The Lord Chancellor & Anor v McCloud & Ors [2018] EWCA Civ 2844

[2018] EWCA Civ 2844Case Nos: A2/2018/0635, A2/2018/0636, A2/2018/0505 & A2/2018/0647IN THE COURT OF APPEAL (CIVIL DIVISION) ON APPEAL FROM THE EMPLOYMENT APPEAL TRIBUNAL THE HONOURABLE SIR ALAN WILKIEVenue Royal Courts of Justice Strand, London, WC2A 2LLDate Date: 20/12/2018
Before
THE RIGHT HONOURABLE LORD JUSTICE LONGMORE THE RIGHT HONOURABLE SIR COLIN RIMERTHE RIGHT HONOURABLE SIR PATRICK ELIAS
Between
THE LORD CHANCELLOR ANDAppellantsSECRETARY OF STATE FOR JUSTICEAppellantTHE MINISTRY OF JUSTICEAppellantCase Nos. A2/2018/0636, A2/2018/0505 & A2/2018/0647THE SECRETARY OF STATE for HOME Appellants DEPARTMENT, THE WELSH MINISTERS &OTHERS-and-R SARGEANT & OTHERS Respondentsand CrossAppellants
Mr John Cavanagh QC, Mr Raymond Hill and Ms Katherine Apps (instructed by theGovernment Legal Department) for Appellants in the first appealMr Andrew Short QC and Ms Naomi Ling (instructed by Leigh Day) for McCloud Group of Respondents in the first appealMr Michael Beloff QC and Mr Ben Jaffey QC (instructed by Bindmans LLP) for theMostyn Group of Respondents in the first appealMr John Cavanagh QC, Ms Katherine Apps and Mr Raymond Hill (instructed by theGovernment Legal Department) for Appellants (the Secretary of State and the Welsh Ministers) in the second appealMr Adrian Lynch QC and Mr Christopher Parkin (instructed by Bevan Brittan LLP) for Appellants (the London Fire & Emergency Planning Authority and the three other Fire and Rescue Authorities) in the second appealMr Andrew Short QC and Ms Lydia Seymour (instructed by Walkers Solicitors) for theClaimants/Cross Appellants in the second appealHearing Hearing dates: 5 th, 6 th, 7 th, 8 th& 9 th November 2018
Judgment Approved
[1]The first of these appeals is by the Lord Chancellor and the Ministry of Justice, from the decision of the Employment Appeal Tribunal of 29th January 2018, affirming the decision of the Employment Tribunal of 16th January 2017, in which the appellants were found to have treated the respondent younger judges less favourably than older judges on the grounds of age by reason of the transitional provisions contained in the Judicial Pension Regulations 2015 and that the appellants had failed to show that such treatment was a proportionate means of achieving a legitimate aim.[2]The second appeal, which was heard on the same occasion as the first appeal, concerns decisions in relation to the transitional provisions of the new Firefighters Pension Scheme and the equivalent Welsh scheme. They raise common or similar issues. Whilst the primary issue in both appeals is whether the respondents to the original claims have unlawfully discriminated on grounds of age, there are also claims that by implementing the transitional provisions, they have in addition breached the principles of equal pay and indirect race discrimination. The structure of this judgment is that it falls into four sections. We deal first with the background, which is common to both schemes; we then consider the age discrimination issue in the judges’ case; then in the firefighters’ case; and in the last section we analyse the equal pay and race discrimination issues in respect of both appeals.

Background to both schemes

[3]In March 2011 the Independent Public Services Pension Commission published a review of Public Sector Pensions, the Hutton Report. It recommended wholesale public sector pension reform in order to place public sector pensions on a more sustainable footing. The Government largely accepted the recommendations of that Report and enacted pension reforms through the Public Service Pensions Act 2013.[4]Paragraph 7.34 of the Report stated:- “The Commission’s expectation is that existing members who are currently in their 50s should, by and large, experience fairly limited change to the benefits which they would otherwise have expected to accrue by the time they reached their current scheme NPA [normal pension age]. This would particularly be the case if the final salary link is protected for past service, as the Commission recommends. This limitation of impact will also extend to people below age 50, proportionate to the length of time before they reach their NPA. Therefore, special protections for members over a certain age should not be necessary. Age discrimination legislation also means that it is not possible in practice to provide protection from change for members who are already above a certain age.”[5]Paragraph 1.132 of the Budget Report of the Government dated 23rd March 2011 read as follows:- “The Government accepts Lord Hutton’s recommendations as a basis for consultation of public service workers, trade unions and others, recognising that the position of the uniformed services would require particularly careful consideration. The Government will set out proposals in the Autumn that are affordable, sustainable and fair to both the public sector workforce and the tax payer.”[6]The Government published a Green Paper on 2nd November 2011 concerning public sector service pensions. It contained a foreword by the Chief Secretary to the Treasury (the Rt Hon Danny Alexander) in which he said:- “I believe it is right that we protect those public service workers who, as of 1st April 2012, have ten years or less to their pension age. It is my objective that these people see no change in when they can retire, or any decrease in the amount of pension they receive at their normal pension age …”[7]The Chief Secretary to the Treasury made a statement in Parliament on 2nd November 2011 recorded in Hansard as follows:- “In addition, I have listened to the argument that those closest to retirement should not have to face any change at all. That is the approach that has been taken over the years in relation to increases to the state pension age and I think it is fair to apply that here too. I can also announce that Scheme negotiations will be given the flexibility, outside the costs ceiling, to deliver.”[8]On the same day, the Chief Secretary to the Treasury wrote to the TUC General Secretary in the following terms:- “9. …I have accepted your argument that there should be transitional protection. It is my objective to ensure that those closest to retirement should not have any detriment either to when they can retire nor any decrease in the amount of pension they receive at their current Normal Pension Age. Over and above the costs ceiling, the Government’s objective is to provide this protection to those who on 1st April 2012 are within ten years of Normal Pension Age. Schemes and Unions should discuss the fairest way of achieving this objective, and for providing some additional protection for those who are just over ten years from their Normal Pension Age. I would be willing to consider tapering of transitional protection over a further three to four years. Full account must be taken of equalities impacts and legislation, while ensuring that costs to the tax payer each and every year should not exceed the OBR forecast for public service pension costs – i.e. those forecasts made before the further reform set out in this letter … 11. … the Government’s offer is conditional on reaching agreement. If agreement has not been reached, we may need to revisit our current proposals.”

The Changes in the Provisions in Respect of Judicial Pensions

[9]Each of the claimants (save for one regional medical member whose circumstances are in all material respects the same) are full time judges appointed before 1st April 2012 and are office holders appointed to public office within the meaning of section 50 of the Equality Act 2010. The first two appellants (as we shall refer to them in this part of the judgment) are the “relevant person” in relation to the claimants for the purposes of sections 50, 51 and 52 of the Equality Act 2010.[10]The salary of each of the claimants was set by the Lord Chancellor having regard to the recommendation of the review body on senior salaries. The compulsory retirement date of each claimant was their seventieth birthday.[11]Each claimant automatically became entitled to benefits under the Judicial Pension Scheme (“JPS”) established under the Judicial Pensions and Retirement Act 1993. Each claimant, as of 31st March 2015, was an active member of the JPS whose current service entitled him or her to benefits under that scheme.[12]In broad terms, the key benefits provided by the JPS were:- a) An annual pension of an amount equal to one fortieth of the Judge’s final “pensionable pay” multiplied by the aggregate length of service in a qualifying judicial office to a maximum of twenty years. b) A lump sum of 2.25 times the annual rate of pension was payable on retirement. c) The normal pension age (the date from which the pension could be taken as of right without actuarial reduction) was sixty five. d) A surviving spouse’s, or civil partner’s, pension was paid at half the rate of the member’s pension. There was also provision for pension in respect of a dependant child.[13]Until 1st April 2012 members were not required to contribute towards their own pension but were required to pay contributions towards survivors’ pensions. From 2006, contributions in respect of survivors’ pensions were 1.8% of the Judge’s pension capped salary.[14]The Pensions Act 2011 empowered the Lord Chancellor, by regulation, to require members of the JPS to make contributions in respect of their own pensions. That power has been exercised a number of times so as to require an increasing level of contributions with effect from 1st April 2012, 1st April 2013 and 1st April 2014 from which date members’ contributions were 3.2% in addition to the 1.8% in respect of survivors’ benefits.[15]The Judicial Pensions Regulations 2015, made pursuant to the Public Services Pensions Act 2013, came into force on 1st April 2015. They established the New Judicial Pension Scheme (“NJPS”).[16]In broad terms, the key benefits provided by NJPS are as follows:- a) Pension is accrued at the rate of approximately 1/43rd of pensionable pay in each year on a career average basis (rather than a final salary basis). There is no limit to the period of service during which pension may be accrued or taken into account in calculating the annual pension. b) No lump sum is payable in addition to the pension calculated in accordance with (a). Instead, a lump sum is available by commuting some of the annual pension entitlement. c) Normal pension age is defined as the same as the State Pension Age (varying according to the member’s date of birth) so as to be the higher of sixty five or the relevant State Pension Age attributable to the individual. d) A surviving adult (spouse, civil partner or nominated partner) pension is paid at the annual rate of three eighths of the members’ pension.[17]Initially the contribution rates under the NJPS were the same as in the JPS until the year 2018 to 2019. They have now (since 6th April 2018) risen to between 4.6% and 8.05% of pensionable pay depending upon the annualised rate of pensionable earnings.

Tax Treatment of Judicial Pension Schemes

[18]On 16th April 2006 a new regime for the taxation of pension schemes was introduced by the Pensions Act 2004 and the Finance Act 2004. The JPS is not a registered pension scheme under the regime. It was treated as an “employer financed retirement benefits scheme”. That meant that neither members’ contributions nor the lump sum attracted favourable tax treatment available in respect of registered pension schemes. However, in practice, this was not a significant disadvantage because of special provisions which were introduced simultaneously. The fact that the JPS was not a registered pension scheme was of significant benefit to its members as benefits accrued within it were not subject to the annual or lifetime allowance charges now imposed on registered schemes by the Finance Act 2004.[19]In this context the Lord Chancellor of the day, Lord Falconer, gave a written assurance to the Lord Chief Justice on 18 March 2004:- “I want you to be assured that the Government’s objective, of enabling Judges to remain in an equivalent financial position in respect of their judicial pension benefits, is settled and clear. I am therefore writing to reassure you and your colleagues that positive steps are being taken to mitigate the effects of the changes on the value of the Judicial Pension. … I want to make it very clear that I consider it extremely important to maintain the attractiveness of the Judicial Pension for the Judiciary, as well as recognising the current entitlements and expectations of serving Judges.” In letters sent by David Staff, Head of Judicial Pay and Pensions at the then Department for Constitutional Affairs, in December 2005 and March 2006 to the Lord Chief Justice and copied to all serving members of the judiciary, it was stated that:- “The Judicial Pension Schemes will fall outside the ambit of the new pension tax regime for registered pension schemes under the Finance Act … This removes the prospect of the value of Judicial Pension benefits being reduced through the imposition of this new charge on Pension Benefits from Registered Pension Schemes.” The Lord Chancellor announced those arrangements to Parliament on 15th December 20. 2005 and concluded by saying:- “I am satisfied that these proposals are in accordance with the terms of the Finance Act 2004. They serve to maintain but not improve the overall remuneration package for the serving Judiciary and to protect the principle of judicial independence in so doing.”

The Lord Chancellor announced those arrangements to Parliament on 15th December

[20]By contrast, the NJPS is a registered scheme and so is subject to the restrictions on accrual of benefits imposed by the Finance Act 2004 by means of the annual allowance and lifetime allowance rules. This is significantly disadvantageous to members of the NJPS as there is the risk of a substantial increase in tax applied to lump sum and/or pension payments.[21]Judges entitled to membership of the NJPS are entitled to opt out of that scheme and, in the alternative, to join a “Partnership Pension Account” (“PPA”) which is a registered stakeholder pension scheme. A judge taking that option terminates the final salary link for benefits accrued under the JPS.[22]The claimants contend, and it is not in dispute, that membership of the NJPS or PPA is considerably less valuable than membership of the JPS, both in terms of the reduction in the benefits paid under each scheme and with regard to the tax treatment respectively of the unregistered JPS and the registered NJPS.[23]By Schedule 2 of the Judicial Pension Regulations, Judges who, on 31st March 2015, were members of the JPS have been affected since 1st April 2015 in the following different ways:- a) Those who were active members of the JPS before 1st April 2012 and were born on or before 1st April 1957 have full protection and remain entitled to continuing active membership of the JPS. b) Those who were active members of the JPS before 1st April 2012 and were born between the 2nd April 1957 and 1st September 1960 are entitled to tapering protection. They have the option of remaining active members of the JPS until their tapered protection closing date, being a date between 31st May 2015 and 31st January 2022, whereupon they fall to be excluded from active membership of the JPS and become entitled to membership of the NJPS (they also had the option to transfer to the NJPS on 1st April 2015). c) Those who were active members of the JPS before 1st April 2012 but were born after 1st September 1960 are not entitled to any protection and have been excluded from active membership of the JPS since 1st April 2015 on which date they were entitled to membership of the NJPS/PPA.[24]It follows that those who fall within (c) are treated less favourably than those who fall within (a) and (b) and those who fall within (b) are treated less favourably than those who fall within (a). The determining factor of whether a person falls within (a), (b) or (c) is their date of birth i.e. their age.

The Relevant Pension Legislation

[25]The Public Service Pensions Act 2013 provides:- “18. Restriction of existing pension Schemes (1) No benefits are to be provided under an existing scheme to or in respect of a person in relation to the person’s service after the closing date.

(b) 31st March 2015 …” (b) 31st March 2015 …”

[26]Schedule 2 to the Judicial Pension Regulations 2015 provides:- “Exceptions to section 18(1) of the Act: full protection members of an existing scheme 8.(1) A person (P) is a full protection member of an existing scheme if sub-paragraph (2) … applies(2) This sub-paragraph applies if – (a) P was an active member of an existing scheme on 31st March 2012; (b) P was an active member of that scheme on the scheme closing date; and (c) unless P dies, P would reach normal pension age under that Scheme on or before 1st April 2022 … PART 3 Exceptions to section 18(1) of the Act: tapered protection members of an existing scheme 12. (1) A Person (P) is a tapered protection member of an existing scheme if sub-paragraph (2) … applies (2) This sub-paragraph applies if (a) P was an active member of an existing scheme on 31st March 2012; (b) P was an active member of an existing scheme on the scheme closing date; and (c) unless P dies, P would reach normal pension age during the period beginning with 2nd April 2022 and ending with 1st September 2025.” (a) P was an active member of an existing scheme on 31st March 2012; (b) P was an active member of that scheme on the scheme closing date; and (c) unless P dies, P would reach normal pension age under that Scheme on or before 1st April 2022 … Exceptions to section 18(1) of the Act: tapered protection members of an existing scheme (a) P was an active member of an existing scheme on 31st March 2012; (b) P was an active member of an existing scheme on the scheme closing date; and (c) unless P dies, P would reach normal pension age during the period beginning with 2nd April 2022 and ending with 1st September 2025.”

EU Directive and the Equality Act 2010 Provisions

[27]Article 1 of the Council Directive 2000/78 sets out a general framework for combating discrimination on grounds of, amongst other things, age, sex and race.[28]Article 2 provides:- “Concept of discrimination 1. For the purposes of this Directive, the “principle of equal treatment” shall mean that there shall be no direct or indirect discrimination whatsoever on any of the grounds referred to in Article 1.”[29]Article 6 provides:- “Justification of differences of treatment on grounds of age 1. Notwithstanding Article 2(2), Member States may provide that differences of treatment on grounds of age shall not constitute discrimination, if, within the context of national law, they are objectively and reasonably justified by a legitimate aim, including legitimate employment policy, labour market and vocational training objectives, and if the means of achieving that aim are appropriate and necessary.”[30]Section 13 of the Equality Act 2010 provides:- “(1) A person(a) (A) discriminates against another(b) (B) if, because of a protected characteristic, A treats B less favourably than A treats, or would treat, others. (2) If the protected characteristic is age, A does not discriminate against B if it can show A’s treatment of B to be a proportionate means of achieving a legitimate aim.”[31]Section 19 of the Equality Act 2010 provides:- “Indirect discrimination(1) A person (A) discriminates against another (B) if A applies to B a provision, criterion or practice which is discriminatory in relation to a relevant protected characteristic of B’s.(2) For the purpose of subsection (1), a provision, criterion or practice is discriminatory in relation to a relevant protected characteristic of B’s if – (a) A applies, or would apply, it to persons with whom B does not share the characteristic, (b) it puts, or would put, persons with whom B shares the characteristic at a particular disadvantage when compared with persons with whom B does not share it, (c) it puts, or would put, B at that disadvantage, and (d) A cannot show it to be a proportionate means of achieving a legitimate aim.(3) The relevant protected characteristics are - age; … race; … sex; …” (a) A applies, or would apply, it to persons with whom B does not share the characteristic, (b) it puts, or would put, persons with whom B shares the characteristic at a particular disadvantage when compared with persons with whom B does not share it, (c) it puts, or would put, B at that disadvantage, and (d) A cannot show it to be a proportionate means of achieving a legitimate aim. … race; … sex; …”[32]Sections 61 and 67 of the Equality Act provide materially:- “61 Non-discrimination rule (1) An occupational pension scheme must be taken to include a non-discrimination rule. (2) A non-discrimination rule is a provision by virtue of which a responsible person(a) (A) - (a) must not discriminate against another person(b) (B) in carrying out any of A’s functions in relation to the scheme; … (3) The provisions of an occupational pension scheme take effect subject to the non-discrimination rule. ... (10) A non-discrimination rule does not have effect in relation to any occupational pension scheme in so far as an equality rule has effect in relation to it (or would have effect in relation to it but for Part 2 of Schedule 7.” 67. Sex equality rule (1) If an occupational pension scheme does not include a sex equality rule, it is to be treated as including one. (2) A sex equality rule is a provision that has the following effect- (a) if a relevant term is less favourable to A than it is to B, the term is modified so as not to be less favourable; (b) if a term confers a relevant discretion capable of being exercised in a way that would be less favourable to A than to B, the term is modified so as to prevent the exercise of the discretion in that way. (3) A term is relevant if it is- (a) a term on which persons become members of the scheme, or (b) a term on which members of the scheme are treated. (4) A discretion is relevant if its exercise in relation to the scheme is capable of affecting- (a) the way in which persons become members of the scheme, or (b) the way in which members of the scheme are treated. (5) The reference in subsection (3)(b) to a term on which members of a scheme are treated includes a reference to the term as it has effect for the benefit of dependants of members. (6) The reference in subsection (4)(b) to the way in which members of a scheme are treated includes a reference to the way in which they are treated as the scheme has effect for the benefit of dependants of members. … (8) A relevant matter is (a) a relevant term; (b) a term conferring a relevant discretion;(c) the exercise of a relevant discretion in relation to an occupational pension scheme. …” ... (a) if a relevant term is less favourable to A than it is to B, the term is modified so as not to be less favourable; (b) if a term confers a relevant discretion capable of being exercised in a way that would be less favourable to A than to B, the term is modified so as to prevent the exercise of the discretion in that way. (a) a term on which persons become members of the scheme, or (b) a term on which members of the scheme are treated. (a) the way in which persons become members of the scheme, or (b) the way in which members of the scheme are treated. (a) a relevant term; (b) a term conferring a relevant discretion; (c) the exercise of a relevant discretion in relation to an occupational pension scheme. …”[33]Section 69 of the Equality Act provides:- “Defence of material factor … (4) A sex equality rule has no effect in relation to a difference between A and B in the effect of a relevant matter if the trustees or managers of the scheme in question show that the difference is because of a material factor which is not the difference of sex. (5) “Relevant matter” has the meaning giving in section 67. (6) For the purpose of this section, a factor is not material unless it is a material difference between A’s case and B’s.”

