[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	United Kingdom Employment Appeal Tribunal
You are here: BAILII >> Databases >> United Kingdom Employment Appeal Tribunal >> Secretary Of State For Trade And Industry v Rutherford & Ors [2003] UKEAT 1029_02_0210 (2 October 2003) URL: http://www.bailii.org/uk/cases/UKEAT/2003/1029_02_0210.html Cite as: [2003] UKEAT 1029_2_210, [2003] UKEAT 1029_02_0210

[New search] [Printable RTF version] [Help]

APPEARANCES

 

For the Appellant	DAVID PANNICK QC MELANIE HALL QC KASSIE SMITH (of Counsel) The Treasury Solicitor (Employment Team) Queen Anne's Chambers 28 Broadway London SW1H 9JS
For the 1st & 2nd Respondents	ROBIN ALLEN QC RACHEL CRASNOW (of Counsel) PAUL TROOP (of Counsel) Messrs Charles Russell Solicitors 8-10 New Fetter Lane London EC4A 1RS and Islington Law Centre 161 Hornsey Road London N7 6DU
For the 3rd Respondent	NO APPEARANCE OR REPRESENTATION BY OR ON BEHALF OF THE 3RD RESPONDENT


 


MR JUSTICE WALL

Introduction: the appeals in broad outline

1. Expressed in broad terms, the important question raised by these two appeals is whether or not the provisions of the Employment Rights Act 1996 (referred to as "the Act" throughout this judgment) which prohibit applications to Employment Tribunals by men and women over the age of 65 who wish to claim unfair dismissal. or redundancy payments, are to be disapplied as being contrary to European Law because they discriminate, indirectly, against male employees.
2. Both Mr. Rutherford and Mr. Bentley were over 65 when their employment came to an end in the circumstances set out below. Both applied to an Employment Tribunal for relief. Their cases were heard together by the Employment Tribunal sitting at Stratford over 10 days between 24 June and 22 August 2002 (including one day in chambers). Both men raised the point identified in paragraph 1 above, which the Tribunal heard, in effect, as a preliminary issue. The Tribunal concluded that the upper age default provisions contained in the Act contravened Article 141 of the Treaty of Rome, and were indirectly discriminatory. They could not, in the Tribunal's judgment, be objectively justified by factors unrelated to sex discrimination, and were therefore to be disapplied. The Tribunal held, accordingly, that it had jurisdiction to entertain the claims of both applicants.
3. These appeals are brought, not by the Applicants' former employers, but by the Secretary of State for Trade and Industry (the Secretary of State). She is involved because both former employers are insolvent, and she is potentially liable in such circumstances under sections 166 and 167 of the Act in respect of claims for redundancy pay, and under Part XII of the Act in relation to any basic award of compensation for unfair dismissal. The Secretary of State therefore elected to appear as if she were a party under rule 10(7) of Schedule 1 of the Employment Tribunals (Constitution and Rules of Procedure) Regulations 2001.
4. The judgment which follows is extremely long. There are several reasons for this. The first is this is the second time Mr. Rutherford's case has been before the Employment Appeal Tribunal (EAT). On the first occasion, as on this, the Employment Tribunal held that it had jurisdiction to entertain his claim. The EAT reversed that decision and remitted the application to the Tribunal to be heard afresh. The Secretary of State's appeal is from the decision reached at that second hearing. This factor requires us to consider in some detail the judgment of the EAT in the first appeal, which is reported as Harvest Town Circle Ltd v Rutherford [2002] ICR 123, and which we will call Rutherford 1.
5. Secondly, we have needed to examine not only the Statutory material (including the relevant Council Directives) but a number of the cases, notably R v Secretary of State for Employment ex parte Seymour-Smith and Perez in its various manifestations. Thirdly, we have decided to set out fully the careful arguments addressed to us on both sides.
6. In order to enable readers of this judgment to find their way around it, we have decided to provide an index, as follows: -

Subject Matter	Paragraph Numbers
Introduction: The appeals in broad outline	1 to 6
The essential facts of each case
(a) Mr. Rutherford	7 to 10
(b) Mr. Bentley	11 to 12
The domestic legislation in outline	13 to 15
The relevant domestic legislation in greater detail
(a) Section 108: the qualifying period	16
(b) The upper age limits	17 to 19
Article 141 (formerly 119) of the Treaty of Rome	20
Section 2(1) of the European Communities Act 1972	21
Relevant Directives from the Council of the European Union	22 to 23
The application of the Equal Pay provisions of the Treaty to Applications for unfair dismissal and redundancy payments	24 to 25
R v. Secretary of State for Employment ex parte Seymour-Smith
(a) Introduction	26 to 27
(b) Seymour-Smith in the Divisional Court and in the Court of Appeal	28 to 31
(c) The reference to the European Court of Justice	32
(d) The judgment of the European Court of Justice	33 to 35
(e) The decision of the House of Lords: disparate impact	36 to 41
(f) The decision of the House of Lords: objective justification	42 to 43
Harvest Town Circle Ltd v Rutherford (Rutherford 1)	44 to 53
Disparate Impact
(1) The reasons given by the Tribunal for the decision under appeal	54 to 61
(2) The argument for the Secretary of State on disparate impact	62 to 89
(3) The argument for the Applicants on disparate impact	90 to 110
(4) Points made for the Applicants in oral argument	111 to 120
(5) The Secretary of State's reply	121 to 126
(6) Analysis and Discussion	127 to 139
Objective Justification
(1) The written evidence presented by the Secretary of State	140 to 148
(2) The reasons given by the Tribunal for the decision under appeal	149 to 160
(3) The argument for the Secretary of State	161 to 178
(4) The argument for the Applicants	179 to 192
(5) The oral evidence given by Mr. Phillips and Mr. Walker	193 to 198
(6) Discussion and Analysis	199 to 216
Footnote: Age discrimination	217 to 225
Conclusion	226

The essential facts of each case

(a) Mr. Rutherford

7. The essential facts are not in dispute, and can be quite shortly stated. Mr. Rutherford was born on 12 May 1931, and is thus 72 at the date of this judgment. He was employed by a company called Harvest Town Circle Limited (Harvest) as a production pattern room controller. Harvest was not an undertaking in which there was a normal retiring age for an employee holding Mr. Rutherford's position. According to his Form IT1, which is dated 14 December 1998, his employment with Harvest began on 5 April 1995, shortly before his 64^th birthday, and continued until 25 September 1998, when he was made redundant. He was then 67. He claimed redundancy and unfair dismissal, and invited the Tribunal to refer his case for a preliminary hearing to determine whether or not it had jurisdiction to entertain his claim.
8. Mr. Rutherford's case was heard by the Employment Tribunal for the first time on 16 July 1999. As already stated, the Tribunal concluded that the upper age default provisions in the Act were indirectly discriminatory and had not been objectively justified by Harvest. Harvest filed a notice of appeal against that decision. However, before the appeal could be heard, the company passed a winding up resolution. This rendered the Secretary of State potentially liable to meet Mr. Rutherford's claim for a redundancy payment and any basic award of compensation made for unfair dismissal. As a consequence, the Secretary of State elected to appear in the case, and was made a party to the appeal.
9. Harvest's appeal was heard in the EAT on 11 and 12 June 2001 by the then President, Lindsay J, sitting with Mr. P. Dawson and Mr. T.C. Thomas. The judgment of the EAT was handed down on 10 July 2001, and. as previously stated, is reported. The EAT allowed Harvest's appeal. It set aside the Tribunal's finding on disparate impact. It also held that the Tribunal had erred in law in failing to invite the Secretary of State to address it on objective justification. It remitted the matter for rehearing. On 25 October 2001, the Employment Tribunal ordered that Mr. Rutherford's case should be heard together with that of Mr. Bentley.
10. Mr. Robin Allen QC, for both Mr. Rutherford and Mr. Bentley described this part of the history as Rutherford 1, as will we. He placed considerable reliance on the judgment of the EAT in Rutherford 1. We will, of course, return to consider the EAT's judgment in due course

(b) Mr. Bentley

11. Mr. Bentley was born on 23 September 1927. He will thus be 76 by the time this judgment is handed down. He was employed by a company called Bodner Elem Ltd from 1978, when he was about 50 to 9 February 2001, when the company went into administrative receivership. He was then 73. In his form IT1 he claimed he was owed unpaid wages, and sought a redundancy payment to reflect his full length of service. As his employer was insolvent, he made his claim fair and square against the Secretary of State on the basis of discrimination contrary to Article 141 of the Treaty of Rome and the Equal Treatment Directive.
12. We were told by Mr. Allen that subsequent to the decision of the Tribunal, a further hearing had taken place on the 15^th January 2003 in relation to quantum, in which it had been held that Mr. Bentley was entitled to £1498.50 from the Secretary of State pursuant to sections 166 – 170 of the Act; and that Mr. Rutherford was found to have been unfairly dismissed and entitled to £990 by way of a basic award and £4500 by way of a compensatory award. Mr. Allen told us that it was unlikely that Mr. Rutherford would be able to recover the compensatory award as his employers, like those of Mr. Bentley, were insolvent. However, subject to the outcome of the appeals, the Secretary of State would be required to pay the basic award pursuant to sections 182 and 184(1)(d) of the Act.

The domestic legislation in outline

13. Part X (comprising sections 94 to 134(A)) of the Act deals with the concept of unfair dismissal. Chapter 1 of Part X deals with the right not to be unfairly dismissed. Section 94(1) establishes the general principle that an employee has the right not to be unfairly dismissed by his employer; and section 111(1) provides that a complaint may be presented to an employment tribunal against an employer by any person claiming that he was unfairly dismissed by the employer. The remaining sections set out the remedies which the Tribunal may grant.
14. Part XI (sections 135 to 181) of the Act deals with redundancy payments. Section 135(1) provides that an employer shall pay a redundancy payment to any employee of his if the employee is dismissed by reason of redundancy or is eligible for a redundancy payment by reason of being laid off or kept on short-time. Chapter VI sets out the circumstances in which application for redundancy payments may be made to the Secretary of State, and Part XII deals with the employee's rights against the Secretary of State where the employer becomes insolvent.
15. Both sections 94 and 135(1) of the Act are subject to an upper age limit. In undertakings where there is no "normal retiring age" (which for this purpose must be the same for both men and women), section 109(1)(b) of the Act provides that the right not to be unfairly dismissed (and hence the right to make a complaint to an Employment Tribunal) is removed if on or before the effective date of the termination of the employee's employment he or she has attained the age of 65. Similarly, by reason of section 156(1)(b) of the Act, employees (both men and women) do not have any right to a redundancy payment if, before the date on which they were made redundant, they had attained the age of 65.

The relevant domestic legislation in greater detail:

(a) Section 108: the qualifying period

16. Section 108 of the Act provides that section 94 (the right not to be unfairly dismissed) does not apply to the dismissal of an employee unless he has been continuously employed for a period of not less than one year ending with the effective date of termination.

(b) The upper age limits

17. Although we are only concerned in these appeals with the upper age limit default provisions contained in sections 109(1)(b) and 156(1)(b) of the Act, it is, we think, relevant to set out the whole of sub-section (1) of each section. Thus section 109 of the Act, which is headed Upper Age Limit reads: -

(1) Section 94 does not apply to the dismissal of an employee if on or before the effective date of termination he has attained –

(a) in a case where –

(i) in the undertaking in which the employee was employed there was a normal retiring age for an employee holding the position held by the employee, and

(ii) the age was the same whether the employee holding that position was a man or a woman, that retiring age, and

(b) in any other case, the age of 65.

18. Sections 155 and 156 of the Act form part of Chapter IV entitled "GENERAL EXCLUSIONS FROM RIGHT". Section 155 provides that an employee does not have any right to a redundancy payment unless he has been continuously employed for a period of not less than two years ending with the relevant date (as defined by section 153 of the Act). Section 156, again headed Upper Age Limit reads: -

(1) An employee does not have any right to a redundancy payment if before the relevant date he has attained –

(a) in a case where –

(i) in the business for the purposes of which the employee was employed there was a normal retiring age of less than 65 for an employee holding the position held by the employee, and

(ii) the age was the same whether the employee holding that position was a man or a woman,

that normal retiring age, and

(b) in any other case, the age of 65

19. In its conclusion, the Tribunal also disapplied sections 119(4) and 162(4) as being contrary to European Community Law. These concern the calculation of the basic award of compensation following successful claims for unfair dismissal and redundancy. The latter provides for a reduction in the payment where the employee was made redundant after his sixty-fourth birthday. We do not think we need to set out these provisions.

Article 141 (formerly 119) of the Treaty of Rome

20. The relevant clauses of Article 141 of the Treaty of Rome read as follows: -

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

2. For the purposes of this Article, "pay" means the ordinary basic or minimum wage or salary and any other consideration, whether in cash or in kind, which the worker receives directly or indirectly, in respect of his employment, from his employer.

Equal pay without discrimination based on sex means –

(a) that pay for the same work at piece rates shall be calculated on the basis of the same unit of measurement;

(b) that pay for work at time rates shall be the same for the same job

Section 2(1) of the European Communities Act 1972

21. Section 2(1) of the European Communities Act 1972 provides that: -

All such rights, powers, liabilities, obligations and restrictions from time to time created or arising by or under the Treaties, and all such remedies and procedures from time to time provided for by or under the Treaties, as in accordance with the Treaties are without further enactment to be given legal effect or used in the United Kingdom shall be recognised and available in law, and be enforced, allowed and followed accordingly; and the expression "enforceable Community right" and similar expressions shall be read as referring to one to which this subsection applies.

Relevant Directives from the Council of the European Union (Council Directives)

22. We were referred to Council Directive 76/207/ EEC of 9 February 1976 on the implementation of the principle of equal treatment for men and women as regards access to employment, vocational training and promotion, and working conditions (the Equal Treatment Directive); to Council Directive 79/7/EEC of 19 December 1978 on the progressive implementation of the principle of equal treatment for men and women in matters of social security (Equal treatment in Matters of Social Security); to Council Directive 97/80/EC of 15 December 1997 on the burden of proof in cases of discrimination based on sex (the Burden of Proof Directive); and Council Directive 2000/78/EC of 27 November 2000 entitled "establishing a general framework for proof in cases of discrimination based on sex" (the Framework Directive).
23. The concept of indirect discrimination is perhaps most conveniently to be found in the definition contained in Article 2.1 of the Burden of Proof Directive (97/80/EC). This states that the principle of equal treatment means that "there shall be no discrimination whatsoever based on sex, either directly or indirectly". Indirect sex discrimination is defined by Article 2.2 as follows:

For the purposes of the principle of equal treatment referred to in paragraph 1, indirect discrimination shall exist where an apparently neutral provision, criterion or practice disadvantages a substantially higher proportion of the members of one sex unless that provision, criterion or practice is appropriate and necessary and can be justified by objective factors unrelated to sex.

The Application of the Equal Pay provisions of the Treaty to applications for unfair dismissal and redundancy payments

24. The manner in which Article 141 of the Treaty applies to Mr Rutherford and Mr Bentley's applications was aptly summarised by Lindsay J when giving the judgment of the EAT in Rutherford 1. He said: -

The argument which succeeded below can be briefly stated. It is this: Article 119 of the Treaty of Rome (now Article 141) requires observance by Member States of the principle that men and women should receive equal pay for equal work. Barber v Guardian Royal Exchange Group [1990] IRLR 240 (ECJ) illustrates that if, by whatever route, be it direct or indirect, any such inequality in pay arises, it is to be prohibited and that national courts have a duty to safeguard individuals' rights to equality in pay and thus, if necessary, to override any national measures which infringe the Article. Awards in respect of unfair dismissal or redundancy payments are 'pay' within the meaning of Article 119; see, respectively, R v Secretary of State for Employment ex parte Seymour-Smith and Perez [1999] IRLR 253 (ECJ) and Barber (supra). Accordingly, if a national measure as to entitlement to unfair dismissal compensation or to a redundancy payment can be seen to be, for example, indirectly discriminatory to a particular extent as between men and women, then it is to be disapplied by the national court, which is obliged to pay proper deference to Article 141. For that purpose such a national measure is to be regarded as indirectly discriminatory to the particular extent if it imposes a requirement which a considerably smaller percentage of women than men, or men than women, is able to fulfil and where the discriminatory measure is not justified by objective factors unrelated to any discrimination on the grounds of sex.

6. Thus far little or no dispute, as we understand it, arises between the parties on the law.

25. We gratefully adopt this summary as a neat and accurate exegesis of the underlying position. The two questions which arise for consideration, therefore, are (1) do the upper age limit default provisions contained in sections 109(1)(b) and 156(1)(b) of the Act have a disparate impact on men?; and (2) if the answer to (1) is "yes" can the provisions be objectively justified? We will henceforth adopt the common shorthand and identify the two concepts as "disparate impact" and "objective justification". As will be apparent, the Tribunal answered the first question "yes" and the second "no".

Seymour-Smith

(a) Introduction

26. Before turning to Rutherford 1 and the reasoning of the Tribunal in the instant case, we think we should examine the leading case on disparate impact and objective justification. Its full title is R v Secretary of State for Employment ex parte Seymour-Smith and another. We shall refer to it throughout as Seymour-Smith. It began as an application for judicial review, which failed in the Divisional Court but succeeded in the Court of Appeal. Both decisions are reported at [1995] ICR 889. The Secretary of State appealed to the House of Lords, and the Applicants cross-appealed. The House of Lords referred a number of questions to the European Court of Justice (ECJ): see [1997] ICR 371. The ECJ's decision is reported at [1999] ICR 447. The case then returned to the House of Lords which, by a majority, allowed the Secretary of State's appeal: [2000] ICR 244.
27. The Applicants in Seymour-Smith were female employees who had been dismissed before they had completed the two year period of continuous employment which, as the law then stood, was the qualifying period for the right not to be unfairly dismissed and thus to be able to apply to an Employment Tribunal. In each case, they were unable to register their applications to a Tribunal, and as a consequence sought judicial review of the statutory instrument which had increased the qualifying period from one year to two.. They contended that the increase in the qualifying period was indirectly discriminatory against women, contrary to the Equal Treatment Directive because fewer women than men were able to comply with it.

(b) Seymour-Smith in the Divisional Court and in the Court of Appeal

28. A critical feature of the case throughout was, of course, the nature and content of the evidence required to establish disparate impact. The Divisional Court (Balcombe LJ and McCullough J) dismissed the application on the ground that it had not been established that the proportion of women who could comply with the qualifying period was considerably smaller than the proportion of qualifying me. Balcombe LJ summarised the arguments placed before the court as follows: - [1995] ICR 889 at 903E to H -

So the first question we have to resolve is: What is the test which the Court should apply in deciding whether the effect of the two-year qualifying period prima facie discriminates against women? On behalf of the Secretary of State Mr Richards submits that the comparison is between the proportion of women who can comply with the qualification and the proportion of men who can comply with the qualification, and the former proportion must be considerably smaller than the latter. On behalf of the applicants Mr Allen submits that the essential comparison is between the proportion of women who cannot comply with the qualification (ie those who are disadvantaged by the qualification) and the proportion of men who cannot comply with the qualification, and then to consider the ratio between these two proportions. An example taken from the evidence will illustrate the difference between these two approaches. In the year 1985, out of a total workforce of some 18.73 million persons, 61% were male and 39% female; 77.4% of the male workers had two years' service or more, but only 68.9% of the female workforce. So the proportion of females to males with two years' or more service was 89:100. If, on the other hand, one looks at those with less than two years' service, 31.1% of the female workforce had less than two years' service compared with 22.6% of the male workforce, ie about 3:2.

