United Kingdom Employment Appeal Tribunal

APPEARANCES

 

For the Appellant	MR MICHAEL SUPPERSTONE (One of Her Majesty's Counsel) Instructed By: Miss Susan A Dauncey Force Solicitor Avon & Somerset Constabulary PO Box 37 Valley Road Portishead Bristol BS20 8QJ
For the Respondent	MR PAUL T ROSE (of Counsel) Instructed By: Messrs Pattinson & Brewer Solicitors Transport House Victoria Street Bristol BS1 6AY


 


MR JUSTICE CHARLES:

1. This appeal is against a decision of an Employment Tribunal sitting at Bristol. The Extended Reasons for which were sent to the parties on 31 January 2000.

Introduction

2. The Applicant before the Employment Tribunal was a Ms Chew. The Respondent was the Chief Constable Avon & Somerset Constabulary (the Chief Constable).
3. The decision of the Employment Tribunal was that the Chief Constable discriminated against the Applicant in breach of Section 1 (1)(b) Sex Discrimination Act 1975 (the SDA). The finding was therefore one of indirect sex discrimination.
4. The Employment Tribunal found against the Applicant on her claim for direct discrimination (see paragraph 39 of the Extended Reasons) and there is no appeal against that decision.

The most relevant sections

5. These are Sections 1 (1)(b) and Section 5 (3) SDA. They are in the following terms:

"1 Sex discrimination against women

(1) A person discriminates against a woman in any circumstances relevant for the purposes of any provision of this Act if ---

(b) he applies to her a requirement or condition which he applies or would apply equally to a man but ---

(i) which is such that the proportion of women who can comply with it is considerably smaller than the proportion of men who can comply with it, and,

(ii) which he cannot show to be justifiable irrespective of the sex of the person to whom it is applied, and

(iii) which is to her detriment because she cannot comply with it.

5(3) A comparison of the cases of persons of different sex or marital status under section 1(1) or 3 (1) ------------------- must be such that the relevant circumstances in the one case are the same, or not materially different, in the other. "

The issues that arise under the sections

6. Before us and the Employment Tribunal it was common ground that the following issues fall to be considered under those sections, namely:

(a) Is there a requirement or condition (see s. 1(1)(b) SDA),

(b) Disparate effect (see s. 1 (1) (b) (i) and 5 (3) SDA),

(c) Justification (see s. 1 (1) (b) (ii) SDA), and

(d) Detriment to the Applicant (see s.1 (1) (b) (iii) SDA).

The Appeal and Cross Appeal

7. These are limited to the conclusions of the Employment Tribunal on "Disparate Effect".
8. The Chief Constable conceded that there was a "requirement or condition". There is no appeal against the findings of the Employment Tribunal on "justification" or "detriment to the Applicant" but some of the evidence and argument before the Employment Tribunal was relevant to those issues and, in our judgment correctly, both parties have referred us to parts of the Extended Reasons dealing with "justification" and "detriment to the Applicant" and have submitted that the Extended Reasons should be read as a whole.

The conduct of the hearing before the Employment Tribunal

9. The hearing lasted 5 days in January 2000 and the parties were represented by the same experienced counsel who represented them before us.
10. During the hearing before us (and notwithstanding the length of the hearing before the Employment Tribunal) it was common ground between counsel that leading counsel for the Chief Constable was given a short and strict time limit for his cross examination of the Respondent's witnesses (i.e. half a day for 4 or 5 witnesses - although leading counsel took about another hour). It also became common ground before us that as a result of this limitation and the compliance therewith a number of disputed points in the evidence of PC Wand, which were relied on by the Applicant both as to (i) the pool that should be taken into account for the purposes of comparison, and (ii) the disparate effect of the requirement or condition on men and women within that pool, was not the subject of cross examination and, as we understand it, was not taken into account expressly in the reasoning of the Employment Tribunal.
11. In our judgment this is unfortunate. But neither side has argued that this time-tabling and its effect constitutes a ground of appeal although both sides did take some points on the evidence and findings of fact made by the Employment Tribunal. In particular these arose in respect of paragraph 5(f) of the amended cross appeal and the answer thereto and in this context the arguments focused on the evidence of PC Wand. But factual issues were also raised before us on other aspects of the appeal and cross appeal.
12. We also understand from what we were told by counsel that much of the evidence before the Employment Tribunal was focused on the issue of justification. As we have already indicated some of that evidence overlapped with issues relating to the pool to be taken for the purposes of comparison and the disparate effect of the condition or requirement on men and women within that pool.

The issues on the Appeal and Cross Appeal

13. In his appeal the Chief Constable argues that:

(a) the Employment Tribunal did not err in its selection of the pool, but

(b) the Employment Tribunal erred in law in concluding that s. 1(1)(b)(i) SDA was satisfied and, further or alternatively, the Employment Tribunal reached a perverse conclusion on this point.

The appeal of the Chief Constable is therefore against the conclusion of the Employment Tribunal on the disparate effect of the requirement or condition on men and women within the pool selected by the Employment Tribunal. Within that argument (and thus point (b)) the Chief Constable's main point is that the percentage difference identified does not warrant the conclusion reached. But in addition he makes points that:

(i) inferences made by the Employment Tribunal, and

(ii) some of the factual conclusions of the Employment Tribunal were not justified. The Chief Constable invites us to allow the appeal and decide ourselves that on the findings of fact made by the Employment Tribunal s. 1(1)(b)(i) SDA is not satisfied. He did not invite us to remit the case to an Employment Tribunal. However an alternative open to us if the Chief Constable persuades us that the Employment Tribunal erred in law would be for us to remit the case to an Employment Tribunal. If we took this course on a further hearing the Employment Tribunal could hear evidence that the limitation imposed by this Employment Tribunal excluded from their consideration.

14. The Applicant takes alternative positions on the appeal:

(a) First she argues that the Employment Tribunal did not err as alleged.

(b) In the alternative, by way of cross-appeal the Applicant argues that the Employment Tribunal erred in law in their selection of the pool. As to this alternative ground the Applicant then goes on to identify what she says the correct pool is and by reference to that pool that the difference in the proportion of men and women who can comply with the requirement is larger than that identified by the Employment Tribunal in respect of the pool they selected. It follows that a purpose of the cross-appeal is to give the Applicant an alternative argument if we were to be in favour of the Chief Constable on his appeal and in particular if we were to conclude (as he argues) that the difference in the proportion of men and women who can comply with the requirement or condition in the pool chosen by the Employment Tribunal is too small to warrant a conclusion that: "the proportion of women who can comply with the requirement or condition is considerably smaller than the proportion of men who can comply with it".

(c) In the further alternative the Applicant asserts (by paragraph 5(f) of her amended cross appeal – which we have referred to in paragraph 11 above) that if the appropriate pool is the entire force of 3,016 officers the Employment Tribunal erred in law in limiting the persons who could not comply with the requirement to 10 (although that was an absolute minimum) and ought to have found that all 31 officers (including the Applicant) identified by PC Wand as operating a different shift or duty rota to the department or group to which they were assigned could not comply with the requirement or condition. The response of leading counsel for the Chief Constable to that assertion was that during the hearing the Employment Tribunal indicated that PC Wand had used the wrong documents on which to base findings as to shift patterns and that as a consequence PC Wand was not cross-examined on each individual he referred to. Leading counsel for the Chief Constable went on to make points of fact in respect of those persons.

Like the Chief Constable the Applicant did not invite us to remit the case but invited us to decide the alternative arguments ourselves. Again remission is a possibility in respect of the arguments in paragraphs 14(b) and (c).