The Claimants’ Claims

[34]The claims brought by V McCloud and others all relate to categories of Judge other than High Court Judges. The claims brought by N Mostyn and others are brought in relation to High Court and Court of Appeal Judges.[35]The claims in the McCloud case are that the claimants were treated less favourably than those falling within the protected and/or taper groups on the grounds of age and that the less favourable treatment was not justified pursuant to section 13(2) of the Equality Act 2010. Thereby they were directly discriminated against on ground of age and in breach of the non-discrimination rule included in the JPS and the NJPS by section 61 of the Equality Act 2010.[36]The female claimants also make a claim for equal pay under the sex equality rule incorporated into the pension schemes by section 67 of the Equality Act on the basis that the transitional provisions disproportionately adversely affect women and the relevant term is not objectively justified. To the extent that the equal pay claims brought by the female judges succeed, then each male claimant makes a “piggy-back” claim to the like effect relying on the success of the female claimants.[37]Certain claimants also brought claims of indirect sex and race discrimination on the basis that the transitional provisions put women and BME claimants at a disadvantage for the purposes of section 19 of the Equality Act 2010 and those provisions are not objectively justified for the purposes of section 19(2)(d) of the 2010 Act.[38]Insofar as it is relevant to this appeal, the Mostyn claimants’ claims are couched in identical terms to those of the claimants in the McCloud litigation.[39]The appellants accept that the claimants have, by virtue of the terms of the NJPS and the contemporaneous tax treatment of the claimants, been directly discriminated against by reason of age but say that the treatment suffered by the claimants is a proportionate means of achieving a legitimate aim.[40]The appellants also accept that they cannot justify disparate discriminatory treatment by reference solely to the saving of cost.

The appellants’ aims

[41]The appellants have not been entirely clear in their assertion of their aims. The pleaded aim was “to protect those closest to retirement from the financial effects of pension reform”. The claimants assert that this suffered from the defect that it said no more than that older members of the judiciary should be treated more beneficially than younger members and can, therefore, be said to be merely a re-statement of the discriminatory treatment. To the extent that it could be said to imply that the pension reforms would affect those closer to retirement most adversely, it was wrong on the facts since the effect of the transitional provisions was to give protection to the older members and not to the younger members.[42]There was a suggestion before the judge that a possible explanation was that older Judges would be more likely to have made fixed or concrete plans for retirement which would be difficult to change. But the judge dismissed that as no more than speculation without any hard evidence and at any rate of only minor consequence to older Judges: see paras 52-55.[43]It was also suggested that the evidence supported a wish by government in the proposed pension reforms that all public sector persons should be treated consistently. 10 year protection was given to older pensioners when the state pension was reformed; the large public sector unions had been given 10 year protection and the Judges should therefore be in the same position.[44]Before the Employment Appeal Tribunal and this Court it was also suggested that the explanation for treating older judges more advantageously than younger judges was a “moral and political aim” of being fair to those closest to retirement. Mr John Cavanagh QC for the appellants described this underlying reason as a “moral one but affected by political aims” and submitted that any reliance on the appellants’ lack of evidence was therefore misplaced.[45]Mr Andrew Short QC for the judges asserted that this “moral and political aim” was never pleaded or argued before Judge Williams. Mr Cavanagh’s riposte was that his new way of putting the matter was no more than the opposite side of the coin of the aim the appellants had already pleaded namely the aim of protecting those closest to retirement from the financial effects of the pension reform. On that basis we are content to consider this alternative way of putting the matter.

The Employment Tribunal decision

[46]Judge Williams began by setting out the largely undisputed facts and we quote some extracts only from his full and careful judgment. In relation to the tax changes he said:- “29. Prior to the introduction of the NJPS there was no public notification of any intended change to the tax status of the judicial pension scheme with the result that applicants continued to seek appointment on the understanding that their terms and conditions, including those relating to the taxation of pensions, would remain as they had been. As Ian Gray, Deputy Director, Pensions and Judicial Reward at the second respondent, wrote in October 2012:- “Switching off this tax advantage has very significant implications for serving judiciary that they could not have anticipated nor reasonably made revised arrangements for and requires this change to be handled differently from the standard pension reform being applied across the public service.”30. The JPS remains unregistered for tax purposes and members are therefore not subject to the annual allowance and lifetime allowance limits. The NJPS is a registered scheme so that the annual allowance and lifetime allowance limits apply with the result that, on transfer into the NJPS, many claimants incur very significant additional tax liabilities compared with their position as members of the JPS. The change from taxunregistered to tax-registered status of their pension scheme affected judges uniquely amongst public servants because theirs was the only scheme which was previously unregistered. Furthermore, as Mr Scanlon explained, these losses have been increased still further by changes to the pension tax regime announced in the summer budget of 2015, after the making of the regulations of 2015 in February of that year. This serves to magnify the disparity between the unprotected and the protected judges.31. The loss sustained by the unprotected and taper-protected judges, including these claimants, was very significantly greater than the loss sustained by other public servants whose pension schemes were reformed. There are two reasons for this. Firstly, the value of a judge’s pension as a proportion of his or her overall remuneration is significantly greater than in the case of other public service employees. Adverse changes to a judge’s pension therefore have a proportionally greater impact.32. The second reason is that judges alone suffered the combined effect of significant adverse changes to their pension scheme in addition to a radical change to the tax treatment of their pensions. The valuation of the losses occasioned to a judge by the taxation reforms inevitably varies from individual to individual. I am not required in this judgment to compute the claimants’ losses, and the figures for individual judges with which I have been provided are illustrative only and are not the subject of precise agreement. Nevertheless, it is clear that the high court judge claimants will incur losses running in many cases into several hundreds of thousands of pounds. It is agreed that the yearly capital investment required to provide a life annuity giving approximately the same benefits on retirement as those lost by the transfer to the NJPS is at least £30,000.” “Switching off this tax advantage has very significant implications for serving judiciary that they could not have anticipated nor reasonably made revised arrangements for and requires this change to be handled differently from the standard pension reform being applied across the public service.”[47]He referred to further features which distinguish Judges from other public servants:- “36. Apart from the uniquely adverse effect on judges of the combination of changes set out above, there are further features which uniquely distinguish appointment as a judge from other public servants. Whilst it is accepted that all public servants accept office or employment on the basis of the terms and conditions offered at the time, and that those terms and conditions may be varied from time to time, in the case of the judiciary there had been explicit and strongly worded assurances from the then Lord Chancellor in 2004 to the effect that it was the government’s settled view that there should be no change to serving judges’ pensions. This assurance was in line with the scheme adopted on the coming into force of Part 1 of the Judicial Pensions and Retirement Act 1993 which established the JPS and was expressed in section 1 to apply “to any person who first holds qualifying judicial office on or after the appointed day”. The changes made by the JPS compared with its predecessor scheme were thus prospective, affecting future appointees only.”[48]He observed (para 40) that the clear and obvious effect of the transitional provision was that the transitional protection was aimed at those who were least affected by the changes, as accepted by the appellants’ witness from the Treasury, Mr Kelly.[49]He then found the real reasons for incorporating transitional provisions into the JPS:- “56. Based on this evidence I consider it proper to find that the government decided to incorporate the transitional provisions into the JPS for no reasons specific to the judiciary, but rather because similar provisions had been agreed with trade unions for other workforces and the government’s preference was for a consistent scheme, and, to a lesser extent, because the state pension age consultation had led to the view that a period of ten years’ notice was appropriate in that case. I found the further arguments based on those nearing retirement having less time to prepare for the effects of reform and having fixed retirement plans lacked cogency for the reasons set out above.”[50]Judge Williams then turned to the law. He put to one side a constitutional argument advanced by Mr Michael Beloff QC for the High Court Judges that, while pensions could not themselves be considered to fall within the prohibition on diminution of salary derivable from the Bill of Rights of 1689 or the Act of Settlement of 1701, the constitutional position of his clients meant that any discrimination by way of pensions and disadvantageous tax treatment should be treated as comparable to a reduction in pay and, for that reason, be rejected. Judge Williams then emphasised that broad matters of public policy were for the government of the day: - “78. … In a wider sense, the aim of the respondents, and of the government as a whole, was to establish public service pension arrangements which were, in the words of the terms of reference for Lord Hutton’s commission:- “Sustainable and affordable in the long term, fair to both the public service workforce and the taxpayer and consistent with the fiscal challenges ahead, while protecting accrued rights”. 79. Those are matters which belong in the realm of public policy and finance for which the government of the day is responsible to the electorate. It is for the government to define its policy objectives, to identify its priorities and to determine what resources it will allocate to them. This tribunal must take particular care not to trespass into areas which are not its proper purview. I am concerned solely with the respondents’ attempt to justify the disparate impact of the transitional provisions contained in Schedule 2 to the 2015 Regulations.” “Sustainable and affordable in the long term, fair to both the public service workforce and the taxpayer and consistent with the fiscal challenges ahead, while protecting accrued rights”.[51]He then set out a number of legal principles derived from both domestic and European law which to a greater or less extent were disputed before us, particularly the proposition at para 80(5):- “I see no basis for saying that the government’s broad discretion in matters of social policy extends beyond that public arena into the arena of private relations between employer and employee.”[52]He then turned to apply the principles of law he had enunciated to the facts of the case, taking both the pleaded aim and the developments of it articulated in argument.[53]He considered first the pleaded aim saying (para 86-7):- “86. The formulation of the respondents’ aim which was most frequently canvassed in evidence and in submissions before this tribunal was taken from the case pleaded in their response to these claims:
“the legitimate aim of protecting those closest [to] retirement from the financial effects of pension reform”
. This adds little to what the Chief Secretary wrote and might suggest that it was thought that the pension reforms would affect most adversely those closest to retirement and/or that they would be in particular need of protection from their financial effects. But it is quite clear from ample contemporary documentation and the unanimous evidence in this case, not only that the opposite is true, but that it was well known to be true long before the enactment of the 2015 Regulations. The older judges are, the less adversely are they affected by the reforms. 87. If an aim is to be described as a legitimate social policy aim, then it must in my judgment be something for which there is a rational explanation. The government has a wide discretion in such matters and its aim does not have to be one with which the tribunal agrees; but its aim may not be, for example, capricious or arbitrary, and it must be capable of being understood. To set out consciously to treat more favourably a group who, as was well known at the time, were the least adversely affected by the reforms appears counter-intuitive and at the very least calls for such a rational explanation. In the absence of such explanation it would be difficult to resist Mr Beloff’s categorisation of the result as bizarre.”[54]He then considered the associated aims that older Judges would have less time to make necessary adjustments and that older Judges would have made fixed plans for retirement which they would find difficult to change, but dismissed these as rational explanations for favouring older Judges, partly because they were not much different to the pleaded aim and partly because there was no evidence to support them. He then considered whether consistency across the public service was a legitimate aim but said that, as government had itself recognised, the position of the Judges was such that consistency between schemes was scarcely attainable.

Judge Williams then summarised his conclusions at para 94 as follows:

[55]- “In summary, my conclusions on the question of the legitimacy of the respondents’ aims are as follows. Description of a group having ‘ten years or less to their pension age’, being ‘closest to retirement’ or having ‘less time to adjust’ all necessarily define that group by reference to the age of those in the group. In my judgment, an aim which amounts to an intention to treat one group more favourably and another less favourably, solely by reference to the age of those in the groups cannot, without further rational explanation of the reason for it, be legitimate. An aim thus expressed amounts to a declaration of intent to do precisely that which the statute prohibits. The respondents have failed to advance any such rational explanation of their reason. Mr Chamberlain’s formulation in his closing submissions, ‘whether … it was lawful for the respondents to introduce these limited and affordable transitional protections, whose aim was to identify a category of scheme members closest to retirement who would see no change at all’, restates the question succinctly, but does not answer the question ‘why?’ The respondents have failed to adduce any evidence of disadvantage suffered by the fully protected and the taperprotected groups of judges which called for redress, or any social policy objective which was served by treating those groups more favourably and the claimant group less favourably. I accept that in implementing their pension reforms the respondents and the government as a whole were entitled in principle to pursue the aim of consistency, and that such consistency could, in a properly evidenced case, be conducive to a social policy objective. However, in my judgment, the respondents have failed to demonstrate beyond the level of ‘mere generalisations’ how consistency in the matter of transitional protection was capable of contributing to their social policy objective, especially since so much else in the JPS was inconsistent with other reformed pension schemes. I find accordingly that it has not been shown in this case that the aim of consistency is capable of justifying derogation from the principle of non-discrimination on the ground of age.”[56]That would have been enough to dispose of the respondents’ case, but Judge Williams went on to consider what the position would be if he had found the suggested aims to be legitimate. He reminded himself (para 107) that he was only concerned with the transitional provisions and not the NJPS itself and said that while, for example, consistency could in some circumstances be a legitimate aim, it fell far short of outweighing the significant derogation from the principle of equal treatment. He concluded:- “116. I have asked myself whether the setting of the limit by the respondents represents a rational attempt to achieve the aim in question. That same question arises whether the aim is said to be to protect those closest to retirement or to achieve consistency across the public sector. In this case the ten-year criterion was “read across” from the other larger public sector schemes with some – albeit that I have found it to be inadequate – support from the state pension age consultation. There was no specific reference to the judiciary at all.117. The nearest the evidence in this case takes me to an answer to the question why ten years was chosen is that it was what was necessary in order to do a deal with the trades unions in the other larger public sector schemes. There was no research or analysis, and nor was there any process of reasoning which led the respondents to consider that – making due allowance for the arbitrariness referred to above – approximately ten years would achieve the desired aim, whereas something like four or five years, for example, would be far too short and something like fifteen to twenty years, for example, would be far too long. I am not, of course, suggesting that any process of precise calculation should, or could, result in such a figure, but I would have expected to see evidence of some thought process which led the thinker to the view that ten years was about right. There was no such evidence and I am satisfied on all that I have heard that there was no such thought process. Rather, the ten-year figure was imported from discussions with trades unions in relation to other schemes where it may have had some rational basis of which I have naturally not heard in these proceedings.118. The transitional provisions initially protected something of the order of 85% of serving judges, many if not most of whom suffered only minor adverse effects from the reforms, whilst leaving the unprotected judges, including the claimants, exposed to a severe adverse impact. In my judgment the balance described by Langstaff P in Seldon (No 2) has not been properly struck in this case. The respondents have failed to provide evidence that a shorter period, or lesser degree, of protection would not have enabled them to achieve their aim, whether of protecting those closest to retirement or of consistency; the respondents adduced no specific evidence – beyond the generalities already referred to – to explain why they chose to set the relevant age limits where they did. One returns repeatedly in this case to the importing of age limits from other schemes and the analogy of the state pension consultation. These transitional provisions were not a reasonably necessary means of achieving the government’s aims because they go beyond what was necessary either to achieve consistency or to protect those closest to retirement.”[57]Judge Williams held, therefore, that the appellants have treated and continued to treat the claimants less favourably than older judges because of their age and that the respondents have failed to show the treatment to be a proportionate means of achieving a legitimate aim.

The appeal to the Employment Appeal Tribunal

[58]The main ground of the appellants’ appeal to the Appeal Tribunal was that in his consideration of both aims and means Judge Williams had erred in law by failing to accord to the appellants’ aims and means the wide margin of discretion mandated by both domestic and European law. They also attacked his decision for focusing on the absence of evidence to conclude that the policy of protecting Judges was irrational when the aims were informed by moral or political or social value judgments. Likewise, in respect of aims, it was said that the judge applied his own judgment or standards of scrutiny rather than giving the government a wide margin of discretion. It was also said that the judge’s decision on means was infected by his wrong conclusion about aims and that he also erred in considering process rather than outcome and that his reasoning on alternative means was directed towards different aims.[59]In relation to the main argument of law, Sir Alan Wilkie detected a tension between decisions of the ECJ/CJEU on the one hand requiring that a wide margin of discretion be accorded to the government in the assessment of social aims and domestic decisions on the other which denied any such wide margin and encouraged judges to adopt their own standards of scrutiny in assessing both legitimacy of aims and proportionality of means. He said, however, (para 134) that the Supreme Court in Seldon v Clarkson Wright & Jakes [2012] UKSC 16; [2012] ICR 716 had reconciled these two approaches by:- 1) distinguishing measures pursuing social policy objectives such as employment policy, the labour market or vocational training of a public interest nature from measures particular to the employer’s situation such as cost reduction or improving competitiveness; and 2) emphasising that the mere fact that a particular aim was capable of being a legitimate aim was only the beginning of the story. Once an aim had been identified, it still had to be asked whether it was legitimate in the particular circumstances of the employment concerned. That required that both the aims and the means be carefully scrutinised to see whether they met the objective and whether there were less discriminatory measures which would do so.[60]In the light of this reading of the law, Sir Alan Wilkie assessed the actual conclusion of Judge Williams in relation to legitimate aims and concluded (para 142) that he could not be criticised for adopting an unduly strict level of scrutiny or demanding a higher level of evidence than was consistent with the proper approach identified by the CJEU and recognised by the Supreme Court in Seldon. He accordingly said that Judge Williams had not erred in law in this respect.[61]In spite of deciding that Judge Williams had not misdirected himself in law on the main legal argument put before him, Sir Alan nevertheless held that Judge Williams had misunderstood and/or misapplied the facts and misdirected himself in a different respect. This was in response to a submission of the appellants that Judge Williams had focused unduly on the need for evidence of disadvantage suffered by the protected group (namely older judges) where it had been common ground that the protected groups were less severely disadvantaged than the unprotected groups. The judge had accordingly ignored “the voluminous evidence … which revealed why it was that the government adopted these transitional provisions”. It was then said that the aim of protecting older judges had been couched from the outset as a moral judgment allied to a process which was highly political and thus was not capable of being supported by hard evidence but was nevertheless a legitimate aim. Reliance was placed on R (on the application of Lumsdon) v Legal Services Board [2015] UKSC 41; [2016] A.C. 697 in which Lords Reed and Toulson (with whom the other members of the court agreed) said (para 56):- “… much may depend on the nature of the justification and the extent to which it requires evidence to support it. For example, justifications based on moral or political considerations may not be capable of being established by evidence. The same may be true of justifications based on intuitive common sense. An economic or social justification, on the other hand, may well be expected to be supported by evidence.”[62]Sir Alan accepted this argument of the appellants saying that Judge Williams had limited his consideration of aim to the statements of the Chief Secretary to the Treasury (which he (Judge Williams) characterised as tautologous) and had dealt with the evidence by focusing on the failure to produce evidence of disadvantage to the protected group of judges or of the social policy objective served by treating that group more favourably. He then said:- “160. In my judgment the Appellants are correct in their contention that this narrow consideration of the material failed to take into account what the EJ had found to be a complex of moral and political judgments informed by the plethora of documentation from different sources. In my judgment, the EJ failed to have regard to that evidence and the complex of reasons which that evidence reveals. He failed to pay due regard to the cautionary words of the CJEU in Fuchs, and the Supreme Court in Lumsdon, on the particular difficulty of producing evidence where moral or political considerations are to the fore. In my judgment, the EJ, in concluding that no legitimate aims had been shown, misunderstood and/or misapplied the facts and thereby misdirected himself.”[63]In the end, however, Sir Alan said this did not matter because no such criticism could be made of the judge’s approach to proportionality of means. He said (paras 168- 170):- “168. … The EJ had regard, as he was entitled, to the unique position of the Judiciary and to the uniquely adverse impact of the pension scheme and tax changes on the unprotected, as opposed to the protected, members of the Judiciary and to come to a view on the issue which was open to him.169. The EJ concluded that, in principle, transitional provisions were appropriate to achieve the aim because that is what they were designed to do (paragraph 112). In so doing he was assuming, contrary to his earlier finding that the pursuit of the policy was a legitimate aim. …170. In the light of his clear and sustainable finding on the question of balance, the EJ’s conclusion on whether the means were reasonably necessary for the achievement of the aim followed as a matter of course.”[64]Thus, it is the appellants who have to challenge Sir Alan’s judgment in this court. They do this by saying:- i) that Sir Alan erred in approving the judge’s self-direction on margin of discretion since this was a case where the very widest margin of discretion should be accorded to government; ii) that Sir Alan was, in any event, correct to say that Judge Williams erred in his approach to the evidence (or lack of it); and iii) that both Sir Alan’s and the judge’s decisions on proportionality of means were affected by wrong conclusions as to legitimacy of aim.