29. After reference to a number of United Kingdom and European authorities, Balcombe LJ continued: [1995] ICR 889 at 904 E to H: -

In my judgment Mr Allen's submission cannot be correct unless it is qualified – which, to be fair to him, he concedes may be necessary. In paragraph 17 of his second affidavit on behalf of the Secretary of State, Mr Peter Parker makes the following point:

'If the question is asked whether a considerably greater proportion of women than of men cannot comply with a given requirement or condition ... the results can be seriously misleading. If a hypothetical requirement is considered which 99.5% of men can comply with but only 99% of women, it follows that the proportion of women who cannot comply with the requirement is twice as great as the proportion of men who cannot comply. The figures relevant to the present proceedings are very much less extreme; nevertheless the arguments advanced ... are of the same kind and are, it is contended, similarly misleading.'

The logic of this argument is unanswerable and, indeed, unanswered by the applicants or their expert witnesses.

30. McCulloch J agreed. He concluded:

My conclusion is that, in considering whether there is considerable disparity, the court should look both at the relative percentages of those who meet the requirement and at the relative percentages of those who do not. Of these, the more important group will be those who qualify. The following example makes the point. If 98% of men qualify and 2% do not, and if 96% of women qualify and 4% do not, it would not be right to conclude that the disparity was considerable. But if only 4% of men and only 2% of women qualified the opposite conclusion might well be correct.

31. The Court of Appeal allowed the applicants' appeal and made a declaration that the Statutory Instrument (SI) introducing the two year qualifying period was incompatible with the Equal Treatment Directive. They declined, however, to adjudicate on the question whether or not the SI infringed what was then Article 119 of the Treaty. The Court of Appeal held, (at [1995] ICR 889 at 953): -

Despite our anxieties, however, we have been forced to the conclusion that on the material before us it has been demonstrated that for the period leading up to the dismissals of the two applicants there had been and continued to be a considerable and persistent difference in the numbers and percentages of men and women in the groups that did comply and the groups that did not comply with the two year qualification period.

(c) The reference to the European Court of Justice

32. The House of Lords reversed the decision of the Court of Appeal to the extent of discharging the declaration, but adjourned the appeal sine die pending a reference to the ECJ under Article 177 of the Treaty of five questions, three of which are material for present purposes namely: -

3. What is the legal test for establishing whether a measure adopted by a Member State has such a degree of disparate effect as between men and women as to amount to indirect discrimination for the purposes of Article 119 of the EC Treaty unless shown to be based upon objectively justified factors other than sex?

4. When must this legal test be applied to a measure adopted by a Member State? In particular at which of the following points in time, or at what other point in time, must it be applied to the measure:

(a) When the measure is adopted.

(b) When the measure is brought into force.

(c) When the employee is dismissed.

5. What are the legal conditions for establishing the objective justification, for the purposes of indirect discrimination under Article 119, of a measure adopted by a Member State in pursuance of its social policy? In particular, what material need the Member State adduce in support of its grounds for justification?"

(d) The judgment of the European Court of Justice

33. In the course of its answer to question 3, the ECJ said: ([1999] ICR 447 at 490E to 491D)

58. As regards the establishment of indirect discrimination, the first question is whether a measure such as the rule at issue has a more unfavourable impact on women than on men.

59. Next, as the United Kingdom Government was right to point out, the best approach to the comparison of statistics is to consider, on the one hand, the respective proportions of men and in the workforce able to satisfy the requirement of two years' employment under the disputed rule and of those unable to do so, and, on the other, to compare those proportions as regards women in the workforce. It is not sufficient to consider the number of persons affected, since that depends on the number of working people in the Member State as a whole as well as the percentages of men and women employed in that State.

60. As the Court has stated on several occasions, it must be ascertained whether the statistics available indicate that a considerably smaller percentage of women than men is able to satisfy the condition of two years' employment required by the disputed rule. That situation would be evidence of apparent sex discrimination unless the disputed rule were justified by objective factors unrelated to any discrimination based on sex.

61. That could also be the case if the statistical evidence revealed a lesser but persistent and relatively constant disparity over a long period between men and women who satisfy the requirement of two years' employment. It would, however, be for the national court to determine the conclusions to be drawn from such statistics.

62. It is also for the national court to assess whether the statistics concerning the situation of the workforce are valid and can be taken into account, that is to say, whether they cover enough individuals, whether they illustrate purely fortuitous or short-term phenomena, and whether, in general, they appear to be significant (see case C-127/92 Enderby [1993] IRLR 591, paragraph 17). It is, in particular, for the national court to establish whether, given the answer to the fourth question, the 1985 statistics concerning the respective percentages of men and women fulfilling the requirement of two years' employment under the disputed rule are relevant and sufficient for the purposes of resolving the case before it.

63. In this case, it appears from the order for reference that in 1985, the year in which the requirement of two years' employment was introduced, 77.4% of men and 68.9% of women fulfilled that condition.

64. Such statistics do not appear, on the face of it, to show that a considerably smaller percentage of women than men is able to fulfil the requirement imposed by the disputed rule.

65. Accordingly, the answer to the third question must be that in order to establish whether a measure adopted by a Member State has disparate effect as between men and women to such a degree as to amount to indirect discrimination for the purposes of Article 119 of the Treaty, the national court must verify whether the statistics available indicate that a considerably smaller percentage of women than men is able to fulfil the requirement imposed by that measure. If that is the case, there is indirect sex discrimination, unless that measure is justified by objective factors unrelated to any discrimination based on sex.

34. The answer the ECJ gave to the fourth question was that it was for the national court, taking into account all the material and factual circumstances, to determine the point in time at which the legality of a rule of the kind at issue is to be assessed ([1999] ICR 447 at 489, paragraph 50). However, the ECJ also pointed out, at ibid paragraph 45 that "the requirements of Community law must be complied with at all relevant times, whether that is the time when the measure is adopted, when it is implemented or when it is applied to the case in point".
35. In answer to question 5 in relation to objective justification, the ECJ stated firstly that objective justification was ultimately for the national court, which has sole jurisdiction to assess the facts and interpret the national legislation (ibid paragraph 67) . The ECJ could, however, provide guidance (ibid paragraph 68). The judgment continues: [1999] ICR 447 at 491H to 492G -

69. It is settled case law that if a Member State is able to show that the measures chosen reflect a necessary aim of its social policy and are suitable and necessary for achieving that aim, the mere fact that the legislative provision affects far more women than men at work cannot be regarded as a breach of Article 119 of the Treaty (see, in particular, case C-444/93 Megner and Scheffel v Innungs-krankenkasse Vorderpfalz [1996] IRLR 236, paragraph 24, and Freers and Speckmann, cited above, paragraph 28).

70. In this case, the United Kingdom Government contends that the risk that the exposure of employers to proceedings for unfair dismissal brought by employees who had only fairly recently been engaged is a deterrent to recruitment, so that extension of the qualifying period for protection against dismissal would stimulate recruitment.

71. It cannot be disputed that the encouragement of recruitment constitutes a legitimate aim of social policy.

72. It must also be ascertained, in the light of all the relevant factors and taking into account the possibility of achieving the social policy aim in question by other means, whether such an aim appears to be unrelated to any discrimination based on sex and whether the disputed rule as a means to its achievement, is capable of advancing that aim.

73. In that connection, the United Kingdom Government maintains that a Member State should merely have to show that it was reasonably entitled to consider that the measure would advance a social policy aim. It relies to that end on case C-317/93 Nolte [1996] IRLR 225.

74. It is true that in paragraph 33 of the Nolte [1996] IRLR 225 case the Court observed that, in choosing the measures capable of achieving the aims of their social and employment policy, the Member States have a broad margin of discretion.

75. However, although social policy is essentially a matter for the Member States under Community law as it stands, the fact remains that the broad margin of discretion available to the Member States in that connection cannot have the effect of frustrating the implementation of a fundamental principle of Community law such as that of equal pay for men and women.

76. Mere generalisations concerning the capacity of a specific measure to encourage recruitment are not enough to show that the aim of the disputed rule is unrelated to any discrimination based on sex nor to provide evidence on the basis of which it could reasonably be considered that the means chosen were suitable for achieving that aim.

77. Accordingly, the answer to the fifth question must be that if a considerably smaller percentage of women than men is capable of fulfilling the requirement of two years' employment imposed by the disputed rule, it is for the Member State, as the author of the allegedly discriminatory rule, to show that the said rule reflects a legitimate aim of its social policy, that that aim is unrelated to any discrimination based on sex, and that it could reasonably consider that the means chosen were suitable for attaining that aim.

(e) The decision of the House of Lords: disparate impact

36. When the case reached the House of Lords for the second time, the applicants succeeded on disparate impact, but failed on objective justification. In each case, the decision was by a majority of 3:2. The appeal was accordingly allowed. The leading speech for the majority was given by Lord Nicholls of Birkenhead, and for the minority by Lord Slynn of Hadley. On disparate impact, it is common ground that the latter summarised the position of the ECJ accurately when he said at [2000] ICR 244 at 250H: -

The European Court has, as I read it, accepted that the starting point is to look at the percentage of those who do qualify and to ask, do these figures cover enough individuals, are they purely fortuitous or significant or are the figures 'relevant and sufficient for the purposes of resolving the case before [the national Court]'? If these figures are not sufficient or significant, it may be necessary to look elsewhere, including, as the Divisional Court accepted, the figures of those who do not qualify.

37. The statistics before the House of Lords, which were presented in tabular form, demonstrated that between 1985 and 1993, the percentage of males who qualified with more than two years continuous employment fluctuated between at its lowest 72% in 1989 and 78.4% in 1993. The corresponding percentages for females over the same period fluctuated from 63.8% in 1989 to 74.1% in 1993. At no point was the percentage of females who qualified greater than that of males, but the difference varied from a minimum of 4.3% in 1993 to 8.8% in 1986. The average difference was 7.45%.
38. For the same period, the percentage of males with less than 2 years continuous experience fluctuated from 21.6% in 1993 to 36.2% in 1989. The corresponding figures for females fluctuated from 25.9% in 1993 and 36.2% in 1989. As with qualified employees, at no point was the percentage of unqualified females less than that of unqualified males. The difference in the percentages varied from a maximum of 8.7% (1986) and 4.3% (1993). The average difference was 7.42%.
39. For the minority, the statistics adduced did not constitute disparate impact. Lord Slynn commented: -

Although the percentage of women who can comply may be 'lesser' than 'considerably smaller' than the percentage of men who can comply if the percentage is persistent and relatively constant, I do not read the Court as saying that any difference between men and women which continues over a period itself establishes discrimination contrary to Article 119. The difference is evidence that 'could' establish such discrimination, but it is 'for the national court to determine the conclusion to be drawn from such statistics'. It seems to me that the disparity, even if not establishing a 'considerably smaller' percentage of women able to qualify must still be significant so as to satisfy the Court that it would be right to find indirect sex discrimination.

Having analysed the figures, Lord Slynn concluded: -

The Court of Appeal [1995] IRLR 464, 476 were impressed by the fact that in 1985 'there were 370,000 fewer women in the advantaged group than predicted', though no figures are relied on by the Court of Appeal for later years and they do not say whether there were less men in the advantaged group than 'predicted'. In any event, the European Court of Justice knew of this figure from the Court of Appeal judgment and did not suggest that it was sufficient 'on the face of it' to require a finding in respect of 1985 that there was a considerably smaller percentage of women who qualify. In my view, in the light of the numbers quoted by the Divisional Court on which the percentages are based, it is clear that the percentages can be relied on for the purposes of considering this issue in accordance with the judgment of the European Court.

In all the circumstances, in the light of the European Court's judgment, I am satisfied that the percentages of those qualifying in 1985 do not establish discrimination contrary to Article 119, and that the percentages of those qualifying in and leading up to 1991 do not establish discrimination at the time of dismissal. Although the 1985 Order did prevent a higher percentage of women than men from claiming compensation, it cannot be said that it 'actually affects a considerably higher percentage of women than men' (paragraph 67 of the European Court's judgment). It would in any event, however much the result could be justified logically, be odd if there was no discrimination in 1985, but in 1991 on a slightly higher percentage of women qualifying (and one as part of a rising trend) there was discrimination.

On these percentages it does not seem to me to be necessary or helpful to consider separately the figures for those who did not qualify.

40. For the majority, Lord Nicholls of Birkenhead, working solely on the statistics of those with more than two years continuous employment, identified the percentage disparity over the 9 years in the following way: -

These figures show that over the period of seven years, from 1985 up to and including 1991, the ratio of men and women who qualified was roughly 10:9. For every 10 men who qualified, only nine women did so. This disparity was remarkably constant for the six years from 1985 to 1990, but it began to diminish in 1991.

These figures are in borderline country. The question under consideration is one of degree. When the borderline is defined by reference to a criterion as imprecise as 'considerably smaller' it is inevitable that in some cases different minds may reach different conclusions. The decisions of the two courts below illustrate this. My own impression differs from a minority of your Lordships. I find myself driven to the conclusion that a persistent and constant disparity of the order just mentioned in respect of the entire male and female labour forces of the country over a period of seven years cannot be brushed aside and dismissed as insignificant or inconsiderable. I agree with the Court of Appeal that, given the context of equality of pay or treatment, the latitude afforded by the word 'considerably' should not be exaggerated. I think these figures are adequate to demonstrate that the extension of the qualifying period had a considerably greater adverse impact on women than men.

The reduction in the disparity, which started in 1991, continued in 1992 and 1993. By 1993 the ratio of men and women qualifiers was about 20:19. But, looking at the overall picture, I do not think the diminished disparity after 1991 is sufficient to displace the message of the figures for the earlier years. Accordingly, it is for the government to show that the extension of the qualifying period was justified, to use the accepted nomenclature, by objective factors unrelated to any discrimination based on sex. On this issue I agree with the Court of Appeal.

41. Before leaving the point, however, Lord Nicholls went on to address the question of qualifiers and non-qualifiers. It needs to be recalled that he had reached his decision on disparate impact by looking only at the qualifiers. He cited paragraph 59 of the judgment of the ECJ (which we have cited at paragraph 33 above) and continued ([2000] ICR 244 at 259D): -

This statement appears to envisage that two comparisons should be made: a comparison of the proportions of men and women able to satisfy the requirement ('the qualifiers'), and a comparison of the proportions of men and women unable to satisfy the requirement ('the non-qualifiers'). Thereafter in its judgment the Court considered only the proportions of men and women who were qualifiers.

Some of the ramifications involved in looking at the composition of the disadvantaged group, as well as the composition of the advantaged group, were explored by the Divisional Court and the Court of Appeal in the present case. Suffice to say, I do not understand the European Court to have rejected use of the figures relating to the non-qualifiers in a suitable case. Indeed, the European Court has looked at the composition of the disadvantaged group in several cases, although in none of them was there an issue on this point: see, for instance, Bilka-Kaufhaus GmbH v Weber von Hartz (case 170/84) [1986] IRLR 317, 320, paragraph 31, Nimz v Freie und Hansestadt Hamburg (case C-184/89) [1991] IRLR 222, 225, paragraph 12, and Kowalska v Freie und Hansestadt Hamburg (case C-33/89) [1990] IRLR 447, 449, paragraphs 13–16. Having regard to the conclusion I have expressed above on the issue of disparate impact, it is unnecessary to reach a firm conclusion on this point. I prefer to leave this question open for another occasion

(f) The Decision of the House of Lords: Objective Justification.

42. On objective justification, the majority found for the Secretary of State. Lord Nicholls identified the test applied by the Divisional Court and the Court of Appeal, namely whether the two year threshold had been proved to result in greater availability of employment than would have been the case without it. At [2000] ICR 244 at 260E-F, Lord Nicholls said: -

The answer given by the European Court to the fifth question referred to the Court by this House has now shown that this test was too stringent. 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 European Court has recognised these practical considerations. If their aim is legitimate, governments have a discretion when choosing the method 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.

43. Later, at [2000] ICR 244 at 261G to 262F, under the heading Objective justification: the continuing operation of the Order of 1985 he added:

The requirements of Community law must be complied with at all relevant times. A measure may satisfy Community law when adopted, because at that stage the minister was reasonably entitled to consider the measure was a suitable means for achieving a legitimate aim. But experience of the working of the measure may tell a different story. In course of time the measure may be found to be unsuited for its intended purpose. The benefits hoped for may not materialise. Then the retention in force of a measure having a disparately adverse impact on women may no longer be objectively justifiable. In such a case, a measure, lawful when adopted, may become unlawful.

Accordingly, if the government introduces a measure which proves to have a disparately adverse impact on women, the government is under a duty to take reasonable steps to monitor the working of the measure. The government must review the position periodically. The greater the disparity of impact, the greater the diligence which can reasonably be expected of the government. Depending on the circumstances, the government may become obliged to repeal or replace the unsuccessful measure.

In the present case, the 1985 Order had been in operation for six years when the two claimants were dismissed from their jobs. The Divisional Court and the Court of Appeal noted there was no evidence that the extension of the qualifying period in 1985 led to an increase in employment opportunities. Ought the government to have taken steps to repeal the 1985 Order before 1991? In other words, had the Order, lawful at its inception, become unlawful by 1991?

Here again, the matter is debatable. As time passed, the persistently adverse impact on women became apparent. But, as with the broad margin of discretion afforded to governments when adopting measures of this type, so with the duty of governments to monitor the implementation of such measures: the practicalities of government must be borne in mind. The benefits of the 1985 Order could not be expected to materialise overnight, or even in a matter of months. The government was entitled to allow a reasonable period to elapse before deciding whether the Order had achieved its objective and, if not, whether the Order should be replaced with some other measure or simply repealed. Time would then be needed to implement any decision. I do not think the government could reasonably be expected to complete all these steps in six years, failing which it was in breach of Community law. The contrary view would impose an unrealistic burden on the government in the present case. Accordingly, I consider the Secretary of State discharged the burden of showing that the 1985 Order was still objectively justified in 1991.