15. As to the arguments referred to in paragraph 14(c) neither side took us to the detail of their arguments (which were essentially arguments on the facts) during the hearing by reference to the documents and there is no finding by the Employment Tribunal concerning the basis upon which findings as to shift patterns should be made. As we have mentioned it also emerged during the hearing that part of the reasons why the evidence of PC Wand on the persons who were the subject of these arguments (i) was not examined in detail in evidence, and (ii) was not dealt with in the findings are those referred to in paragraphs 10 and 11 above.
16. In the above circumstances we are of the view that:

(a) we cannot determine the factual elements of the arguments referred to in paragraph 14(c) and thus those arguments,

(b) we have to proceed on the basis that the Employment Tribunal decided that it was not necessary for them to determine those arguments and therefore did not do so, and thus

(c) if we conclude that such arguments are relevant to the determination of the issues in this case and therefore that the Employment Tribunal were wrong not to make findings on them that would be an error of law by the Employment Tribunal which would lead to the result that this case should be remitted.

The Background Facts

17. These are set out in paragraphs 1 to 11 of the Extended Reasons and we will not repeat those paragraphs in this judgment.

The requirement or condition

18. In August 1994 the Avon & Somerset Constabulary introduced a policy for part time working. The expressed object of that scheme was that:

"It is anticipated that there will be a reduction in the premature wastage of experienced and capable officers ……………… "

However the policy for part time working made clear that:

"A part time officers duty roster for the 28 day period will be divided into the following categories:

A duty day – a day within the "normal" duty week when the officer will be carrying out duty

A free day ……….

A rest day …….…

The "normal" duty week will be defined as the cycle of duty allocated on the shift pattern which has been adopted by the district/department in which the officer is serving."

As appears from paragraph 13 of the Extended Reasons the Chief Constable accepted that the requirement or condition for the purposes of s. 1 (1)(b) SDA was the requirement or condition set out within the Policy for Part Time Working that the Applicant must accord to "the cycle of duty allocated on the shift pattern which has been adopted by the district/department in which the officer is serving".

19. Paragraphs 3 to 10 of the Extended Reasons deal with applications made by the Applicant for part time working. In the round these show that those applications were turned down because they did not conform with the above requirement or condition. This is indicated by, for example, paragraph 7 of the Extended Reasons which is in the following terms:

"Without seeing her, the district Commander, Supt Nelson, turned the application down. We are satisfied that the principal reason for turning the application down was that her duty pattern did not conform with the requirements of the Part-time Policy as set out above. Supt Nelson made clear in his evidence that he enforced this policy strictly. We consider his other reasons for rejecting the application as specious and do not accept their genuineness. Even though she was not fully qualified he knew that she could start in the job subject to a tutorship basis. We also consider his suggestion that the "hours" available were not a "vacancy" was obtuse and obstructive."

The most relevant paragraphs in the Extended Reasons

20. The paragraphs which deal expressly with "disparate effect" are paragraphs 14 to 21 of the Extended Reasons. They are in the following terms

"14 We next have to decide on the disparate effect. To do this we must first decide the pool We are warned in London Underground Ltd v Edwards (1998) IRLR 364 at para 24 that we have to get it right. So we approach this decision with some trepidation. Since the condition or requirement applied to the whole of the respondent Force, we consider that this is the pool we have to look at. This pool consists of 3,016 officer of whom 2,581 are men and 435 are women.

15 We next have to see how many of each sex cannot comply with the requirement. Evidence of this is inevitably subject to uncertainty. There may be people who have left the force because they cannot comply. There may be people who are only able to comply with the greatest difficulty but have decided not to bring proceedings. It is however, clear that the requirement is not applied uniformly across the whole force. The evidence indicates that in Somerset East District (based in Yeovil) there are six officers who do not fully comply with the requirement and that there is at least one other (PC Worrall) in another district. Furthermore we have had evidence from two part-time PC's Butt and Young, from Somerset West, that the requirement has caused them considerable difficulty and has resulted in their husbands having to change their jobs.

16 We consider that we are entitled to draw the inference that all officers who have part-time agreements which allow them to work hours which do not comply with the requirement do so because they cannot comply with that requirement. We are also entitled to take the view that people who are forced to comply with the requirement at the expense of obliging their husbands to alter their jobs can be said to be unable to comply. We are also entitled to take the view that the number set out is the absolute minimum.

17 All of the people set out above are women. There has been some evidence of a man who works full-time but can only comply with difficulty. We are fully willing to accept that this requirement is one which can cause difficulties for men as well as women. It would include, for example, people with childcare obligations and with sick or elderly relatives, as well as people whose health preludes shift working.

18 The evidence we have indicates that 10 out of 435 women cannot comply against 1 out of 3016 men. This is a difference of less than 2%. We are, however satisfied that it is not a fortuitous difference, that it is statistically significant and that it represents the bare minimum of people who cannot comply.

19 We consider that the difference is not only statistically significant, but also is explicable in terms of sex discrimination. The majority of the people who cannot comply have childcare responsibilities and we accept that as a general rule, the overwhelming burden (if such it be) of childcare responsibilities fall on women.

20 Even if we are wrong about the pool, similar considerations apply to all other pools. If we look only at Somerset West, we have evidence of one officer out to 31 women who cannot comply against 243 men all of whom can comply. If we include those who, to our knowledge, comply with difficulty it is 3 women out of 31 against 1 man out of 243. Similar figures appear if we restrict the pool to 24 hour officers in Somerset West, the figures then being 243 men who can all comply against 1 or 3 women out of 19 who cannot comply.

21 In those circumstances we have come to the conclusion that the disparate impact provision of s.1(1)(b)(ii) has been shown by the applicant."

21. It was common ground before us that paragraph 18 of the Extended Reasons contains a mathematical, or typing, error because the force numbered 3016 and the number of men in it numbered 2581. So on the analysis of the Employment Tribunal the correct figures should be 10 out of 435 women cannot comply against 1 out of 2581 men. Thus 99.96% of men could comply and 97.70% of the women could comply: a difference of 2.26%.
22. The Employment Tribunal go on to deal with "justification" and then "detriment to the Applicant". As to those matters in our judgment the following paragraphs of the Extended Reasons are of particular relevance namely 22 to 25, 27, 28 and 32. They are in the following terms (with our emphasis):

"22 The respondent's contention is that the requirement is justifiable under s.1(1)(b)(ii) Sex Discrimination Act 1975. This contention is undermined by the fact that the condition at issue has been removed from the new draft policy, which was already in preparation in March 1999 and was disseminated in November 1999 (though it has not yet been formally approved). The contention is further undermined by the fact that it has not been strictly applied in all areas. In particular in the neighbouring Somerset East district there are a number of officers who have entered into part-time agreements which do not comply with this requirement. We accept, however that this is not an absolute answer.

23 The basis on which we assess justification is that set out in the judgment of Balcombe L J in the Court of Appeal in Hampson v Dept of Education (1989) IRLR 69:-

"In my judgment 'justifiable' requires an objective balance between the discriminatory effect of the condition and the reasonable needs of the party who applies the condition."

There is a singular absence in the evidence of the respondent's witnesses of any acceptance or appreciation of the discriminatory effect of the condition or, therefore, of any attempt to carry out the balancing exercise.

24 There is no dispute that the overwhelming obligation of the respondents is to provide adequate police cover 24 hours a day every day of the year. Much of the respondent's evidence concentrates on showing that this is best met by a rotating shift system. This, again, is not in dispute. The issue is whether there should be a requirement that all officers should comply with this system.

25 On this issue the respondent's first contention is that there would be operational problems were this requirement not strictly enforced. We accept that the presence in any section of an officer who works days and not weekends means a corresponding increase on the burden on other officers to work unsocial hours. This could, in certain circumstances, make it more difficult to maintain minimum manning levels at these times. There is no evidence that this has in fact been a problem in Somerset East where there are several such officers, but nevertheless this is a relevant factor to be balanced against the discriminatory effect.