Margin of discretion/level of judicial scrutiny

[65]When the case was opened to us, this seemed to be the main point of law at issue and, as a point of law, was therefore a proper matter for this court to consider on an appeal from judgments which were otherwise based largely on questions of fact, evidence and judicial evaluation of what are now well recognised concepts of legitimacy of aim and proportionality of means.[66]Mr Cavanagh opened the respondents’ appeal by accepting that in a case of direct discrimination by virtue of age, a legitimate aim had to be one of social policy rather than an aim adopted for what he called “operational reasons”, but he submitted that once a social policy aim (such as protecting the older judges from the effect of the reforms) was established, the court should afford the widest margin of discretion to government when it determined whether the aim was legitimate or not and that a different level of judicial scrutiny was appropriate for social policy aims from that to be applied to operational decisions.[67]Mr Short for the Judges submitted in his skeleton argument, by contrast, that a two stage approach was necessary. First, the appellants had to show that their measures were consistent with the government’s social policy aims and should be accorded a margin of discretion in so doing but, once that had happened, the court had to decide both legitimacy of aim and proportionality of means without any allowance for margin of discretion. He even went so far as to say that, to the extent that the EU directive or the CJEU case law said anything different, that did not matter because the terms of the Equality Act 2010 imposed a higher standard of scrutiny than European law required. The directive provided a floor but not a ceiling. Sir Alan was correct to discern a difference between two distinct lines of authority in the domestic and the European case law but wrong to say that Seldon had reconciled them by saying that margin of discretion had any part to play beyond helping to establish the existence of a social aim.[68]Mr Short may not have needed to go as far as this in defending Sir Alan’s upholding Judge Williams on proportionate means since Sir Alan had held that Judge Williams had afforded a sufficient margin of discretion on that topic. But he did need to go that far in relation to the firefighters’ appeal (with which we deal below) since Sir Alan had held in terms that a wide margin of discretion should be afforded in relation to legitimate aim, which required a lesser degree of judicial scrutiny, while in relation to proportionality of means a lesser margin of discretion and a greater judicial scrutiny should be applied (see paras 80-83 of the judgment in the firefighters’ case).[69]It is accordingly necessary for this court to come to a conclusion on the width of any appropriate margin of discretion in relation to both aims and means and indeed to decide whether there are two distinct lines of authority which needed to be (and were) reconciled in Seldon.[70]We find it helpful to consider first the pre-Seldon domestic authorities.[71]It is convenient to begin with R v Secretary of State for Employment ex parteSeymour-Smith [2000] 1 WLR 435. The issue was whether the Unfair Dismissal (Variation of Qualifying Period) Order 1985, which increased the qualifying period from one year to two years for unfair dismissal claims, was inconsistent with the Equal Treatment Directive and/or section 119 of the EC Treaty because fewer women than men were able to comply with it. The aim of the Order was the legitimate one of encouraging employment; on the question whether the means were proportionate, the Court of Appeal held that the defendant had to show that the threshold of two years had been proved to result in greater availability of employment than would have been the case without it. After the House of Lords had referred a number of questions to the European Court of Justice, the House decided that the test applied by the Court of Appeal was too stringent and that the Order was a reasonable and proportionate response unrelated to any discrimination based on sex. Lord Nicholls of Birkenhead said, in a passage, aptly quoted by Judge Lewzey in the Sargeant case, at 450 F-H:- “… The burden placed on the government in this type of case is not as heavy as previously thought. Governments must be able to govern. They adopt general policies, and implement measures to carry out their policies. Governments must be able to take into account a wide range of social, economic and political factors. The Court of Justice has recognised these practical considerations. If their aim is legitimate, governments have a discretion when choosing the methods to achieve their aim. National courts, acting with hindsight, are not to impose an impracticable burden on governments which are proceeding in good faith. Generalised assumptions, lacking any factual foundation, are not good enough. But governments are to be afforded a broad measure of discretion. The onus is on the member state to show(1) that the allegedly discriminatory rule reflects a legitimate aim of its social policy,(2) that this aim is unrelated to any discrimination based on sex, and(3) that the member state could reasonably consider that the means chosen were suitable for attaining that aim.” This shows that in an appropriate case the government is to be accorded a margin of discretion when it comes to assessing proportionate means. Seymour-Smith was a sex discrimination case, but in our view the same principle must be applied whatever the ground of discrimination relied upon.[72]In Hardy & Hansons Plc v Lax [2005] ICR 1565, much relied on by Mr Short, a private employer had not permitted a female employee, returning to work after giving birth, to work part-time. This prohibition of part-time work had to be objectively justified, a matter on which the tribunal had to make up its own mind rather than to decide whether the ban on part-time work came within a range of reasonable responses open to an employer to adopt. In paragraph 14 Pill LJ (with whom Thomas and Gage LJJ agreed) recorded the employer’s submission that it ought to have been granted a margin of discretion in deciding whether to permit job sharing; he rejected that submission in para 32 saying:- “The employer has to show that the proposal, in this case for a full-time appointment, is justified objectively notwithstanding its discriminatory effect. The principle of proportionality requires the tribunal to take into account the reasonable needs of the business. But it has to make its own judgment, upon a fair and detailed analysis of the working practices and business considerations involved, as to whether the proposal is reasonably necessary. I reject the employers’ submission (apparently accepted by the appeal tribunal) that, when reaching its conclusion, the employment tribunal needs to consider only whether or not it is satisfied that the employer’s views are within the range of views reasonable in the particular circumstances.” It is fair to say that this authority appears to equate the phrases “a range of reasonable responses” and “margin of discretion”.[73]This authority was followed in the age discrimination case MacCulloch v ImperialChemical Industries Plc [2008] ICR 1334 in which older employees were given greater benefits on redundancy than were given to younger employees. Elias P in the Employment Appeal Tribunal summarised the legal principles applicable to justification in para 10 including:- “(4) It is for the employment tribunal to weigh the reasonable needs of the undertaking against the discriminatory effect of the employer’s measure and to make its own assessment of whether the former outweigh the latter. There is no “range of reasonable response” test in this context: Hardy & Hanson vLax.” It is to be noted in both these cases that the dispute was a private dispute between employer and employee without any reference to the position of government.[74]By comparison with these domestic authorities Mr Cavanagh cited a number of authorities from the European Court of Justice and the Court of Justice for the European Communities which decided (not unlike Seymour-Smith) that governments should be awarded a margin of discretion in the field of social and employment policy.[75]Thus in Palacios de la Villa v Cortefiel Servicios S.A. case C-411/05 [2008] 1 C.M.L.R. 16 in which the question was whether a compulsory retirement age in a collective agreement (which Spanish courts unlike the English courts would regard as legally binding) infringed the Equality Directive, the Grand Chamber of the Court of Justice of European Communities said:- “68. It should be recalled in this context that, as Community law stands at present, the Member States and, where appropriate, the social partners at national level enjoy broad discretion in their choice, not only to pursue a particular aim in the field of social and employment policy, but also in the definition of measures capable of achieving it (see, to that effect, Mangold v Rudiger Helm (C-144/04) [2005] E.C.R. I- 9981; [2006] 1 C.M.L.R. 43 at [63]).”[76]In the similar case of Rosenbladt v Oellerking case C-45/09 [2011] IRLR 51, the Grand Chamber of the Court of Justice of the European Union cited Palacios at para 58 and said at para 69:- “Accordingly in the light of the wide discretion granted to the social partners at national level in choosing not only to pursue a given aim in the area of social policy but also in defining measures to implement it, it does not appear unreasonable for the social partners to take the view that [the measure in question] may be appropriate for achieving the aims set out above.”[77]Similar statements can be found in Hütter v Graz Technische C-88/08; [2009] 3 CMLR 35, Fuchs v Land Hessen C-159/10 [2012] ICR 93 and HK Danmark vExperian SA C-476/11 [2014] ICR 27; this latter case was a case in which the CJEU did say that the measure was unreasonable.[78]So far, we detect no inconsistency between the domestic and the European authorities. The parties agreed, however, that the most significant authority was that of Seldon vClarkson Wright & Jakes [2012] UKSC 16; [2012] ICR 716 in the Supreme Court which, although a case about a compulsory retirement age contained in a solicitors’ partnership, contains a comprehensive statement of the law in relation to age discrimination. Baroness Hale of Richmond JSC who gave the leading judgment (with which Lord Hope of Craighead DPSC, Lord Brown of Eaton-under-Heywood, Lord Mance and Lord Kerr of Tonaghmore JJSC agreed) emphasised recital 25 to the Equality Directive in para 2 of her judgment:- “However, differences in treatment in connection with age may be justified under certain circumstances and therefore require specific provisions which may vary in accordance with the situation in member states. It is therefore essential to distinguish between differences in treatment which are justified, in particular by legitimate employment policy, labour market and vocational training objectives, and discrimination which must be prohibited.” She then conducted a comprehensive survey of the European jurisprudence citing (inter alia) Palacios in which the encouragement of recruitment was considered to be a legitimate aim but which required the means employed to achieve that aim to be both appropriate and necessary, “although member states enjoyed a broad discretion in the choice both of the aims and of the means to pursue them” (para 33)[79]Having conducted that survey Lady Hale (in para 50) set out the messages which can be taken from the European case law, of which numbers (2), (4) and (6) are particularly relevant for the purposes of this case:- “(2) If it is sought to justify direct age discrimination under article 6(1) the aims of the measure must be social policy objectives, such as those related to employment policy, the labour market or vocational training. These are of a public interest nature, which is “distinguishable from purely individual reasons particular to the employer’s situation, such as cost reduction or improving competitiveness” (Age Concern [2009] ICR 1080 and Fuchs [2012] ICR 93). … (4) A number of legitimate aims, some of which overlap, have been recognised in the context of direct age discrimination claims:(i) promoting access to employment for younger people (Palacios de la Villa, Hütter and Kücükdeveci);(ii) the efficient planning of the departure and recruitment of staff (Fuchs);(iii) sharing out employment opportunities fairly between the generations (Petersen, Rosenbladt and Fuchs);(iv) ensuring a mix of generations of staff so as to promote the exchange of experience and new ideas (Georgiev and Fuchs);(v) rewarding experience (Hütter and Hennigs);(vi) cushioning the blow for long serving employees who may find it hard to find new employment if dismissed (Ingeniørforeningen i Danmark);(vii) facilitating the participation of older workers in the workforce (Fuchs; see also Mangold v Helm (Case C-144/04) [2006] All ER (EC) 383);(viii) avoiding the need to dismiss employees on the ground that they are no longer capable of doing the job, which may be humiliating for the employee concerned (Rosenbladt); or(ix) avoiding disputes about the employee’s fitness for work over a certain age (Fuchs). … (6) The gravity of the effect on the employees discriminated against has to be weighed against the importance of the legitimate aims in assessing the necessity of the particular measure chosen (Fuchs).” …[80]The judgment then addressed the particular issues which arose in that case namely whether the three aims of the retirement clause identified by the employment tribunal were capable of being legitimate aims and whether the clause was a proportionate means of achieving any or all of those identified aims. In para 54 Lady Hale recorded the acceptance by the Secretary of State for Business Innovations and Skills as intervener that there was a distinction between aims such as cost reduction and improving competitiveness, which would not be legitimate, and aims relating to employment policy, the labour market and vocational training, which would be legitimate. At para 59 Lady Hale said that the fact that a particular aim was capable of being a legitimate aim is only the beginning of the story. It was still necessary to enquire whether it was in fact the aim being pursued. Then (para 61) it had to be asked whether the aim being pursued was legitimate in the particular circumstances of the employment concerned. Then at para 62:- “Finally, of course, the means chosen have to be both appropriate and necessary. It is one thing to say that the aim is to achieve a balanced and diverse workforce. It is another thing to say that a mandatory retirement age of 65 is both appropriate and necessary to achieving this end. It is one thing to say that the aim is to avoid the need for performance management procedures. It is another to say that a mandatory retirement age of 65 is appropriate and necessary to achieving this end. The means have to be carefully scrutinised in the context of the particular business concerned in order to see whether they do meet the objective and there are not other, less discriminatory, measures which would do so.”[81]Mr Short emphasised this last paragraph as being intended to refer to the requirement in the domestic cases that the tribunal make up its own mind, at any rate in relation to means, without according any margin of discretion to the employer, whether the employer was or was not a government department.[82]Mr Cavanagh relied on further decisions of the CJEU since Seldon which have reiterated paragraph 68 of Palacios which we have already cited; see in particular Specht v Land Berlin C-541/12; [2014] ICR 966 and Unland v Land Berlin C-20/13; [2015] ICR 1225 para 57.[83]Mr Short, by contrast, relied on Lockwood v Department of Work and Pensions [2014] ICR 1257 in which redundancy payments increased with age so that employees under 35 received significantly smaller sums than those aged 35 or over with the same length of service. This court held that Ms Lockwood had indeed suffered less favourable treatment on the grounds of her age but that the employment tribunal had been entitled to be satisfied that the disparate treatment was a proportionate method of achieving the aim underlying the scheme of providing a financial cushion until alternative employment could be found. There was no sign, said Mr Short, of any margin of discretion being afforded to the Department; the tribunal applied its own mind to the matter.

The law applicable to this case

[84]For our part we see no difference between the domestic authorities and the European authorities, let alone any attempt by the Supreme Court in Seldon to reconcile them. Nor do we consider the Equality Act 2010 imposes any greater burden on an employer in relation to age discrimination than does the Equality Directive. The terms of the statute derive straightforwardly from the Directive and they should be construed in the same way.[85]In that context it is axiomatic that the state or the government (if it is employer) must be accorded some margin of discretion in relation to both aims and means. As Lord Nicholls said in Seymour-Smith, governments must be able to govern. But it is for the tribunal in any particular case to determine what the appropriate margin is. This approach gives full force to Lady Hale’s statement in para 62 of Seldon that the means of achieving any particular aim must be carefully scrutinised by the factfinding tribunal. There is, in our judgment, no inconsistency between a tribunal carefully scrutinising a decision of government but nevertheless according government a margin of discretion when it comes to both aims and means. We do not see that Lockwood is in any way inconsistent with this approach since no argument seems to have been addressed to the court in that case in any way similar to the arguments we have received in this case.[86]But, as Lady Hale said in para 59 of Seldon, establishing that an aim is capable of being a legitimate aim is only the beginning of the story. It is for the tribunal then, according an appropriate margin of discretion, to decide whether it is legitimate in the circumstances of the case. For this purpose, an aim must at least be rational and, if it is not, the employment tribunal is entitled to say so. Margin of discretion cannot rescue an aim that is irrational.[87]We would therefore hold that, where government has a legitimate interest in any issue which arises in a discrimination claim, it is to be afforded a margin of discretion but it is for the fact-finding tribunal to assess both whether the government has such a legitimate interest and the amount of discretion it should be afforded and then the tribunal should decide the case itself in accordance with ordinary principles.

Application to the judgments below

[88]We would therefore respectfully differ from Sir Alan insofar as he held (as we think he did at any rate in the Sargeant case) that there was a margin of discretion in relation to aims but not in relation to means. But it is, of course, the decision of Judge Williams that has to be considered in the light of the principles we have set out above.[89]As to that it may first be observed that Judge Williams did accord the respondents a margin of discretion. This appears both from para 87 (cited in para 53 above) and from para 93 of his judgment. In the latter paragraph he is considering the claim of consistency which, of course, he ultimately rejects since the judicial pension scheme and other schemes were different in the respects he sets out. But he recognises that the advantages of certainty, fairness in the eyes of the public and ease of communication might all “be considered to be within the broad discretion accorded to member states to set the aims of their social policy.” He was thus completely alive to the need to accord a discretion to government to set aims of social policy in relation to employment. He just did not consider that the aims relied on stood up to scrutiny, whatever margin of discretion was to be afforded to government. His conclusions in paragraphs 86-87 which we have quoted betray no error of law and cannot be successfully impeached.[90]Secondly and more importantly Sir Alan did not detect any error of law on the part of Judge Williams in relation to any margin of discretion. He held that the approach of Judge Williams to the legitimacy of the aims was consistent with the ECJ/CJEU authorities and Seldon and upheld his adoption of the standard of rationality. He concluded that the appellants had not succeeded in demonstrating that the judge erred in law. We agree and endorse Sir Alan’s conclusion on this point, even though we have to an extent departed from Sir Alan’s analysis of the authorities.

Other errors of law detected by Sir Alan in the employment tribunal judgment

[91]We have already observed that Sir Alan did accept Mr Cavanagh’s submission that Judge Williams ignored the evidence which revealed why the government adopted the transitional provisions and wrongly criticised the appellants for failing to provide evidence in support of their aims which were moral or political and could not, in the nature of things, be supported by evidence. Sir Alan was particularly critical of Judge Williams’ approach to the expressed wish for consistency, saying that the judge characterised the evidence as no more than generalisations and that that was “a misdirection and/or misunderstanding of or misapplication of the facts and evidence”.[92]With every respect to Sir Alan we cannot agree with him about this. In our view Judge Williams was entitled to say that an aim which protected older judges rather than younger judges when the older judges needed it least was irrational and that (para 51) it did not help to expand that by saying that older judges would have less time to prepare for pension reform. It does not seem to us that it adds anything either to say that it was somehow fairer to older judges to give them protection which they needed less than younger judges. Judge Williams’ references to lack of evidence have to be seen in this light. The point is that there was just no evidence that older judges did need that protection more than younger judges. He did not criticise absence of evidence on the question of fairness because he was not asked to treat “fairness” as a separate issue, so no question of lack of evidence could arise. We permitted Mr Cavanagh to advance the argument on the basis that it was the other side of the coin of the perceived wish to protect the older judges rather than the younger ones. But if, as the judge held (and we agree) that wish was irrational, the point does not go anywhere.[93]As far as the wish for consistency is concerned while it is true that the judge did hold that the contention was based on generalisation rather than hard evidence, his real point was that, while consistency requires like cases to be treated alike, the Judges’ position was so different from those of other public servants that true comparisons could not be made.[94]Finally and importantly these matters are essentially for the employment tribunal judge to assess and this court (and indeed the Employment Appeal Tribunal) should, in our judgment, be slow to substitute our own judgment about the mass of evidence which there was before the tribunal for that of the employment judge. Moreover, as Lady Hale observed in Essop v Home Office [2017] UKSC 27; [2017] ICR 640, para 47, we must be able to detect an error of law. The quotations we have, at some length, made from the judgment of the employment tribunal in the previous paragraphs are a testament to his approach and the evidence; they speak for themselves as to the care and diligence with which he approached his task. We can detect no error of law.[95]We would therefore uphold Judge Williams’ conclusions on legitimate aim and it is, strictly speaking, unnecessary to go to consider whether Sir Alan was, in any event, right to uphold his judgment on proportionate means. We will, however, express our views shortly.

Proportionate means

[96]Mr Cavanagh submitted first that the judge’s errors in relation to margin of discretion in relation to legitimate aims meant that his decision on means should be set aside. Since we have not accepted that Judge Williams made any material error in his legal approach to aims, that no longer arises.[97]Secondly Mr Cavanagh submitted that the other errors which Sir Alan did find in the judge’s decision on aims affected his decision on means. That point does not now arise either but even if we had upheld Sir Alan’s decision on aims, we do not consider those other errors affected the judge’s decision on means. On this we agree with Sir Alan. Judge Williams made clear that the decision on means had to be made on the basis that his decision on aims was incorrect (see paras 112 and 113). There is no reason to suppose that Judge Williams was unable to do what he said he was doing.[98]Mr Cavanagh then said that there were in any event errors in the judge’s decision on means. In particular he focused on paras 119-121 in which the judge considered whether there were less discriminatory ways of achieving the aim of protecting older judges from the effects of the pension reforms. It was said that the judge fell into the trap of analysing different aims, such as treating all judges in the same way. There is perhaps a little force in Mr Cavanagh’s criticism but we do not consider it vitiates the substance of the decision. Like Sir Alan, we regard the criticised paragraphs as no more than an addition to the judge’s fundamental reason for finding disproportionate means namely that the discriminatory effect of the reforms on the younger judges went beyond what was reasonably necessary either to achieve consistency or to protect those closest to retirement (paras 114 and 118). This was based on all the factors contained in para 166 of Sir Alan’s judgment and referred to earlier in the judgment of Judge Williams.[99]It was then said that the judge took account of the process by which the decisions on pension reforms were reached as a reason for saying that the means adopted were not proportionate. It is true that the judge did criticise some ministers for failing to give separate consideration to the separate position of the judges but that was not of the essence in his decision on means which, as we have said, was based on the means used to favour the older judges and discriminate against the younger ones. To take but one example, the addition to the reforms, which in any event made the younger judges worse off in comparison with the older judges, of the requirement that those younger judges participate in a registered (rather than an unregistered) scheme with its extremely serious tax consequences was in itself a disproportionate means of achieving any legitimate intended aim. Judge Williams reached a decision on means which was open to him on the evidence and it should not be disturbed.

Conclusion on first appeal

[100]We would therefore uphold Judge Williams’ (and indeed Sir Alan’s) decision that the respondents have failed to show that their treatment of the claimants was a proportionate means of achieving a legitimate aim and dismiss this appeal.