Harvest Town Circle Limited v Rutherford (Rutherford 1)

44. Before we can examine the reasoning of the Tribunal in the instant case it is necessary to look carefully at the judgment of the EAT in Rutherford 1, the history of which we have given in paragraphs 7 to 10. We do not have the reasons of the Employment Tribunal at Stratford from which the appeal was made. We do, however, have the judgment of the EAT given by Lindsay J. Having set out the underlying law in the passage we have already quoted in paragraph 24, Lindsay J summarised the decision of the Tribunal in the following way: -

6. …….However, the employment tribunal continued, and seems to have held (although never in terms precisely saying this), that, on their analysis of the figures presented to them, the statutory provisions we have mentioned did, indeed, affect to their disadvantage more than double the percentage of relevant men than of relevant women and that accordingly our domestic legislation is indirectly discriminatory on the grounds of sex unless it could properly be regarded as objectively justified by reference to features other than a difference of sex. The employment tribunal felt unable to conclude there was such objective justification. Accordingly (although, again, this is not clearly said in terms) ss.109 and 156 were to be overborne and the employment tribunal accordingly had jurisdiction to hear Mr Rutherford's claims for unfair dismissal and redundancy payments notwithstanding his age and those statutory provisions.

7. At the risk of oversimplification, we see the principal questions arising on the appeal to be related, firstly, to whether the statistics presented to the employment tribunal and the tribunal's analysis of them provided an adequate base in law for the tribunal's conclusion that the national measures in issue were indirectly sexually discriminatory and, secondly, to whether the tribunal's handling of the issue of objective justification was appropriate. We need to ask ourselves whether there were errors of law in either of those areas.

45. Having set out the terms of Article 141, Lindsay J conducted an extensive review of the European authorities and concluded: -

Thus, even before coming to Seymour-Smith [1999] ICR 447 in the European Court of Justice, it can be seen, upon looking at European Court of Justice cases, that there had been no constancy in cleaving to a percentage or proportion as opposed to numbers or to the difference between the two (whether expressed as the one being higher or lower than the other) as needing to be 'substantial' as opposed to 'considerable'.

46. Lindsay J then cited paragraphs 59 and 60 from the judgment of the ECJ in Seymour-Smith which we have set out at paragraph 33 above, and commented ([2002] ICR 123 at 128F-G:

The ECJ was there going in more detail than ever before into the issues and the reference there to the 'best approach' suggests that there are more than two possible approaches; one sees also that a reference to numbers alone will not be sufficient. The passage also shows that at this point in the judgment the Court was looking at the respective percentages of men and women, firstly, able to comply and, secondly, unable to comply with the requirement in issue. Four figures were thus involved. That possible enlargement of the approach to the question was then, as it seems, shrunk back to a possibly more limited test as, in its paragraph 65 on p.278 the ECJ concluded:

'... The national court must verify whether the statistics available indicate that a considerably smaller percentage of women than men is able to fulfil the requirement imposed by that measure. If that is the case, there is indirect sex discrimination, unless that measure is justified by objective factors unrelated to any discrimination based on sex.'

47. Lindsay J then turned to examine the relevant English authorities, notably the decisions of the House of Lords in Barry v Midland Bank plc [1999] ICR 859 (Barry) and Seymour-Smith [2000] ICR 244, following the reference to the ECJ. The outcome of this analysis of both the European and English authorities was that the EAT laid down the following guidelines for the Tribunal to which it was remitting Rutherford 1.

18. Where does this leave us? Again leaving aside the cases where a smaller but persistent and constant disparity appears, we believe the authorities are to be synthesised and may be extended as follows:

(i) There will some cases where, on the statistics, a disparate impact is so obvious that a look at numbers alone or proportions alone, whether of the advantaged (qualifiers) or disadvantaged (non-qualifiers), will suffice beyond doubt to show that members of one sex are substantially or considerably disadvantaged in comparison with those of the other.

(ii) However, in less obvious cases it will be proper for an employment tribunal, as the national court of fact, to use more than one form of comparison, no one of which is necessarily to be regarded as on its own decisive.

(iii) In such less obvious cases it will be proper for the employment tribunal to look not merely at proportions (as proportions alone can be misleading) but also at numbers, and to look at both disadvantaged and non-disadvantaged groups and even to the respective proportions in the disadvantaged groups expressed as a ratio of each other.

(iv) It will never be wrong for a tribunal to look at more than one form of comparison, if only to confirm that the case remains as obvious as it had first appeared. Moreover, if there is any doubt as to the obviousness of the case, the tendency should always be to look at a second or further form of comparison.

(v) As more cases of indirect discrimination are heard, a better feel, a more soundly based assessment of what is or is not properly to be regarded as a considerable or substantial disparity will develop. For example, in cases similar to it, regard will doubtless be had to the fact that in Seymour-Smith [1999] IRLR 253 the ECJ found the 1985 disparity of 8.5% to be not considerable - see, supra [1999] IRLR 278 - and, when the case returned to the House of Lords, see Lord Goff at [2000] IRLR at p.268 and Lord Nicholls at pp.269-270, Unfortunately, those seeking a simple and universal touchstone must be disabused; as numbers as well as proportions will be likely, save in the most obvious cases, to need to be taken into account and as different forms of comparison, of their nature, throw up different scales of difference, it would be a mistake (even leaving aside the cases of persistent but constant smaller disparities) to think that any disparity of less than 8.5% must necessarily be found not substantial or not considerable.

(vi) No distinction is to be drawn between a considerable and a substantial disparity. That being so, it would be a mistake to conclude that anything that was merely not trivial or de minimis sufficed.

(vii) The employment tribunal, in such less obvious cases, after looking in detail at such figures as should have been laid before it, must then stand back, as it were, and, assimilating all the figures, is then to judge whether the apparently neutral provision, criterion or practice in issue has a disparate impact, be it on men or women, that could fairly be described as considerable or substantial.

48. The EAT then turned to examine the statistics which had been available to the Tribunal in Rutherford 1 and the manner in which the Tribunal had treated those statistics. This is what the Tribunal had said:

We believe that the statistics (which we are told to do not go back beyond 1993) show that throughout a considerably higher percentage of men over 65 being economically active in all the years from 1993-1998 inclusive. Contrariwise there is a considerably smaller percentage of economically active women over 65 during the same period. In 1998, the year of dismissal, 3% of just over 5,000,000 women over 65 were economically active, that is in employment or available to work within two weeks. This compares with 8% of just over 3,500,000 men over 65 who were economically active. Looking in particular at the percentages, we note that throughout the entire period the percentage of economically active males are [sic] more than double the percentage of economically females. Accordingly in our decision the domestic legislation is indirectly discriminatory on grounds of sex unless it can objectively justified by other factors than a difference of sex.'

49. The EAT commented: -

20. There are, as it seems to us, serious flaws in that approach. To illustrate by reference only to the figures for 1998, the tribunal's figures can be tabulated as follows:

Females Males

All in employment (a) 168,000 266,000

Economically active (b) 172,000 275,000

Economic activity rate (b/b+c) 3% 8%

Economically inactive (c) 4,834,000 3,345,000

Total 5,006,000 3,620,000

These figures are, firstly, mistaken; the true computation of the 'Economic activity rate' is 3.4% (not 3%) for the female 'economic activity rate' and 7.6% (not 8.0%) for the male. The comparison - 3.0% to 8.0% - on which the tribunal relied was thus not a correct one to make. The 5% gap (8%-3%) was truly one of 4.18% (7.60%-3.44%). With only small figures being in issue, that difference (itself of a reduction of 16.4%) is not to be overlooked.

21. Secondly, although this point was not taken below and is not relied upon before us, we find it hard to see how it can be correct, when assessing the possible disparate effect of unfair dismissal and redundancy provisions on the over-65s, to have in mind all over-65s describable as economically active, a figure which (on the definitions used) will include, for example, numbers of self-employed, numbers of those - directors and partners perhaps - in practical control of their own employment and numbers of persons on fixed term contracts to whom contractual relief would exceed anything statutorily available. It could not be assumed that such classes, unlikely to be affected by unfair dismissal or redundancy legislation, would fall equally or in any other as yet ascertained ratio between men and women.

22. Thirdly, reference to the totals, 5,006,000 women and 3,620,000 men, will surely include literally millions to whom unfair dismissal or redundancy is utterly meaningless; to include, for example, 80 and 90 year olds who have no wish or who have no longer the physical and mental ability to work, amongst the comparison serves only to distort the picture, especially since, as women in general live longer than men, the figure for women over 65 either unable or unwilling to work at any one time is likely to be larger than the corresponding figure for men.

23. Fourthly, these figures throw no real light on the impact of ss.109 and 156 as they look only at those who have survived to 65 and have remained in or have taken up employment. All men and women unfairly dismissed or made redundant upon their attaining 65 will not appear in these figures. Those who retired because they knew that they might otherwise be dismissed, will not appear in the figures. One is thus attempting to judge the impact of the legislation by looking only at those upon whom it has not, at the time of the statistics, had an impact but upon whom it might later have an effect. One is leaving out those upon whom its effect has, by the same date, perhaps already been crucial.

24. These factors, taken together, illustrate, in our view, the inutility of the figures laid before the tribunal: they cannot serve the purpose for which it was intended they should be used.

25. We recognise that in Seymour-Smith [1999] IRLR 253 the figures upon which all courts relied were, as to those who could not comply with the qualification there in issue (of having put in at least two years' service) were figures not of persons actually affected to their disadvantage by the legislation then in issue but of persons who might be, in the sense that it could transpire that they would be dismissed without remedy before they had served the two-year period then required. It thus cannot be said that in all cases figures for those who might be put to a disadvantage are outside the range of that which has to be considered. However, in Seymour-Smith [1999] IRLR 253 the relevant comparisons, as Mr Reade emphasises, included figures for the total workforce. No corresponding figures (ie of those both under and over 65) play any part in the employment tribunal's reasoning in the case at hand.

50. Lindsay J then referred to the well known statement of Sedley LJ in Allonby v Accrington and Rossendale College [2001] ICR 1189, 1198, commenting on a statement of Waite J (as he then was): -

26. Mr Allen, underlining to us that we can only deal with errors of law, reminds us, by analogy with domestic cases as to the composition of 'pools', of the decision of Waite J, President, in Kidd v DRG (UK) Ltd [1985] IRLR 190 EAT paragraph 20 as follows:

'The choice of an appropriate section of the population is our judgment an issue of fact (or perhaps strictly a matter for discretion to be exercised in the course of discharging an exclusive fact-finding function) ...'

That approach could perhaps lead to a conclusion that the employment tribunal's approach was invulnerable as being entirely within its province as a matter of fact. However, that passage has recently attracted some comment in the Court of Appeal in Allonby v Accrington and Rossendale College [2001] EWCA Civ 529; [2001] IRLR 364 at 368, where Sedley LJ says:

'I would sound a strong note of caution about this. As the EAT's excellent analysis of the possible pools shows, once the impugned requirement or condition has been defined there is likely to be only one pool which serves to test its effect. I would prefer to characterise the identification of the pool as a matter neither of discretion nor of fact-finding but of logic. This was the approach adopted by this court in Barry v Midland Bank plc [1998] IRLR 138, 144 and endorsed by Lord Slynn on further appeal [1999] IRLR 581, 584. Logic may on occasion be capable of producing more than one outcome, especially if two or more conditions or requirements are in issue. But the choice of pool is not at large.'

We respectfully adopt that view. Here, in our judgment, logic suggests that the impact, be it even or disparate amongst men and women, of ss.109 and 156 cannot be fairly assessed in any informed way by reference to the figures which the tribunal had laid before it. The tribunal below, through no fault of its own, in our view erred in law in trying to do, and, as it seemed, succeeding in doing, the impossible. Furthermore, the case, to use the language of our earlier analysis of the law, would undoubtedly fall into the 'less obvious' category such that the tribunal should have looked into comparisons other than the (mistaken) one between 3% of just over 5,000,000 and 8% of just over 3,500,000, the only one to which they refer.

51. The EAT declined Mr. Allen's invitation to make a preliminary reference to the ECJ. It allowed the appeal and remitted the matter to the Employment Tribunal.
52. On the question of objective justification, the EAT in Rutherford 1 was of the opinion that the Tribunal had dealt with this in far too summary a fashion, and was critical of the fact that Harvest, a small company, had been left to address this issue on its own, without assistance from the Secretary of State. Lindsay J said: -

28. …..Where the validity of primary legislation affecting literally hundreds of thousands of people is in issue, it seems to us essential, if the parties have not themselves made adequate arrangements for evidence being given on the subject of objective justification, that the tribunal itself should take pains to see that it is sufficiently informed on the subject. The matter cannot properly be left simply on the basis that there is an onus on one party or another and that the onus has been left undischarged. The obvious course, in a case such as the one at hand, was for the employment tribunal to have requested an invitation to be sent to the Secretary of State in order that he might consider what arguments there were by way of objective justification. Given that State and other pensions are commonly paid from age 65 or even earlier and that in a sense it may be that every employee over 65 who remains in work might thereby exclude some younger (unpensioned) person from the same job, the existence of possible objective justifications irrespective of sex seems to us so obvious, and informed argument on the point to be so essential to a just decision, that for the employment tribunal to fail invite the Secretary of State to be joined represents, in our view, an error of law. Of course, he could decline to be joined or, even if joined, could decline to take any active part. If necessary, witness summonses could be used and the Secretary of State may, if necessary, be joined as a party against his will. Where the validity of important primary legislation is in issue it would seem to us little short of irresponsible on the Secretary of State's part for him, without giving reasons, to do nothing to explain the policy considerations which may have lain behind the will of Parliament as expressed in the primary legislation in issue. Where literally hundreds of similar cases could arise, it would plainly be unjust for some cases, indistinguishable from one another, to be decided in his absence to one effect and then others to be decided otherwise because of his presence, nor would it be helpful if he was bound by the result in one (having been joined as a party) but not in others (because he was not so joined).

29. The way in which the subject of objective justification was dealt with below provides, in our view, a second ground for remitting the matter afresh, this time with an invitation being sent to the Secretary of State. He may be joined as a party and is to be invited to draw attention to any factors which he believes may properly be regarded as objective justification for the primary legislation here in issue or, alternatively, to explain, if he chooses not to attempt to justify the legislation which Parliament had passed, why that is so.

53. The EAT then declined to tell the Tribunal what statistics it was to require, largely because the EAT did not know what statistics were available, although Lindsay J commented: -

30. At first blush it seems to us that those put at a disadvantage by the primary legislation in issue would consist of or would need to include all those who, on arriving at age 65, would have wished, and would have been physically and mentally able, to continue in employment properly so called but who either were then dismissed or made redundant by reason of the relative freedom which the legislation conferred upon their employer or who were so fearful of that freedom being exercised against them that they accepted retirement. We would not wish to include in any statistics reference to persons who were neither physically nor mentally able, nor wished, to be employed. However, we have no means of knowing whether statistics of such kinds could be made available or whether adequate inferences sufficient for the task could be drawn from other statistics. Moreover, as the argument before us has largely been, on the one hand, that the statistics presented below were sufficient and, on the other, that what was drawn from them was inappropriate, what range of alternative statistics might have been possible to have been laid before the tribunal below has not been explored. We thus shrink from telling the employment tribunal what statistics it is to require upon the remission if adequate consideration is to be given to the questions before it; that will be a primary subject which, upon hearing argument on the point, and upon having the range of possible alternative statistics explained, it will need to grapple with.

Disparate Impact

(1) The reasons given by the Tribunal for the decision under appeal:

54. We propose to address the two aspects of disparate impact and objective justification separately. We address disparate impact first. We are conscious that this is paragraph 54 of our judgment, and we have only just reached the reasons of the Tribunal in relation to the decision under appeal. We hope, however, that our extensive citation of authority and our recitation of all that has happened before will fall into place when we examine the arguments advanced to us on the appeal.
55. Having set out the history, the facts, the relevant statutory provisions, and Article 141 of the Treaty; having stated that "pay" in Article 141 covers redundancy payments and having recited paragraphs 59 to 61 from the decision of the ECJ in Seymour-Smith, the Tribunal referred to Rutherford 1 and stated its intention of following the proper approach to statistics set out in the directions given by the EAT in that case. Dealing first with disparate impact, the Tribunal recorded the submission made on either side and then stated: -

15. The first issue the Tribunal has to address is who is the "pool" of individuals chosen to assess the impact of the legislation. Neither party has contended that it should be the entire population. Mrs. Hall (leading counsel for the Secretary of State) submits that the pool should comprise all employees aged between 16 and 79 (or possibly some earlier age because of the danger of the statistics being imprecise from about 68 onwards due to the small numbers involved) who have one year's continuous service (compensation for unfair dismissal) or two years service (redundancy payment). Mr. Allen submits that the pool should comprise employees, those actively seeking work and those not in work but who would like to work. Mr. Allen then submits that the Tribunal should consider a number of age bands between 16 and 79……

56. There follows a discussion of the difference between the experts due to the fact that one had used figures for the United Kingdom and the other for Great Britain. The reasons then continue: -

Mrs. Hall submitted that the Tribunal should adopt a very similar approach to the statistics to that adopted in the Seymour-Smith case. There the courts looked at figures for men and women employees who had less than two and more than two years' service. The House of Lords referred to the qualified and unqualified groups. The qualified group were protected employees in that, having two years' continuous service, if dismissed they could apply (whether ultimately successful or not) for compensation for unfair dismissal. Thus one event – dismissal – gave them the right of access to these Tribunals. The difference in the present case is that a number of events are required before a claim can be made to these Tribunals. At the least the employee must have the requisite period of service, must be 65 or over and must be dismissed. We are leaving out of this analysis a class identified by Lindsay J of those who retire in anticipation of being dismissed as there are no statistics available for such a class.

57. The Tribunal then recited paragraph 30 of the judgment of the EAT in Rutherford 1, which we have set out at paragraph 53 above and continues: -

It accordingly appears to this Tribunal that to focus on employees with e.g. one year's continuous service from 16 to 79 is not correct. Those aged 16 to 17, who happen to have one year's continuous service (somewhat surprisingly 37,000 in Spring 2001) may by the time they reach 65 (if they ever do) be in employment with more or less than one year's continuous service, be self-employed, be too ill to work at all, be unemployed through lack of a job, be in retirement or have emigrated. Although one can minimise the effect of death and emigration by looking at the figures as a percentage of the population, there are still, in the Tribunal's opinion too many factors which invalidate this "snapshot approach". We therefore consider that Professor Elias' (the expect called by the Applicants) figures or person working, seeking work or wanting to work more appropriate when looking at an age band that stretches over nearly 50 years.

58. The Tribunal then sets out the guidance given by the EAT in paragraph 18 of Lindsay's judgment, which we have recited at paragraph 47 above, and records its intention to attempt to follow that guidance. The reasons continue:

It is clear from the House of Lords' judgments in the Seymour-Smith case that the relevant date for determining disparate impact is the year in which the dismissal took place (1998 for Mr. Rutherford and 2001 for Mr. Bentley) although statistics for earlier years are of assistance and later years if they confirm any trend shown by earlier figures. We shall look at a number of figures in a number of different ways, but we consider that our starting point should be those for whom retirement by 65 has some real meaning. We believe that a span of ten years prior to 65 is appropriate, although clearly one can made a case for eight or twelve years or something else. The band is therefore from 55 to 64 which takes one up to one's sixty-fifth birthday. A ten year band thereafter is from 65 to 74 and we consider that this too should be studied as should the entire band from 55 to 74. Table A5 annexed sets out the relevant statistics.