-------------------------

27 The respondents also contend that having officers working atypical hours would have an effect on morale. They have no direct evidence of this, either from the situation in Somerset East or from their Staff Survey. In the absence of any such evidence we do not think that we can assume that officers would be demoralised by having to support other officers who are seeking to juggle the demands of childcare with their chosen careers.

28 The respondents contend that to make an exception of the applicant would lead to a large number of applications from other officers for similar advantages. To some extent this is double-edged, because it assumes that there are a considerable number of officers who are finding difficulties in meeting the pattern of hours required. In any event, we consider that this does not justify a blanket condition. If the condition were removed, this would not oblige the respondents to accept all applications for atypical hours. The person would first have to show a strong case for working such hours and it would then be open to the respondents to show that, in the particular circumstances, to allow the officer to work atypical hours would have a deleterious operational effect. This would be the case, for example, where there is already one or more officers in the same sections working such hours.

-------------------------

32 The next issue is whether the applicant can comply with the condition. We are surprised that this is still in issue. The applicant is a single mother and is dependent on professional childcare except for one weekend in five when her ex-partner has the children. Her evidence, which we accept, is that nursery schools only open weekdays and only during daytime hours. She has a two-bedroomed house which precludes a live-in nanny or au pair. We find the suggestion that she should be able to swap houses with her ex-partner offensive. We are struck by the fact that the respondents are maintaining this contention despite the fact that they made no efforts to assist the applicant in finding appropriate childcare and they have no policy of supporting childcare places for their officer, given the unsocial working patterns of most Force staff."

Some further preliminary points

23. In our judgment paragraph 32 of the Extended Reasons is significant outside the issue with which it deals expressly namely "detriment to the Applicant". This is because although this paragraph refers to particular difficulties of the Applicant it is also a demonstration, and confirmation, of what we would regard as obvious, namely that the requirement or condition creates difficulties in respect of child care and in particular difficulties for persons who have primary responsibility within their families for child care, or who are single parents. For example the point made about nursery schools clearly leads to problems in finding any, or any affordable, child-care during the evenings, or at week ends if a member of the family is not readily available to provide it. It seems to us obvious that these problems are compounded if, as the requirement or condition would require, the person employed to provide child-care would have to work on different days, nights and weekends to match the part time working rota set by the relevant district or department.

Authorities

24. As appears from the Extended Reasons the Employment Tribunal were referred to London Underground Ltd v Edwards [1998] IRLR 364 and to Hampson v Department of Education [1989] IRLR 69.
25. The hearing before the Employment Tribunal took place in January 2000. We were told and accept, that the Employment Tribunal were not taken to the report of the decision of the European Court of Justice in R v Secretary of State for Employment ex parte Seymour-Smith [1999] IRLR 253 but were referred to a passage in Harvey on Industrial Relations and Employment Law which covered that case. We confess that given the representation of the parties we find it surprising that the Employment Tribunal were not referred to the case itself. Not only were we referred to it but we were also referred to the decision of the House of Lords in that case whose decision was given on 17 February 2000 (i.e. after the hearing before the Employment Tribunal) and is reported as R v Secretary of State for Employment ex parte Seymour-Smith and Perez (No 2) [2000] IRLR 263.
26. We were also referred to other cases but will not mention them at this stage.
27. It is plain that the Employment Tribunal in large measure based their decision on the reasoning of Court of Appeal in London Underground Ltd v Edwards (No 2) [1998] IRLR 365. As appears from paragraph 3 of the judgment of Potter L J in that case in the Edwards (No 1) case it was found by the Employment Appeal Tribunal that the Employment Tribunal had erred in law by selecting the wrong "pool" of employees for comparison. It seems that this decision, paragraph 24 of the judgment of Potter LJ in Edwards (No 2) and, we were told, submissions made by counsel found the comment made by the Employment Tribunal in paragraph 14 of the Extended Reasons that they have to get the pool right. We return to this point when we consider the selection of the pool.
28. We were referred in particular to the following paragraphs in the judgments of the Court of Appeal in the Edwards (No 2) case namely:

(a) paragraphs 23 to 30 of the judgment of Potter L J which are as follows (with our emphasis):

"23 I accept the submissions of Mr Allen. In my view, there is a dual statutory purpose underlying the provisions of s.1(1)(b) and in particular the necessity under subparagraph (i) to show that the proportion of women who can comply with a given requirement or condition is 'considerably smaller' than the proportion of men who can comply with it. The first is to prescribe as the threshold for intervention a situation in which there exists a substantial and not merely marginal discriminatory effect (disparate impact) as between men and women, so that it can be clearly demonstrated that a prima facie case of (indirect) discrimination exists, sufficient to require the employer to justify the application of the condition or requirement in question: see subparagraph (ii). The second is to ensure that a tribunal charged with deciding whether or not the requirement is discriminatory may be confident that its disparate impact is inherent in the application of the requirement or condition and is not simply the product of unreliable statistics or fortuitous circumstances. Since the disparate impact question will require to be resolved in an infinite number of different employment situations, well but by no means comprehensively exemplified in the arguments of Mr Allen, an area of flexibility (or margin of appreciation), is necessarily applicable to the question of whether a particular percentage is to be regarded as 'substantially smaller' in any given case

24 The first or preliminary matter to be considered by the tribunal is the identification of the appropriate pool within which the exercise of comparison is to be performed. Selection of the wrong pool will invalidate the exercise; see, for instance, Edwards (No1) and Jones v University of Manchester (1993) IRLR 218, and cf the judgment of Stephenson LJ in Perera v Civil Service (No 2) (1983) IRLR 166 at 170, 26-27 in the context of racial discrimination. The identity of the appropriate pool will depend upon identifying that 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 LU workforce namely train operators, to whom the new rostering arrangements were to be applied (see paragraph 3 above). It did not include all LU employees. Nor did the pool extend to include the wider field of potential new applicants to LU for a job as a train operator. That is because the discrimination complained of was the requirement for existing employees to enter into a new contract embodying the rostering arrangement; it was not a complaint brought by an applicant from outside complaining about the terms of the job applied for. There has been no dispute between the parties to this appeal on that score. However, Mr Bean has placed emphasis on the restricted nature of the pool when asserting that the industrial tribunal were not entitled to look outside it in any respect. Thus he submitted they should not have taken into account, as it apparently did, its own knowledge and experience, or the broad national 'statistic' that the ratio of single parents having care of a child is some 10:1 as between women and men.

25 In my view, Mr Bean was incorrect in that last respect. An industrial tribunal does not sit in blinkers. Its members are selected in order to have a degree of knowledge and expertise in the industrial field generally. The high preponderance of single mothers having care of a child is a matter of common knowledge. Even if the 'statistic', ie the precise ratio referred to is less well know, it was in any event apparently discussed at the hearing before the industrial tribunal without doubt or reservation on either side. It thus seems clear to me that, when considering as a basis for their decision the reliability of the figures with which they were presented, the industrial tribunal were entitled to take the view that the percentage difference represented a minimum rather than a maximum so far as discriminatory effect was concerned.