Sargeant and Others v. The Secretary of State for the Home Department and Others

[101]We turn to the appeals concerning the transitional protection provisions of the new Firefighters’ Pension Scheme 2015 and the Firefighters’ Pension Scheme (Wales) 2015 (together “the 2015 Scheme”). Like those in relation to the NJPS, these provisions similarly gave full or tapering protection to firefighters approaching their Normal Pension Age (“NPA”) from all or part of the disadvantage they would otherwise suffer in moving from their current pension scheme to the 2015 Scheme. They provoked challenges on the grounds of age and other discrimination by more than 5,000 firefighters in England and Wales who were given either less protection than others or none. Five lead claims were heard by the London Central Employment Tribunal. All the claims were dismissed but the firefighters achieved limited success on their appeals to the Employment Appeal Tribunal. That led to further appeals to this court by both the firefighter claimants and the respondents.

The facts

[102]The claimants are members of the Fire Brigade Union (“the FBU”) and were formerly also members of either the Firefighters’ Pension Scheme 1992 (“the FPS”) or a special section of the New Firefighters’ Pension Scheme 2006 (“the NFPS”). The members of that section were “retained” (part-time) firefighters who were in service prior to April 2006 and we need say no more about them than that their membership of it gave them the like pension benefits as those provided by the FPS. (The main membership of the NFPS was of firefighters whose service commenced after 1st April 2006, when the FPS closed to new members. Their pension terms were less favourable than those of the FPS: they were comparable to those of the 2015 Scheme, which in some respects are more favourable. None of the main members of the NFPS have challenged either the 2015 Scheme or its transitional protection provisions).[103]The terms of the FPS were in the Firefighters’ Pension Scheme Order 1992. It was a defined benefit final salary scheme and a registered scheme for tax purposes. Its essential benefits were these:(a) an annual pension of 1/60th of final pensionable pay accrued during the first 20 years’ service and thereafter 2/60ths of such pay up to a maximum of 40 years’ accrual, so giving most members a full pension after 30 years’ service and an effective accrual rate of 1/45th of final pensionable pay for each year of active membership;(b) a lump sum of the lower of one quarter of the value of the pension by HMRC or Scheme Rules or 2.25 times the annual rate of pension payable on retirement between 50 and 55 without having accrued 30 years’ service, which was commutable on retirement at rates applicable according to the member’s age;(c) an NPA of 55, with an ability to retire from age 50 with no penalties provided that the member had accrued 25 years’ service (so giving an effective NPA for most members of between 50 and 55: a “rule of 75”); and(d) a deferred pension age of 60, when a member who had opted out of active membership, or left service before retirement, could take an immediate and unreduced pension.[104]The 2015 Scheme came into force on 1st April 2015. Like the NJPS, it was introduced under the Public Service Pensions Act 2013 which followed the Hutton Report’s recommended reform of public sector pensions. It was established by the Firefighters’ Pension Scheme (England) Regulations 2014 and like Regulations for Wales. Its terms are materially less favourable than the FPS. The main changes are that:(i) pension benefits are calculated on the basis of career average earnings rather than final salary;(ii) the annual accrual rate is 1/59.7 as opposed to 1/45th;(iii) the NPA is 60 as opposed to 55; and(iv) the deferred pension age is 65 or the state pension age if higher.[105]Paragraph 7.34 of the Hutton Report (see para 4 above) expressed the view that special transitional protection from the impact of a reformed pension scheme for older members of a public sector scheme should not be necessary and that age discrimination legislation anyway precluded it. The Government, however, took the different view expressed on 2nd November 2011 by the Chief Secretary to the Treasury in the Green Paper, in Parliament and in his letter to the TUC General Secretary (see paras 6 to 8 above). Its intention was instead to provide that those who, on 1st April 2012, were within ten years of their NPA under their current pension scheme should be protected from any change in their pension arrangements and that tapering protection should be given to those then within 14 years of their NPA.[106]As with the NJPS and other new public sector schemes, this intention was carried out in relation to the 2015 Scheme. The Scheme’s Regulations included transitional protection provisions giving preferential treatment to certain members of the FPS (and of the special section of the NFPS), the measure of protection depending on the date they would reach their NPA of 55 under the FPS. The effect of the provisions was to create three groups of FPS (or special section NFPS) members who, as from 1st April 2015, were destined to be treated differently according to their age. They were as follows:-(a) Active members born on or before 1st April 1967. They would have been 45 or older on 1st April 2012 and so within ten years of their NPA under the FPS. They were given full protection, meaning they enjoyed continuing active membership of the FPS without limit of time.(b) Active members born after 1st April 1967 but before 2nd April 1971. They would have been between 41 and 45 on 1st April 2012 and so within 14 years of their NPA under the FPS. They were given tapered protection, meaning they remained active members of the FPS for an additional 53 days for each month by which their age on 1st April 2012 was over 41.(c) Active members born after 1st April 1971. They received no protection and were automatically transferred to the 2015 Scheme in respect of all pensionable service from 1st April 2015.[107]These provisions were discriminatory on age grounds as between the three groups. Those in group(a) were treated manifestly more favourably than those in groups (b) and (c). Those in group(b) were treated more favourably than those in group (c), but disadvantageously compared with those in group (a). Those in group (c), with no protection, were treated disadvantageously compared with those in group (b) and even more so in comparison with those in group (a). Such discrimination was unlawful unless the Governments of England and Wales (its promoters, “the Governments”) could show that the treatment of groups (b) and (c) was a proportionate means of achieving a legitimate aim (section 13(2) of the Equality Act 2010).

The claims

[108]Five lead cases were heard by the Employment Tribunal. Ms Sargeant, Mr Bebbington and Mr Bygrave were the lead English claimants; Mr Dodds and Ms McEvoy, the lead Welsh claimants. The claims were against(i) their respective Fire and Rescue Authority (“FRA”) employers,(ii) by the English claimants, also against the Secretary of State for Communities and Local Government, and(iii) by the Welsh claimants, also against the Welsh Ministers. The Secretary of State for the Home Department was later substituted for the original English Government respondent. The Secretary of State and the Welsh Ministers are the ministers respectively responsible under the Public Services Pensions Act 2013 for establishing schemes for the payment of pensions to firefighters in England and Wales.[109]Ms Sergeant, born on 1st September 1976, had no transitional protection. She is of minority ethnic origin. Her claims were for(i) direct age discrimination,(ii) equal pay,(iii) indirect sex discrimination, and(iv) indirect race discrimination. Her age discrimination claim is self-explanatory. Her equal pay claim asserted that, compared with older male firefighters, she was doing equal work but receiving less pay by reason of her reduced pension entitlement. Her indirect sex and race discrimination claims were based on the assertion that women and those of ethnic minorities were formerly under-represented in firefighting, but had increased over time, and that the use of age as a determinant of the right to transitional protective benefits was disproportionately to exclude women and those of ethnic minorities from their enjoyment.[110]Mr Bebbington, born on 21st March 1983, also had no transitional protection. His claims were for direct age discrimination and equal pay. The latter was a so-called “piggy-back” claim: its success depended upon a comparator woman succeeding in her equal pay claim. Mr Bygrave, born on 17th August 1972, was given tapered protection. He was a retained member entitled to the benefits of the special section of the NFPS. His claims were for age discrimination and a piggy-back equal pay claim.[111]Mr Dodds, born on 29th May 1980, had no transitional protection. He is of minority ethnic origin. His claims were for direct age discrimination, equal pay (on a piggyback basis) and indirect race discrimination. Ms McEvoy, born on 4th July 1976, also had no transitional protection. Her claims were for direct age discrimination, equal pay and indirect sex discrimination.

The decision of the Employment Tribunal

[112]The claims were heard over five days in January 2017 by Employment Judge Lewzey. Her reserved judgment and reasons were dated 14th February 2017. She held that the transitional protection provisions were a proportionate means of achieving a legitimate aim and dismissed the age discrimination claims. She also dismissed the equal pay and indirect discrimination claims. Judge Williams’s decision (to different effect) in the judges’ case had been given on 16th January 2017 and Judge Lewzey was referred to it but said that she had disregarded it in deciding the firefighters’ claims. The appeals before us are against the decision of the Employment Appeal Tribunal, but as our primary focus must be on whether Judge Lewzey made any error of law we shall explain her decision at some length. We shall deal first with her decision on age discrimination and give our decision on the appeals relating to that. We shall then deal with those relating to the equal pay and indirect discrimination claims.[113]Judge Lewzey said there were no significant disputes of fact and much common ground. She quoted para 7.34 of the Hutton Report (para 4 above). She summarised the English Government’s different view that, in implementing the reform to public sector pensions, it should give transitional protection to those closest to retirement, and she quoted from the three 2nd November 2011 references to that by the Chief Secretary to the Treasury (paras 6 to 8 above). She summarised the terms of the FPS and referred to the Government’s discussions with TUC representatives from 2011, which were focused on the four largest public sector pension schemes (for the NHS, Local Government, Teachers and the Civil Service and representing 82% of the membership of the Public Service Pension Schemes in Great Britain and Northern Ireland). She referred to the discussions between the Government and the FBU. The latter’s preferred scheme was that:- “31. … members of the FPS should maintain their current entitlement. They sought protection for all members of [the FPS]. The FBU did not take a direct part in the central negotiations which focussed on the four largest public sector schemes. The FBU wanted the best possible deal for its members. A particular concern of the FBU was that firefighters who had been recruited on the basis that they could take their pension between 50 and 55, might not be able to work to the new pension age of 60. Mr Starbuck [an FBU national officer, who had made two witness statements] explained that as firefighters get older it becomes harder to maintain their cardiovascular fitness. Mr Starbuck said that a person who cannot maintain that fitness, but who does not meet the criteria for an ill health retirement, is left with the choice of leaving the FRA with a deferred pension or drawing their pension early with an actuarial reduction. 32. When the NPA was fixed at 60 by the [Public Service Pensions Act 2013], the FBU sought to obtain the best possible deal for its members by seeking additional protection for those who were within ten years of their expected retirement and mitigating the detriment for unprotected firefighters who were in fact obliged to retire early. The FBU attempted to deal with their concerns in their negotiations with the DCLG and the devolved administrators.”[114]Judge Lewzey noted that the FBU at no stage agreed with the proposed transitional provisions, which led to a trade dispute and industrial action. At paras 45 to 48, she gave a short summary of counsel’s submissions, including Mr Cavanagh’s for the Governments that as the aims were social policy aims implemented by them, there should a less intrusive review by the tribunal, that the standard of scrutiny in age discrimination was lower than in other types of discrimination and that aims similar to those in question had been approved by European and United Kingdom courts. She set out the legislation relating to the age discrimination claims, including section 18(5) of the Public Service Pensions Act 2013, which permitted the making of protective measures of the type made in this case (no-one suggests, however, that it permitted measures that were unlawfully discriminatory). She then explained her conclusions on the age discrimination claim, her discussion including references to counsel’s respective submissions.[115]Judge Lewzey noted (para 52) that it was agreed that the claimants were being paid less than their comparators (pension payments are deferred pay) and that the complaint was only about the transitional provisions, not that the 2015 Scheme was itself discriminatory. She said (para 55) that there was an issue of fact as to why the transitional provisions of the 2015 Scheme were adopted. The answer turned on the documentary evidence and she made her findings (in favour of the appellants’ i.e. the Governments and FRAs) in paras 66 to 68, to which we shall come.[116]At para 57, Judge Lewzey embarked upon the first issue before her, namely whether the transitional provisions had a “legitimate aim” (section 13(2) of the Equality Act 2010). She recorded the difference between Mr Short (for the claimants) and Mr Cavanagh as to the degree of scrutiny she had to apply in considering the justification for the discriminatory treatment as between the three groups of firefighters. Mr Short’s submission was that it was as explained in Hardy & Hansons plc v. Lax[2005] ICR 1565, which required the court to make its own judgment, without according the employer any margin of appreciation or range of reasonable responses (we referred to that authority at para 72 above). Mr Cavanagh’s submission, relying on decisions of the CJEU and of the Supreme Court in Seldon v. Clarkson Wright &Jakes [2012] ICR 716 (which we discussed at paras 78 to 80), was that as the relevant aim was a social policy decision by a state, the exercise of objective justification did require the court to recognise a margin of discretion on the part of the state.[117]At para 61, Judge Lewzey identified what the Governments and FRAs said were the aims of the transitional protection provisions. She said:- “The aims have been identified in a number of ways. In the list of issues set out at paragraph 11.1 the aim is expressed to be:- “… to protect those closest to pension age and to retirement from the effects of pension reform.” “… to protect those closest to pension age and to retirement from the effects of pension reform.” In his closing submissions, Mr Cavanagh identifies the aims 119. as:-(1) “To protect those closest to pension age from the effects of pension reform, since they would have least time to rearrange their affairs before retirement, by making lifestyle changes or alternative financial provision (or by finding alternative employment);(2) To take account of the greater legitimate expectation that those closer to retirement would have that their pension entitlements would not change significantly when they were close to retirement.(3) To have a tapering arrangement so as to prevent a cliff edge between Fully Protected and Unprotected Groups.(4) In achieving these substantive aims behind the transitional provisions, the UK Government sought to ensure that a clear and simple message could be communicated, and that there was consistency across the public sector.” 120. Mr Lynch for the FRAs identifies the legitimate aim in their ET3 at paragraph 9 … as:- “The transitional provisions recognise that the nearer in time a firefighter was to reaching his or her Normal Pension Age, the more difficult it was likely to be to adjust to the move to the 2015 scheme. This is because the firefighters who were near a Normal Pension Age had less time to make the necessary changes to lifestyle and less time to put in place appropriate financial adjustments to accommodate the transfer to different pension provisions than was the case for firefighters whose Normal Pension Age was temporally more distant”. 121. On 26 October 2016, the FRAs filed voluntary further and better particulars adopting the aims in the agreed list of issues … as the aim.”[118]At paras 66 to 68, Judge Lewzey made her finding as to why the transitional protection was adopted. She said:- “66. I have considered whether there were real aims. The Hutton Report sets out the aims but did not recommend transitional protections. The Chief Secretary to the Treasury articulated the transitional protections, in his letter to Brendan Barber, TUC General Secretary dated 2 November 2011 … that those closest to retirement should not suffer any detriment, either as to when they can retire, or any decrease in the amount of pension they receive at NPA. The protection was provided to those who were within ten years of NPA on 1 April 2012 and there was also scope for tapering for three to four more years. The cost of the transitional protections was outside the costs ceiling and therefore did not need to be offset by reductions elsewhere in the pension schemes. 67. It was the decision of the Chief Secretary to the Treasury who took the decision to provide protection across the public sector for those within ten years of NPA, with a taper for three to four years. The policy originates from concessions within the Treasury concerning changes to the State Pension Age, in respect of which a ten year notice period was applied after extensive consultation. Mr Kelly gave evidence about the Treasury decision making process at paragraphs 44 to 56 of his witness statement. The Command Paper entitled “Public Service Pensions: Good Pensions That Last” (6/4851 – 4879) explained the rationale. In the Forward the Chief Secretary records:
‘I believe it is right that we protect those public service workers who as of 1 April 2012 have ten years or less to their pension age. It is my objective that these people see no change in when they can retire, nor any decrease in the amount of pension they receive at their current Normal Pension Age. On presentation of the Command Paper to the House of Commons on 2 November 2011 further detail was provided. The extract from Hansard sets out the matter in more detail as set out in paragraph 23 above [the material additional detail was as to the willingness of the Chief Secretary to the Treasury ‘to consider tapering of transitional protection over a further three to four years.]’ 68. The Welsh situation is slightly different. Mr Pomeroy [Head of the Fire Services Branch of the Education and Public Services Group within the Welsh Government] explained this. The [Public Service Pensions Act 2013] constrained the Welsh Government. The Welsh firefighters had more advantageous early retirement factors but a worse accrual rate. The decision was taken to adopt the same transitional protections.” ‘I believe it is right that we protect those public service workers who as of 1 April 2012 have ten years or less to their pension age. It is my objective that these people see no change in when they can retire, nor any decrease in the amount of pension they receive at their current Normal Pension Age. On presentation of the Command Paper to the House of Commons on 2 November 2011 further detail was provided. The extract from Hansard sets out the matter in more detail as set out in paragraph 23 above [the material additional detail was as to the willingness of the Chief Secretary to the Treasury ‘to consider tapering of transitional protection over a further three to four years.]’

In his closing submissions, Mr Cavanagh identifies the aims

[119]Mr Short was critical of what Judge Lewzey said in the second sentence of para 67, in particular of her failure to examine the aptness of the lesson said to be derived from the changes to the state pension age. He said that, unlike Judge Williams in paras 48 and 49 of his reasons in the judges’ case, Judge Lewzey applied no scrutiny to the point and did not, as Judge Williams there did, identify its inaptness.[120]At para 69, Judge Lewzey recorded Mr Short’s submission, supported by references to six authorities, that in identifying what is said to be a legitimate aim, it is not enough merely to point to a decision favouring one group over another; it must be shown why a particular age group is being favoured. The submission was that:- “… the Respondents cannot establish a legitimate aim corresponding to a real social need to a high standard of proof, unless they can also show that those nearer retirement and Normal Pension Age were in greater need of protection. … it is insufficient to say that those nearest to retirement have less time to adjust. The amount of time that a person has to adjust is the period of time until Normal Pension Age and the older the person is, the closer they are to that age. It is also common ground that the closer someone was to retirement, the less change he or she would face, and less adjustment would be required. He argues that the Respondents must explain with precision the nature of the lifestyle changes and alternative financial provision and establish why the more limited amount of time to make those changes gives rise to a real social need.”[121]Judge Lewzey did not suggest that the Governments’ evidence provided an explanation of the sort Mr Short said was required. Nor did it. She responded to the submission in more general terms. She said:- “70. … The protected group were treated more favourably because of proximity to retirement. Whilst retirement is age related, and proximity to retirement is connected with age, there may be good reasons for treating different age groups differently. Mr Cavanagh relies on Seldon, where the measure complained about was a compulsory retirement age. Those below that retirement age were treated more favourably than others because of their age, but the objective justification defence succeeded. 71. In the case of firefighters, the decision maker was the Chief Secretary to the Treasury who decided to make a more generous provision to the public sector workers than had been recommended by Hutton. … It is clear to me that the transitional provisions that were envisaged by primary legislation were age related transitional provisions which protected those closest to Normal Pension Age. The evidence is that the decisions were taken with great care and after negotiations with the representatives of the Unions. There were detailed negotiations with the TUC and, in relation to the firefighters’ pension schemes, the FBU was involved in negotiations with the DCLG and the Welsh Government. The evidence is that the DCLG and Welsh Government took seriously the representations made by the FBU. 73. Mr Short argues that the reforms have less impact upon older firefighters than younger firefighters and that the suggestion that younger firefighters can make good the effect of the pension reforms by applying some of their salary towards retirement is in fact saying that that the younger firefighters can choose when they experience the disadvantage of being paid less than the older firefighters. He argues that in any event it is wholly unrealistic given the sums in question. The suggestion that older firefighters would be less able than the younger firefighters to make changes in their expenditure leading up to retirement is unsupported by evidence. His argument is that the closer the scheme members are to retirement, the less they would be affected by the reforms.”[122]Mr Short’s point as to the inability of younger firefighters to apply part of their salary to acquiring investments that would make up the difference between their pension expectations and the provision that would be enjoyed by the protected groups is important. The agreed expert evidence was that a full-time firefighter too young for transitional protection would need to make a yearly capital investment of between about £16,000 and £19,000 to provide an annuity giving approximately the same benefits as those earned each year by older comparators with transitional protection. As the basic gross pay of such a firefighter was just under £30,000, such provision would be unachievable.[123]Judge Lewzey turned, at para 73, to consider the case law. She referred to three CJEU authorities affirming the broad discretion that member states enjoy in choosing to pursue a particular aim in the field of social policy and in defining the measures to implement it: Mangold v. Helm Case-144/04 [2006] 1 CMLR 43; HK Danmark v.Experian A/S Case-476/11[2014] ICR 27; and Rosenbladt OellerkingGebaudereinigungsges Case-45/09[2011] IRLR 51. She said (para 75) that “the decisions … under examination in the present case are decisions of the elected Government. They are social policy choices which may well have a political element.” She cited from Palacios de la Villa Case-411/05[2009] ICR 1111, including para 71 of the court’s judgment:- “It is, therefore, for the competent authorities of the member states to find the right balance between the different interests involved. However, it is important to ensure that the national measures laid down in that context do not go beyond what is appropriate and necessary to achieve the aim pursued by the member state concerned.” She said (para 76) that on these authorities, “it [is] for the Member state to balance the different interests and I must be careful not to substitute my own view for that of the Government.” She said it was clear that member states enjoy a broad discretion in the choice of both aims and means, and she cited the passage from Lord Nicholls of Birkenhead’s speech in R v. Secretary of State for Employment Secretary, Ex ParteSeymour-Smith and Another (No 2)[2000] 1 WLR 435, at 450F, that we cited at para 71 above. She also cited from paras 28, 33, 50, 53 and 55 of Lady Hale’s judgment in Seldon. She did not, however, cite from paras 59 and 61. As we shall explain, that was an unfortunate omission.[124]At para 80, Judge Lewzey agreed with Mr Cavanagh that, for the reasons he had given, the Hardy & Hansonsstandard of objective justification was not applicable to the nature of the scrutiny a court must apply in relation to an issue as to the objective justification of a social policy measure adopted by a state; and, at para 83, that “the correct test to be applied is the test set out in Seldonin social policy cases following the CJEU approach.” She continued:- “… This is a situation where a Member State was introducing a measure as a result of having made a social policy decision to protect those within 10 years of retirement. The Government has a wide discretion in social policy matters. The standard of scrutiny involves granting a wide margin of discretion to the Member State. I am satisfied that this is the correct standard in the present case and the stricter test which applies to operational matters of a private sector employer is not the correct test.” At para 88, after referring to R (Lumsdon and Others) v. Legal Services Board [2016] 129. AC 697, she said “I must take care not to substitute my view of the social policy issues for a view of the Member State.”[125]At para 90, Judge Lewzey recorded this submission by Mr Cavanagh:- “Mr Cavanagh argues that those closest to retirement have a greater legitimate expectation that things would not change in a significant way when they are only a few years away from retirement as compared with those who are earlier in their career. His submission is that a person in the early part of his or her career is not focused on, or concerned about their pension because retirement is a long way off and there may be changes to their careers or personal circumstances. He suggests that someone closer to Normal Pension Age is focused on their pension entitlement and has a legitimate expectation that their pension will stay as it is with no sudden changes in the last years before retirement.” That passage provoked criticism by Mr Short before us. He said (correctly) that there was no evidence supporting the submission in the first sentence; and that what was said in the second sentence as to the supposed attitudes of the young towards pension provision was unjustified stereotyping.[126]At para 95, Judge Lewzey returned to the question of the extent to which the state’s declared social policy aim needed to be based on solid evidence. She recorded Mr Short’s submission that the Governments’ social policy choice in the present case was not based on precise or concrete factors. All members of the FPS were given three and a half years’ notice of the impending changes, yet the respondents had not identified longer term plans that would have been made by the older members, but not by the younger ones, which would have been disrupted by the changes. Further, as the transitional protections were tied to the NPA rather than the expected date of retirement, the Governments’ policy gave little weight to actual expectations in any event (Mr Short’s point there was that the FPS’s “rule of 75” enabled many firefighters to retire between 50 and 55). Judge Lewzey said, however, that Mr Kelly’s evidence had been that in the consultation concerning the state pension agethe evidence was that people would finalise their plans the closer they got to retirement.[127]Between paras 91 and 94 Judge Lewzey referred to two authorities that she regarded as of assistance in considering the legitimacy of the aims. The first was R (Unison) v.First Secretary of State [2006] IRLR 926, although it is not clear what principle she derived from it that assisted her conclusion on the legitimacy issue. The other, Commission v. Hungary Case C-286/12, is one that Judge Lewzey regarded as demonstrating that “EU law recognises those nearest retirement who face a sudden change require transitional provisions to require time to adjust.”[128]As for the need for a social policy aim to be supported by evidence, Judge Lewzey cited para 56 from the judgment of Lords Reed and Toulson in Lumsdon [2016] AC 697, which we cited at para 61 above but here repeat:- “The justification for the restriction tends to be examined in detail, although much may depend upon the nature of the justification, and the extent to which it requires evidence to support it. For example, justifications based on moral or political considerations may not be capable of being established by evidence. The same may be true of justifications based on intuitive common sense. An economic or social justification, on the other hand, may well be expected to be supported by evidence ….”