59. There then follows a detailed commentary on the figures, to which we shall return in paragraph 60. For present purposes we take up the reasons later in the same paragraph, when the Tribunals states: -

Mrs. Hall submitted that we should concentrate on the qualifiers (by which she meant those between 16 and 64 who had 12 months continuous service) and then calculate the percentage they were of the combined age group (16-79) for both male and female employees. In Table 1 annexed it will be seen that in 1998 males were 98.79% and females 98.69% and in 2001 males were 98.88% and females 99.01%. The percentage point difference is very small as it is for the other years in the table. Sometimes the males form the higher percentage and sometimes it is the females. However, for various reasons already stated, we do not think that these statistics and this breakdown throws up relevant figures for us to consider. (emphasis supplied) Looking now at what we consider to be the relevant statistics and standing back from the detail and forming an overall view, we are convinced that both in 1998 and 2001 the relevant legislative measures disadvantaged a substantially higher proportion of males than females. In forming this conclusion we have considered the numbers, the percentages, the percentage point differences and especially the ratios, not only for the years in question but also for previous years. Where a ratio is over 1.30 (130 of one class compared with 100 of the other class) we conclude that there is a substantially higher proportion of one class than the other. However, as stated this is just one factor because we have considered the other factors mentioned all of which point in our judgment to the same conclusion.

60. The figures on which the Tribunal relied are summarised in the Tables attached to the reasons. Table A3(3) demonstrates that in the age group 50 to 79, for each of the years between 1994 and 2001, the number of male employees persons looking for work or who would like to work exceeded the number of females in the same categories (the average (in thousands) was 2,922.2 men to 2,530.6 women), whereas in the total population, the numbers of women exceeded the number of men (the average (again in thousands) was 8,358.7 women to 7,541.5 men). Expressed as percentages, the average over the eight year period by which male employees, persons looking for work or who would like to work exceeded females in the same categories was 8.5% (38.7% as against 30.2%), and the ratio average was 1.28 (128 men to every hundred women).
61. We do not think it necessary to analyse the figures any further, since the thrust of the Secretary of State's attack on the Tribunal's reasons in relation to disparate impact was not on its analysis of the statistics presented to it (although Mr. Pannick was critical of any analysis based on the small figures for those over 65) but on the Tribunal's approach, and in particular its rejection of the Secretary of State's argument that it was necessary to look at those who qualified for the right to apply to a Tribunal. We thus turn to examine the arguments advanced to us.

(2) The argument for the Secretary of State on disparate impact

62. We propose to set out the arguments addressed to us on both sides in detail. However, Mr. Pannick's principal argument was both concise and simple. It was that in the highlighted sentence from the Tribunal's reasons which we have reproduced at paragraph 59 above, the Tribunal had committed two fundamental errors. It had not only selected the wrong pool; it had also erred in law by excluding the correct pool from consideration.
63. Mr. Pannick developed the argument in the following way. The Tribunal, he submitted, focused on the disadvantaged groups of men and women, but should have focused on what are the almost identically high proportions of men and women workers generally who can qualify for unfair dismissal and redundancy payments and who satisfy the test of being under the age of 65. The statistics relating to the qualifiers were significant and sufficient. The Tribunal was wrong to focus on the proportions of men and women who cannot satisfy the requirement, where the figures are very small and not significant.
64. Mr. Pannick argued that the impact which a measure has on each sex can only be assessed fairly and properly by considering the pool of people who are liable to be affected by that measure. The composition of the pool is dictated by the nature of the measure being challenged: see the statement by Sedley LJ 75in Allonby v Accrington and Rossendale College [2001] ICR 1189, 1198 which we have recited in paragraph 50
65. The error made by the Tribunal was its failure to identify the correct pool, which invalidated its statistical analysis from the outset: As Potter LJ stated in London Underground v Edwards (No. 2) [1999] ICR 494, 505C-D at paragraph 23, selection of the wrong pool within which the exercise of comparison is to be performed will invalidate the exercise. The error was that the Tribunal focused its attention on employees for whom retirement at aged 65 "has some real meaning", which it identified as the age bands 55 to 64; 65 to 74; and 55 to 74. The case for the Secretary of State was that the Tribunal erred in law in failing properly to consider disparate impact by reference to the pool of employees aged 16-79. Had it considered that pool, it is plain that the age requirement of being under 65 does not have a disparate impact on men:
66. The figures, he submitted, plainly demonstrated the proposition contained in the final sentence of the preceding paragraph. The appropriate test was to consider for men and then for women what proportion of employees in the pool (16-79) are able to satisfy the criterion of being aged 16-64. In 1998 (when Mr Rutherford was dismissed), 98.79% of men in the pool could satisfy the requirement, and 98.69% of women in the pool. In 2001, when Mr Bentley was dismissed, 98.88% of men in the pool could satisfy the requirement, and 99.00% of women in the pool. In all other years from 1995-2001, there is also a very tiny difference between the proportion of men in the pool who qualify and the proportion of women in the pool who qualify. Sometimes a marginally higher proportion of men qualify, and sometimes a marginally higher proportion of women. Applying the standards in Seymour-Smith, if the correct pool is as the Secretary of State submitted, there was plainly no disparate impact for the purposes of Article 141.
67. The Secretary of State's contention was that the Tribunal erred in its identification of the pool for the following reasons. Firstly, the concept of indirect discrimination is concerned with the impact of a requirement which is said to affect one sex considerably more than the other. It is therefore vital to identify correctly who is affected by the requirement to which objection is made. It would frustrate the purpose of the concept of indirect discrimination if the Tribunal were artificially to narrow the group of employees on whom the concept has an impact.
68. Secondly, the identification of the correct pool cannot be divorced from the right or advantage to which a claimant seeks access but to which he claims to have been denied on grounds of his sex. In this case, Mr Rutherford seeks compensation for unfair dismissal and a redundancy payment. Mr Bentley claims a redundancy payment. Each of those statutory rights, subject to certain conditions, is available to all employees. Neither claims employment rights which are in any way different from the rights that are available to all other employees. Just as the rights are not available to (for example) those who are not in fact employees but self employed, so the rights are not available to those who exceed the upper age limit. For this reason, all employees aged 16-79 are affected by the criteria to which Mr Rutherford and Mr Bentley object.
69. In paragraph 59 of the ECJ's judgment in Seymour-Smith (which we have set out at paragraph 33 above) the court had held that the correct approach was to focus on the respective proportions of men and women "in the workforce" who could satisfy the criterion of two years' employment. Since the criterion here being challenged is similarly liable to affect the workforce as a whole, the question of disparate impact should be assessed by reference to the whole of that workforce and not merely selected parts of it. The Tribunal was wrong to distinguish Seymour-Smith on the basis that one event, dismissal, gave a right of access to the tribunals in that case and because

The difference in the present case is that a number of events are required before a claim can be made to these Tribunals. At the least the employee must have the requisite period of service, must be 65 or over and must be dismissed

70. Mr. Pannick submitted that the Tribunal presumably meant that to bring a claim, the employee must (a) have the requisite period of service, (b) be under the age of 65, and (c) be dismissed. The same conditions, he submitted, were applicable in Seymour-Smith.
71. In relation to the concept of a pool comprising employees aged 65-74, Mr. Pannick pointed out that the Applicants' primary case before the Tribunal had been that the pool should consist of those employees who had exceeded the upper age limit. However, it was wrong in principle to assess disparate impact solely by reference to those unable to fulfil the impugned requirement or condition. An assessment of disparate impact depends on the effect of the impugned criterion on men and women employees generally, not just on the effect on those who cannot comply with it. It was therefore necessary to define the pool by reference to all those who could comply with it, but for the criterion which is being challenged. This is not a case where the statistics relating to qualifiers (those who can comply with the requirement) are not sufficient or significant.
72. The Secretary of State relied on a number of authorities Firstly, Mr. Pannick pointed out that in the decision of the House of Lords in Seymour-Smith at [2000] ICR p.259G-260A, Lord Nicholls (for the majority) had left open the question of whether it is permissible to consider the disadvantaged group. But he was not there referring to considering only the disadvantaged group as the sole members of the pool, but was referring to whether it was appropriate to have regard to the disadvantaged group as part of the wider pool. Lord Slynn of Hadley (for the minority), at pp.250H-251A, explained that the starting point must be to look at the percentages of those who do qualify, and only if those figures are not sufficient or significant then to look at the percentages of those who do not qualify.
73. Mr. Pannick pointed out that neither the ECJ nor the House of Lords accepted the approach for which the applicants had contended in Seymour-Smith. In the Divisional Court in that case, [1995] ICR 889, 903F-G (cited in paragraphs 28 and 29 above) Balcombe LJ had rejected the proposition advanced by Mr. Allen, as had McCulloch J.
74. Secondly, Mr. Pannick relied on University of Manchester v Jones [1993] ICR 474, a case under the Sex Discrimination Act 1975. The University had advertised for a careers adviser, stating a preference for a graduate aged 27-35 years. The applicant was aged 46 and claimed to have been indirectly discriminated against because she, along with many other women, had taken her degree as a mature student. The Tribunal agreed with the applicant that the relevant pool consisted of graduates who had taken their degrees as mature students, and that the proportion of women in that category who could comply with the advertisement's age condition was considerably smaller than the proportion of male graduates in that category who could comply with it. The EAT, by a majority, had reversed the tribunal on the basis that the Tribunal had, in effect, redrafted the advertisement so that the relevant pool would read "a graduate preferably aged 27-35 who undertook their degree as a mature student with a record of successful experience". That was an impermissible approach. The advertisement was directed to the graduate population at large and the only permissible pool was graduates aged 27-35 years. There would be many, the majority of the EAT said, who could demonstrate disadvantage if they could elect their own parameters
75. The Court of Appeal in Jones v Manchester University had endorsed the EAT's decision, finding that the relevant pool consisted of all those men and women who satisfy all the criteria for selection other than the impugned criterion. Mr. Pannick relied on a number of passages from the judgments of all three members of the court. At .493H, Ralph Gibson LJ had said: -

In order to compare the proportion of women who can comply with the requirement with the proportion of men who can comply with it, it is necessary to determine the relevant pool. In my judgment, the relevant total is the number of men and women referred to in the sub-section, i.e. those men and women to whom the person – in this case, the employer, applies or would apply the requirement. In this case, that means all men and women graduates with the relevant experience. I do not accept that the relevant total is all men and women: the employer would have no occasion to apply the requirement to any men or women other than those who are able to comply with the requirement in question….

76. Evans LJ, at pp.501A-E, had expressed a similar view: -

If therefore the numbers of women and of men, respectively, remaining after the requirement is applied are to be compared as 'proportions' of something other than the total number of those who can comply, then the question arises, as proportions of what? One possibility is, as proportions of 'all men' and 'all women', even of 'all humanity' subdivided in this way (R v Secretary of State for Education ex parte Schaffter [1987] IRLR 53 per Schiemann J). The other possibility is what Mustill LJ called 'the relevant population', meaning all persons who satisfy the relevant criteria apart from the requirement or condition which is under consideration (Jones v Chief Adjudication Officer [1990] IRLR 533). The latter approach is supported by the race relations case Perera v Civil Service Commission and another [1983] IRLR 166 and by obiter dicta in Price v Civil Service Commission and another (Phillips J) [1977] IRLR 291. In my judgment, it is much to be preferred. This means that the proportion of women in the 'group' (those who can comply) must not be considerably smaller than the proportion of women in the 'relevant population' or 'pool'. (Section 5(3), which requires comparisons of women with men to be such that the 'relevant circumstances' are the same, supports this construction, in my view.)

It follows that the statutory concept, in my judgment, is that of a 'pool' or 'relevant population', meaning those persons, male and female, who satisfy all the relevant criteria, apart from the requirement in question. It is, in effect, the total number of all those persons, men and women, who answer the description contained in the advertisement, apart from the age requirement. Here, that means all graduates with the relevant experience………

It follows that, in my judgment, the Industrial Tribunal erred in law in having regard to a 'pool' which consisted of mature graduates with relevant experience only, and that the EAT majority was correct in holding that there should be no subdivision in the present case. This means that discrimination contrary to s.1(1)(b)(i) cannot be established; the statistics only support the applicant's case if comparison is limited to mature graduates aged between 25 and 29 (or perhaps 30) years of age.

77. Mr. Pannick pointed out that at 502B-D Evans LJ had emphasised the artificiality of seeking to convert a complaint about age discrimination into a case of sex discrimination (a point to which we will return in the section headed Footnote: Age Discrimination at paragraphs 217 et seq. below); and that Sir David Croom-Johnson at p.506A-B and F-H, had cited with approval the observation of Schiemann J in an earlier case that to reduce the size of the pool involves "a very real risk that you have incorporated an act of discrimination into your definition".
78. Thirdly, Mr. Pannick referred us to London Underground v Edwards [1995] ICR 574, in which the EAT had considered a complaint of indirect sex discrimination in relation to the rostering arrangements for train operators. The Applicant was a single parent with a young child who complained that the hours were not compatible with her child care responsibilities. The EAT held that it would be wrong to define the pool as consisting only of those train operators who were single parents. The correct pool consisted of all train operators, male and female, to whom the rostering arrangements were applied. In London Underground v Edwards (No. 2) [1999] ICR 494, the Court of Appeal approved this statement of principle. Potter LJ at 505C-G (paragraph 23) had said:

The identity of the appropriate pool will depend upon identifying the sector of the relevant workforce which is affected or potentially affected by the application of the particular requirement or condition in question and the context or circumstances in which it is sought to be applied. In this case, the pool was all those members of the employer's workforce, namely train operators, to whom the new rostering arrangements were to be applied ...".

79. Fourthly, in Barry [1999] ICR 319, Mrs Barry had complained that the method of calculating redundancy payments by reference to current salary indirectly discriminated against part-time workers. Mrs Barry was a clerk whose grade was described as S4. The bank argued that the pool should comprise S4 clerks. The Court of Appeal disagreed on the basis that the redundancy scheme did not distinguish between various grades. The correct pool was all employees to whom the relevant parts of the scheme applied. On dismissing an appeal by Mrs Barry, the House of Lords did not consider the pool [1999] ICR 859
80. Mr. Pannick submitted that the pool suggested by the Applicants was therefore impermissible because it did not focus (as it should have done) on all those persons who could claim for unfair dismissal and redundancy but for the age requirement of which complaint was made. The pool suggested by the Applicants was impermissibly confined to persons who were excluded by reason of the impugned criterion. The vast majority of employees are under the age of 65. Focusing on the tiny pool of those who cannot comply with the age requirement distorted the analysis by ignoring the larger picture.
81. The pool suggested by the Secretary of State (all employees 16-79) included employees who were not directly affected by the upper age limit, and who may never become affected by it, as the Tribunal had pointed out at the end of paragraph 15 of its reasons (set out at paragraph 57 above). But, submitted Mr. Pannick, the criticism of the Secretary of State's pool was without substance. This was because, firstly, all employees are potentially affected by the upper age limit. And they are relevant to the assessment of disparate impact because they can comply with the requirement to which objection is made. Secondly, the mere fact that a pool contains individuals who are not and may never be affected by the impugned requirement cannot determine the adequacy of the pool. If that were the case then each of the various pools chosen by the Tribunal would be invalid for the same reason. Many of those in the pool had not been dismissed, let alone on the ground of age.
82. Thirdly, the Secretary of State's pool did not exclude any group of individuals who are not actually or potentially affected by the upper age limit. It included all employees for whom reliable statistics are available. By contrast, the Tribunal's pool arbitrarily excludes, for example, those 54 year old employees who may have directly benefited from the upper age limit or may have suffered a detriment by, for example, leaving employment and becoming self employed in anticipation of the consequences of the upper age limit.
83. Fourthly, Mr. Pannick submitted that the case law did not support the contention that the correct pool must exclude those who are not directly affected by the impugned requirement. If the Tribunal's criticism were correct, the pool in Seymour-Smith was wrongly selected. The pool consisted of all men and women in the workforce, including those who had not been dismissed, those who had been employees for less than 2 years and those who had exceeded the upper age limit. So it included many people who would not have been in a position to claim unfair dismissal. Furthermore, in University of Manchester v Jones (supra) the correct pool was composed of all those individuals who satisfied all the criteria for selection other than the impugned criterion. That pool included those who were not interested in applying for the advertised position and those who did not read the advertisement.
84. Mr. Pannick submitted that it was also wrong in principle, for the Tribunal to focus on a pool consisting of employees aged 55-64. Neither Mr Rutherford nor Mr Bentley fell within this age group. So membership of this age group was not the reason for the detriment of which they complained. Secondly, there was no evidence before the Tribunal as to the age at which the upper age limit begins to have any meaning and whether that differs as between men and women. Why, Mr. Pannick asked, choose 55 ? Why not 54 or 45 or 35 ? There may be many young employees for whom the age limit of 65 is relevant, for example those making pension investments. And there may be many people aged 55-64 for whom the age limit of 65 is a benefit: for example, the 55 year old hoping to succeed the 64 year-old Chief Executive.
85. Neither party had argued for an analysis based upon a pool comprising those aged 55-64. The principal issue between the parties on the composition of the pool was whether it should comprise either, as the Applicants contended, those employees aged 65-79 years or, as the Secretary of State contended, those employees aged 16-79 years. A pool consisting of employees aged 55-74 represented a combination of the two pools criticised above: employees aged 65-74 and employees aged 55-64. So this pool was invalid for the same reasons.
86. Finally, on disparate impact, Mr. Pannick addressed the argument that the Secretary of State was bound by the EAT decision in Rutherford 1 by reason of precedent and (in Mr. Rutherford's case) by issue estoppel. The Secretary of State responded, firstly, that the EAT in Rutherford 1, having "synthesised" and "extended" the previous authorities on disparate impact in paragraph 18, (reproduced at paragraph 47 above) had made it very clear at paragraph 30 (reproduced at paragraph 53 above) that although it considered the approach of the Tribunal to the statistics to have been inadequate, it was not telling the Tribunal what statistics to consider, and that it was offering guidance merely "at first blush". It could not then be suggested that the EAT was determining these issues. That Mr Justice Lindsay did not regard his comments in Rutherford 1 as definitive was confirmed by the judgment of the EAT, given by him, in Gidella and others v Wandsworth Borough Council and another [2002] 3 CMLR 1083 (Gidella) at paragraphs 10 and 18 – a case to which we will return.
87. In any event, Mr. Pannick argued, the parties to the proceedings before the EAT in Rutherford 1 were Mr Rutherford and his employer. The Secretary of State was not a party. So even if there were to be an issue estoppel between the parties in that appeal, it could not apply to the Secretary of State. In any event, there was no issue estoppel in the case of Mr Bentley.
88. Furthermore, the EAT was not bound by its previous decisions, although Mr. Pannick of course accepted that a previous ruling would be departed from only in exceptional circumstances, for example if the EAT was satisfied that it is plainly wrong.
89. The Secretary of State therefore submitted, in summary, that the Tribunal erred in law in its identification of the relevant pool because :

(1) The relevant pool does not consist of employees aged 55-74.

