submission is that the Tribunal explains the basis for its conclusion in para. 69 itself, namely that it was in the nature of a pay freeze of the kind imposed in 2010 that it was likely to be a temporary measure because, as it says, it would reasonably be assumed that “years of below inflation pay settlements are politically unsustainable”. In my view that is a sufficient basis for the Tribunal’s conclusion that NOMS regarded the reduction in the rate of progression as a temporary, or stopgap, measure. I see nothing in its conclusion inconsistent with the use of the phrase “going forwards” in the ET3.

I would dismiss the appeal.

I agree.

I also agree.