At para 88, after referring to R (Lumsdon and Others) v. Legal Services Board [2016]

[129]Judge Lewzey, at para 97, summarised as follows her conclusion as to the need for supporting evidence:- “The authorities suggest that need for precise and concrete factors depends on the nature of the justification. The government relies on the fact that those in the protected group were closer to retirement. Political considerations may have played a part in the Government’s decision. For those reasons I reject the criticism that the Government’s decision was not based on precise or concrete factors. The fact that the Scottish Government adopted a different measure, that is drew the line in a different place, shows that these are social policy matters for which there is no right or wrong answer and the choice is that of the Government.”[130]The final matter to refer to before coming to Judge Lewzey’s conclusions on legitimate aim is what she said about a fitness issue in relation to firefighters:- “101. There is a final matter to be considered in relation to the legitimacy of the aims and that relates to the fitness issue. Mr Lynch has made a number of submissions concerning fitness and the report of Dr Williams. He argues that a central reason for refocusing protection on the older firefighters is that they have the least ability to change their lifestyles and circumstances to accommodate the changes to the Normal Pension Age. This involves a consideration that it is the older firefighters who face the greatest difficulties in maintaining their fitness and weight. Mr Lynch argues that the new fitness regime is something that should be taken into account because the protection arrangements focusing on older firefighters allow for the new fitness structures to come into effect before the change is made to the firefighters’ NPA. Retaining health and fitness is more difficult for older firefighters who will soon be 55 and otherwise facing being obliged to work until 60. The evidence for this is contained within the fitness report of Dr Williams. … The FBU was extremely concerned about issues concerning fitness and the possible disadvantage to a firefighter who became unfit and thus no longer able to work and needed to take early retirement. 102. The Williams Report … used a standard of fitness based on cardiorespiratory figures and VO2 42 max. The Williams Report noted that there were limitations to this standard. The report noted that a number of FRAs used a lower standard of VO2 35 max and expressed the view … that 100% of firefighters would be able to work until 60 years of age with such a standard. Whilst the report also says that if the VO2 42 max standard was used there would be some firefighters who could not meet it, although the majority would be able to regain their fitness with appropriate training. Mr Lynch submits that the protective arrangements mean that those who might be in difficulty in terms of benefiting from new policies and structures because the new structures are not in place and will need to be effective are protected from the need to work beyond 55.”

Judge Lewzey’s conclusion on “legitimate aim” was as follows:

[131]- “104. Having undertaken the analysis set out above, I am satisfied that the correct test for me to apply in determining the legitimate aims is to be determined by the approach to scrutiny laid down by the ECJ and the Supreme Court in Seldon. There is a wide margin of discretion for the Member State. On the evidence before me I am satisfied that the Respondents have demonstrated that the aims were to protect those closest to pension age from the effects of pension reform; to take account of the greater legitimate expectation that those closer to retirement would have that their pension entitlements would not change significantly when they were close to retirement; to have a tapering arrangement so as to prevent a cliff edge between fully protected and unprotected groups; and that there was consistency across the public sector. 105. It is my decision that the Respondents have demonstrated these aims.” She thus found the aims to be essentially as Mr Cavanagh had submitted them to be, as recorded by her at para 61 (see para [117] above); and, apparently, that they were also legitimate aims.[132]Judge Lewzey moved to the proportionality of the “means” adopted by the transitional protection provisions for achieving the aims. She referred to the threestage test for such an issue that was adopted as correct by the Privy Council in Elloyde Freitas v. Permanent Secretary of Ministry of Agriculture, Fisheries, Lands and Housing and others [1999] 1 AC 69, at 80G:- “… whether:(i) the legislative objective is sufficiently important to justify limiting a fundamental right;(ii) the measures designed to meet the legislative objective are rationally connected to it; and(iii) the means used to impair the right or freedom are no more than is necessary to accomplish the objective.”[133]As to the first consideration, Judge Lewzey noted that it involves balancing the need to achieve the aim against the impact of the means used to achieve it. She referred to Lady Hale’s observation in Seldon, at para 50(6), that:- “The gravity of the effect upon the employees discriminated against has to be weighed against the importance of the legitimate aims in assessing the necessity of the particular measure chosen (Fuchs).”[134]As to the second consideration, she referred to the Opinion of the Advocate General in Age Concern England Case C-388/07 [2009] 3 CMLR, at paras 86 and 87, indicating that it is for the member state “to find the right balance between the interests involved, provided the requirements of proportionality are respected”, and suggesting that member states are left a relatively wide discretion in identifying the means; and cited from para 51 of the court’s judgment in Age Concern England:- “In that connection, it must be observed that in choosing the means capable of achieving their social policy objectives, the Member States enjoy broad discretion (see to that effect Mangold paragraph 63). However, that discretion cannot have the effect of frustrating the implementation of the principle of non discrimination on grounds of age. Mere generalisations concerning the capacity of a specific measure to contribute to employment policy, labour market or vocational training objectives are not enough to show that the aim of that measure is capable of justifying derogations from that principle and do not constitute evidence on the basis of which it could reasonably be considered that the means chosen are suitable for achieving that aim.”[135]As for the third element of the inquiry, Judge Lewzey noted that the right approach had been formulated in different ways in the CJEU authorities.[136]In coming to her decision on proportionality of means, she said that, having found that it was a legitimate aim to protect those closest to retirement:- “111. …it follows that the place where the line was drawn was a matter of social policy choice. … The line was drawn ten years from Normal Pension Age with a four year taper. This was consistent with the rest of the public sector. The FBU would have preferred the line to be drawn elsewhere so that all member of the FPS were protected. That is a preference. The Government made a social policy choice which it applied across the public sector that those within ten years of NPA had protection to which was added a four year taper. …115. It is clear to me on the case law that there has to be a line drawn at some point. That is a social policy choice and inevitably some individuals will be disadvantaged. The FBU put forward the arguments in negotiation that the starting point for the transitional provisions should have been when a firefighter would have qualified for a full unreduced pension. Had this been agreed, the transitional provisions would still have protected those closest to retirement with a different cutoff date. Indeed, the Scottish Government took account of the length of service in their transitional provisions, but nonetheless the FBU is pursuing claims for discrimination in Scotland. Mr Starbuck’s evidence was that the FBU had noted that the Police had obtained an improved position and, thus, sought to obtain improvements for the firefighters.116. It was reasonably necessary for the Government to draw the line at some point. I am satisfied that the Respondent have demonstrated a legitimate aim and having considered the three stage test, I am satisfied that that aim was proportionate.117. In these circumstances it is my judgment that the treatment of the Claimants by the transitional provisions included in the Firefighters Pension Scheme 2015 are a proportionate means of achieving a legitimate aim and the claims of direct age discrimination fail.”[137]Judge Lewzey then dealt with the separate equal pay and indirect sex and race discrimination claims, all of which she also dismissed.

The decision of the Employment Appeal Tribunal

[138]The claimants’ appeals to the Employment Appeal Tribunal against Judge Lewzey’s decision were heard by Sir Alan Wilkie. He heard them immediately after the appeals in the judges’ case. Sir Alan delivered separate judgments in each case on 29th January 2018. His judgment in the firefighters’ case refers to certain of his holdings in the judges’ case and was apparently written after it. It included a thorough summary of Judge Lewzey’s reasoning and decision on legitimacy of aims and proportionality of means.[139]The firefighters’ argument was that Judge Lewzey had been wrong, in relation to both issues, not to apply an objective test mirroring that applied in Hardy & Hansons. She was said to have been wrong to have had regard to the Governments’ margin of discretion rather than to carry out the justification assessment herself. She was anyway said to have been wrong in failing to apply any proper scrutiny to the claimed legitimacy of the aims when the evidence in support was no more than mere generalisations. She was said similarly to have failed to apply any proper scrutiny in relation to proportionality.[140]In deciding the correct approach for Judge Lewzey to have adopted, Sir Alan referred (paras 67 and 68) to his consideration of the like question in the judges’ appeals. We have expressed our view (para 88 above) that, in the firefighters’ appeals, Sir Alan held that there was a margin of discretion in relation to aims but not in relation to means. As to Judge Lewzey’s decision on aims, he held that she had directed herself correctly, had exercised sufficient scrutiny and had come to a decision to which she was entitled to come. He said:- “75. In my judgment, the Employment Judge was correct in following Seldon in that part of her decision. It involved considering whether there were legitimate aims by reference to, and giving effect to, the state’s margin of discretion in pursuing and implementing social policy. The Employment Judge, in her analysis of the evidence, the facts and the arguments, was entitled to conclude that the Respondents had established legitimate aims. In my judgment looking at this part of the judgment as a whole, the Employment Judge understood the facts, considered and applied the correct legal principles and came to a conclusion to which she was entitled to come. I reject the contention that, looking at his part of the judgment as a whole, she failed to exercise sufficient scrutiny. She had well in mind the fact that it was a high test, but there was a margin of discretion and, in her careful exposition of the facts, the law and her conclusions, she did not, in my judgment, err in coming to the conclusion that she did.”[141]Having upheld Judge Lewzey’s decision on “legitimate aims”, Sir Alan considered her decision on proportionality of means. He said:- “80. In my judgment, in considering whether the means adopted by the Government were proportionate in order to achieve the legitimate aims, the Employment Judge was entitled to have regard to the fact that the Government was seeking to implement a social policy and that questions of consistency of application were significant. In so doing she was following what she took to be the approach identified in Seldon and to that extent she cannot be faulted.81. She did, however, have to grapple with the issue posed by the Claimants, namely that, comparing the protected group with the unprotected group, the differential between the two was said to be catastrophic and unfair to the unprotected group. This contention was made in the context of the Claimants’ submission of law that, when considering the issue as applied to them, the ET ought, additionally, to make up its own mind on the question of proportionality, applying the established domestic law principles described in Hardy and Hansons, MacCulloch and Lockwood.82. The Employment Judge did not do so. Her conclusion was that it was sufficient for her to consider the issue of proportionality by reference solely to the approach identified by the ECJ/CJEU line of authorities and, as she understood it, approved by the Supreme Court in Seldon. It also appears that she did not fully appreciate that, although the differential of which the complainants complained amounted to their being subject to the changes in the pension scheme by reason of the Government’s implementation of the pension reforms, which were not in issue, their complaint was of differential treatment by not being granted the full protection against such changes granted to those who were older than they were but denied to them by reason of their age. It followed that, in practice, their complaint was that they were subject to the changes in the pension scheme whereas they should have been protected from those changes and that this failure was unlawful age discrimination.83. In my judgment, in this limited respect, the Employment Judge erred in law. She failed to appreciate that in Seldon, whilst the Supreme Court had given effect to the approach of the ECJ/CJEU in applying article 6(1), both in respect of legitimate aim and means (see paragraph 55 of Baroness Hale in the Supreme Court) the Supreme Court had gone on, in paragraph 59 and following, to require that the means be carefully scrutinised in the context of the particular business concerned, in order to see whether they met the objective and whether there were not other, less discriminatory, measures which would do so. In particular, she failed to recognise that, on the issue of proportionate means of achieving a legitimate aim, the Court of Appeal in Lockwood had said that the judgment in MacCulloch had provided comprehensive guidance on the application of the test and the rigour with which tribunals must apply it (at paragraph 46). In my judgment, the Employment Judge erred in law in failing to consider whether, in the context of the Firefighters’ Pension Schemes, the application of the transitional provisions and the differential treatment on the grounds of age was a proportionate means for achieving what she had concluded, were, and was entitled to conclude, legitimate aims of social policy.”[142]The outcome was, therefore, that Sir Alan upheld Judge Lewzey’s decision on legitimate aim but held she had erred in law in the proportionality assessment. He then dealt with the claimants’ appeals against her dismissal of their equal pay and indirect sex and race discrimination claims. He dismissed the appeals against the dismissal of the equal pay and associated piggy-back claims and dismissed the indirect sex discrimination claims. He remitted for re-hearing by Judge Lewzey the proportionality of means issue in the age discrimination claims and in the indirect race discrimination claims, but he stayed such remittal until after the final determination by this court of the Governments’ and the FRAs’ appeals (which he permitted) against his decisions on proportionate means in the age discrimination claims and on the indirect race discrimination claims. He also permitted the claimants to appeal against his decision on the equal pay claims. He refused them permission to appeal on legitimate aim, saying they could file a respondents’ notice on that in response to the Governments’ and FRAs’ appeals on proportionate means. In the event, this court permitted the claimants to appeal on the legitimate aim issue. We have, therefore, had cross-appeals against Sir Alan’s decision on the age discrimination issues. We now consider them.