(2) The relevant pool consists of employees aged 16-79.

(3) Applying that relevant pool, the loss of protection against unfair dismissal and the loss of a right to a redundancy payment at age 65 does not have a disparate impact on men. The proportion of working men aged 16-79 who can comply with the requirement to be under the age of 65 is almost identical to the proportion of working women aged 16-79 who can comply with the requirement to be under the age of 65.

Therefore, he argued, the complaint of indirect sex discrimination must fail, whether or not the age requirement is objectively justifiable.

(3) The argument for the applicants on disparate impact

90. For the applicants, Mr. Robin Allen QC's first and overarching submission was that there was no error of law made by the Tribunal. There was no dispute on the facts. The Tribunal had carefully analysed the statistical evidence placed before it, followed the guidelines laid down by Lindsay J in Rutherford 1 and reached a conclusion which it was plainly entitled to reach. The pool which it had identified in paragraph 16 of its reasons (set out in paragraph 58 of this judgment) was correct, and the suggestion that it had not considered other age bands was wrong. It had considered a number of age bands, including a wider age band from 50 to 79 and even the age bands from 16 to 79 broken down into the age bands from 16 to 64 and 65 to 79. The Tribunal had then, after looking at the relevant statistics, stood back from the detail and formed an overall view, that both in 1998 and 2001, the relevant legislative measures disadvantaged a substantially higher proportion of males than females.
91. Mr. Allen began his argument on what he described as the "broad pool" point by noting that there had been a dispute before the Tribunal as to whether only employees should be considered, or whether employees and also those not employed but who were also willing to work or actively seeking work should be considered in the pool. This dispute had been resolved by the Tribunal in favour of his clients and was now no longer live.
92. The inclusion of those who were also willing to work or actively seeking work as well as actual employees, however, accorded fully with the view expressed in Rutherford 1 by Lindsay J. and had been adopted by the Tribunal (see paragraph 15 of the reasons):
93. In his judgment in Rutherford 1 Lindsay J. had recognised that the default normal retirement age would exercise a chill effect on a wider group than merely those who were in fact employed. Indeed one of the grounds for allowing the Harvest's appeal had been the focus on employees alone, as could be seen from paragraph 23 of Rutherford 1 (paragraph 49 above). The argument now adopted by the Secretary of State focused only on the age range of the pool. It was argued that as the measure in question applied to all those in work from the age of 16 to 79 the pool should comprise this group. However the same reasoning as lead the EAT in Rutherford 1 to say that it was an error to look only at employees led to the conclusion that in deciding what pool to use the relevance of the normal retirement age must be considered.
94. Normal retirement age, however, is obviously barely relevant for many in employment - as was recognised by the EAT in Rutherford 1. The Tribunal had put it succinctly when it said that the Secretary of State was arguing for a pool which "…stretche[d] over nearly 50 years…" prior to the critical date: - see paragraph 15 of the Reasons, set out at paragraph 57 above.
95. Mr. Allen submitted, accordingly, that the Tribunal had taken a wholly sensible and measured approach to the problem. They had stated that they would look at a "…number of figures in a number of different ways", but had reasonably taken as a "starting point" those for whom retirement by 65 had "some real meaning…" (paragraph 16 , set out a paragraph 58 above) words which Mr. Allen emphasised. It had then sensibly acknowledged that a "span of ten years prior to 65 was appropriate although clearly one can make a case for 8 or 12 years or something else…" (ibid.).
96. Mr. Allen submitted that the Tribunal had rightly concluded that the measure has real significance for those aged 55 to 64 and that this was both consistent with the approach in Rutherford 1 and with common sense. It disclosed no error of law, whereas the Secretary of State's approach was "both counter-intuitive and inconsistent with the previous EAT judgments".
97. Mr. Allen submitted that the Secretary of State's reliance on paragraph 59 of the ECJ's judgment in Seymour-Smith (set out at paragraph 33) ignored the fundamental difference between the nature of the provision at issue in Seymour-Smith and the nature of the provision at issue in the present appeal as discussed and addressed in Rutherford 1. The two-year qualifying requirement for unfair dismissal at issue in Seymour-Smith could potentially affect anyone in the population who was working, regardless of their age. In the present case, though all those over the age of 20 are in principle entitled to a redundancy payment or to claim unfair dismissal, the provision at issue only prevents those working over the age of 65 from claiming these rights. It was, accordingly unrealistic and illogical to claim that the provision has any effect on those aged 20. This was correctly recognised by the Tribunal at para 15 of its decision.
98. Mr. Allen dealt succinctly with a number of arguments in the Secretary of State's notice of appeal relating to the "broad pool" point. The Secretary of State had alleged that the Tribunal erred in law as there "was no sociological or psychological studies before it as to the impact the legislation had on any particular age groups. The choice of the age band 55 – 64 was therefore capricious". That was not correct. The Tribunal had been presented with expert evidence to the effect that the default retirement age appeared to have a chilling effect on recruitment and employment that increased as the employee's age grew. Further, there were voluminous amounts of statistical evidence showing the employment rates at all different age groups, as was apparent from the statistics annexed to the Tribunal's decision. In any event, this group was not the only pool for comparison that the Tribunal considered. It also considered the pool from 50 to 79 and even from 16 to 79 as suggested by the Secretary of State. Even without such evidence, however, Mr. Allen invited us to accept that such a decision came within the knowledge of the Tribunal as an expert tribunal.
99. The Secretary of State had also argued in the Notice of Appeal that "It was no part of the Applicants' case that the disadvantage suffered comprised the psychological, social or economic effect of the legislation on those who had not reached the upper age limit". This seemed to refer to the starting point of 55. This point confused disadvantage and impact or relevance. The choice of 55 as the starting point for the heightened focus for the Tribunal's enquiry related to the relevance or impact of the default normal retirement age rule. Nobody suggested that between 55 and 64 an employee could not claim unfair dismissal or a redundancy payment. The disadvantage was always recognised as beginning in relation to quantum at 64 and crystallising in an exclusionary rule at 65. As to this Mr. Rutherford accepted wholly the decision and judgment in Rutherford 1.
100. Moreover the Applicants had never argued that 55 was the only permissible age from which to start. They advanced evidence in relation to earlier age bands but asserted that 55 was about the right age from which to focus the enquiry as to adverse impact. They would willingly have engaged in analysing any accurate information advanced by the Secretary of State as to individual ages rather than age bands and did so in their closing submissions, until the Secretary of State alerted the Tribunal to the fact that the information she had submitted was inaccurate.
101. Moreover in written closing submissions to the Tribunal for the Applicants it was expressly stated that, although it was impossible to identify a precise "bright line" at which the exclusionary rule will begin to have an impact on an employee, it was clear that somewhere between the early and late 50s a person who is employed, officially unemployed or willing to work will be affected by retirement considerations whether they accept them or not. Such considerations were likely to affect in some way his or her decisions about the strategy to adopt in relation to employment and would affect the way in which employers looked at employment. Accordingly, the emphasis in the case of the Applicants had been throughout on those for whom exclusion of these protections has some relevance - in short those aged 50+s.
102. Moreover Professor Elias (generally recognised as the best statistical expert on these matters had pointed out the "drowning effect" of using statistics in relation to 16 – 79. Professor Elias had stated in evidence:

"The problem I see is when one looks at qualifiers as a percentage of the whole pool aged up to 79, the advantaged group is so much bigger than the non-advantaged. I focused on the size of men and women who were disadvantaged. Your [the Respondent's] way expresses one number as a percentage of a much bigger number which ends up with very small differences, and doesn't convey the fact that this difference is quite significant. I agree with the Chairman that there can be some kind of drowning effect... 99% of women can qualify and 99.5% of men can. But with non-qualifiers twice as many men as women cannot qualify. The better solution to focus on each group and give each its own analysis…"

103. There was, accordingly, no confusion or error of approach in the Tribunal's approach in relation to this point. The Tribunal had taken the correct approach, and the Secretary of State's argument on the broad pool should be rejected
104. Mr. Allen dealt next with what he described as the "disadvantage – confusion" point. The Secretary of State argued that the Tribunal confused the identification of the correct pool with an attempt to describe the disadvantaged persons. This was simply wrong. The Tribunal had broken down the relevant pool (those 55 to 74) into two pools; age 55 to 64 and 65 to 74. The age band 65 to 74 was the disadvantaged group. There was no confusion by the Tribunal, indeed in order to carry out the fourfold comparison urged by Lindsay J. in Rutherford 1 this was a necessary step to take.
105. There was no merit in the suggestion in the notice of appeal that there was no evidence as to whether a higher proportion of men than women within the first sub-group was adversely affected. The point simply misunderstood what was being argued. Table A5 appended to the Tribunal decision showed that the proportion of men and women within the Secretary of State's first sub-group were fully analysed.
106. Paragraph 11(f) of the Notice of appeal had argued that "a disadvantage cannot not be inferred from the mere fact that someone is in employment" and that. "the Tribunal was not concerned with a requirement that was intrinsically discriminatory for those within the first sub-group. The Tribunal failed to have regard to the fact that for some within that sub-group, the upper age limited could be an advantage". In response, Mr Allen repeated what was in the Respondent's notice: -

"The first sentence was never asserted. The Tribunal could legitimately look at the effect of the default statutory normal retirement age of 65 on those within the first sub-group. The argument that upper age limit could be an advantage for someone in the 55-64 group (the "dead person's shoes" argument) is of little significance, since the test for adverse impact is not concerned to assess whether there may be collateral advantages arising from an impugned requirement or condition. It is whether there is disparate adverse impact."

107. Finally, Mr. Allen addressed what he described as the "Breaking Down" point. The Secretary of State had argued that the Tribunal erred, firstly, in breaking down the broad pool and secondly that in making the comparison, the advantaged (16 to 64) should be compared against the advantaged and a comparison of the disadvantaged (65 to 74) against the disadvantaged should be ignored. He submitted that the Tribunal did in fact undertake both forms of comparison, and its results were tabulated on page 10 of its decision.
108. Mr. Allen commented that in arguing that only the advantaged groups should be compared, the Secretary of State had again relied on the passage already quoted from Seymour-Smith. However, the Secretary of State's argument was inconsistent with the judgment of the EAT in Rutherford 1 which provided the guidance as to the correct way to approach such cases, and which was recited by the Tribunal in the present case. It effectively summarised the jurisprudence available, including the numerous cases of the ECJ that had been referred to. Sub-paragraphs 18 (i) to (iii) of Mr. Justice Lindsay's judgment were the most relevant (see paragraph 47 above):
109. Mr. Allen pointed out that in the same case, the EAT at paragraph 26, (reproduced at paragraph 50 above) had expressed the view that the statistics in Mr Rutherford's case "would undoubtedly fall into the 'less obvious' category such that the Tribunal should have looked into other comparisons..." (Mr. Allen's emphasis). The Secretary of State had not appealed against that EAT decision. The Secretary of State was thus bound by issue estoppel in relation to Mr Rutherford's case, and by precedent in relation to Mr. Bentley's case. Moreover the Secretary of State declined to argue against this approach in Gidella.
110. The Tribunal in the present case has followed the guidance of the Mr. Justice Lindsay's EAT decision to the letter. Both the case law and the related legislation recognised that a comparison should be made between the disadvantaged groups. The Tribunal carried out the "four way comparison" set out in Rutherford 1 as can be seen from the Reasons. There had been no error of law.

(4) Points made for the Applicants in oral argument

111. As one would expect, both counsel developed their written submission skilfully in oral argument over the two days of the hearing, and Mr. Allen helpfully produced further written submissions on second day in answer to the arguments advanced by Mr. Pannick on behalf of the Secretary of State.
112. Both in his oral and in his further written argument, Mr. Allen laid considerable emphasis on the guidelines laid down by the EAT in the judgment of Lindsay J in Rutherford 1, and pointed to the fact that Lindsay J has also chaired the EAT (albeit with different lay members) in Gidella. He submitted that we should be extraordinarily careful about reaching conclusions which were inconsistent with those two cases.
113. In Gidella, the applicants were bus drivers employed by the local authority. They were all over 65 when they were dismissed. Their applications for unfair dismissal were rejected by the Tribunal pursuant to section 109 of the Act. The Tribunal rejected their submission that section 109 could be disapplied because it breached Article 141 of the Treaty. Gidella in the Tribunal was decided prior to the decision of the EAT in Rutherford 1, although the Secretary of State was represented on the appeal.
114. The only statistics to which the Tribunal had had regard in Gidella were those of the entire workforce which, in 1998 when the claimants were dismissed, comprised 11.578 million women workers under 65 and 168,000 over that age and of 14.261 million men under 65 and 266,000 men over 65. The Tribunal found, accordingly, that section 109 of the Act applied to 98.17% of men and 98.57% of women, and that the difference of 0.4% was not significant. There was, accordingly, no disparate impact.
115. The EAT allowed the claimants' appeal and remitted the matter of reconsideration by the Tribunal. It did so, it seems, on the basis that the Tribunal had failed to consider other statistics, including figures provided by Professor Elias which, it was argued, demonstrated disadvantage and took the case into the "less obvious" category. If those figures were correct, there was, arguable, an alternative lesser but relatively constant disparity over a number of years. No overall assessment had been made as was required by Rutherford 1. The Secretary of State agreed that it was not an "obvious" case. The EAT was at pains to stress that it was not to be thought to be saying that the Tribunal's decisions was necessarily wrong, but equally could not say what the result would have been had there been a full Rutherford 1 analysis undertaken. Mr. Allen, not surprisingly, relied on this decision.
116. Mr. Allen was also at pains to emphasise that the phrase "those for whom retirement has some real meaning" was only a starting point for the Tribunal's analysis, as could be seen from the decision and from the tables attached to it. Furthermore, focusing on the 50 + age group.
117. As to the argument that (1) the Tribunal should have asked itself what proportion of the working population were men under 65 and what women and (2) that its decision that these figures were irrelevant was a fundamental error, Mr. Allen argued that there were two point involved here. The first was a consideration or proportions; the second was a consideration of which population of persons was under scrutiny. He agreed that the Tribunal had said that a consideration of the working population as defined by the Secretary of State was wrong. However, it had not said that consideration of proportions was either wrong or irrelevant. What it did was to process the numbers and proportions. This was in accordance with Rutherford 1 and showed that although there was a rough equality of numbers in the advantaged groups of males and females, in the disadvantaged groups there were more than 3 males for every two females. Moreover, the Tribunal had not just looked at the non-qualifiers.
118. Mr. Allen defended the use of the phrase "for whom retirement at age 65 has some meaning". It was what had been discussed before Lindsay J. Moreover, it was entirely sensible to look at the way the composition of the population changed as it approached the significant age. It gave an important added perspective on the context for the smaller figures in the disadvantaged groups. It was a way of making sense of very large figures which, without further analysis would mask relevant trends.
119. The phrase had not, he argued, introduced a sociological concept of considerable subjectivity. Each case was about the social impact of rules which on their face were neutral but which ceased to be so because of the way society operated. To use the phrase "some real meaning" is no more than to say that scrutiny of the figures to establish whether or not there has been disparate impact involved looking particularly at what happened in the period leading up to as well as after 65.
120. Mr. Allen finally submitted that, if we were against him, it would be wrong not to remit the matter back to the Tribunal for further consideration. Any other course, he submitted, would involve overruling Rutherford 1, and ignoring the speeches of Lord Nicholls in both Seymour-Smith and Barry.

(5) The Secretary of State's reply

121. In reply, Mr. Pannick reverted to his central theme. It was vital to focus attention on what the Tribunal had actually decided. It had said in terms that it did not think the statistics relating to those who qualified were relevant. That was a fundamental error. The "various reasons already stated" for rejecting any consideration of these statistics were insubstantial. If the Tribunal was basing itself on what Lindsay J had said in Rutherford 1 it was mistaken. There was nothing in Rutherford 1 which suggests that the Tribunal should ignore qualifiers. Ignoring the qualifiers was a clear breach of the guidelines in Seymour Smith – see in particular paragraph 59 of the ECJ's judgment (set out at paragraph 33 above) and the speech of Lord Nicholls in the House of Lords at [2000] ICR 244 at 258-9 (set out at paragraphs 40 and 41 above). Lindsay J, Mr. Pannick ventured to suggest, would be astonished if he were told that Rutherford 1 purported to over-rule or was inconsistent with the decision in Seymour-Smith.
122. Mr. Pannick submitted that it was not possible to argue, as Mr. Allen had, that the Tribunal had had regard to qualifiers. It had mentioned the group only to reject it. It has not addressed the argument that the percentages of men and women who qualified was almost identical. Furthermore, Mr. Pannick argued, the Tribunal had not addressed the point that the disadvantaged group was a tiny proportion of the workforce as a whole. Dr. Hotoff (the expert called by the Secretary of State) had produced statistics which were not controversial and which showed that in 1998 qualifiers in the age group 16-64 comprised 13,589,000 men and 12,546 women, whereas those in the disadvantaged group aged 65-79 comprised 195,000 men and 125,000 women. The comparable figures for 2001 were 13,842,000 men aged 16-64 and 12,813,400 women: 190,500 men aged 65-79 and 115,200 women.
123. Mr. Pannick was at pains to make clear that he was not submitting that the figures as to the non-qualifiers were irrelevant. However, because they represented so small a proportion of the workforce the difference between what he described as the male success rate and the female success rate was so small as make it unsound and unsatisfactory as a basis for an assessment of disparate impact. In Seymour-Smith the ECJ at paragraphs 63 and 64 of its judgment (reproduced at paragraph 33 above) had described a difference between 77.4% and 68.9% as not appearing, on the face of it, to show disparate impact ("a considerably smaller percentage"). Equally, Lord Nicholls in his speech for the majority in the House of Lords described the figures identified in paragraphs 37 to 39 above as "borderline country" (see the extract cited at paragraph 40 above).
124. Given the Tribunal's erroneous approach to the pool, and its reliance on a very small percentage of the workforce, which produced at best borderline statistics, Mr. Pannick invited us to find not merely that the Tribunal had committed an error of law in ignoring the qualifiers, but to go on to find that, on the correct test, that of the qualifiers, there plainly was no disparate impact. We should accordingly allow the appeal and dismiss the claims, not remit them to the Tribunal for yet another hearing.
125. This was, Mr. Pannick submitted, a simple case. There was a small differential between male and female non-qualifiers, but that was all. Allowing this appeal would not be to overrule Rutherford 1, for the reasons already given. Lindsay J had not regarded Rutherford 1 as definitive, as his judgment in Gidella made clear: see paragraph 18 of his judgment. Nothing in any of the cases was authority for the proposition that it was proper for the Tribunal to focus exclusively on the one-qualifiers. To the contrary, all the cases required the Tribunal to look at both the statistics for the advantaged group and for the disadvantaged group. The Tribunal had not done this. It had said it terms that the wider group of qualifiers was irrelevant. That was fatal to its reasoning.
126. We should not send the matter back. If we applied Seymour Smith we should find that the figures were so striking as to the near parity in the proportions of men and women in the qualifying group that we should find no disparate impact.