26 Equally, I consider that the industrial tribunal was entitled to have regard to the large discrepancy in numbers between male and female operators making up the pool for its consideration. Not one of the male component of just over 2,000 men was unable to comply with the rostering arrangements. On the other hand, one woman could not comply out of the female component of only 21. It seems to me that the comparatively small size of the female component indicated, again without the need for specific evidence, both that it was either difficult or unattractive for women to work as train operators in any event and that the figure of 95.2% of women unable to comply was likely to be a minimum rather than a maximum figure. Further, if for any reason fortuitous error was present or comprehensive evidence lacking, an unallowed-for increase of no more than one in the women unable to comply would produce an effective figure of some 10% as against the nil figure in respect of men; on the other hand, one male employee unable to comply would scarcely alter the proportional difference at all. Again, I consider Mr Allen is right to point out in relation to Mrs Quinlan that, albeit the industrial tribunal lacked the evidence to find as a fact that she could not comply, the reference to her indicates that they had her uncertain position in mind when assessing the firmness of the figure of only 4.8% as the basis for a finding of prima facie discrimination.

27 Finally, although the industrial tribunal did not state that they placed reliance upon the LU's original proposal for the single-parent link scheme, which was designed to deal with precisely the kind of difficulties which faced the applicant, they might legitimately have done so in my view; it was a recognition by LU of the likely adverse impact of the proposed rostering arrangements on those members of the workforce who were single parents and thus afforded support for the industrial tribunal to take the view that the effect of the rostering arrangements, unalleviated by such a scheme, was discriminatory in effect.

28 Thus I do not consider that the decision or reasoning of the industrial tribunal is open to attack unless Mr Bean can make good his main argument that their conclusion was contrary to the wording of the stature, or alternatively a decision which no reasonable tribunal could have reached. As to the first, in the light of the previous expressions of view of this court to which I have referred, I have not found this an easy question. However, I have concluded that the argument must fail in the circumstances of this case.

29 In many respects, no doubt, it would be useful to lay down in relation to s.1(1)(b) a rule of thumb or to draw a line defining the margin within, or threshold beyond, which, in relation to small percentage differences, the lower percentage should not reasonably be regarded as 'considerably smaller' than the higher percentage. However, it does not seem to me appropriate to do so. For the various reasons discussed in this judgment, and because of the wide field and variety of situations in which the provisions of the section are to be applied, the circumstances and arguments before the adjudicating tribunal are bound to differ as to what in a particular case amounts to a proportion which is 'considerably smaller' for the purposes of determining the discriminatory or potentially discriminatory nature of a particular requirement or condition. If a figure were to be selected in the field of employment, it would be likely to vary according to the context, and in particular as between a case where the requirement or condition is applied on a national scale in respect of which reliable supporting statistics are available and those where it is applied in relation to a small firm or an unbalanced workforce where the decision may have to be made on far less certain evidence and to a large degree upon the basis of the industrial tribunal's own experience and assessment as applied to such figures as are available. The difficulties are well illustrated by this case.

30 Plainly, a percentage difference of no more than 5% or thereabouts is inherently likely to lead an industrial tribunal to the conclusion that the requirements of s.1(1)(b) have not been made out, but I am not prepared to say that such a conclusion must inevitably follow in every case. Nor, for essentially the same reasons as those expressed by Knox J in Greater Manchester Police Authority v Lea, am I prepared to say that such a conclusion necessarily followed in this case."

(b) paragraphs 40, 41, 43, 48 and 49 of the judgment of Simon Brown L J are in the following terms (with our emphasis):

"40 The critical question arising on the appeal is this: On these facts can it properly be said that 95.2% is a 'considerably smaller' proportion than 100% within the meaning of s.1(1)(b)(i) of the Act?

41 Contending for the answer no, Mr Bean QC submits in effect that the question answers itself: a difference of under 5% is simply not sufficient. It would, he says, be perverse to describe 95."% as 'considerably smaller' than 100%. On this approach, of course, the fact that only one in 21 women was unable to comply with the new roster requirement was of itself fatal to the respondent's cause: it mattered not whether any of the men were affected, nor what proportions of men and women are lone parents in the population as a whole. Those further facts, submits Mr Bean, are simply irrelevant.

------------

43 Mr Allen's second and principal argument is that, even assuming the 5% figure to be correct (as I think we must), ie that 95% of the women in the group could comply with the new roster requirement, that proportion could properly be held to be considerably smaller than the 100% proportion of men. It is not to be overlooked, he submits, that at the other end of the scale 5% of the women were disadvantaged indeed. In the final analysis, he argues, equality of treatment is the paramount consideration.

48 As to how percentages of this sort have been regarded in the past, one finds only limited assistance, and indeed some inconsistency, in the cases. Whereas, for example, in Staffordshire County Council v Black (1995) IRLR 234, the EAT (presided over, be it noted, by Morison J) upheld the industrial tribunal's finding that 89.5% was not 'considerably smaller' that 97% - stating that 'the figures speak for themselves' in Greater Manchester Police Authority v Lea (1990) IRLR 375, the EAT (presided over by Knox J) upheld the industrial tribunal's finding that 95.3% was 'considerably smaller' than 99.4%.

49 I can state my conclusions really quite shortly. Given that this legislation is concerned essentially to contrast the impact of a given requirement or condition as between men and women rather than as between the women in the group, it would seem to me wrong to ignore entirely the striking fact here that not a single man was disadvantaged by this requirement despite the vast preponderance of men within the group. Looked at in the round, this requirement clearly bore disproportionately as between men and women, even though only one woman was affected by it. Had there been an equal number of women drivers to male drivers and the same 5% proportion of them been affected, ie 100, Mr Bean's argument would remain the same, namely that too large a proportion of women were able to comply with the requirement to leave room for a finding that such proportion was 'considerably smaller' than the proportion of men who could comply. It is not an argument I am ultimately prepared to accept. The approach to s.1(1)(b)(i) must, I conclude, be more flexible than this argument allows. Parliament has not, be it noted, chosen to stipulate, as it could, just what difference in proportions would be sufficient. Once, then, one departs from the purely mechanistic approach contended for by the appellants, and has regard to other facts besides merely a comparison between 95% and 100%, the respondent's argument becomes compelling: no other fact could be more relevant than that, whereas 5% of the women were disadvantaged, not one of the 2,023 men was. That further consideration in my judgment supports the industrial tribunal's finding her."

Swinton Thomas L J indicated that he too had found the case to be a difficult one.

29. During the hearing it struck me that there was a similarity in approach between that taken by the Court of Appeal in the Edwards (No 2) case and that taken by the House of Lords in a case in which I appeared as counsel namely R v Monopolies & Mergers Commission ex parte SYT [1993] 1 WLR 23 on the question whether for the purposes of Section 64 of the Fair Trading Act 1973 South Yorkshire was a "substantial part of the United Kingdom". The issue arose in connection with the provision of local bus services. In that case at page 32 B in his speech Lord Mustill said this:

"Accordingly I prefer to state that the part must be "of such size, character and importance as to make it worth consideration for the purposes of the act." To this question an enquiry into proportionality will often be material but it will not lead directly to a conclusion."

That passage has to be read with the passage at page 31 A where Lord Mustill said this:-

"My Lords, although I agree that the relationship of the part to the whole is not to be ignored, I am unable to accept that proportionality is the beginning and end of the matter".

That passage and the second sentence of the passage at page 32 which I have cited found an argument that the approach of Lord Mustill does not apply in this case where section 1 (1)(b) SDA expressly refers to proportions.

30. Before leaving the SYT case the passage at page 32 C through to 33 A of Lord Mustill's speech is also relevant to issues as to the choice of the pool and perversity raised in this case. That passage is in the following terms:

"Applying this test to the present case one will ask first whether any misdirection is established, and secondly whether the decision can be overturned on the facts. As to the first it is quite clear that the approach of the commission was in general accord to what I would propose. It is true that matters such as academic and sports activities, mentioned by the commission, are of marginal importance at the most, but I do not regard their inclusion in the list of features to which the commission paid regard as vitiating an appreciation of 'substantive' which was broadly correct. On the second question the parties are at odds as to the proper function of the courts. The respondents say that the two stages of the commission's inquiry involved wholly different tasks. Once the commission reached the stage of deciding on public interest and remedies it was exercising a broad judgment whose outcome could be overturned only on the ground of irrationality. The question of jurisdiction, by contrast, is a hard-edged question. There is no room for legitimate disagreement. Either the commission had jurisdiction or it had not. The fact that it is quite hard to discover the meaning of section 64(3) makes no difference. It does have a correct meaning, and one meaning alone; and once this is ascertained a correct application of it to the facts of the case will always yield the same answer. If the commission has reached a different answer it is wrong, and the court can and must intervene.