The cross-appeals on the age discrimination issues

[143]We start by referring to what we said earlier (paras 84 to 87) as to the correct approach in claims of the present nature to the determination of legitimacy of aims and proportionality of means. In brief repetition, where, as here, the decision giving rise to the alleged discriminatory treatment is made by a state’s Government, an employment tribunal must, first, when determining whether the aim was a potentially legitimate one, accord an appropriate margin of discretion to the decision-making authority. But, as Lady Hale said in Seldon, at para 59, that a particular aim is capable of being a legitimate one “is only the beginning of the story”: as she then also said, at para 61 (to which Judge Lewzey did not refer), it still has to be asked whether the aim “is legitimate in the particular circumstances of the employment concerned.”[144]In our judgment, that latter part of the exercise requires the tribunal to make an objective assessment; whilst according the appropriate margin of discretion to the Government in its decision to pursue a particular social policy aim, the tribunal must still be satisfied itselfas to the legitimacyof the aim in the particular circumstances of the employment affected by it. If so satisfied, when the tribunal comes to assess the proportionality of the means of achieving such aim, it must again itself be satisfied as to their proportionality, although in making its assessment it must again accord a margin of discretion to the Government in its decision as to means.[145]With respect to Sir Alan, we consider that his self-direction in the firefighter appeals in relation to aims overstated the accord to be given to the margin of discretion and overlooked the need for the tribunal to be objectively satisfied as to their legitimacy; and, in relation to means, we consider he did not sufficiently recognise that the tribunal must, whilst determining for itself their proportionality or otherwise, again accord an appropriate margin of discretion.[146]We turn to the age discrimination appeals, upon which we had arguments from Mr Short for the claimants, Mr Lynch for the FRAs and Mr Cavanagh for the Governments. It is common ground that the Governments’ aim to provide protection (full or tapered) to older firefighters and none to younger firefighters was a social policy aim and so was potentially capable of being a legitimate aim whose implementation, if objectively justified, would not constitute direct discrimination against younger unprotected (or less protected) firefighters. The first task for Judge Lewzey was to decide whether the Governments’ aim to protect the older firefighters was a legitimate aim. We deal first with her decision as to that, which is challenged by the firefighters as having involved an error of law.[147]Judge Lewzey identified the aims asserted by the Governments at para 61. For convenience, we repeat them:- “(1) To protect those closest to pension age from the effects of pension reform, since they would have least time to rearrange their affairs before retirement, by making lifestyle changes or alternative financial provision (or by finding alternative employment); (2) To take account of the greater legitimate expectation that those closer to retirement would have that their pension entitlements would not change significantly when they were close to retirement. (3) To have a tapering arrangement so as to prevent a cliff edge between Fully Protected and Unprotected Groups. (4) In achieving these substantive aims behind the transitional provisions, the UK Government sought to ensure that a clear and simple message could be communicated, and that there was consistency across the public sector.”[148]Aim (1) was broadly as pleaded by the Governments in their Grounds of Resistance. The pleaded case as to the full ten year protection was, in para 31:- “… because those who are 10 years or less away from their [NPA] will have less time to make any necessary changes to their lifestyle and plans for eventual retirement, than those with longer still to serve. So, for example, those closest to [NPA] will have less time in which to make additional provision to supplement their pension entitlements and might find it difficult to do so.”[149]The aim (2) case as to the older firefighters’ “expectations” was, however, pleaded differently. Para 32 of the Grounds of Resistance asserted that there was “also a fairness consideration in that older firefighters will have spent a greater proportion of their careers with the expectation that they would be able to retire at age 55 with a full pension than their younger comparators.” The fairness point might be controversial, but otherwise the pleaded assertion was perhaps a statement of the tolerably obvious. By the time of the hearing, however, aim (2) had departed from the pleaded case in two respects:(i) that those closer to retirement had a “greater” expectation than their younger colleagues that their pension expectations would not change; and(ii) that such expectation was a “legitimate” one. The latter pointwas a forensic embellishment by Mr Cavanagh (first evinced upon the exchange of skeleton arguments for the hearing before Judge Lewzey). We do not regard it as introducing a point of material substance, nor did Mr Cavanagh suggest otherwise. But the other change did put a different slant on the Governments’ case; and, if it was to be made good, we consider it needed to be supported by evidence as to such claimed “greater” expectation, which it was not.[150]Aim (3) was a separate one directed at providing special tapering financial protection to a group of firefighters whose ages put them between those enjoying full protection and none. It did not purport to assert a justification for doing so. Aim (4), a consistency aim, reflected a well-recognised virtue in public law and policy, that of treating like cases in a like way, but it was not one that could give legitimacy to the other aims if they were not legitimate on their own merits. The decision in the judges’ case illustrates that.[151]Having identified an evidential problem with aim (2) (we shall come to what we also regard as evidential problems about aims (1) and (3)), we proceed on the basis that, as regards aims, the heart of the question that Judge Lewzey had to consider and assess was whether the Government’s aims (1) to (3) were not just proper social policy aims (which is not disputed) but were legitimate such aims (which is). As Lady Hale explained at para 61 of Seldon, the question as to legitimacy had to be answered by reference to the particular circumstances of the employment concerned. In this case we consider that must mean by way of an objective analysis in light of the particular circumstances of the pension arrangements (current and past) of the firefighting service and of the relative potential impact of the transitional protective provisions on scheme members of different ages.[152]Once the legitimacy of aims question is so identified, there is an apparent problem with Judge Lewzey’s reasoning towards her conclusion on it at para 104. By inference, she there found that the aims asserted were legitimate. But whether or not they were legitimate required an objective analysis by her of the nature we have described. Instead, however, of carrying out such an analysis, she proceeded from a finding that the claimed aims were social policy aims straight to the conclusion that they were also legitimate ones. She at no point engaged in any objective assessment as to their legitimacy. That was an exercise that required her to ask herself why the oldest members of the FPS were being so generously preferred over younger members; and why those entitled to tapered protection were being so preferred over even younger members. Having asked and answered those questions, she would then have to consider whether the answers pointed to the aims being legitimate. She did none of those things.[153]Judge Lewzey’s omission to consider, and answer, the why questions was not for want of submissions that to do so was an essential part of her task. Her response to Mr Short’s submission to that effect amounted to no more than a recognition that the older groups were being treated more favourably because of their proximity to retirement and that “whilst retirement is age related, and proximity to retirement is connected with age, there may be good reasons for treating different age groups differently” (para 70). That was merely to re-state the aims, not to explain why they might be legitimate; and that “there may be good reasons” for them is hardly a sound basis for a finding of their legitimacy.[154]It appears to us that the primary reason Judge Lewzey did not engage in an objective assessment of the legitimacy of the aims was because she considered she did not have to. She regarded the chosen aims as a decision of the Governments, perhaps of a moral nature and/or as one that had a political element, and that it was not for her to substitute her view for that of the Governments. She considered that the standard of scrutiny she was required to apply involved granting a wide margin of discretion to the Governments: see her paras 75, 76, 83 and 88, quoted at paras 123 and 124 above. She was also probably influenced by the decision in Commission v. Hungary, from which she derived that the CJEU regarded transitional protective provisions of the type in question as legitimate. Importantly, she also concluded, at para 97 (see para 129 above), that the justification the Governments were advancing was not one that needed to be “based on precise or concrete factors”. In so concluding, she was apparently influenced by the observations in Lumsdon that “… justifications based on moral or political considerations may not be capable of being established by evidence.” Ultimately, however, we consider that her reason for concluding that the Governments’ aims were legitimate social policy aims was because she considered it 160. was not for her to second-guess their policy decisions. Her duty was to defer to them unquestioningly.[155]In our judgment, and with respect to what was manifestly a conscientious set of reasons, Judge Lewzey was wrong so to approach the issue of legitimacy of aims. She failed to carry out the objective assessment exercise that, in Seldon at para 61, Lady Hale had identified as the second part of the “legitimate aim” inquiry. She did not apply the guidance that Lady Hale there gave.[156]In defence of Judge Lewzey’s approach, it can fairly be said that the presentation of the Governments’ case made it impossible for her to engage in an objective assessment of the legitimacy of the aims. That is because there was no evidence as to the reasons underlying the aims. The Governments’ case was advanced to her, as to us, on the basis that the unproved assertions as to the need for, and virtue of, the aims were all she needed in order to rule on their legitimacy. Judge Lewzey was no doubt aware of that; and her view was that their legitimacy did not need to be supported by evidence: see again her para 97.[157]We respectfully disagree with such view. We consider that the Governments’ rationale for the protective provisions did need to be supported by evidence. The Hutton Report’s opinion was that the implementation of its proposed reforms to public sector pensions would not require the provision of special protection for members over a certain age and, moreover, that it was anyway precluded by age discrimination legislation. The Governments then took a different view. They proposed different transitional treatment between three groups of members of a nature that was manifestly discriminatory. Treating the accrual of rights to a retirement pension as pay, their proposals meant that those with full or tapered protection were being paid significantly more than unprotected younger firefighters for doing equal work. If their proposal was to be upheld as justifiable, the Governments had to show why it was justifiable. Yet they provided no evidence to substantiate the reasons for such discriminatory treatment. Their claimed belief, as Mr Cavanagh put it, that “it felt right” so to protect older firefighters, and that the decision to do so “was a moral decision” and so did not need to be evidentially substantiated, are in our view not good enough. If the Governments’ opinion as to the need to protect the older firefighters was based on something more than visceral instinct, they needed to explain what it was so that the tribunal could assess it when considering the legitimacy of the chosen aims.[158]What were the concerns that the Governments sought to meet by the transitional provisions? We recognise what might be regarded as a primary concern, namely that firefighters close to their retirement age under the FPS should not suddenly be transferred to a scheme requiring them to work until 60. Given, however, that the FPS’s rule of 75 meant that many firefighters in practice retired at between 50 and 55, it is unclear why the Governments chose to adopt the blunt instrument of a 10 year rule counting back from an NPA of 55. They have not explained that. But what were the other considerations that they had in mind in proposing the full and tapered protection? Those entitled to it were, by reason of their greater accrued rights under the FPS, already materially better off than their younger colleagues. Why were they favoured? What were the types of lifestyle changes that the Governments considered the younger members could make in preparation for retirement that their older colleagues could not? If, as aim (1) asserted, the Governments assessed that younger members could make “alternative financial provision” that older members could not, what did they have in mind? Were they really suggesting that one option for younger members was to leave the fire service and find a better job and pension arrangements elsewhere? They were certainly asserting (para 5 of the Grounds of Resistance) that “scheme members wishing to maintain the same level of income as they would have expected under previous arrangements may wish to use part of their earned income to make investments.” But the expert evidence showed that this idea was in practice unrealistic, not least because for a firefighter to use part of his salary to fund his retirement would be likely to result in an immediate reduction of his standard of living. There appears to us to be a real question as to the rationality of this suggestion.[159]We would be disposed to accept that if, in principle, the factual position was that older firefighters close to retirement were likely to face financial or other difficulties that, with more time to prepare for it, younger firefighters could somehow avoid or overcome, that might provide a justification for some sort of transitional financial protection for older firefighters. But if that was the Governments’ case, they needed to demonstrate it by evidence.

Governments’ aims were legitimate social policy aims was because she considered it

[160]We have noted that Judge Lewzey was probably influenced in her conclusion that the Governments did not need to provide supporting evidence by the observations of Lords Reed and Toulson in Lumsdon, at para 56, which we have cited. We disagree that that authority justified her conclusion. Mr Cavanagh’s submission was indeed that the aim was either a moral or political one that did not need to be supported by evidence. We would be disposed to agree that, once the Governments had decided that their chosen aim was the right one, it may be that it could be characterised as a moral aim. But they could only first arrive at the decision to pursue it by making an assessment as to the justification for paying three groups of workers materially differently for doing equal work. That required an analysis of an economic nature, namely as to whythe pay differential was justified. It was that analysis, and the conclusions from it, that required to be supported by evidence but were not.[161]We therefore agree with and accept Mr Short’s submission that the Governments’ aims were ones whose claimed justification had to be supported by evidence. It was for the Governments to show that, despite the apparently discriminatory effect of their transitional protective measures as between the three groups of FPS members, their measures were a legitimate aim of social policy. In the event, they sought to do so by nothing more than assertions and generalisations. Even though governments are entitled to be afforded a broad measure of discretion, “Generalised assumptions, not based on any factual foundation, are not good enough” (Seymour Smith, per Lord Nicholls of Birkenhead). We noted, at para 71 above, that Lord Nicholls was there addressing himself primarily to means. But the ECJ made the same point about aims in the Age Concern EnglandCase C-388/07 [2009] ICR1080, at paras 51 and 65:- “51. Mere generalisations concerning the capacity of a specific measure to contribute to employment policy, labour market or vocational training objectives are not enough to show that the aim of that measure is capable of justifying derogation from that principle … 65. … However, it is important to note that the latter provision [Article 6(1) of Directive 2000/78] is addressed to the member states and imposes on them, notwithstanding their broad discretion in matters of social policy, the burden of establishing to a high standard of proof the legitimacy of the aim pursued.” The ECJ reminded itself of those observations in Fuchs v. Land Hessen Case C160/10 [2012] ICR 93, at paras 77 and 78. They show that the burden of proof of the legitimacy of a claimed aim is a high one. So it should be. If the promoter of a policy that is directly discriminatory on age grounds wishes it to be recognised as legitimate, it must prove why it is.[162]As Mr Short pointed out, it is not as though it would not be possible to adduce such evidence, if the aims can be substantiated. He referred to this court’s decision in Lockwood v. Department of Work and Pensions and another [2014] ICR 1257. The issue there was as to the justification of paying former civil service employees who were over 35 a higher level of redundancy payment than was paid to those under that age. The different treatment of younger employees was discriminatory on the grounds of age and had to be justified if it was not to be held unlawful. Rimer LJ’s judgment, at paras 14 and 15, explained the full and careful evidence adduced before the employment tribunal by the DWP in explanation and justification of the redundancy scheme that had been devised. There is, in principle, no reason why the Governments could not have adduced evidence in this case directed at explaining and justifying the discriminatory provisions of the transitional protective provisions.[163]In our judgment, the absence of supporting evidence as to the claimed legitimacy of the Governments’ aims meant that there was no basis upon which Judge Lewzey could properly find that the aims were legitimate. It may perhaps be compared with Judge Williams’s contrary (and correct) conclusions in para 94 of his judgment in the Judges’ case (see para 55 above). Her finding to that effect was an error of law on her part.[164]Whilst Sir Alan Wilkie concluded that Judge Lewzey’s finding on legitimacy of aims was unimpeachable, we therefore respectfully disagree. We allow the firefighters’ appeals against his decision on legitimacy of aims. We have considered whether we should remit the age discrimination claims to the employment tribunal for a reconsideration of the issue of legitimacy of aims but have concluded that there is no point in doing so. The only conclusion to which the tribunal could properly come is that, in the absence of evidence supporting the claimed legitimacy of the aims, the respondents’ case as to justification must fail. In those circumstances, we consider that this court should deal finally with the issue of liability in the age discrimination claims by upholding the firefighters’ claims that they have been the victims of unlawful age discrimination and substituting an order to that effect for the order for their dismissal made by Judge Lewzey in paragraph (i) of her reserved judgment. We remit to the employment tribunal the question of the remedies to which the claimants are entitled in consequence of our decision.[165]It follows that we see no reason to extend this judgment by considering the Governments’ and FRAs’ appeals on proportionality of means. They do not now arise. We dismiss them and set aside Sir Alan Wilkie’s order remitting the proportionality issue for a re-hearing by the employment tribunal.

Equal Pay and Race Discrimination

[166]We now turn to the claims made in each of these cases that in addition to amounting to direct age discrimination, the adoption of the transitional provisions constituted a breach of the principles of equal pay and gave rise to indirect race discrimination. The equal pay claims arose out of the operation of the sex equality rule which is incorporated into every occupational pensions scheme; and the indirect race claims arose out of the non-discrimination rule which is likewise incorporated into all such schemes (see ss 67 and 61 of the Equality Act respectively reproduced at para 32 above) In view of our conclusion that there was direct age discrimination in each case, the resolution of these claims is of no real practical significance. However, we heard relatively detailed argument on these issues, albeit not as full as the submissions on age discrimination, and so we will state our conclusions upon them, albeit relatively briefly.[167]Initially, in addition to the equal pay and indirect race claims, a separate complaint of indirect sex discrimination was advanced. However, all parties now accept that this was always misconceived, as indeed Judge Lewzey recognised in the firefighters’ case. A free standing indirect sex discrimination claim could not succeed because pensions constitute pay and where the alleged sex discrimination relates solely to pay, as in this case, the claim can only be made under the equal pay provisions of the Equality Act (whether pursuant to the sex equality clause or the sex equality rule) and not those relating to other forms of sex discrimination: see section 70. The equal pay provisions allow for a complaint of indirect discrimination in relation to pay, but only in the manner there stipulated.

The relevant law

[168]The relevant law is set out in detail in paras 28-33 above. For convenience, we will set out again the critical provisions bearing on these two claims.[169]With respect to the equal pay claims, the Equality Act draws a distinction between contractual claims and pension rights under an occupational pension scheme. A “sex equality clause” is incorporated into every work relationship whose effect is to modify any contractual term which confers less favourable terms or benefits on members of one sex compared with another (section 66). In relation to occupational pensions a “sex equality rule” applies whose effect, put broadly, is to modify any term less favourable to one sex compared with the other or the less favourable exercise of a discretion so as to remove the inequality (section 67).[170]In each case there is a defence which provides that the difference in treatment is not unlawful if there is a material factor other than sex which explains the difference in pay (the “material factor defence”). Curiously, this is drafted differently with respect to the sex equality clause and the sex equality rule. The defence is set out in section 69 of the Act which, so far as is relevant, is as follows:- “(1) The sex equality clause in A's terms has no effect in relation to a difference between A's terms and B's terms if the responsible person shows that the difference is because of a material factor reliance on which—(a) does not involve treating A less favourably because of A's sex than the responsible person treats B, and(b) if the factor is within subsection (2), is a proportionate means of achieving a legitimate aim. (2) A factor is within this subsection if A shows that, as a result of the factor, A and persons of the same sex doing work equal to A's are put at a particular disadvantage when compared with persons of the opposite sex doing work equal to A's. … (4) A sex equality rule has no effect in relation to a difference between A and B in the effect of a relevant matter if the trustees or managers of the scheme in question show that the difference is because of a material factor which is not the difference of sex.”[171]The effect of subsections (1)(b) and (2) read together is that in relation to the sex equality clause, this defence cannot be established if the clause operates in either a directly or indirectly discriminatory way. In relation to occupational pensions and the sex equality rule, subsection (4) merely states that the material factor must not be “the difference of sex” without in terms thereby embracing indirect sex discrimination. Nevertheless, for reasons we give below, in our view this provision must be interpreted so as to exclude the operation of the material factor defence where the pay arrangements constitute either direct or unlawful indirect discrimination. In other words, the defence must operate in the same way whether the equality clause or the equality rule is in issue.[172]With respect to the race discrimination claim, the relevant definition of indirect discrimination is found in section 19 of the Equality Act:- “Indirect discrimination(1) A person (A) discriminates against and (B) if A applies to B a provision, criterion or practice which is discriminatory in relation to a relevant protected characteristic of B’s.(2) For the purpose of subsection (1), a provision, criterion or practice is discriminatory in relation to a relevant protected characteristic of B’s if – (a) A applies, or would apply, it to persons with whom B does not share the characteristic, (b) it puts, or would put, persons with whom B shares the characteristic at a particular disadvantage when compared with persons with whom B does not share it, (c) it puts, or would put, B at that disadvantage, and (d) A cannot show it to be a proportionate means of achieving a legitimate aim.” (1) A person (A) discriminates against and (B) if A applies to B a provision, criterion or practice which is discriminatory in relation to a relevant protected characteristic of B’s. (2) For the purpose of subsection (1), a provision, criterion or practice is discriminatory in relation to a relevant protected characteristic of B’s if – (a) A applies, or would apply, it to persons with whom B does not share the characteristic, (b) it puts, or would put, persons with whom B shares the characteristic at a particular disadvantage when compared with persons with whom B does not share it, (c) it puts, or would put, B at that disadvantage, and (d) A cannot show it to be a proportionate means of[173]By section19 (3) the relevant protected characteristics include sex and race.[174]Unlike direct discrimination, which applies where an individual is treated less favourably on grounds of a protected characteristic, the concept of indirect discrimination is concerned with group disadvantage: the provision, criterion or practice (“PCP”) must put persons who share a protected characteristic at a disadvantage when compared with persons who do not share that characteristic. In addition, the PCP must place the particular claimant at the same disadvantage as that shared by the group. It is only where these conditions are satisfied, so that there is what we will call “prima facie indirect discrimination”, that there is an obligation to justify the difference of treatment by demonstrating that the adoption of the PCP is a proportionate means of achieving a legitimate aim.[175]The concept of justification requires a consideration of aims and proportionality of means as in the case of direct age discrimination. A point of difference is that the aims are not limited to matters of social policy as they are with direct age discrimination cases. In other respects, however, the concept of justification will in principle apply in the same way whatever the protected characteristic relied upon. Where government policy is in issue, some leeway will be given to government as regards aims and means irrespective of the protected characteristic in issue (see para 71 above).[176]Where the same PCP gives rise to a prima facie case of indirect discrimination with respect to more than one protected characteristic, it does not necessarily follow that a defendant who successfully establishes a defence of justification with respect to one of the protected characteristics will be able to do so with respect to another, or vice versa. This is because the extent of the group disadvantage may differ with respect to each of the different protected characteristics andit is obviously easier to justify a PCP where relatively few people are disadvantaged than where that number is large. Indeed, a differential impact as between different protected characteristics is almost inevitable in a situation where, as alleged here, direct age discrimination gives rise to indirect race or sex discrimination. All those who do not satisfy the age criterion will be disadvantaged but, save in a very exceptional case, only a sub-set of that disadvantaged group will also be further disadvantaged by virtue of their sex or race (assuming that the evidence justifies a finding of prima facie indirect discrimination at all).[177]However, whilst in principle the same PCP may be justified with respect to one characteristic but not another, in each of these appeals it was (in our view realistically) accepted by all counsel, at least by the end of the hearing, that if we were to conclude - as we have done - that the same test for justification applies to the various forms of discrimination arising out of the PCP adopted here (save for the more limited aims permitted in direct age discrimination), it will either provide a defence to all the forms of discrimination alleged or to none of them. Accordingly, in view of our finding that there is no objective justification to the claims of direct age discrimination in either the judges’ or firefighters’ cases, it follows that there is no justification defence with respect to the equal pay or indirect race discrimination claims either.[178]The only question, therefore, is whether a prima facie case of either form of discrimination has been established. If it has, then given the lack of any defence of justification, the claims must succeed. The appellants contend that it has not been established in either the judges’ or firefighters’ case although for reasons which are not identical in each appeal. We turn to address the findings on these issues in the courts below and to consider the arguments with respect to them advanced in this court.

The Judges’ case

[179]In the Judges’ case the issues of equal pay and indirect race discrimination were dealt with extremely succinctly by Judge Williams. No doubt this was because, as the judge pointed out, the submissions on these aspects of the case “took the form of a brief postscript” to the primary submissions on age discrimination. The respondents (i.e. the appellants before us) had conceded that as a result of the increased number of women and BME judges appointed in recent years, these groups were disproportionately in the younger age group who were put at a particular disadvantage by the operation of the transitional provisions. The judge noted that although the concept of justification was narrower in the context of direct age discrimination, because of the need to establish social policy aims, it had not been suggested that for any other reason the application of the test of justification could lead to different results with respect to the equal pay and race discrimination claims. Accordingly, having found that there was no justification with respect to the age claim, the logic of the judge’s analysis was that these other claims must succeed also, although he made no formal order to that effect.[180]In the Employment Appeal Tribunal Sir Alan Wilkie accepted a submission that Judge Williams had failed to consider an argument which had been advanced with respect to the equal pay issue, namely that the difference in pay was explained by a material factor other than sex, that factor being age. (In fact this argument, if successful, would have applied equally to the indirect race claim but it does not seem to have been relied upon in that context.) Sir Alan did not deal with that material factor issue in this judgment because it had been agreed that he need not do so if he upheld the Employment Tribunal decision on direct age discrimination, which in the event he did. However, he noted that precisely the same issue arose in the firefighters’ case and he said that if at any stage the material factor defence were to be in issue in the judges’ case, his reasoning in the firefighters’ judgment would apply likewise to the judges’ equal pay claim. For reasons we explain below, he did in the firefighters’ judgment conclude that the material factor defence applied and was an answer to the equal pay claim. It follows that in his view the defence must likewise have been a complete answer to the equal pay claim in the judges’ case also. The claimants have challenged that analysis.[181]So far as the issue of indirect race discrimination was concerned, the only ground of appeal was that the Employment Tribunal ought to have found that the admitted group disadvantage was justified. This submission was not addressed in terms by Sir Alan because of his conclusion that there had been unlawful direct age discrimination. In any event, as we have explained, it is accepted that no justification defence can succeed in view of our conclusions that the direct age discrimination was not justified.