(6) Analysis and Discussion

127. We have set out the competing arguments in such detail because we take the view that they are quite finely balanced. Nonetheless, having considered the matter, we have come to the clear conclusion that we prefer the arguments advanced on behalf of the Secretary of State.
128. The first and crucial point seems to us to relate to the pool. There seem to us two, related issues here. Firstly, did the Tribunal select the correct pool? Secondly, if there was more than one pool to be considered, was the Tribunal's decision to reject the pool proposed by the Secretary of State without analysis an error of law?
129. In our judgment, the Tribunal selected the wrong pool. But even if the pool it selected was the correct one, its rejection of the pool proposed by the Secretary of State without any proper form of reasoned analysis was an error of law sufficient to vitiate its decision.
130. There is in our judgment nothing either in the guidance given by Lindsay J in Rutherford 1 or in Gidella which renders permissible the rejection of an arguable pool without the expression of good reasons for doing so. Indeed, the appeal was allowed in Gidella for precisely this reason, that the Tribunal had failed to deal with the alternative pool proposed by the Applicants.
131. In our judgment, the Tribunal in the instant case failed to follow the guidance given by the ECJ and the House of Lords in Seymour-Smith: see the speech of Lord Slynn of Hadley reproduced at paragraph 36 and that part of the speech of Lord Nicholls reproduced at paragraph 41. We agree with Mr. Pannick that in the instant case, where the court is dealing with an entitlement to redundancy payments and unfair dismissal, the proper pool to be examined is the entire work force. To deal with any particular segment of it on the basis that it represents those "for whom retirement has some meaning" is to introduce a subjective element which, as the Tribunal itself recognised, was capable of being expanded or reduced without the application of any measurable criteria.
132. Mr. Allen made the point that the inclusive approach was appropriate in Seymour-Smith because what was at issue in that case was the qualifying period of continuous employment before the right to apply to the Tribunal arose. That had nothing to do with age, and accordingly affected the entire population. The upper age limited default provisions of the Act only affected that part of the workforce which was or was approaching 65, and it was therefore inappropriate to apply the Seymour Smith approach.
133. We do not accept that argument. In our judgment, Seymour Smith is the benchmark case. We do not think that the guidance given by the ECJ and the majority opinions of the House of Lords was intended to be limited to the facts of the particular case, or that the case can properly be distinguished because it does not deal with indirect discrimination in the context of the age default provisions of the Act. This does not mean, as Mr Pannick accepted, that it would be wrong, in appropriate cases, to consider the disadvantaged group. But to limit the application of the decision of the European Court and the speeches in the House of Lords in Seymour Smith to the question of the period of continuous employment required for the qualification for the right to apply to a Tribunal for wrongful dismissal is, in our judgment, both to diminish the decision and to misunderstand its implications.
134. It seems to us, moreover, that a broad, expansive concept of the "pool" is a concept not only embraced by Seymour Smith but by the domestic authorities, of which Jones v Manchester University is a good example.. The dangers of building artificiality into any concept is self-evident when one is dealing with statistics. Furthermore, it is quite clear that an error as to the choice of pool can vitiate the decision making process and thus the decision itself: see London Underground v Edwards It is not a straightforward issue of fact and discretion: it is, as Sedley LJ said in Allonby a matter of logic. Logic usually requires one pool, but where there is arguably more than one, analysis is required to identify the preferred pool. That process, in our judgment, is manifestly lacking in the instant case.
135. We wish to make it quite clear that in forming these views we do not regard ourselves as either purporting to overrule Rutherford 1 or departing from the approach of the EAT in Gidella. Nothing in the guidance given by Lindsay J in Rutherford 1 warranted the Tribunal in the instant case ignoring the pool selected by the Secretary of State and / or saying it was not relevant. We agree with Mr. Pannick that Lindsay J would be surprised to be told that Rutherford 1 was inconsistent with Seymour Smith.
136. We also agree with Mr. Pannick that the consequences of the failure to look at the advantaged group are compounded in the instant case by the very small numbers in the disadvantaged group, and the consequentially small differentials which emerge. We think this of particular relevance when looking, for example, at Table A3(3), summarised at paragraph 60 above.
137. Mr. Pannick was, we think, minded to accept that these figures could – if the pool selected by the Tribunal was correct – come within paragraph 61 of the ECJ's judgment in Seymour-Smith (reproduced at paragraph 33 above) as comprising small figures which represented, nonetheless, a "persistent and relatively constant disparity" (see also the extract from Lord Nicholls' speech in the same case reproduced at paragraph 40 above).
138. However, in our view, the evidential value and relevance of Table A(3)(3) is, once again, vitiated by the failure to consider the wider pool. The wider pool (the qualifiers / advantaged group) gives a clear and unequivocal answer to the question to disparate impact, namely that there is no disparate impact. If one then looks at the disadvantaged group, the figures are very small and are in "borderline country". Thus, by comparison with the qualifiers / advantaged group, the disadvantaged group represents, in our view, an unsound basis on which to find disparate impact.
139. For all these reasons, we have come to the conclusion that Mr. Pannick's primary submission is correct. The Tribunal erred in law both in choosing the wrong pool and in excluding the correct pool from its considerations. If it had chosen the correct pool, it would have found no disparate impact. Its order must accordingly be set aside.

Objective justification:

(1) The written evidence presented by the Secretary of State

140. We approach the second part of the case on the basis that we are wrong on the first part, and that disparate impact has been made out. The question then becomes whether or not the statutory age default provisions can be justified on objective criteria unrelated to sex.
141. Evidence on the policy considerations underlying the age 65 limit for bringing proceedings for unfair dismissal and redundancy payments was given to the Tribunal by two officials, Mr. Charles Phillips, Assistant Director, Individual Employment Rights in the Secretary of State's Department and by Mr. Stephen Walker, Assistant Director responsible for redundancy payments policy issues in the Employment Relations Directorate of the Department. Mr. Phillips dealt with unfair dismissal: Mr. Walker dealt with redundancy payments.
142. In his witness statement, Mr. Phillips identified the three-fold test set out in Seymour-Smith (see the judgment of the ECJ at paragraph 77 reproduced at paragraph 35 above) and that part of the speech of Lord Nicholls reproduced at paragraph 42 above. Exhibited to his statement was an Appendix giving the historical background to the upper age limit for claiming unfair dismissal which also gave, he said, some indication of the policy thinking behind the measure at the time of its inception. He stated that he had found relatively little contemporaneous documentation on the Government's view of the provisions over time. Inevitably, thinking on policy issues moved on, and the government was currently working on the implementation of the EC employment directive designed to prohibit age discrimination, in relation to which the Government was committed to legislate by 2006. He described the lengthy process required for the change, and the Government's belief that it would be premature to take a view on the social policy merits of abandoning the default provisions in the Act until the Government was ready to make all the necessary age-related changes, by 2006, in a planned and systematic way
143. Mr. Phillips identified three of the social policy aims of the default provisions. The first was to ensure that employers can meet the legitimate expectations of younger employees for advancement. The second was that the employer knows, to a large measure, what his future recruitment needs are likely to be. A broader aspect of the second policy was that the measure also helped to ensure that across the workforce as a whole, employers were able to predict their recruitment and human resource management needs. The policy thus achieved the socially desirable aim of enabling employers to plan ahead for their recruitment needs according to the needs of their business.
144. The third social policy aim identified was that. whilst the Government wanted to encourage employment amongst older workers, giving the right to claim unfair dismissal to employees over 65 might prove counter-productive, since if the performance of an employee declined to an unsatisfactory level, compulsory retirement could result in a claim for unfair dismissal, whereas the current provision allowed an employee over 65 to be removed from his or her job with dignity.
145. Mr. Phillips repeated the point that the Government was currently working on implementing the EC's Employment Directive agreed at the end of 2000, which introduced provisions to combat discrimination on (inter alia) the new ground of age. The Government would be introducing legislation to prohibit age discrimination in employment and occupation by 2006. Mr. Phillips described the process of implementing such a policy, which starts with an initial mapping exercise, and is followed by a period of research and consultation. Mr. Phillips stated that there were many complex issues which needed to be addressed and added: -

Against this background, the Government believes it would be premature in the extreme to take any view on the social policy merits of abandoning a legislative measure in this area, which has been upheld for many years, and before the Government is ready to make all necessary age-related legislative changes, by 2006, in a planned and systematic way.

146. Mr. Walker's statement, dealing with redundancy payments, identified the social policy aims of the upper age limit in the following way:

The basic objective of the statutory redundancy payments scheme is to ensure that where employees who have demonstrated commitment to their employer (by completing at least two years' continuous service) are dismissed through no fault of their own, they are compensated for the loss of expected continued employment. The main justification for the "upper age limit" provisions has been that, as employees approach state pension age, or the normal retirement age for their job if that is lower, their expectation of continued employment diminishes and eventually ceases and the therefore, as the approach the age in question, their entitlement to compensation should similarly diminish and eventually cease.

A secondary justification has been that the provisions remove a potential disincentive to employers keeping employees on beyond the age at which they would normally be expected to retire i.e. 65 (or as the case may be) the lower normal retirement age for their job. In the absence of these provisions, employers would be faced with the prospect of having to make disproportionately large payments in the event of subsequent redundancy (given that age and length of service are factors in the calculation) if they kept such employees on rather than replacing them with younger ones. The provisions can therefore be said to favour older employees, at least in this respect.

147. Attached to Mr. Walker's statement was a document entitled Redundancy Payments Scheme Age Cut-Off: Historical Background, which identified similar points being made in a Government background brief in 1985, and a number of letters from Ministers in the Department of Trade and Industry between 1998 and 2001. The Tribunal quoted from the background document, recording this question and answer given around September 1965:

Why have the categories in section 16 (age) been excluded? "Workers over 65 (60 for women) are excluded because, having passed the age at which they become eligible for a State Pension, they have not the same expectation of continued employment as younger workers, however desirable it may be in their own and the country's interests that they should remain at work if they wish"

148. Mr, Walker also confirmed he agreed with the contents of Mr. Phillips' statement. Both gave oral evidence, the content of which was canvassed before us, and formed part of Mr Allen's argument on objective justification. We propose, however, to examine extracts from the oral evidence once we have set out the Tribunal's reasoning.

(2) The reasons given by the Tribunal for the decision under appeal

149. The Tribunal dealt with this topic under three headings: Justification generally, Age discrimination and Justification. Under the first heading it began its consideration with the Redundancy Payments Act 1965, pointing out that it differentiated between men and women. Women between the ages of 18 and 60 qualified: men qualified between 18 and 65. At the time, women could draw a State Pension at 60 but men had to wait to 65. In September 1985 the Government had been asked about these age limits and had replied in the answer set out in paragraph 147.
150. In paragraph 18 of its reasons, the Tribunal then made its first reference to Mr. Walker's evidence in these terms:-

Mr S J Walker, assistant director in the Department of Trade and Industry, gave evidence as to the policy behind the redundancy payments legislation, but except when referring to documents it appeared to the Tribunal that Mr Walker was giving generally his own impression of what he considered the policy consideration to be. We are not saying that Mr Walker was wrong in what he said – rather that we would wish to rely on documents where they exist. Helpfully, we do have some documentary assistance.

151. The Tribunal then referred to four letters from Ministers in the Department and from a previous Secretary of State between April 1998 and August 2001 which essentially made the same point as that set out in the answer to the September 1965 question. The Tribunal then commented: -

It appears from the documents that the purpose of the redundancy payments scheme was from the outset through to 2001 to compensate employees for the loss of expected continued employment and that the cut off provision was because that expectation on approaching 65 is diminished. It also appears that the original cut off provision was directly linked to the State Pension scheme (thus the different cut off points for men and women) and that this rationale still existed by 1998 and through to 2001. Mr. McCartney (then a Minister of State) in 1998 refers to eligibility for the State Pension as being a consideration at the outset, but in no way suggests that it is no longer a consideration. Mr. Johnson (then a Minister of State) in 2001 explains that employees who approach personable age have a diminished expectation of continued employment. It is also to be noted that at the time of both dismissals there were no plans to amend the redundancy payments scheme.

152. The Tribunal then turned to unfair dismissal. It referred to the Official Notes on Clauses written when what became the Industrial Relations Act 1971 was going through Parliament. This connected the age default provisions with the respective personable ages for 60 and 65 respectively for women and men, and the absence of an expectation to work after reaching those ages. The extract from the Official Notes cited by the Tribunal ends: -

In some occupations, the normal retiring age is reached earlier; the legislation makes special provision for this different expectation of retirement age. It would be difficult to justify a requirement on employers that made it necessary for them to establish that every retirement was a justified dismissal.

153. The Tribunal then recorded that the reason given in the Official Notes for the qualifying period and upper age limited was that unless special measures were taken, the initial caseload for Tribunals was likely to be extremely heavy - although the emphasis here was more on the qualifying period.
154. The Tribunal then commented on Mr. Phillips' evidence in these terms: -

Mr C R Phillips, assistant director in the Department of Trade and Industry, gave evidence and frankly admitted that he had found little documentation as to policy considerations. Like Mr Walker, we believe Mr Phillips when not relying on documents was giving his own understanding of what the policy was."

155. The Tribunal then dealt with the legislative history of the age default provisions. It referred to Marshall v Southampton and South West Hampshire Area Health Authority [1986] ICR 335 (Marshall), in which the ECJ which held that dismissal of a woman solely because she had attained the qualifying age for a State pension, which was lower than the qualifying age for a man constituted unlawful discrimination. The Government, as a consequence, had changed the law to equalise the age default provisions at 65 for women and for men. The Tribunal accepted that the purpose of the change was to bring domestic law into line with European Community law. It then commented:

Mr. Phillips suggested (we believe this is the correct word) that the upper age limit applying to unfair dismissal was: to enable employers by retiring employees without having to face unfair dismissal claims to advance younger employees; to enable employers to dismiss employees whose performances had declined with dignity.

156. However, the Tribunal followed this citation from Mr. Phillips' evidence with this comment: -

We should perhaps here mention that both Mr Walker and Mr Phillips stated that their witnesses statements were endorsed by a Minister in the Department, but we conclude that that is approval only of what they are stating and not that what they are stating is Government policy. It would have been easy for each witness to have stated 'I have spoken to the Minister (name) and (s)he has informed me that the Government policy behind (the relevant provision) in (the year in question) was…' No such statement was made by either witness."

157. The Tribunal then referred to another Tribunal case, Nash v Mash / Roe Group Limited [1998] IRLR 168, in which the South London Tribunal had held that the age default provisions were indirectly discriminatory. It stated that the Government had considered this decision and referred to a letter written by the President of the Board of Trade to a Member of Parliament stating that the Government would give careful consideration to the outcome of the case. On the evidence, however, the Tribunal found it difficult to determine any change in the policy behind the cut of provision from that stated at the outset in the Official Notes on Clauses.
158. Under the heading Age Discrimination, the Tribunal then noted that the Labour Party in 1995 had committed itself to introduce legislation to combat age discrimination, but once in power had preferred to bring in a Code of Practice. The Tribunal also noted that the Government was now committed as a consequence of the Age Directive to bring in legislation prohibiting age discrimination by 2006. It noted, however, that neither Mr. Walker nor Mr. Phillips had been able to say that such legislation would necessarily deal with the cut off provisions for unfair dismissal and redundancy payments, and it also noted that the equalisation of the ages at which men and women could draw State pensions would not start until 2010 and would not be completed before 2020.
159. Under the heading Justification, the Tribunal identified the dates on which the legislative measures fell to be justified (1998 for Mr. Rutherford and 2001 for Mr. Bentley). It then noted that the burden of proving justification was on the Secretary of State set out the extract from the speech of Lord Nicholls in Seymour-Smith (which we have reproduced at paragraph 42 above) propounding the three-fold test.
160. The Tribunal's reasoning is then set out. We think we should give this in full:

It is first to be noted that the original measures became law prior to 1 January 1973 when the United Kingdom became a member of the European Union. Accordingly, they were not unlawful as being in conflict with European Community law at the outset. However, each provision was drafted so as to fit in with the drawing of the State Pension (different for men and women) reflecting the social policy behind it. This means that a man who is dismissed between 60 and 65 can, if redundant, obtain a redundancy payment or if unfairly dismissed, compensation. However a woman dismissed between 60 and 65 has the same rights - but can also draw on her State Pension. Both of the sections under attack were the result of policies inextricably linked to the State Pension age and the redrafting of these did not alter this link. Accordingly the justification for the two provisions is tainted with sex discrimination. The State could have addressed this issue this post Marshall, but chose instead to make the revisions it did. Twelve years after Marshall, Mr. Rutherford was dismissed; fifteen years after Marshall, Mr. Bentley was dismissed. This, in the opinion of the Tribunal, is sufficient time for legislation to have been considered and brought into effect. Mrs. Hall (leading counsel for the Secretary of State) submitted that the State had no obligation to justify any legislative provisions until discrimination was proven to have taken place. Mrs. Hall further submitted that proof could be provided by clear statistics or other evidence which the State accepted or by a court finding which the State accepted. Accordingly Mrs Hall said that , as the State did not accept there was any disparate impact in the Nash case, Mr. Rutherford's initial case or this case, there was no onus on her to justify at all. Mrs. Hall relies on a judgment of Lindsay J sitting alone in the EAT in Lewis v Airflow Streamlines plc (26 July 2001). However it appears to us that Lindsay J was saying that it would be procedurally sensible to deal with disparate impact alone because if the Applicants failed in that the issue of justification did not need to be addressed. We do not take Lindsay J to be saying that the two issues cannot be tried together as a matter of law. In short, we do not find Mrs. Hall's argument attractive or persuasive. We see no reason why the State should not address its need to justify legislative measures as soon as it is aware they are or reasonably might be tainted by sex discrimination. We believe the State were in effect put on notice of possible disparate impact in having a default age of 65 in 1986 when the Marshall case was reported and should then have realised that it could not justify the relevant measures because they were the result of a social policy tainted by sex discrimination We believe the State were in effect put on notice of possible disparate impact in having a default age of 65 in 1986 when the Marshall case was reported and should then have realised that it could not justify the relevant measures because they were the result of a social policy tainted by sex discrimination. Accordingly, in the Tribunal's view, the State should then have considered appropriate amending legislation. Lastly Mrs. Hall submitted that the Government did not wish to deal with the age matters piecemeal and that after consultation law prohibiting age discrimination would be brought in by 2006. However as stated above there is no evidence that the present Government – even assuming it will be in power in 2006 – will necessarily abandon the 65 default age. It is not in our judgment reasonable for the State to further put off rectifying what we have held to be unlawful measures.