I agree with this argument in part, but only in part. Once the criterion for a judgment has been properly understood, the fact that it was formerly part of a range of possible criteria from which it was difficult to choose and on which opinions might legitimately differ becomes a matter of history. The judgment now proceeds unequivocally on the basis of the criterion as ascertained. So far, no room for controversy. But this clear-cut approach cannot be applied to every case, for the criterion so established may itself be so imprecise that different decision-makers, each acting rationally, might reach differing conclusions when applying it to the facts of a given case. In such a case the court is entitled to substitute its own opinion for that of the person to whom the decision has been entrusted only if the decision is so aberrant that it cannot be classed as rational: Edwards v Bairstow [1956] AC 14. The present is such a case. Even after eliminating inappropriate senses of 'substantial' one is still left with a meaning broad enough to call for the exercise of judgment rather than an exact quantitative measurement. Approaching the matter in this light I am quite satisfied that there is no ground for interference by the court, since the conclusion at the which the commission arrived was well within the permissible field of judgment. Indeed I would go further, and say that in my opinion it was right."

31. In the Seymour-Smith case the European Court of Justice answer the third question posed to it by the House of Lords in paragraphs 51 to 65 of the judgment which are in the following terms (with our emphasis):

"51 By its third question, the national court seeks to ascertain the legal test for establishing 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.

52 Article 119 of the Treaty sets out the principle that men and women should receive equal pay for equal work. That principle excludes not only the application of provisions leading to direct sex discrimination, but also the application of provisions which maintain different treatment between men and women at work as a result of the application of criteria not based on sex where those differences of treatment are not attributable to objective factors unrelated to sex discrimination (see joined cases C-399/92, C-409/92, C-425/92, C-34/93, C-50/93 and C-78/93 Helmig and others [1994] IRLR 216, paragraph 20).

53 It is common ground that the disputed rule does not entail direct sex discrimination. It must therefore be considered whether the rule may constitute indirect discrimination incompatible with Article 119 of the Treaty.

54 The applicants in the main proceedings maintain that where there is an intrinsic risk that a measure adopted by a Member State will have a disparate effect on pay as between men and women, and/or such disparate effect is actually demonstrated by reliable and significant statistics, Article 119 of the EC Treaty is infringed unless that measure can be shown to be based upon objectively justified factors unrelated to sex.

55 In particular, they claim that where there are statistics which are significant, cover the entire workforce, and demonstrate long-term phenomena that cannot be explained as fortuitous, anything more than a minimal difference in impact would infringe the obligation to give effect to the principle of equal treatment.

56 According to the United Kingdom Government, the terms used by the Court in its case law on indirect discrimination clearly show that it has in mind a markedly different impact.

57 For its part, the Commission proposes a 'statistically significant' test, whereby statistics must form an adequate basis of comparison and the national court must ensure that they are not distorted by factors specific to the case. The existence of statistically significant evidence is enough to establish disproportionate impact and pass the onus to the author of the allegedly discriminatory measure.

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 (stet) ] 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."

32. The House of Lords in the Seymour-Smith (No 2) case applied that test and by a majority (3 to 2) decided as stated in the head note that:

"The applicants had shown that at the time of their dismissal in 1991 the two-year qualifying period to bring an unfair dismissal complaint had a disparately adverse impact on women so as to amount to indirect discrimination contrary to Article 119 of the EC Treaty.

In a case of indirect discrimination, the obligation is to avoid applying unjustifiable requirements having a considerable disparity of impact. The approach adopted by the European Court is similar to that provided in s.1(1)(b) of the Sex Discrimination Act. A considerable disparity can be more readily established if the statistical evidence covers a long period and the figures show a persistent and relatively constant disparity. In such a case, a lesser statistical disparity may suffice to show that the disparity is considerable than if the statistics cover only a short period or if they present an uneven picture.

In the present case, from 1985 up to and including 1991, the ratio of men and women who qualified was roughly 10:9. A persistent and constant disparity of that order in respect of the entire male and female labour forces was adequate to demonstrate that the extension of the qualifying period had a considerably greater adverse impact on women than men."

In paragraphs 51 to 61 of his speech Lord Nicholls said (with our emphasis):

"51 This is not a case of direct discrimination. The 1985 Order drew no distinction between men and women. The contention of Ms Seymour-Smith and Ms Perez is that the order was indirectly discriminatory, because it introduced a qualifying condition which bore more hardly on women than men. Women were treated less favourably than men, the two-year qualifying period was smaller than the proportion of men. Before your Lordships two issues arise for decision. The first is whether, at the time of the dismissal of Ms Seymour-Smith and Ms Perez in 1991, the 1985 Order did have a disparately adverse impact on women. If it did, the second issue calling for decision is whether the differential impact was objectively justified.

52 Disparately adverse impact

One of the questions referred by your Lordships' House to the European Court sought guidance on 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. The view of the European Court is set out in paragraphs 60-65 of its judgment: (1999) IRLR 253, 266-267. These paragraphs are reproduced in the speech of my noble and learned friend Lord Slynn of Hadley. These paragraphs, especially the relationship between paragraphs 60 and 61, were much analysed and debated before your Lordships. The European Court, it was submitted, wavered uncertainly between what must be established as a substantive criterion and the evidence needed for that purpose. In paragraph 60 the court referred to statistical evidence that 'a considerably smaller percentage of women than men is able to satisfy the condition', and treated this as evidence of apparent sex discrimination calling for justification. However, in the next paragraph the court observed that statistical evidence revealing '' lesser but persistent and relatively constant disparity over a long period' could also be evidence of apparent sex discrimination. In paragraph 61, unlike paragraph 60, the court gave no guidance on the extent of statistical disparity required to establish apparent sex discrimination. Nor did the court spell out, in so many words, how these two approaches fit together. Nor, on its face, does the answer given by the Court to the referred question in paragraph 65 leave any room to apply the 'lesser disparity' approach envisaged in paragraph 61.

53 As I see it, the reasoning underlying these paragraphs is that, in the case of indirect discrimination, the obligation to avoid discrimination does not consist of applying requirements having precisely the same impact on men and women employees. The obligation is to avoid applying unjustifiable requirements having a considerable disparity of impact. In this regard the European court has adopted an approach similar to that provided in s.1(1)(b) of the Sex Discrimination Act 1975. A considerable disparity can be more readily established if the statistical evidence covers a long period and the figures show a persistent and relatively constant disparity. In such a case a lesser statistical disparity may suffice to show that the disparity is considerable than if the statistics cover only a short period or if they present an uneven picture.

54 Having set out the applicable principles, the European Court addressed the facts in the present case. In doing so the Court focused exclusively on the 1985 statistics: see paragraphs 62-65. This may have been because the Court had foremost in mind the contention that the making of the 1985 Order was ultra vires. The relevant date for the purpose of deciding that issue was 1985, being the date when the Order was made. The 1985 statistics showed that 77.4% of men and 68.9% of women satisfied the extended qualifying period introduced the 1985 Order. The Court commented that, on its face, those statistics did not appear to show that a considerably smaller percentage of women than men was able to fulfil the extended requirement. In its formal answer to the referred question in paragraph 65, the Court repeated the considerably smaller percentage test already stated in paragraph 60. However, I cannot believe that in doing so the Court intended to repudiate the approach stated in paragraph 61. Nor do I think this can be taken as an indication by the Court that the paragraph 61 approach was inapplicable in this case if a date later than 1985 was the appropriate date.