The firefighters’ case

[182]In the firefighters’ case the equal pay and race discrimination claims took on a greater significance than in the judges’ case once Judge Lewzey had dismissed the age discrimination claim. The appellants advanced three reasons before the judge why these additional claims should fail, as they did before us. One of these was the justification defence which Judge Lewzey in fact accepted but which we have rejected. The other two arguments were designed to show that there was no prima facie discrimination at all. First, it was not accepted, as it had been in the judges’ case, that the extent of the disadvantage was significant enough to demonstrate a relevant group disadvantage necessary to establish a prima facie case (the “no group disadvantage” submission.) The Secretary of State (but not the FRAs or the Welsh Government) had conceded that the statistics demonstrated a disproportional impact on younger women and BME firefighters compared with older ones but nonetheless argued that the impact was still too limited to amount to a relevant group disadvantage. In that context the appellants relied upon a decision of the EAT in Tyne and Wear Passenger Transport Executive v Best [2007] ICR 523 in which the Employment Appeal Tribunal held that there was no prima facie indirect discrimination in that case because the overwhelming majority of the disadvantaged group was male.[183]Second, the appellants submitted that even if there were the relevant group disadvantage, that would still not be enough to constitute a prima facie case of discrimination requiring justification. It remained open to the employers to avoid such a finding by showing that the disadvantage was not causally connected in any way with the sex or race of the particular claimants (the “no causal connection” submission). The submission was that the disadvantage suffered by any individual complainant must have some connection with the protected characteristic in issue, and in this case, it had none. It was solely the result of age. In the context of the equal pay claim it therefore constituted a material factor other than sex which wholly explained the difference in treatment. Although there is no material factor defence identified in terms in the definition of indirect discrimination, the argument was that essentially the same principle applies here too. It is said to be implicit in any indirect discrimination claim that if a defendant can show that there is no causal connection whatsoever between the protected characteristic and the disadvantage suffered by the group, there is no discrimination and the claim must fail.[184]Judge Lewzey accepted both these submissions. As regards the no group advantage submission, she noted that both the women and the BME firefighters formed “only a small minority of the protected and unprotected group” and adopted the approach taken by the Employment Appeal Tribunal in the Tyne and Wear case. This was reinforced in her view by the fact that the numbers of women and BME firefighters were so small.[185]As to the no causal connection argument, the judge accepted that there was a factor other than sex or race which explained entirely the difference in treatment, namely age. There was no other legally relevant cause of the disadvantage suffered by the women or BME workers and therefore no causal connection with sex or race. In the context of the equal pay claim, this was the material factor defence relied upon. In arriving at this conclusion, the judge relied upon two recent decisions of the Court of Appeal, Essop v Home Office [2015] EWCA Civ 609; [2015] ICR 1063 and Naeem vSecretary of State for Justice [2015] EWCA Civ 1264; [2016] ICR 289. The attempt to treat age discrimination as a form of sex or race discrimination was, in the judge’s view, “entirely artificial.” The judge cited approvingly a comment of Lord Scott of Foscote in the case of Secretary of State for Trade and Industry v Rutherford(No.2) [2006] UKHL 19; [2006] ICR 785 where, in another case where it was alleged that 192. there was sex discrimination arising from the imposition of an age rule (in that case one which denied unfair dismissal or redundancy compensation to persons over 65) Lord Scott rejected the claim and commented that (para16):- “….a difference in treatment of individuals that is based purely on age cannot be transformed by statistics from age discrimination, which it certainly is, to sex discrimination.”[186]In the Employment Appeal Tribunal Sir Alan Wilkie accepted that whilst there was some confusion in the Employment Tribunal’s judgment on the equal pay issue, the Employment Tribunal had been right to dismiss that particular claim on the grounds that a material factor defence had been established. But the reason he gave bore no relation whatsoever to the material factor defence accepted by Judge Lewzey. He said this (paras.93-94):- “93. Paragraph 129 is less than clear - the Employment Judge seemed to think that, because the material factor was age and not sex, "no material factor defence was necessary". Upon a proper analysis, her finding of fact meant that the material factor defence under section 69(4) arose. Notwithstanding that element of confusion, in my judgment, the Employment Judge did not err in law in concluding, on the basis of that finding of fact, that the material factor defence had been made out. The equal pay claim arose undersection 67which imposes a sex equality rule. The statutory defence to such a claim is found insection 69(4). By contrast, a claim undersection 66, based on a sex equality clause, is subject to a different, and more complicated, statutory defence pursuant to the provisions ofsection 69(1), (2) and (3)involving, as one element, the issue of justification. That issue does not arise for decision undersection 69(4). 94. It follows that the appeal against the equal pay claim fails. So too does the associated piggy-back claim.”[187]The assumption underlying this analysis is that because of the different wording of section 69(4) compared with sections 69(1) and (2) read together, the material factor defence available with respect to pensions must be treated as being wider in scope. It is excluded where the PCP is directly discriminatory but can still be relied upon where it gives rise to unlawful indirect discrimination. Since there was no direct discrimination on grounds of sex here, the material factor defence was applicable.[188]It is on the face of it puzzling why there should be different formulations of the material factor defence with respect to the equal pay clause and the equal pay rule. Nevertheless in our view Sir Alan was wrong to give them a different scope. He misconstrued the effect of section 69(4). If his analysis were correct, the provision would fail fully to implement the EU principle of equal treatment into domestic law. Article 2 of Directive 2000/987 requires that there should be “no direct or indirect discrimination whatsoever” (see para 28 above). It would also provide an unprincipled distinction between the operation of equal pay principles with respect to occupational pensions and other employment rights. Moreover, section 69(4) is framed in virtually the same way as the original material factor defence in section 1(3) of the Equal Pay Act 1970 which provided that the defence applies where the pay differential “is genuinely due to a material factor which is not the difference of sex”. In Glasgow CityCouncil v Marshall[2000] ICR 196, 202H, Lord Nicholls held that the phrase “not the difference of sex” was apt to embrace “any form of sex discrimination, whether direct or indirect”.[189]In our judgment that must be the appropriate construction of section 69(4) also and indeed no counsel sought to argue otherwise, nor did they seek to rely upon Sir Alan’s analysis.[190]Sir Alan did not, however, accept the other two grounds which Judge Lewzey had relied upon in determining that there was no prima facie case of indirect discrimination requiring justification. He did not refer at all to the submission, made with respect to both the race and sex claims, that there was no group disadvantage. However, the fact that he remitted the indirect race discrimination claim to the Employment Tribunal to consider afresh the issue of justification shows that he could not have been persuaded by the argument. Nor did he accept the submission that there was no prima facie case because of the lack of any causal connection whatsoever between the group disadvantage and the protected characteristic. This was in fact how the material factor defence argument was advanced; it was never based on a narrow construction of section 69(4).[191]As far as the no causal connection argument was concerned, the law had changed between the decisions in the Employment Tribunal and the Employment Appeal Tribunal. Before the hearing in the Employment Appeal Tribunal, the Supreme Court had reversed the decision in Essop, and whilst it affirmed the decision in Naeem, it did so on different grounds: Essop v Home Office; Naeem v Secretary of State forJustice[2017] UKSC 27; [2017] ICR 640. We consider this judgment in some detail below. Suffice it to say that Sir Alan held that in his view the decision of Lady Hale established that there was no need for any causative link at all between the group disadvantage and the protected characteristic; indirect discrimination could be established without it. Accordingly, even if it were the case that the only cause of the discrimination was age, this did not preclude a finding of race discrimination. It was enough that the PCP in issue caused the disadvantage, as it plainly did here. Absent justification, therefore, the indirect race discrimination was unlawful.

[2006] UKHL 19; [2006] ICR 785 where, in another case where it was alleged that

[192]But for the misconstruction of section 69(4), it is clear that Sir Alan would have held that there was no material factor defence in play with respect to the equal pay claims and that prima facie indirect discrimination requiring justification had been established there also.

The issues in the appeal

[193]If, therefore, the appellants in the firefighters’ case are to succeed in their appeal, then given that for reasons already explained they cannot show that any prima facie indirect discrimination was justified, they will have to show that there was no prima facie discrimination which required justification. This involves establishing that at least one or other of the arguments accepted by Judge Lewzey but rejected by Sir Alan (one explicitly and one implicitly) is correct. That is indeed what the appellants allege. They seek to restore the conclusion of Judge Lewzey on each of these points.[194]The position with respect to the judges’ appeal is different. Only the no causal connection argument was in play, and then only in connection with the equal pay claim. For some reason it does not seem to have been run in relation to the race claim although it is now being advanced as a potential defence to that claim also.

The “no group disadvantage” submission

[195]This argument relates solely to the firefighters. The submission was that even though there was a disproportionate adverse impact on women and BME firefighters, the statistics were not sufficiently cogent or telling to be able to infer that these women or BME firefighters suffered a particular disadvantage when compared with men or white firefighters so as to justify the further inference of prima facie indirect discrimination.[196]The first necessary step when considering whether there is prima facie indirect discrimination is to identify the pool from which to test whether there is the relevant disadvantage. As Sedley LJ observed in Grundy v British Airways plc [2008] IRLR 74 (para 27) this should be selected with the aim of suitably testing the discrimination of which complaint is made. That is sometimes a difficult and controversial question but here it is relatively straightforward. The claimants are for the most part the disadvantaged firefighters in the FPS (although they also include some retained firefighters in the NFPS who were in post before April 2006). The imposition of the age barrier affects all firefighters in the FPS; either they retain their pension benefits entirely, or they are given tapered protection, or their benefits are unprotected. The question is whether within the group disadvantaged by age, there are sub-groups selected by a protected characteristic of sex or race who suffer a particular disadvantage.[197]The relevant figures we have seen with respect to the firefighters in England demonstrate that if one takes the members of the FPS who were aged 45 or over on the 31st March 2012 and therefore benefited fully from the transitional provisions, there were 9,348 men and 81 women – the latter constituting well under 1% of the advantaged group. For those aged under 45 who would not benefit from the transitional provisions or would benefit to a lesser extent, the numbers were 12,401 men and 566 women, which means that women constituted a little over 4% of the disadvantaged group. So some 99% of the advantaged group and 96% of the disadvantaged group are men. On the other hand, the proportion of women who are advantaged is around 12% of the total number of women employed whereas the proportion of men is much higher at around 43%.[198]If we turn to consider the position of BME firefighters in England, it is more difficult to assess the effect because there are no statistics which break down the ages of the workforce by reference to their ethnicity. The figures suggest that they constitute about 4% of the workforce but it is impossible to know what proportion are in the advantaged and disadvantaged groups.[199]In Wales also there is no age breakdown for either sex or race. Such limited information as we were shown suggests that in Wales about 4% of the firefighters are women and fewer than that, it seems some 2% (amounting to only thirty in total), are BME. We do not know how they are distributed amongst the advantaged and disadvantaged groups. Having said that, it is almost inevitably the case - and indeed Mr Cavanagh did not dispute this - that the proportions of women and BME firefighters who do not benefit from the transitional provisions will be greater than the proportion who do, and probably significantly so, if only because in recent years active steps have been taken to encourage these groups to join the force. Historically firefighting has been perceived as a male job, and for reasons which are less obvious, the jobs have been performed by a disproportionate number of white men. Recent recruitment practices have sought to change that.[200]The Employment Tribunal did not address the statistics even in the relatively cursory way which we have done above, and there does not appear to have been any discussion or debate about them in the oral hearing - no doubt because the focus was on age discrimination. The principal submission in support of the proposition that the necessary disparate impact had not been established was based on the Tyne and Wear case where HH Judge Serota in the Employment Appeal Tribunal expressed the view that the disadvantage cannot be significant “where the overwhelming majority of the disadvantaged group is male.” On this analysis, even if the proportion of (say) men who are advantaged is significantly greater than the proportion of women who are advantaged, that will not suffice to constitute prima facie indirect discrimination if the overwhelming majority of the disadvantaged workers are male. That is plainly the case here.[201]The definition of indirect discrimination does not identify how serious that disadvantage needs to be. In an earlier incarnation of the definition, the test was whether “a considerably smaller proportion” of the disadvantaged group could comply with the requirement (i.e. the PCP in the current definition). In LondonUnderground Ltd v Edwards (No.2) [1999] ICR 494, 504 Lord Justice Potter said this:- “22. In my view there is a dual statutory purpose underlying the provisions of section 1(1)(b) of the Act of 1975 and in particular the necessity under sub-paragraph (i) to show that the proportion of women who can comply with a given requirement or condition is “considerably smaller” than the proportion of men who can comply with it. The first is to prescribe as the threshold for intervention a situation in which there exists a substantial and not merely marginal discriminatory effect (disparate impact) as between men and women, so that it can be clearly demonstrated that a prima facie case of (indirect) discrimination exists, sufficient to require the employer to justify the application of the condition or requirement in question: see sub-paragraph (ii). The second is to ensure that a tribunal charged with deciding whether or not the requirement is discriminatory may be confident that its disparate impact is inherent in the application of the requirement or condition and is not simply the product of unreliable statistics or fortuitous circumstance. Since the disparate impact question will require to be resolved in an infinite number of different employment situations, well but by no means comprehensively exemplified in the arguments of Mr. Allen, an area of flexibility (or margin of appreciation), is necessarily applicable to the question of whether a particular percentage is to be regarded as “substantially smaller” in any given case.”[202]In that case there was a change in the shift system which one woman out of twenty one could not meet because she was a single parent. There were over two thousand men who could all without difficulty work the new system. Notwithstanding that 95% of the women could comply with the new requirement, the Court of Appeal held that the employment tribunal had been entitled to find that the necessary disparate impact was established. It was relevant that among the population at large women were more likely than men to be single parents with childcare responsibilities.[203]The current formula no longer has the comparatively smaller proportion test, and it certainly would not be more stringent that that test. Edwards is in our view quite inconsistent with a rule of the kind which was adopted in the Tyne and Wear case.[204]Furthermore, the facts in the Tyne and Wear case were very different from this case. As HH Judge Serota noted, it was not a case where the disadvantage had resulted from the application of a PCP. Rather it was a case like the well-known case of Enderby v Frenchay Health Authority Case C-127/92 [1994] ICR 112 where one group was treated less favourably than another but as the result of entirely separate pay arrangements relating to each group.[205]In our view, and consistently with the Edwards judgment, it is important to have regard to all potentially relevant factors when considering whether the necessary disparate impact has been established. It is not legitimate to adopt a rule of thumb, as Judge Lewzey did, and treat it as decisive in all cases. In this case the proportion of women who benefited fully from the transitional arrangements was significantly lower than the proportion of men who did so; and the figures were not trivial. Exceptionally that may be the case, an example being Nelson v Carillion Services Ltd[2003] IRLR 428, where the relevant pool had only eight members. But save arguably with respect to the BME firefighters employed by the Welsh FRA, we are a long way from that situation.[206]It is also important to bear in mind that here the appellants knew precisely how the transitional arrangements would affect different categories of the workforce. Moreover, the employer also knew that firefighting had traditionally been seen as men’s work, and that historically few BME people had chosen or been selected to work in the force. In those circumstances it is hardly surprising that there was a disproportionate number of women and BME workers in the disadvantaged younger group. It cannot be said that it is the result of chance.[207]However, although these factors point strongly - perhaps very strongly - in favour of the conclusion that there is the necessary particular disadvantage, we are reluctant finally to determine this question ourselves for a number of related reasons. First, whilst we are satisfied that the Employment Tribunal reached the conclusion it did on an erroneous basis, it does not follow that the conclusion was wrong. Second, we heard virtually no argument on this issue and indeed we were only directed to some of the relevant statistics after some prompting during the course of the hearing. A party may well have a sense of unfairness if we were now to rule on the matter. Third, the information we have is sketchy, particularly with respect to the Welsh FRA. Fourth, the proper body to decide this question, at least at the first instance, is the employment tribunal.[208]Although we are inclined to the view that the appellants will have real difficulty in showing that there is no relevant group disadvantage, we are not prepared to say that it is obvious that only one outcome is possible in relation to both characteristics and in both England and Wales. Accordingly, had this still been a live issue, we would have remitted it to the Employment Tribunal, assuming that the appellants still wished to pursue the matter. However, given that the claimants have succeeded on the age discrimination claims, there is no point in taking that step.

The “no causal connection submission”