(3) The argument for the Secretary of State

161. Mr Pannick submitted that the Tribunal had erred in law in concluding that there could be no objective justification in this case because the legislative aims were linked to direct sex discrimination in pension ages for men and women. The Tribunal, he submitted, failed to focus on the fact that both men and women enjoy the employment benefits in issue in this case at the age of 65; that the case is not concerned with the State pension age of 65 for men and 60 for women; and that in any event sex discrimination in the State pension age was lawful under EC Directive 79/7.
162. Mr. Pannick began by setting out the principles. These were to be found in the decision of the ECJ in Seymour-Smith at pp.491-492, paragraph 69, which for ease of reference we reproduce here :

"It is settled case law that, if a Member State is able to show that the measures chosen reflect a necessary aim of its social policy and are suitable and necessary for achieving that aim, the mere fact that the legislative provision affects far more women than men at work cannot be regarded as a breach of Article 119 of the Treaty ... ".

163. The ECJ added at p.492, paragraphs 73-76 that the Member State enjoys a broad measure of discretion in deciding on appropriate measures to achieve the aims of social and employment policy, but this cannot be allowed to frustrate the implementation of the fundamental principle of equal pay without sex discrimination, and so mere generalisations will not suffice. Therefore, the Court concluded (at p.492, paragraph 77), if there is disparate impact

"it is for the Member State, as the author of the allegedly discriminatory rule, to show that the said rule reflects a legitimate aim of its social policy, that that aim is unrelated to any discrimination based on sex, and that it could reasonably consider that the means chosen were suitable for attaining that aim".

164. Mr. Pannick relied on that part of the speech of Lord Nicholls of Birkenhead in the House of Lords in Seymour-Smith, which we have set out at paragraph 42, and which we will not repeat. He submitted that the House of Lords had considered the position in relation to measures which were objectively justified when introduced but which may not remain objectively justified.
165. Mr. Pannick summarised the outcome thus:

(1) The retention of a measure having a disparate impact may no longer be objectively justified if the measure is found in the course of time to be unsuited for its intended purpose or the benefits hoped for have not materialised. See pp.261G-262A.

(2) If the Government introduces a measure which proves to have a disparate impact on one sex, it has a duty to take reasonable steps to monitor the working of the measure and to review the position periodically. See p.262A-E.

(3) In monitoring the implementation of measures, the Government has a broad measure of discretion. The practicalities must be borne in mind. The Government is entitled to allow a reasonable period to elapse before deciding whether a measure has achieved the social policy aim in question and if not, whether it should be replaced with some other measure or simply repealed. Time is then needed to implement any decision. See p.262C-E.

(4) With regard to the 1985 Order being considered in that case, the Government could not reasonably have been expected to complete all the above steps in six years. A contrary decision would place an unrealistic burden on the Government: p.262E.

166. Mr Pannick then turned to the reasoning of the Tribunal. He summarised it as follows:

(1) In giving evidence which was not reflected in any documents, Mr Charles Phillips (who dealt with unfair dismissal) and Mr Stephen Walker (who dealt with redundancy payments) were not explaining the Government's policy but merely expressing their own personal impression or understanding as to the policy aims of the upper age limit.

(2) The initial and continuing policy behind the upper age limit for unfair dismissal claims was that the State pension age was generally regarded as the normal retirement age after which employees no longer had an expectation to work. On that basis, it would be difficult to justify a requirement on employers to establish that every retirement was a justified dismissal.

(3) The initial and continuing purpose of the redundancy payments scheme was to compensate employees for the loss of expected continued employment. The upper age limit was imposed because that expectation diminished when an employee approached 65, an upper age limit which was directly linked to the State pension scheme.

(4) Both the Redundancy Payments Act 1965 (which introduced a redundancy payments scheme) and the Industrial Relations Act 1971 (which first gave to employees protection from unfair dismissal) imposed upper age limits of 60 for women and 65 for men. Those upper age limits reflected policies which were inextricably linked to the differential State pension ages for women and men. The justification for the upper age limits was therefore tainted by sex discrimination. Following the ECJ's decision in Marshall v Southampton and SW Hampshire Area Health Authority (Teaching) [1986] ICR 335 (Authorities 2/33), the legislation was amended to equalise the upper age limits for men and for women at 65. However, that amendment did not alter the Tribunal's conclusion that both section 109(1) and section 156(1) of the ERA were the result of policies inextricably linked to the State pension age and that accordingly the justification for the two provisions was tainted by sex discrimination.

(5) The State was put on notice of the possible disparate impact in having a default age of 65 in 1986 when the Marshall case was reported and should then have realised that it could not justify the relevant measures because they were the result of a social policy tainted by sex discrimination. It should then have considered appropriate amending legislation.

(6) The fact that the Government is presently consulting on the introduction of legislation to prohibit age discrimination was immaterial since there was no evidence that the default age of 65 would necessarily be abandoned.

167. Mr. Pannick attacked this reasoning in the following way. He said the first question which the Tribunal should have asked itself was whether the Secretary of State had identified a legitimate social policy aim. He submitted that in relation to unfair dismissal, there were, as at 1998 (when Mr Rutherford was dismissed), three social policy aims advanced by the age limit. They were:

(a) Enabling employers to meet the expectations of younger employees for advancement

(b) Assisting employers to identify their future recruitment needs.

(c) Enabling an employer to dismiss an older and less capable employee without the need to justify the dismissal, and so damage the dignity of the employee, in a tribunal.

168. These aims were all set out in the witness statement of Charles Phillips, filed on behalf of the Secretary of State.
169. By 2001 (when Mr Bentley was dismissed), the continued existence of the upper age limit was also justified on the basis that the issue was to be addressed as part of the planned and systematic age-related legislative changes required by 2006 in the light of Directive 2000/78 of 27 November 2000. This was also set out in the witness statement of Charles Phillips
170. In relation to redundancy payments, Mr. Pannick submitted that:

(1) In 1998 and 2001, the social policy aim advanced by the age limit was that as employees approach State pension age, their expectation of continued employment diminishes and eventually ceases, and so their entitlement to compensation should similarly diminish and eventually cease. This was stated in the witness statement of Stephen Walker and in a letter from Ian McCartney, Minister at the DTI.

(2) By 2001, the continued existence of the upper age limit was also justified on the basis that the issue was to be addressed as part of the planned and systematic age-related legislative changes required by 2006 in the light of Directive 2000/78 of 27 November 2000. This was explained in the witness statement of Charles Phillips, , approved in the witness statement of Stephen Walker and also referred to in a letter from Alan Johnson, the Minister at the DTI).

171. Mr. Pannick submitted that the Tribunal ought to have concluded that each of the above aims is a legitimate aim of social policy. Moreover, the Tribunal's finding in paragraphs 18 and 19 that the witness statements relied upon did not reflect the view of Ministers was perverse. The witness statements were the evidence for the Secretary of State. Both witnesses confirmed that their statements had been read and endorsed by the relevant Minister. Furthermore, their statements were supported by the material to which reference has been made. In any event, the issue was whether (when disparate impact is alleged) there was an objective justification for the impugned practice. The question was whether the reasons set out above provided an objective justification, not whether they have featured in the reasoning of the Government (however that is to be assessed) at all times.
172. Mr. Pannick argued that the Tribunal did not suggest (nor could it reasonably have done so) that fixing an upper age limit did not advance the aims identified. The Tribunal did not suggest that the choice of an upper age limit was disproportionate or otherwise outside the ambit of discretion available to the Secretary of State. The Tribunal had rejected the Secretary of State's argument that there was objective justification because (and only because) the Tribunal found that the aims were related to sex discrimination.
173. Were the aims unrelated to any discrimination based on sex ? Mr. Pannick submitted that the Tribunal's conclusion (at paragraph 21) that the justification was based on sex discrimination was unsustainable for a number of reasons: -

(1) It was not in dispute that the age limits for protection from unfair dismissal and for receipt of redundancy payments were originally 65 for men and 60 for women, and that this involved unlawful sex discrimination : see Marshall v Southampton and SW Hampshire Area Health Authority (Teaching) [1986] ICR 335

(2) Following Marshall, the upper age limit for claiming unfair dismissal was amended by the Sex Discrimination Act 1986 so as to remove the sex discrimination and introduce the age of 65 for men and women.

(3) After the EAT decision in Hammersmith & Queen Charlotte's Special Hospital Authority v Cato [1988] ICR 132, the upper age limit for redundancy payments was amended by the Employment Act 1989 so as to remove the sex discrimination and introduce the age of 65 for men and women.

(4) The Tribunal's reasoning (paragraph 21) was :

"This means that a man who is dismissed between 60 and 65 can, if redundant, obtain a redundancy payment or if unfairly dismissed, compensation. However a woman dismissed between 60 and 65 has the same rights - but can also draw on her State Pension. Both of the sections under attack were the result of policies inextricably linked to the State Pension age and the redrafting of these did not alter this link. Accordingly the justification for the two provisions is tainted with sex discrimination".

(5) The reasoning of the Tribunal was erroneous because :

(a) There is no sex discrimination in the ages at which men and women have a right to compensation for unfair dismissal and redundancy.

(b) The ages at which men and women are entitled to receive a State pension remain different (65 for men, 60 for women). But that was not a matter about which the Applicants were complaining. They were both over the age of 65 when dismissed. (See paragraphs 2 and 3 of the Tribunal decision). So they were both eligible for a State pension by the time they were dismissed. The issue in the present case was not whether a woman aged 60-64 should enjoy a State pension as well as employment rights, but whether it is justifiable to say to all employees that they lose the relevant employment rights after the age of 65.

(c) Even if the Applicants had been complaining that men are not entitled to receive a State pension until the age of 65, it is not a matter in respect of which the Tribunal has jurisdiction.

(d) In any event, the sex discrimination on which the Tribunal focused (that women aged 60-64 are eligible to receive a State pension as well as employment rights) involves sex discrimination in favour of women not in relation to employment rights but in relation to receipt of a State pension. And it is lawful for the State to provide that women are eligible to receive a State pension at the age of 60, men at the age of 65. See Article 7.1(a) of Directive 79/7 of 19 December 1978. The United Kingdom is working towards the equalisation of State pension ages for men and women, with equality to be progressively introduced from 2010 so that equality will be achieved by 2020. It takes time to secure equality because of the expectations of women who planned for retirement at the age of 60.

(e) The equalisation of unfair dismissal ages for men and women, despite the lawful difference in treatment as to State pensions, was required by the ECJ in Marshall.

174. In any event, Mr. Pannick submitted, the principles stated by the House of Lords in Seymour-Smith entitle the Government to maintain the age limit of 65 for these employment rights until they reach decisions about age discrimination following the general consultation on age discrimination currently being undertaken in relation to Directive 2000/78. This was explained in the evidence of Charles Phillips
175. The Tribunal stated (end of paragraph 21) that

"there is no evidence that the present Government - even assuming it will be in power in 2006 - will necessarily abandon the 65 default age".

That, Mr. Pannick submitted, missed the point. The point was not whether the Government will ask Parliament to amend the law. The point was that the Government is entitled carefully to consider the issues raised by age discrimination before it decides whether to invite Parliament to amend the law.

176. The Tribunal had also concluded (near the end of paragraph 21) that

"We see no reason why the State should not address its need to justify legislative measures as soon as it is aware they are or reasonably might be tainted by sex discrimination. We believe the State were in effect put on notice of possible disparate impact in having a default age of 65 in 1986 when the Marshall case was reported and should then have realised that it could not justify the relevant measures because they were the result of a social policy tainted by sex discrimination"

.

177. The Secretary of State's response to that was:

(1) For the reasons already given, the age limits are not "tainted by sex discrimination".

(2) If the age limits are not tainted by sex discrimination, then the principles stated in Seymour-Smith recognise that the State enjoys a broad margin of discretion in considering whether and when to amend the policies because of disparate impact.

(3) In any event, there was no evidence that the Government were aware of any disparate impact at the time of Marshall in 1986. That case was concerned with direct sex discrimination arising out of different upper age limits for men and women claiming unfair dismissal. And even if the Government were aware of the disparate impact, they were still entitled then to take the view that the age limits were objectively justified.

(4) The Applicants also suggested that the Government was aware of the disparate impact after the Employment Tribunal decision in Nash v Mash/Roe Group Ltd [1998] IRLR 168. But the statistical approach there taken to disparate impact was similar to that adopted in the first Tribunal decision in the present case of Mr Rutherford, which the EAT found to be inadequate. Indeed, the Tribunal in Nash had wrongly focused on the numbers of men and women adversely affected, rather than proportions of a pool.

178. Mr. Pannick submitted, finally, that because the Tribunal had focused its attention on what it wrongly considered to be the links between the legislative aims and sex discrimination in the State pension age, the Tribunal failed to consider whether the age limit of 65 was objectively justifiable. Had it asked the right questions, the Tribunal would have concluded that the age limits are well within the discretion of the Government, especially when it is considering reform as part of the consultation on age discrimination generally.

(4) The argument for the applicants

179. Mr. Allen submitted that the Tribunal had carefully considered all the evidence on this issue. He asked us to note the Tribunal's careful consideration of the evidence set out in paragraphs 18 to 21 of its reasons, the bulk of which we have set out above.
180. Mr. Allen, in defending the Tribunal's reasoning, was highly critical of the evidence advanced on behalf to the Secretary of State. In practice, he argued, the Tribunal was asked to carry out a process of divination not estimation in relation to the position of the Government on these provisions as at the time of the relevant dismissals.
181. At the hearing, a number of potential social policy justifications were identified by the Secretary of State beyond those contained in the documents. The burden was on the Secretary of State to prove that the measures in question were objectively justified. The importance of clear evidence in relation to such policies had recently been re-asserted by the ECJ: Kutz – Bauer v. Freie und Hansestadt Hamburg (Case C-187/00) Judgment 20 March 2003 at paragraphs 52 - 63. The Secretary of State had failed to discharge this burden.
182. The Secretary of State's challenge to the Tribunal's rejection of some of the assertions made by Messrs Phillips and Walker on the grounds of perversity was unsustainable. The Tribunal had been entitled to take the view that some of the evidence given by these witnesses was merely their own impression of what they considered the policy considerations to be. In coming to this view, the Tribunal had taken into account, inter alia:

(a) the fact that these witnesses were unable to find any documents to support their assertions, and that

(b) the witnesses' statements omitted a section to the effect that they had spoken to the minister concerned and the minister had confirmed that the departmental policy was as Messrs Phillips and Walker had understood it. In giving evidence at the hearing, both witnesses claimed that the relevant minister had previously "endorsed" their witness statements. The Tribunal considered that this could only be taken to mean that the minister had read the statements, and not that the minister was agreeing that the statements reflected departmental policy - see para 19 of decision (set out at paragraph 153 above);

(c) the other reasons put forward in oral evidence and not contained in the witness statements were not approved by the Government and were offered as speculation or opinion.

183. Accordingly, Mr. Allen submitted, the Tribunal had given valid reasons for its decision, and the evidence supported the decision. In no way could the decision be labelled perverse. In any case, Mr. Allen submitted, the evidence of Mr. Walker was insufficient to prove objective justification. Section 109 and 156 were default provisions operating in the absence of the imposition by the employer of a normal retirement age below the age of 65. The Secretary of State had produced no evidence to show why the default was necessary in circumstances where employers were free to impose a normal retirement age of their own choosing.
184. Mr. Allen submitted that the Tribunal had been entitled to find that the Secretary of State's were tainted by sex discrimination. It had been entitled to find as it had in paragraph 18 of its reasons that the original cut-off provisions in relation to redundancy payments were directly linked to the state pension age. The different cut-off ages for men and women clearly constituted direct discrimination and the provision was amended as a result of Marshal. The justifications for the measure thereafter remained the same, but after 1986 the cut-off age became 65 for both men and women. As age 65 was the state pension age for men, so after that date, it could only be said (on the same reasoning) that men no longer had a continued expectation of work. The point where a woman's continued expectation of work ceased 5 years earlier, at age 60. The cut off provision was chosen at 65 as the expectation of men in employment of continued employment diminished on reaching this age, as it is the age at which they draw their State Pension. Mr. Allen argued that this rationale cannot apply to women that are in employment at the age of 65. The Tribunal had therefore decided that the provision was tainted with sex discrimination in that the choice of age to equalise the default cut off was chosen because it was the male state pension age. The Tribunal illustrated the discriminatory effect by pointing out the absurdity that women aged between 60 and 65 were able to claim both a state pension and a redundancy payment whereas men aged between 60 and 65 could only claim the latter.
185. The Secretary of State had argued that the different pension ages are lawful by virtue of Article 7 of Directive 79/7. Mr. Allen accepted that this was correct in that although the different state pension ages constitute direct discrimination, the directive permits this pending the equalisation of the state pension age for men and women. He argued, however, that the directive only applied to the very narrow circumstances relating to pensions, and did not permit discrimination in the employment field. For that reason, it could not be relied upon in support of the Secretary of State's case.
186. The Tribunal had not, as the Secretary of State argued, confused historic social policy aims with the social policy aims as at 25 October 1998 and 9 February 2001. That argument was unsustainable considering the Tribunal's finding of fact that the social policy aims remained the same from the outset through to 2001.
187. It was, moreover, wrong in law for the Secretary of State to argue that the decision in Marshall only triggered an obligation to amend the legislation to remove the direct discrimination arising out of different upper age limits for men and women; that it was not concerned with indirect discrimination; that the obligation to amend the legislation was fully discharged by the equalisation of the cut off age for both men and women at age 65. Mr. Allen cited to us paragraph 45 of the judgment of the ECJ in Seymour- Smith (which we have quoted at paragraph 34) making it clear that the requirements of Community law must be complied with at all relevant times. Moreover, Mr. Allen repeated his reliance on the fact that the Secretary of State never advanced a justification or need for there to be a statutory default retirement age in the absence of a decision by the employer as to what is to be the normal retirement age.
188. As was explained fully to the Tribunal and as is clear from the statute in relation to unfair dismissal, it is a simple matter for the employer to determine what is to be the normal retirement age. In such a case the employer can fix it at an age, which is above or below 65. Accordingly the issue was: What is the necessity for a state imposed default age? Neither the written statements of the witnesses called by the Secretary of State nor the oral evidence ever supplied an answer to that question. The provisions permitted employers to set their own normal retirement age if they wished, and it was clearly the law that they could impose such a retirement age and that it did not even need to be the subject of agreement: see, for example, Hughes v. Department of Health and Social Security [1985] ICR 419. Accordingly if for no other reason these provisions were never justifiable under EC law.
189. Mr. Allen in oral argument accepted that the three stage approach identified by Lord Nicholls in Seymour-Smith applied. However, as the ECJ had made clear in Seymour-Smith the second and third propositions contained essential issues of necessity. In Seymour-Smith (paragraph 69) the phrase used was that the State had to show that the measures chosen reflected a "necessary aim of its social policy and are suitable and necessary for achieving them". This had been recently re-emphasised in Kutz-Bauer.
190. Secondly, Mr. Allen argued, the manner in which the onus of proof was to be discharged was important. It had repeatedly been said that generalisations were not enough. Why was 65 chosen as the default age? Was it because it was the male pensionable age? The argument that it was an age where there was a diminished expectation of future employment was itself is directly connected with the receipt of a pension. Women's expectations of employment diminish at 60 when they receive a pension: men's expectations diminish at 65. Mr. Allen submitted that the statistical evidence showed this clearly. Accordingly, to choose 65 as the age at which there is a diminished expectation of employment is to choose an age when only men's expectation of employment diminishes. The difference in pensionable age quite obviously discriminates between men and women. The only reason it is not actionable is the derogation in Article 7 of Directive 79/7. But that is a very narrow derogation and has no relevance to matters outside the scope of the Directive itself.
191. The Applicants did not assert that the male pensionable age was unlawful. What they did assert was that the use of it for other social policy reasons outside the scope of Directive 79/7 is unlawful. To put the matter another way, it was not unlawful, Mr Allen submitted, to have differential pension ages; it was unlawful to use pension age as a basis for other social policy aims within the scope of Article 141.
192. In essence, Mr. Allen submitted, the Secretary of State never answered the question: why it was necessary to set a default at 65, if Parliament was content to let employers fix the default age altogether both in relation to unfair dismissal and redundancy. Neither Mr. Phillips nor Mr. Walker addressed this issue. Such evidence as was given consisted of generalisations based on personal opinion and quite insufficient to meet the test set by the cases.