55 Before your Lordships it was common ground that 1991, not 1985, was the relevant date for the purpose of the issue now being considered. The position at this later date was not considered by the European Court. I turn to the available statistics, covering the period from 1985 to 1993, extracted from the annual labour force surveys.

Percentage of men Percentage of women

with more with more

Year than 2 years than 2 years Disparity

1985 77.4 68.9 8.5

1986 77.2 68.4 8.8

1987 75.3 67.1 8.2

1988 73.4 65.6 7.8

1989 72.0 63.8 8.2

1990 72.5 64.1 8.4

1991 74.5 67.4 7.1

1992 77.9 72.1 5.8

1993 78.4 74.1 4.3

56 These figures show that over a 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.

57 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 the majority 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.

58 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.

59 Qualifiers and non-qualifiers

Before turning to justification I must mention a point which was debated before your Lordships. In paragraph 59 of its judgment the European Court described the approach which should be adopted to the comparison of statistics:

'…….. the best approach to the comparison of statistics is to consider, on the one hand, the respective proportions of men 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 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.

61 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 –184/89) (1991) IRLR 222, 225, paragraph 12, and Kowalska v Freie und Hansestadt Hamburg (case –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."

We pause to comment that contrary to what is said in paragraph 57 of his speech Lord Nicholls was in the majority and it therefore appears (and we understand that it was the case) that one of their Lordships changed his mind after Lord Nicholls had written his speech. The minority view (Lords Slynn and Steyn) is set out in the speech of Lord Slynn who was satisfied that having regard to the judgment of the European Court the percentages of those qualifying did not establish discrimination (see paragraph 35 of his speech). Lord Goff gave a reasoned speech saying why he preferred the reasoning of Lord Nicholls to that of Lord Slynn and he concluded that the approach of the Court of Appeal who were impressed by the "persistency and consistency of the figures" was in accordance with paragraph 61 of the judgment of the European Court (see paragraphs 42 and 43 of his speech).

33. Lord Slynn gave the minority speech. After setting out the third and fourth questions in paragraphs 12 and 13 of his speech he said:

"12 In answer to the third question, the Court set out the rival positions adopted before it. Thus, (a) the applicants contended that where there are significant statistics covering the entire workforce which cannot be regarded as fortuitous, 'anything more than a minimal difference' in impact could be a breach; (b) the United Kingdom contended for 'a markedly different impact'; (c) the Commission proposed a test by which 'the existence of statistically significant' evidence is enough to establish disproportionate impact.

13 The Court then indicated its conclusion as to the proper approach. The first question is whether the measure has 'a more unfavourable impact on women than on men'. The best approach to the comparison of statistics relevant to that is 'to consider, on the one hand, the respective proportions of men in the workforce able to satisfy the requirement of two years' employment under the disputed rule and those unable to do so, and on the other, to compare those proportions as regards women in the workforce': [1999] IRLR 253, 278, paragraph 59."

In paragraph 14 he sets our paragraphs 60 to 65 of the judgment of the European Court and in paragraphs 22 to 27 says:

"22 As far as 1985 is concerned, I do not consider that it has been established that these statistics are insufficient or fortuitous; on the contrary, I accept, as the Divisional Court accepted, that these figures were appropriate. In my view, these figures are 'relevant and sufficient for the purposes of resolving the case' (paragraph 62). They do not show that in 1985 a considerably smaller percentage of women than men was able to fulfil the requirement.

23 The European Court contemplates that a percentage which is 'lesser' than 'considerably smaller' may yet be evidence of indirect discrimination if it is 'persistent and relatively constant'. There thus remains the important question as to whether 'the statistical evidence revealed a lesser but persistent and relatively constant disparity over a long period between men and women'. The European Court did not give guidance on this question as it had on the 1985 figures.

24 Starting from the basis that the 1985 percentage for women was not in itself 'considerably smaller' for the purposes of the rule, it must thus be asked whether it was such that, if it continued, it could be said to be 'consistent and relatively constant'? This is no doubt the most difficult aspect of this part of the case, in view of the numbers (rather than the percentage) of women involved. As has been seen, the Divisional Court was satisfied that, looking at the period from 1985 to 1993 as a whole, the figures were not 'considerably smaller'. They do not expressly consider whether, although less than considerably smaller, the figures were sufficiently persistent and constant as to indicate indirect discrimination.

25 The Court of Appeal, on the other hand, were impressed by the 'persistency and consistency' of the figures from 1985 to 1991, though they accepted that there was a narrowing of the gap in 1992 and a 'more marked narrowing of the gap in 1993'. They concluded that it had been shown 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 between the numbers and percentage of men and women in the group' who did and did not qualify.

26 It seems to me clear, from their conclusion that in the period 'there had been' a considerable difference, that the Court of Appeal were proceeding on the basis that the percentage of women who qualified in the beginning was considerably smaller; that considerable difference 'continued'.

27 If the starting point is, on the other hand, that the percentage of women qualifying in 1985 was not 'considerably smaller', the conclusion does not necessarily follow. Accordingly, although your Lordships are reluctant to interfere with the assessment of the Court of Appeal on issues of fact, it seems to me that the matter has to be looked at from the starting point indicated in the European Court's judgment. Since the figures are agreed, it seems to me open to your Lordships to resolve this issue without the matter being referred back to the Court of Appeal or the Divisional Court."

34. In paragraph 42 Lord Goff disagreed with Lord Slynn's view as expressed in paragraph 26 in the following terms:

"42 My noble and learned friend concludes that the Court of Appeal were proceeding on the basis that the percentage of women who qualified in the beginning was considerably smaller, and that that considerable difference 'continued'; and comments that, if in the beginning that was not so, the conclusion does not necessarily follow. I very respectfully suggest that this is reading too much into the Court of Appeal's judgment. The Court of Appeal stressed that they were impressed by the 'persistency and consistency of the figures from 1985 to 1991 inclusive' (see p.476, 101). If they had believed that the statistics as at 1985 were sufficient of themselves to justify the considerable disparity, they would doubtless have said so; moreover, it would have been unnecessary for them to invoke the persistence and consistency of the figures from 1985 to 1991 unless they considered that, in such circumstances, a lesser disparity over that period would satisfy the 'considerably smaller' requirement. It is therefore my understanding that the Court of Appeal regarded the persistence and consistency of the disparity as essential to their decision and so concluded, on the statistics over the period leading up to the dismissals, that (to adopt the words of the European Court) a 'lesser but persistent and relatively constant disparity' existed over that long period. I strongly suspect (I can say no more) that the European Court interpreted the Court of Appeal's judgment in the same way."

35. In paragraph 28 Lord Slynn sets out the statistics and in paragraphs 29 and 34 to 36 he said:

"29 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."

…

34 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.

35 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 woman 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.

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

Comments and conclusions by reference to those authorities

36. We state and make the following views and comments:

(1) In considering the decision of the ECJ it is important to remember that the Seymour-Smith case was concerned with a measure adopted by a Member State which affected the workforce of the State and there was no question posed as to the pool to be taken into account. Reliable statistics were available and on the approach taken by the parties and both the ECJ and the House of Lords they were the material used to assess the extent of the disparate effect of the measure.

(2) The ECJ base their decision on their existing case law.