[209]We turn to the second submission which, if sustained, is in principle an answer to the equal pay and indirect race discrimination claims in both appeals, (although curiously it was not relied upon with respect to the claims of BME judges in that appeal). The submission is that if there was no causal relationship of any kind between the disadvantage suffered by the individual and the particular protected characteristic, there can be no prima facie indirect discrimination to support either an equal pay or race discrimination claim. It is not asserted that the protected characteristic must be the direct cause of the discrimination; if it were, that would amount to direct discrimination. But the argument is that there must be a causal link of some kind, however limited, between the protected characteristic and the creation of the disadvantage. As it is sometimes put (in the context of sex discrimination claims) the disadvantage resulting from the imposition of the PCP must not be “tainted by sex” and if the employer can show that it is not, no question of prima facie discrimination arises.[210]This submission has a perfectly respectable pedigree. Lord Justice Underhill identified some of the authorities which support this principle in the Naeem case, paras 25-28. In particular he referred to two decisions of the House of Lords, Strathclyde Regional Council v Wallace[1998] ICR 205and Glasgow City Council vMarshall[2000] ICR 196 which were concerned with the construction of the Equal Pay Act and in particular section 1(3) which (like section 69(4) of the Equality Act) created the material factor defence where the pay differential “is genuinely due to a material factor which is not the difference of sex”. In Wallace Lord Browne- Wilkinson said, at p. 213 B-D:- “… in considering section 1(3) of the Equal Pay Act 1970, the only circumstances in which questions of ‘justification’ can arise are those in which the employer is relying on a factor which is sexually discriminatory. There is no question of the employer having to ‘justify’ … all disparities of pay. Provided that there is no element of sexual discrimination, the employer establishes a subsection (3) defence by identifying the factors which he alleges have caused the disparity, proving that those factors are genuine and proving further that they were causally relevant to the disparity in pay complained of.”[211]Likewise, in Marshall Lord Nicholls said, at p. 203, “if the employer proves the absence of sex discrimination he is not obliged to justify the pay disparity”.[212]These authorities are not decisive of the point, however, because it may be said that if as a matter of fact a PCP has a disparate impact on women or BME workers as the case may be, that of itself provides the necessary element of sex or race discrimination referred to in those judgments even if there is no causal connection between the disadvantage and the protected characteristic.[213]However, the principle was adopted unequivocally in two Court of Appeal authorities: Armstrong v Newcastle Upon Tyne NHS Trust Hospital [2006] IRLR 124 which was in turn approved by the Court of Appeal in Gibson v Sheffield City Council [2010] EWCA Civ 63; [2010] ICR 708. Armstrong is a controversial case and in Gibson Lord Justice Pill considered himself bound by the decision but doubted whether it was consistent with either the decision of the European Court in Enderby or the House of Lords’ decision in Marshall. However, Lady Justice Smith, in a judgment with which Lord Justice Maurice Kay expressly agreed, considered that not only was Armstrong binding but it was correct in principle (para 66):- “My conclusion is that whether the alleged indirect discrimination arises in the field of pay or non-pay, it is always open to a defendant to demonstrate that, notwithstanding the appearance that the practice puts women at a particular disadvantage, in fact the apparent disadvantage has arisen due to factors which are wholly unrelated to gender.”[214]Her reason for reaching this conclusion was that a defence of this nature, although not expressly identified in the definition of indirect discrimination, was inherent in the very concept of discrimination itself (para 63):- “There must be such an implied possibility because the purpose of the legislation is to prevent sex discrimination, including unjustifiable indirect discrimination. A respondent is not to be held to have discriminated—and be put to justification of his practice—merely because it has given rise to a statistical imbalance.”[215]Gibson was decided under the law as it was prior to the Equality Act but given that the rationale for this approach lies in an assumption of what the concept of discrimination involves, there would be no reason to read the current definition differently. This analysis was followed in Naeem which did involve considering the concept of indirect discrimination as it is now defined under the Equality Act.[216]However, as we have said, in the conjoined appeals in Essop and Naeem, the Supreme Court upheld the appeals and held that prima facie indirect discrimination can arise even if the disadvantage suffered is wholly unconnected with the protected characteristic in issue. The mere fact of group disadvantage is enough. In Essop the facts were unusual in that it was conceded that there was group disadvantage but nobody could identify why. The Home Office required staff to pass a skills assessment before they could be promoted. A statistical report showed that BME candidates and those aged over 35 performed significantly less well than white or younger candidates. It was not able to explain why the assessment had this impact. A number of claimants who had taken but failed the assessment alleged that there was indirect race and age discrimination which required objective justification. The employment judge decided as a preliminary issue that each claimant had to show why he or she was disadvantaged and that this was for the same reason as the group was disadvantaged. This required an analysis of why the PCP in question disadvantaged the group. The Employment Appeal Tribunal reversed that decision and held that it was sufficient that the PCP in fact disadvantaged the group and if the individual with the relevant protected characteristic failed the assessment, that would be sufficient to satisfy prima facie discrimination requiring justification. If there was evidence that someone within the group might have failed for some particular reason, such as not properly preparing for the assessment, that might be reflected at the remedy stage, for example in the amount of compensation payable. The Court of Appeal in turn overturned the decision of the Employment Appeal Tribunal.[217]In the Court of Appeal Sir Colin Rimer, giving a judgment with which Sir Terence Etherton, Chancellor, and Lord Justice Lewison agreed, accepted that read literally, the definition of indirect discrimination in section 19 did not require members of the disadvantaged group to show why they were disadvantaged. Nevertheless, it was necessary under section 19(1)(c) that the individual should show that he or she had suffered the same disadvantage as the group and this could only be done once it was known why the group itself had suffered the disadvantage. As Sir Colin recognised, that posed a problem for a claimant because it was not known by anyone why the assessment had the effect it did. However, that difficulty was in practical terms mitigated by Sir Colin’s acceptance that in principle the statistical report might be relied upon both to demonstrate the group disadvantage and to justify an inference that the particular claimant had suffered the same disadvantage as the group. This would or might provide facts from which a tribunal, in the absence of any other explanation, could conclude that there was indirect discrimination: see section 136 of the Equality Act. It would then be open to the employer to show that the reason why any particular candidate failed was for a particular reason which distinguished his or her case from that of the group as a whole.[218]In the unusual circumstances of the case, the Court of Appeal did not, therefore, decide that it was necessary for a claimant to know, let alone prove, why the PCP disadvantaged the group. The Supreme Court appears, however, to have understood that the Court of Appeal did so hold (see [2017] ICR 640, at 646F) and rejected the analysis that it so attributed to it, although the outcome of the decision of the Supreme Court was in fact substantially the same. The leading judgment was given by Lady Hale (with whom Lords Clarke, Wilson, Carnwath and Hodge agreed). She began by summarising the essential nature of, and the difference between, direct and indirect discrimination as follows (para 1):- “The law prohibits two main kinds of discrimination—direct and indirect. Direct discrimination is comparatively simple: it is treating one person less favourably than you would treat another person, because of a particular protected characteristic that the former has. Indirect discrimination, however, is not so simple. It is meant to avoid rules and practices which are not directed at or against people with a particular protected characteristic but have the effect of putting them at a disadvantage. It is one form of trying to ‘level the playing field’”.[219]Lady Hale then set out in some detail the various definitions of indirect discrimination, both domestic and European, which have been adopted over the years and she identified five points, which she described as “salient points”, inherent in the concept. The first three were as follows (paras 24-26):- “24. The first salient feature is that, in none of the various definitions of indirect discrimination, is there any express requirement for an explanation of the reasons why a particular PCP puts one group at a disadvantage when compared with others. Thus there was no requirement in the 1975 Act that the claimant had to show why the proportion of women who could comply with the requirement was smaller than the proportion of men. It was enough that it was. There is no requirement in the Equality Act that the claimant show why the PCP puts one group sharing a particular protected characteristic at a particular disadvantage when compared with others. It is enough that it does. Sometimes, perhaps usually, the reason will be obvious: women are on average shorter than men, so a tall minimum height requirement will disadvantage women whereas a short maximum will disadvantage men. But sometimes it will not be obvious: there is no generally accepted explanation for why women have on average achieved lower grades as chess players than men, but a requirement to hold a high chess grade will put them at a disadvantage.25. A second salient feature is the contrast between the definitions of direct and indirect discrimination. Direct discrimination expressly requires a causal link between the less favourable treatment and the protected characteristic. Indirect discrimination does not. Instead it requires a causal link between the PCP and the particular disadvantage suffered by the group and the individual. The reason for this is that the prohibition of direct discrimination aims to achieve equality of treatment. Indirect discrimination assumes equality of treatment—the PCP is applied indiscriminately to all—but aims to achieve a level playing field, where people sharing a particular protected characteristic are not subjected to requirements which many of them cannot meet but which cannot be shown to be justified. The prohibition of indirect discrimination thus aims to achieve equality of results in the absence of such justification. It is dealing with hidden barriers which are not easy to anticipate or to spot.26. A third salient feature is that the reasons why one group may find it harder to comply with the PCP than others are many and various (Mr Sean Jones QC for Mr Naeem called them “context factors”). They could be genetic, such as strength or height. They could be social, such as the expectation that women will bear the greater responsibility for caring for the home and family than will men. They could be traditional employment practices, such as the division between “women's jobs” and “men's jobs” or the practice of starting at the bottom of an incremental pay scale. They could be another PCP, working in combination with the one at issue, as inChiefConstable of West Yorkshire Police v Homer[2012] ICR 704, where the requirement of a law degree operated in combination with normal retirement age to produce the disadvantage suffered by Mr Homer and others in his age group. These various examples show that the reason for the disadvantage need not be unlawful in itself or be under the control of the employer or provider (although sometimes it will be). They also show that both the PCP and the reason for the disadvantage are “but for” causes of the disadvantage: removing one or the other would solve the problem.” removing one or the other would solve the problem.”[220]In our view this judgment makes it clear that it is the fact that the group disadvantage exists and is caused by the PCP which suffices to give rise to prima facie indirect discrimination; why the PCP should have that effect may be unclear, as in the Essop case itself. Contrary to the view of the Court of Appeal, it is not necessary to ask why the PCP should have that effect, and nor need the reason be connected in any way with the protected characteristic. It is true that the reference to “hidden barriers” is a little puzzling because if the reason is not in fact connected with the protected characteristic, there may be no barriers as such operating at all (although almost always there will be). But taken as a whole, this judgment provides a clear answer to the question how prima facie indirect discrimination is established (subject to a further argument which we consider below) and resolves earlier uncertainties. Unlike Lady Justice Smith in Gibson, Lady Hale does not allow that some preconception of what the concept of discrimination might involve should influence the construction of the section. The definition should be read as it is; there is no express requirement for the reason why the PCP has the detrimental effect it has to be determined, and no justification for implying such an obligation. Moreover, this approach better achieves the objective of achieving equal effects.[221]It is doubtful, however, whether the different approaches are of any great practical significance as indeed Lady Justice Smith noted in Gibson (para 71). Given that the group disadvantage must be significant, it will be a very unusual case where it is not possible to explain why the PCP has the disparate impact which it does. It will also be a rare case where there is substantial disparate impact which can be shown to be wholly unconnected to any sex tainting. That arises in many ways; it may, for example, be the result of physical barriers (e.g. height), cultural barriers (e.g. child care causing more women to work part-time), historical discrimination or stereotyping.[222]Notwithstanding the categorical statement by Lady Hale in the paragraphs we have set out to the effect that indirect sex or race discrimination can arise without sex or race tainting, the appellants have argued that a careful reading of the judgment as a whole shows that this is not the whole story and that Lady Hale did not, and did not intend to, make the protected characteristic wholly irrelevant to the causation issue. The argument focuses on the relationship between the group disadvantage and the disadvantage suffered by the individual which was the issue in the Essop case itself. In any indirect discrimination claim, the individual claimant must show that he or she is disadvantaged by the PCP in the same way as the group. But in the context of the Essop case, what is the position of someone who fails the assessment for reasons wholly unconnected to his or her age or race, for example because he simply did not prepare properly? How can it be justified to give such person a remedy at all? Lady Hale dealt with this argument in the following way (para.32):- “That leads to the second argument—that “undeserving” claimants, who have failed for reasons that have nothing to do with the disparate impact, may “coat tail” upon the claims of the deserving ones. This is easier to answer if the disadvantage is defined in terms of actual failure than if it is defined in terms of likelihood of failure (because only some suffer the first whereas all suffer the second). But in any event, it must be open to the respondent to show that the particular claimant was not put at a disadvantage by the requirement. There was no causal link between the PCP and the disadvantage suffered by the individual: he failed because he did not prepare, or did not show up at the right time or in the right place to take the test, or did not finish the task. A second answer is that a candidate who fails for reasons such as that is not in the same position as a candidate who diligently prepares for the test, turns up in the right place at the right time, and finishes the tasks he was set. In such a situation there would be a “material difference between the circumstances relating to each case”, contrary tosection23(1): para 4 above.”[223]Mr Lynch QC, counsel for the FRAs, has focused on this aspect of the judgment, and specifically the first answer given by Lady Hale, to mount an argument that the disadvantage must in some way be connected to the protected characteristic. He states that Lady Hale has recognised in these paragraphs that it is open to a defendant to show that “there is no causal connection between the PCP in issue, the relevant characteristic, and the disadvantage suffered by the claimant”. Accordingly he submits that if, as in Essop itself, the disadvantaged group are identified by either age or race, and if the defendant can show that the reason any particular claimant suffered the disadvantage was not connected with either of those characteristics, the claim must fail. That, he submits, was the position with the firefighters. The only factor which separated the advantaged and disadvantaged groups was age: 100% of the older firefighters qualified for the advantage, whatever their sex or race, and 100% of the younger firefighters did not qualify, again irrespective of sex or race. There was no hint or taint of sex or race discrimination. As Judge Lewzey rightly observed, it was wholly unreal and artificial to convert what was undoubtedly age discrimination into sex or race discrimination.[224]If this analysis were right, it would fundamentally undermine the earlier passages in Lady Hale’s judgment. Every firefighter claimant would be defeated not on the grounds that there was not an identifiable disadvantaged group, but because he or she did not fall into it. One by one the members of the group would be disqualified ostensibly because of their particular individual circumstances but in fact for precisely the same reason in each case, namely that their treatment was not tainted by sex or race, as the case may be.[225]It is, we think, implicit in this argument that Lady Hale has made an assumption that where there is group disadvantage, there is a causal link between the PCP and the protected characteristic such that the latter in some way causes the disadvantage. Accordingly, a defendant who can show that there is a different causal link giving rise to the disadvantage which is not connected to the protected characteristic (e.g. in Essop failing the assessment) would then defeat the claim.[226]We do not accept that there is any such assumption underlying, or implicit in, the analysis. Nowhere in the discussion of the undeserving claimant does Lady Hale refer to the protected characteristic at all. She is focusing solely on the PCP and the disadvantage resulting from it. The cause of the disadvantage in Essop for someone who has not prepared for the assessment is not the requirement to pass the assessment but rather the failure to prepare for it. This is a factor which takes the individual outside the group, not because the cause of the disadvantage is not the protected characteristic, but because the cause is not the PCP itself but a material factor which is sufficiently independent of it (and unrelated to the protected characteristic). It is true that it will therefore be necessary in an individual case to ask a reason why question, but that question is not why the PCP causes the group disadvantage in order to discover whether it is connected to the protected characteristic or not. Rather it is asking why the individual who shares the protected characteristic suffers the disadvantage. If a defendant can show that it is for some reason other than the application of the PCP, the necessary relationship between the PCP and the disadvantage will not be established in the particular case and the claimant’s treatment will not be unlawful. (This was essentially where the Court of Appeal ended up, albeit by a slightly different route.)[227]In this case the cause of the disadvantage is age. Why that particular policy gave rise to a disparate impact is irrelevant; it is enough that it did. Only if it could be shown that an individual has suffered a disadvantage for a reason other than age would his or her claim be defeated, but no such reason has been shown. A theoretical example, suggested by Mr Short, would be persons who had opted out of the pension scheme altogether; even if they were in the younger age group, they would not suffer any disadvantage because of their age and would not be in the protected group. The disadvantaged group encompasses those who are disadvantaged by the application of the PCP and who share a common protected characteristic. If a particular claimant is not disadvantaged by the application of the PCP, he cannot claim to be part of that group even though he shares the protected characteristic.[228]The appellants placed weight on the dictum of Lord Scott in the Rutherford decision, which was relied upon by Judge Lewzey, to the effect that age discrimination cannot artificially be converted into sex or race discrimination (see para 187 above). But Rutherford was a wholly different case. That was a complaint by male claimants who alleged that they were subject to indirect sex discrimination because persons over the age of 65 were unable to claim unfair dismissal or redundancy compensation. It was asserted that this had a disparate impact on men because proportionately more men than women worked beyond the age of 65. A majority of their Lordships (Lords Scott, Roger and Lady Hale) held that the claim was misconceived and that the issue of discrimination did not even arise. Each gave a separate judgment although the reasoning was similar in each case. The critical feature in that case was that there was no obligation to work over the age of 65; it was a voluntary act of the worker. As Lord Scott put it:- “The composition of the respective groups would not depend upon an individual’s ability or inability to satisfy particular condition. It would depend, of course, on the individual’s decision whether or not to continue in employment after the age of 65 and, also, on whether he or she survived to that age. The latter condition is essentially non-discriminatory, otherwise than on the ground of age. Age discrimination cannot be turned by statistics into sex discrimination.”[229]Lady Hale considered that the pool should be defined by reference to those who wanted the benefit, which consisted of those over 65 who wanted to continue to work. But the rule had no disparate impact at all upon grounds of sex with respect to that group; all were treated equally, male and female alike, in being denied these benefits.[230]Here there is the obvious distinction that the age bar does act as a barrier to qualification for the benefit of the transitional provisions. The composition of the groups does depend on the ability or inability to satisfy the age condition. It is not, therefore, inconsistent with Rutherford to say that if the composition demonstrates a relevant disparate impact with respect to a protected characteristic, this gives rise to prima facie indirect discrimination.

Was there in fact sex or race tainting?

[231]We have so far assumed for the purposes of the no causation argument that there is no taint of sex or race discrimination in the application of the age requirement. But in our view that assumption is not sustainable. We do not accept that age discrimination gives rise to no taint of such discrimination. The reason that there is a higher proportion of women in the younger age group is the common perception, until relatively recently at least, that being employed in the fire service was men’s work. It seems that it was the same for BME workers also; either that, or there may have been historic discrimination which explains the shortage of BME staff. The fact that there have more recently been positive efforts to increase the proportions of these two categories of worker necessarily exacerbates the disparity between the proportion of one sex or race in the younger group compared with the older. Accordingly, even if the law were that there needs to be some degree of causal connection between the disadvantage and the protected characteristic, we would find such a connection here. The fact that women and BME firefighters are disproportionately affected by the rule is not pure chance.

Piggy back claims

[232]Finally, the appeals also raised the question whether, if the female claimants had succeeded in their equal pay claims, the men would be able to “piggy back” on their success and in turn claim that the equality rule gave them the right to the benefit of the transitional provisions. Mr Lynch, who was advancing this aspect of the appeal, became indisposed before the completion of his oral submissions and did not develop any arguments on this issue. In the circumstances counsel requested that we should not deal with this point in our judgment and we do not do so.

Conclusions

[233]We have found that in both the judges’ and firefighters’ cases the manner in which the transitional provisions have been implemented has given rise to unlawful direct age discrimination. In neither case could the admitted direct age discrimination be justified. In the Judges’ case we see no error in the reasoning of Judge Williams either in his assessment of aims or means. In the firefighters’ case we take the view that there were no legitimate aims and since we are satisfied that the contrary conclusion would not be open to an employment tribunal, we have made that determination ourselves and not remitted the case, save for the determination of remedy.[234]So far as the equal pay and indirect race discrimination claims are concerned, we are satisfied that these claims are made out in the Judges’ case. The only difference in the firefighters’ case is that, had it been necessary (and we see no reason why it should be) we would have remitted the question whether the disadvantage was sufficiently substantial in the circumstances to establish a prima facie case of indirect discrimination, both in the equal pay and the race claims. - - - - - - - - - - - - - - - - - - - - - ORDER -. - - - - - - - - - - - - - - - - - - - - UPON HEARING COUNSEL FOR THE APPELLANTS, RESPONDENTS AND CROSS APPELLANTS AT A HEARING ON 5, 6, 7, 8, 9 NOVEMBER 2018 IT IS ORDERED THAT Appeal A2/2018/0635: the Judges’ claims[1]The Appellants’ appeal in A2/2018/0635 is dismissed.[2]Accordingly, the Judicial Claimants’ claims of age discrimination succeed.[3]Further, where a claim of equal pay or indirect race discrimination is made by a female or BAME Judicial Claimant, such a Claimant’s claim of equal pay and/or race discrimination, succeeds.[4]By consent, the Court of Appeal makes no ruling on the ‘piggy-back’ issue.[5]The age discrimination claims by the Judicial Claimants are remitted to the Employment Tribunal for a remedy hearing to be dealt with by an Employment Judge who need not be the same judge as dealt with liability, such remission to be stayed pending the outcome of an application for permission to appeal and, if granted, appeal, to the Supreme Court.[6]The equal pay and indirect race discrimination claims are remitted to the Employment Tribunal on the piggy-back issue and remedy, such remission to be stayed pending the outcome of an application for permission to appeal and, if granted, appeal, to the Supreme Court.[7]The Appellants shall pay the Respondents and Cross-Appellants’ costs of the appeals, on the standard basis, to be subject to a detailed assessment if not agreed.[8]The Appellants shall pay the solicitors for the McCloud Judicial Claimants £50,000 (inclusive of VAT of £8,333) on account of their costs within 49 days, such payment to be stayed for so long as there is an outstanding application for permission to appeal to the Supreme Court or, if such an application is granted, any outstanding appeal.[9]The Appellants shall pay the solicitors for the Mostyn Judicial Claimants £50,000 (inclusive of VAT of £8,333) on account of their costs within 49 days, such payment to be stayed for so long as there is an outstanding application for permission to appeal to the Supreme Court or, if such an application is granted, any outstanding appeal.[10]Permission to appeal to the Supreme Court is [granted/ refused].

Appeals A2/2018/0636, A2/2018/0505 & A2/2018/0647: the Firefighters’ claims

[9]The appeals in relation to age discrimination by the Secretary of State for the Home Department and the Welsh Ministers in A2/2018/0636, and the appeal of the Fire and Rescue Authorities in A2/2018/0647 are dismissed.[10]The Firefighter Claimants’ appeal in A2/2018/0505 is allowed.[11]The Firefighter Claimants’ cross-appeals in A2/2018/0636 and in A2/2018/0647 are allowed.[12]The Court of Appeal sets aside the ruling of Employment Judge Lewzey on liability in relation to age discrimination, and substitutes a finding that the Firefighter Claimants’ claims of age discrimination are successful.[13]By consent, the Court of Appeal makes no ruling on the ‘piggy-back’ issue.[14]The age discrimination claims by the Firefighter Claimants are remitted to the Employment Tribunal for a remedy hearing to be dealt with by an Employment Judge who need not be the same judge as dealt with liability, such remission to be stayed pending the outcome of an application for permission to appeal and, if granted, appeal, to the Supreme Court.[15]The equal pay and indirect race discrimination claims are remitted to the Employment Tribunal on the prima facie indirect discrimination and piggy-back issues and remedy, such remission to be stayed pending the outcome of an application for permission to appeal and, if granted, appeal, to the Supreme Court and any remitted hearing if so ordered by the Supreme Court. Thereafter, the claims are to remain stayed in the Employment Tribunal, but with permission to the Firefighter Claimants to apply to the Employment Tribunal to lift the stay in the event that they may become material.[16]The Appellants shall pay the Respondents and Cross-Appellants’ costs of the appeals, on the standard basis, to be subject to a detailed assessment if not agreed. The obligation is joint and several.[17]The Secretary of State for the Home Department shall pay the solicitors for the Respondents and Cross-Appellants’ £50,000 on account of their costs within 49 days, such payment to be stayed for so long as there is any outstanding application for permission to appeal to the Supreme Court or, if such an application is granted, an outstanding appeal and remitted hearing if so ordered by the Supreme Court.[18]Permission to appeal to the Supreme Court is granted/ refused.

Cited in 29 later judgments

and 20 more of the most senior