(5) The oral evidence given by Mr. Phillips and Mr. Walker

193. The reasons for the manner in which the Tribunal expressed itself in relation to the evidence of Mr. Phillips and Mr. Walker becomes clearer when we look at the note of their oral evidence, and in particular their cross-examination. We bear in mind, of course, that we do not have a transcript, but a very full note prepared by junior counsel. We were shown a number of passages from Mr. Phillips' evidence which, Mr. Allen submitted, demonstrated the defective nature of the Secretary of State's case, and justified the submissions he makes, as set out above.
194. Mr. Phillips, for example, was asked in cross-examination what the default provisions achieved in relation to unfair dismissal. His answer, as recorded, is that "it allows the unthinking employer to rely on the Statute rather than to think for themselves. It is not the case that every business wants to think for itself" He was then asked: "In any event, the Statutory default is inconsistent with the government policy in the Code?" Answer: "The Code said that you should think about it, and it would be better if you did".
195. Mr. Phillips put the matter slightly differently the next day, when it was put to him that if employers can choose a normal retirement age it was unnecessary for the State to impose a default. He replied:

I disagree. The employer does not have complete freedom. He must agree the normal retirement age in his contract of employment. The current default allows a company to take advantage of the social policy benefits that I outline in my statement without having to go through the process of agreeing a normal retirement age. The default removes that burden on the employer.

196. Later in his cross-examination Mr. Phillips agreed with Mr Allen that there was no reason why an employee who was older than his retirement age should not be treated fairly. It was put to him that if employers chose to have a normal retirement age they could impose it on their employees, As a consequence they could achieve the policy aim of meeting the legitimate expectations of younger employees for advancement and what their recruitment needs were likely to be without the need for an age default provision. Mr. Phillips agreed.
197. Later again, Mr. Phillips was asked out the status of his statement, and whether or not it his statement had been endorsed by the Secretary of State. Mr. Phillips answered that the Secretary of State and seen his witness statement and endorsed it. Later in his evidence, the point was taken up again in a series of exchanges with the Chairman, which we will set out in full: -

Chairman: The Minister has seen and endorsed your statement. What does that mean? Does it mean that she agree with your views or that you have stated government policy?

Witness: The question would not have been put to the Minister in such explicit terms. It would have been put to her in the context of this litigation. This is the evidence from the Department on the social policy justification, do you agree to this evidence?

Chairman: When you express a view that a particular policy leads to a particular result, and the Minister says: I agree, she may be giving her personal view rather than that of the government?

Witness: No, I got approval from Alan Johnson, the Employment Minister. He was reading the submission and giving his agreement in his capacity as such and knowing that he was responsible for employment legislation. Ministers do not respond to submissions in a personal capacity.

Chairman: The basis of some of your statements is your own view. Is that endorsed by the minister or (sic) a statement of government policy?

Witness: They are the same. The endorsement equates to the Minister saying I, as Minister responsible for the legislation, believe that what you say is true as government policy.

Chairman: You are now stating current government policy?

Witness: Yes, I am also asking the ministers to endorse, which they have, that this has always been a reasonable view for governments to hold back to 1997.

Chairman: The Tribunal must look at the date of the initial legislation and see if there has been any change in policy to October 1998 (the date of dismissal), we may then want to look at what happened thereafter. When you are stating the government policy at the date of your witness statement, is that the policy for now or has it been the policy for some years?

Witness: That is a very difficult question. I have been considering it and have not been able to come to any real conclusion. My evidence is cast as being that these are the sorts of arguments that a reasonable government might have adduced. It is my view that the ministers in endorsing these argument are endorsing them from 1997 when this government was first in place. This was not explicitly put to them, but I suspect they would have yes if they were asked whether this was the case.

Chairman: There has been a change. I have to distinguish between the policy behind the default age and age discrimination. This government has a different view on age discrimination from previous governments

The Witness: Yes, on age discrimination.

The Chairman: But you cannot determine the precise policy behind the default clause

Witness: Agreed.

198. One other passage in particular was relied upon by Mr. Allen, when he returned to the theme addressed in the previous paragraph. Mr. Phillips was asked why it was unacceptable for the employees of an employer who had chosen not to have a normal retirement age to have rights. There followed this exchange:

Witness: There is a disparity from the employee's perspective

Counsel: You said it was unacceptable from the Government's point of view. If an employer has chosen not to have a normal retirement age why is it unacceptable for his employees to have rights over 65? Would you like to withdraw your statement?

Witness: I cannot withdraw the sentence it has been endorsed.

Counsel: You cannot defend it:

Witness: I would say that the situation would be unfavourably disfavourable to hose employers who do not have a normal retirement age. The socially desirable results – the ability of the employers to predict their workforce needs – are currently available to those employers who impost (sic) a normal retirement age. Also, due to the default, they are available to those employers without a normal retirement age. If there were no default the latter employers would not benefit.

Counsel: By their own choice.

Witness: But the choice imposes burdens. I am talking about the lazy employer.

(6) Discussion and analysis

199. On this part of the case, we prefer the arguments of the Secretary of State. The principal syllogism upon which the Tribunal appears to rely seems to us to be the following, namely (1) the default provisions relating to both redundancy payments and unfair dismissal are directly related to the State retirement age; (2) the State retirement age is different for and thus discriminates between women and men; (3) therefore the default provisions are tainted with sex discrimination and unjustifiable. We think this analysis unsound.
200. Firstly, we do not think it strictly accurate to say that the default provisions are inextricably linked to the State retirement age. We accept that we are, or course, proceeding in this part of the case on the premise that disparate impact has been established, notwithstanding the equalisation of the ages after which both women and men are prevented from seeking redundancy payments or unfair dismissal, consequent upon the decision of the ECJ in Marshall v Southampton and South West Hampshire Area Health Authority [1986] ICR 335 (Marshall). However, we do not agree with the Tribunal when it says in paragraph 21 of the reasons that "both of the sections under attack were the result of policies inextricably linked to the State Pension age and the redrafting of those did not alter this link". We do not think that the changes in the Statute, designed as they were to meet the Marshall point, can be dismissed so easily.
201. Nor, with respect, do we think that Mr. Allen rescues the point by submitting that all the government did was to substitute the male retirement age as the default provision age for both sexes. As the extracts from the evidence set out above demonstrate, Mr. Allen was seeking to make the point with Mr. Phillips that no default age was needed in a field in which the employer could dictate the normal retirement age for his employees. The Tribunal does not seem to have endorsed that point.
202. The policy arguments being advanced by the Secretary of State (perhaps, in the oral evidence, not as elegantly as they might have been) are those which we have set out in the section headed The written evidence presented by the Secretary of State above. In our view those constitute reasonable policy objectives, and we agree with Mr. Pannick that the government was perfectly entitled to say that it would not be right to leave it to individual employers to address this issue, given the complex issues which surround it.
203. Similarly, in relation to redundancy payments, Mr. Walker identified the policy aims. We have set these out above. Once again, we regard these as rational. We do not read the Tribunal's reasons (this being a sex discrimination, not an age discrimination case) as stating the government is not entitled to impose an upper limit.
204. Mr. Pannick argued that there were many situations in which the upper age limit did not apply and that Parliament had been astute only to impose it where it was justified. We do not think it necessary to pursue this line of argument. It is, we think, sufficient for the purposes of this appeal to find that on the evidence available to the Tribunal, the government, in the statements of Mr. Phillips and Mr. Walker described policies reflecting legitimate aims of the State's social policy.
205. The policy aims described by Mr. Phillips and Mr. Walker do not seem to us to be related to any discrimination based on sex. Mr. Allen argues that they were initially based on the differences in the Statement retirement age, and that the redrafting makes no difference. As previously stated, we do not agree with that proposition.
206. It is clear the differential between the State retirement ages for men and women is not unlawful under Community Law. Article 7 of Counsel Directive No 79/7/EEC of 17 December 1978, which provides in Article 1 for the progressive implementation of equal treatment for men and women in the field of social security and other elements of social protection, provides in terms that:

This directive shall be without prejudice to the right of Member States to exclude from its scope: -

(a) the determination of pensionable age for the purpose of granting old-age and retirement pensions and the possible consequences thereof for other benefits

207. It is, accordingly, we think incorrect to identify the current default provisions preventing men and women claiming redundancy payments or unfair dismissal after the age of 65 as "tainted with sex discrimination" because of the existence of different (and lawful) State pension ages between women and men. The present case has nothing to do with social security and other benefits, which is the subject matter of Directive 79/7/EEC. As Mr. Pannick pointed out, these appeals are not about the State pension. Neither Mr. Rutherford nor Mr. Bentley is complaining about the fact that a woman under 65 could claim her State pension as well as seeking redundancy payments and unfair dismissal. Both Mr. Rutherford and Mr. Bentley are over 65 and eligible for their State pensions. Their complaint is that the apparently equal age default provisions affect more men than women and are indirectly discriminatory. That, in our view, is a different question.
208. In Marshall, Miss Marshall had a normal retirement age of 60, but was kept on to 62 when she was dismissed. Hers was a case of direct sex discrimination, and the government was not permitted to rely on Article 7 to justify the lawfulness of her dismissal. In the instant case, the Secretary of State is not trying, as we understand it, to defend the policies behind the default provisions by reliance on Article 7: it is the claimants who are attempting to say that the respective pension ages for men and women are discriminatory. In our judgment, it is a perfectly legitimate response to that argument for the Secretary of State to argue that his policy on the age default provisions cannot be said to be tainted with sex discrimination when (a) Article 7 does not apply to cases involving employment and (b) in any event, where it does apply, it renders the age difference lawful.
209. Reliance was placed by the Tribunal and Mr. Allen on the decision of the Employment Tribunal sitting at South London in Nash v Mash/Roe Group Limited [1998] IRLR 168. Mr. Allen's reliance on this case was, we think, primarily based on the proposition that as it was decided in January 1998, the Secretary of State had been on notice for a considerable period that the age default provisions had been regarded as indirectly discriminatory as contrary to Article 119.
210. Mr. Nash had no normal retirement age and he was dismissed when he was 69. On a preliminary issue, the Tribunal found that it had jurisdiction to hear his claim. He produced statistics based on the 1991 Census and subsequent Labour Force Surveys demonstrating that in 1991, 6.8% of the male population aged 65-74 were economically active, compared to 3.9% of females. Figures for 1996 prepared by the Equal Opportunities Commission showed that 7.5% of men aged 65 or over were economically active in Great Britain as opposed to 3.1% of women.
211. The Secretary of State was not a party to the case, and had not responded to an invitation to make representations as to the social policy implications justifying the original enactment and retention of the relevant provisions. As a result, the Tribunal said that in the absence of this information it could not find objective justification.
212. The basis upon which the Tribunal appears to reach its conclusion on disparate impart is contained in paragraph 23(iv) of its reasons:

(iv) the tribunal is satisfied from the statistics and the arguments put forward by Professor Elias that at least from 1991 the legislative provisions had an adverse impact upon male employees aged 65 plus in that the number of such male employees who are thereby disqualified from the right not to be unfairly dismissed and to receive a redundancy payment is considerably larger than the corresponding number of female employees. Accordingly, the tribunal finds that the domestic legislation is indirectly discriminatory.

213. We do not find this case of any assistance. In the first place, the Secretary of State was not represented and had addressed no argument on objective justification to the Tribunal. Secondly, as Mr. Pannick pointed out, the statistical approach taken to the statistics relating to disparate impact was similar to that adopted in Rutherford 1 which the EAT had rejected.
214. We agree with Mr. Pannick that the Tribunal was wrong in law to find that the Secretary of State had not fulfilled the three limbs of the Seymour-Smith test for objective justification. The Tribunal was, in our view, wrong in particular to find that Secretary of State's justification for policies was tainted with sex discrimination. We also think that, in the context of amending legislation, the Tribunal failed to give any weight to the consultation process currently under way in relation to age discrimination, and to allow the government a reasonable margin of appreciation when striking the balance between the need to legislate and the need to ensure that proper processes have been gone through before legislation is placed before Parliament
215. In addition to the reasons advanced by Mr. Pannick, we think the Tribunal's statement in paragraph 20 of its reasons that neither Mr. Phillips nor Mr. Walker could state that the legislation to ban age discrimination due to be enacted by 2006 would necessarily deal with the cut off provisions for unfair dismissal and redundancy payments missed the point for a further reason. The legislation must fulfil the government's obligations under the Directive. If it does not do so, it will, undoubtedly, be the subject of challenge and liable to be disapplied for that very reason..
216. The appeal will, accordingly, be allowed on this basis also. In the light of our conclusions, we see no point in remitting the matter to the Tribunal for further consideration. The order of the Tribunal will be set aside, and the claims of both Mr. Rutherford and Mr. Bentley will be dismissed.

Footnote: Age discrimination

217. Even if we had concluded this judgment at the point at which we had reached our conclusion on disparate impact, we would not have remitted these two cases to the Tribunal for reconsideration. This is for two reasons. The first, is that there were no issues of fact on this appeal, and we had reached the conclusion that Mr. Pannick's argument on the law in relation to disparate impact were correct. Secondly, however, we would have been reluctant to commit the parties to a further lengthy argument about statistics in a case which, we feel bound to say, is not about sex discrimination but about age discrimination.
218. In this respect, we agree with a submission made by Mr. Pannick in his written argument, to which we have not hitherto adverted. Mr. Pannick submitted:

The errors of law made by the Tribunal are, in part, the consequence of the Applicants seeking artificially to fit a case which is really about age discrimination into the provisions concerned with sex discrimination. The artificiality of the exercise is emphasised by Directive 2000/78 of 27 November 2000 which prohibits age discrimination in employment and vocational training. That prohibition must be transposed into domestic legislation by December 2006.

219. We also note that in giving the judgment of the EAT in Gidella, Lindsay J in paragraph 1 used colourful language to describe the application this process. He said: -

This appeal concerns the domestic provision which, broadly speaking, deprives employees of the right to complain of unfair dismissal where they are aged 65 of over at the dismissal. However, at this juncture, discrimination on account of age is not unlawful and so complaint against the provision, if to be raised, has to be tortured to fit the shape of an equal pay complaint on the footing it is then said) that the domestic provision involves such disparity between its effect on women and men as to make it a breach of Article (141) of the EC Treaty.

220. Lindsay J made the same point, in somewhat milder language, in Rutherford 1 (which, of course, precedes Gidella) when he commented in paragraph 1, that the case was not a "somewhat premature one of ageism, or age discrimination, as one might first think, but depends on an argument as to men being thereby put to such a disadvantage in relation to equality of pay that those sections are to be overridden by Article 141 of the EC Treaty".
221. These are not the only comments on the artificiality of the process. It will be recalled that in the case of University of Manchester v Jones, Evans LJ made two comments, both of which we feel to be of direct relevance here. At [1993] ICR 474 at 501H to 502D: -

I will add two general comments. First, the words of the subsection make some statistical evidence inevitable, but I have wondered throughout this appeal whether Parliament can have envisaged the kind of detail which has been produced in this case. Even these figures involve a considerable amount of approximation; for example, the numbers and ages of those attending university are taken apparently as the numbers and ages of those who obtain degrees, and there are no precise figures for those who graduate and are able to acquire the relevant experience, before or after graduating, below the age of 35. These and similar added complications are themselves an additional reason, in my judgment, for rejecting the submission that subdivision is permissible as a matter of law.

Secondly, I cannot help feeling that this is essentially a case about discrimination by age – indeed, that is obvious – and it is unsurprising that attempts to convert it into a case of indirect sex discrimination should meet obstacles which are difficult and, in the present case on the evidence given, impossible for the applicant to overcome. If indirect discrimination could be shown by reference to the effects of an age limit, then there would always be unlawful discrimination against either women (s.1) or men (s.2) subject only to the safeguard of 'considerably smaller' in s.1(1)(b)(i). Moreover, discrimination against older women, even if in favour of younger women, would be regarded as sex discrimination against women. For example, the applicant's submissions here would be the same, even if the shortlists and the chosen applicants consisted entirely of (younger) women.

222. We wish to make it clear that we have considerable sympathy for both Mr. Rutherford and Mr. Bentley. We understand the sense of injustice under which each labours. Each of the members of the EAT has had direct experience of cases in which the inability to apply to a Tribunal after the age of 65 has caused injustice. An obvious example is that of the man, perhaps in a second marriage, who has no normal retirement date, but who has planned his life around retiring at 68 or 69 when his mortgage will be paid off, or when his children finish their tertiary education. If he is made redundant or dismissed after his 65^th birthday he is without redress. Such a man is the counterpart to the man identified by Mr. Pannick in paragraph 84 of this judgment, who has benefited from the existence of the age provisions.
223. Sympathy is, of course, a poor aid to Statutory construction. We do not, of course, criticise either Mr. Rutherford or Mr. Bentley for bringing these applications. By pointing up the difficulties, they and their lawyers have performed a public service, of which we hope the government will take note. But that said, we see no purpose in there being a Rutherford II. Subject, of course, to the Court of Appeal taking a different view, it is our judgment that this litigation should stop now.
224. We have come to the conclusion that, as the law currently stands, Mr. Rutherford and Mr. Bentley have no remedy. It will be no comfort to them, but is, we think, of considerable importance, that the government is obliged, under Directive 2000/78 to legislate in relation to age discrimination by 2006.
225. We cannot, of course, speculate about the outcome of the current consultations and the terms of the future legislation. What we can, however, properly say is that Mr. Pannick's able arguments which have succeeded before us on indirect sex discrimination, would, we think, be unlikely to succeed before a Tribunal hearing an application for age discrimination, It is to be hoped, however, that after 2006 no such application will need to be made.

Conclusion

226. The appeals will be allowed and the claims dismissed.