(3) The ECJ confirm that there are two stages to the overall determination of whether there is indirect sex discrimination for the purposes of Article 119 which accords with subsections 1 (1)(b)(i) and (ii) SDA. The two stages are "disparate effect" and "justification". The first stage is a "trigger or threshold stage" (also see Edwards (No 2) paragraph 23). It involves a consideration of the question whether the provision has a disparate effect as between men and women. If it does have that effect to a sufficient degree there is indirect sex discrimination unless and the second or "justification stage" the provision is justified by objective factors unrelated to any discrimination based on sex (see paragraph 65 of the judgment of the ECJ and, for example, paragraph 120 of the opinion of the Advocate General).

(4) The decision of the ECJ (and the House of Lords) is not authority for the proposition that the first stage can only be established by reference to statistics relating to the effect of the relevant measure or provision. Rather the decision of the ECJ is directed to the approach to be taken to available and reliable statistics and in this context it confirms paragraph 17 of the judgment in Enderby v Frenchay Health Authority [1993] IRLR 591 at 595

(5) The ECJ does not directly answer the point made by the Applicants that the first stage will be satisfied if there is an intrinsic risk that the measure adopted will have a disparate effect as between men and women. This was based on the O'Flynn case (see the summary of the arguments at [1999] IRLR 258). However the answers that they do give indicate that that they did not accept that the Seymour- Smith case was one in which the answer could be based on an intrinsic risk and the correct approach was to consider the available and reliable statistics.

(6) The judgment in O'Flynn v Adjudication Officer is reported at [1996] 3 CMLR 124. It is a case concerning the effect on migrant and national workers of a provision introduced by the United Kingdom relating to funeral payments. The condition in question was a territorial condition. The respondent authorities argued that the condition was only discriminatory if it was shown that it was substantially more difficult for migrant workers than for national workers to satisfy it. It was held that the first stage is satisfied if the measure or provision is intrinsically liable to affect migrant workers more than national workers and in such a case it is not necessary to find that the provision does in practice affect a substantially higher proportion of migrant workers (see paragraphs 17 to 21 of the judgment).

(7) It seems to us that by paragraph 60 (and paragraphs 63 and 67) of their judgment the ECJ reject the argument of the Applicants (recorded in paragraph 55 thereof) that where there are statistics which are significant and demonstrate a long term phenomena anything more than a minimal impact would infringe the obligation to give effect to the obligation of equal treatment. This view accords with the application of the ECJ decision by the House of Lords and with the view of the Court of Appeal in Seymour-Smith (see Edwards (No 2) paragraph 15 and 18).

(8) Much of the argument in the House of Lords focused on paragraphs 60 and 61 of the judgment of the ECJ in Seymour-Smith. As the highlighted passages from the speeches set out above show the majority concluded that if the available statistics in the relevant year (or years) did not as a matter of a free standing percentage indicate that a considerably smaller percentage of women than men is able to satisfy the condition nonetheless the first stage could be established if there was a persistent and lesser disparity over a long period. The size of that disparity is not identified by the ECJ who say that it is for the national courts to decide in those circumstances whether the first stage is satisfied. In contrast to the minority (the speech of Lord Slynn) the majority in the House of Lords were of the view that the persistent disparity they (and the Court of Appeal) relied on was in percentage terms less than "considerably smaller" (see in particular Lord Goff at paragraph 42) or "border line" (see Lord Nicholls at paragraph 57).

(9) It follows that in a case where there are relevant, reliable and available national statistics upon which a decision as to the establishment of the first stage can be based the House of Lords have interpreted the decision of the ECJ in Seymour-Smith as a decision that does not require, or establish, that a purely mathematical or mechanistic should be taken to deciding whether the first stage is satisfied. This is an approach that introduces flexibility even in that situation, and therefore confirms the need for flexibility in other situations.

(10) However both the House of Lords and the ECJ (i) reject an approach that something more than a de minimis divergence is enough, and (ii) indicate that in pure percentage terms the percentages under consideration in the Seymour-Smith were not as percentages large enough to satisfy the description "considerably smaller". Also Lord Nicholls expressly comments that the latitude afforded by the word "considerably" should not be exaggerated (paragraph 57). Both decisions therefore confirm and reinforce the point made in Edwards (No 2) that a percentage difference of no more than 5%, or thereabouts, is inherently unlikely to lead to the conclusion that the first stage has been made out. Indeed in the Seymour-Smith case the percentages were bigger (in 1985 8.5%) and both the ECJ and the House of Lords were of the view that that percentage by itself (or on the face of it – see paragraph 64 of the ECJ judgment) did not satisfy the first stage by demonstrating the existence of a "considerably smaller" percentage or proportion.

(11) Edwards (No 2) was a case where there were no relevant national statistics and the proportions under consideration were from the relevant workforce. This will not be an uncommon situation. Indeed in some cases, for example, on the introduction of a condition or requirement, there will be no experience of its operation within the relevant workforce upon which statistics could be based. In our judgment in such a situation an employer could not be heard to say that no claim for indirect sex discrimination could succeed until the requirement or condition had been in force long enough to produce information on which statistics could be based. In those circumstances the courts would consider different methods of assessing whether the first stage was satisfied. It may be that this could not be demonstrated without some experience of the operation of the relevant condition or requirement (or an equivalent one in a different workforce) but this is not the same thing as saying that it could never be demonstrated without that experience or statistical evidence

(12) In situations such as that envisaged in the example given in sub-paragraph (11), and in cases where there is not a large workforce on which reliable statistics can be based or relevant national statistics, in our judgment the points set out in (i) to (v) below all support the view that in assessing whether the statutory test which involves the application of a word of degree with an inexact meaning (and what was described in SYT as a chameleon word) the court can, and should, take a flexible approach having regard to the circumstances of the given case, namely:

(i) the flexible approach taken in Seymour Smith by reference to disparity over a number of years shown by national statistics (see subparagraph (9) above),

(ii) the fact that the provision under consideration is a trigger or threshold (see for confirmation Edwards (No 2) paragraph 23 and see also (with the qualification that here the section refers to proportions) SYT),

(iii) the underlying purpose of the legislation and the European legislation upon which it is based,

(iv) the nature of the test which is, or is essentially, one of fact and the wide variety of situations that the legislation (and thus the threshold provision) is intended to cover, and

(v) by analogy the approach taken in O'Flynn.

(13) In our judgment this flexibility would enable a court to have regard to amongst other things (i) the make up and overall numbers of the workforce under consideration, (ii) as in Edwards (No 2) the point that no man was disadvantaged, (iii) more generally the effect of a change in numbers of men and women e.g. when the numbers of women are small a reduction could have a dramatic effect on the proportion who are disadvantaged, (iv) history in respect of or relevant to the relevant workforce, and (v) the inherent or inherently likely effect of the condition or requirement, and thus for example as in Edwards (No 2) (paragraphs 24, 25 and 27) the adverse or likely adverse effect of a rostering arrangement on a single parent and the common knowledge that more women than men are single parents (or the parent with primary responsibility for day to day child care).

(14) It follows that in our judgment in a case such as this one, and with the qualification and confirmation referred to in sub-paragraph (10) above, the approach in Edwards (No 2) remains good law.

37. The pool. For the purposes of determining the trigger or threshold stage a comparison that satisfies s. 5(3) SDA has to be made. This section also contains words of degree and there is thus room for different decision makers applying the right criteria, test or approach to reach different conclusions. The same applies to the overall conclusion reached on the trigger or threshold stage (see the passages from the speech of Lord Mustill in SYT cited in paragraph and Lord Nicholls in Seymour-Smith paragraph 57). Thus different pools could be within the range of options or decisions open to the decision maker (here the Employment Tribunal) applying the correct criteria, test or approach. It follows in our judgment that the Employment Tribunal:

(a) were not confined to choosing one and only one pool on the basis that only one would be correct,

(b) in reaching their conclusion the Employment Tribunal could therefore have considered and weighed the position in respect of different pools within the range of decisions open to them, and

(c) in reaching their conclusion it was open to Employment Tribunal to consider and refer to other pools and thus to Somerset West (as they did in paragraph 20 of the Extended Reasons).

In this context paragraph 24 in Edwards (No 2) refers to the choice of an appropriate pool and in our judgment the reference therein to the wrong pool is to one that is outside the permissible range of choices applying the correct criteria, test or approach.

The Appeal and Cross Appeal

38. It was argued by reference to Effa v Alexander Health Care Trust and another (unreported transcript 5 November 1999), Nagarajan v London Regional Transport [1999] 3 WLR 425 and Zafar v Glasgow City Council [1998] ICR 120 that the evidence before the Employment Tribunal did not justify the inferences they drew in the first and third sentences of paragraph 16 of the Extended Reasons and thus that:

16 We consider that we are entitled to draw the inference that all officers who have part-time agreements which allow them to work hours which do not comply with the requirement do so because they cannot comply with that requirement. ------------- We are also entitled to take the view that the number set out is the absolute minimum.

We do not agree. The cases relied on relate to the drawing of an inference as to whether a person has been less favourably treated on racial grounds and thus the answer to a question posed by the statute. They point out that the inference needs to be based on primary fact. The inferences criticised are not answers to a statutory question. Nonetheless we naturally accept that they cannot be "plucked out of the air" and have to be based on evidence or a permissible use by an Employment Tribunal of their general knowledge and expertise. As to the latter point it is pointed out in Edwards (No 2) that an Employment Tribunal does not have to sit in blinkers (see paragraph 25). The second inference criticised is effectively the same as that made and found to be justifiably made in Edwards (No 2) (see again paragraph 25) and in our judgment this Employment Tribunal did not err in law in making this inference. In our judgment the criticism of the first inference is stronger but nonetheless in our judgment the Employment Tribunal were entitled to make it having regard to the evidence they heard and their general knowledge and expertise. It relates to seven female officers identified before the Employment Tribunal and the evidence they heard included evidence as to the difficulties for 24 hour patrol officers and the shifts worked by those female officers. We also agree with the point made on behalf of counsel for the Applicant, Ms Chew, that the position taken by the Chief Constable that the requirement or condition was very important supports an inference that it would only be relaxed on good grounds, rather than simple preference.

39. It was argued that although the Employment Tribunal were entitled to draw the inference in the second sentence of paragraph 16, namely:

We are also entitled to take the view that people who are forced to comply with the requirement at the expense of obliging their husbands to alter their jobs can be said to be unable to comply.

nonetheless PCs Butt and Young did comply and their evidence did not support the finding that the requirement had caused them considerable difficulty and resulted in their husbands having to change their jobs. From that it was argued that the Employment Tribunal ought to have found that 8 not 10 women could not comply. This is an attack on a primary finding of fact. We had a bundle of statements and some of the documents before the Employment Tribunal and notes of some of the evidence. Albeit that the passages we were referred to by counsel for Ms Chew do not contain assertions that the husbands of Pcs Butt and Young had to change their jobs in our judgment they do show that the Employment Tribunal had evidence before them that compliance caused these female officers considerable difficulty and were only resolved by them working opposing shifts to their husbands. In our judgment the Employment Tribunal had evidence before them on which they could reach the conclusion that 10 women (including Pcs Butt and Young) could not comply, and in any event the Chief Constable has failed to establish before us that this was not the case.

40. Leading counsel for the Chief Constable argued before us that as the decision reached by the Employment Tribunal on the pool was one of fact it could only be challenged on grounds of perversity. In doing so he relied on Kidd v D.R.G. (U.K.) Ltd [1985] ICR 405 in particular at 408 C/F, 413 C/E, 414H – 415A and 415 C/F, and Highfield v Lander Carlisle (transcript 19 October 2000) in particular at paragraphs 22 (where Kidd is relied on) and paragraph 41. We agree that as these cases demonstrate the decision on the pool is properly categorised as one of fact. However again as they show it is not one of primary fact and the decision is made by the Employment Tribunal applying the statute and their judgment and discretion. It follows in our view that the only permissible attack is not one of perversity in the Wednesbury sense (and thus strictly on the basis that the decision is outside the range of options on an application of the correct approach to its determination) and by so stating in paragraph 41 the Highfield case this Tribunal used the phrase "perversity in the Wednesbury sense" loosely. In our judgment a permissible attack includes one that the Employment Tribunal applied the wrong approach (criteria or test) (see paragraph 37 above) and we note that in paragraph 42 of the Highfield case this was in fact the ground on which this Tribunal allowed the appeal and therefore this Tribunal must be taken to have been including it within their use of the phrase "perverse in the Wednesbury sense".
41. During his submissions leading counsel for the Chief Constable, in our judgment correctly, accepted that the decision of the Employment Tribunal on the pool could be attacked on the basis that they had applied the wrong criteria, test or approach. Indeed this was the basis of one of his grounds of appeal against the conclusion reached by the Employment Tribunal on "disparate effect" and we note that in fact the paragraph he initially relied on in the Highfield case (paragraph 41) was directed to "disparate effect" and not to the pool.
42. Ms Chew in her cross appeal argued that in their selection of the pool the Employment Tribunal applied the wrong test because he submitted in reliance upon Meer v London Borough of Tower Hamlets [1988] IRLR 399 that "condition or requirement" means an absolute bar. That was a case concerning racial discrimination. During his oral submissions counsel for Ms Chew, in our judgment correctly, accepted that the Meer case was not authority that founded that argument in a case of sex discrimination having regard (amongst other things) to the Equal Treatment Directive (see also Harvey L paragraph 113). This argument was therefore correctly not pursued.
43. In our judgment the result of the argument based on Meer not being pursued is that the pool chosen by the Employment Tribunal was within the permissible range of options open to them. However the abandonment of that argument does not mean that other choices including G District (Somerset West) were not also permissible options on the basis that in them the requirement or condition was applied strictly, or with particular vigour.
44. In this context we comment that we accept that, as leading counsel for the Chief Constable submitted, the Employment Tribunal found that the requirement or condition applied to the whole force. However as we read paragraphs 13, 15 and 22 of the Extended Reasons they include findings that (i) Superintendent Nelson applied it strictly in Somerset West, (ii) it was not applied uniformly across the force, and (iii) it was not applied strictly in Somerset East and one other unidentified district (but which appears from paragraphs 3.2.3 and 3.2.4 to the answer to the cross appeal to be North Bristol).
45. Further we agree with paragraph 3.2.4 of that answer to the cross appeal that the Extended Reasons do not contain findings of (or, we add, as to) non-compliance in the other six regions. At one stage during submissions it was asserted on behalf of the Chief Constable that there was strict compliance in 6 (including Somerset West) out of the 8 regions. In our judgment there is no such finding by the Employment Tribunal. We are unclear whether it was a point raised below. We understand that it is disputed. We are not in a position to decide this point and if we concluded that it was necessary for it to be decided we would have to remit.
46. In our judgment (see paragraph 37 above) permissible options as the pool would have been (i) Somerset West on the basis of the application of the requirement in that District by Superintendent Nelson, or (ii) that District together with others with a similarly strict, or vigorous, application of the requirement. As mentioned above the Employment Tribunal do not make findings of fact in respect of the issue whether there are other such Districts.
47. However in paragraph 20 of the Extended Reasons the Employment Tribunal make findings in respect of Somerset West in the following terms:

20 Even if we are wrong about the pool, similar considerations apply to all other pools. If we look only at Somerset West, we have evidence of one officer out to 31 women who cannot comply against 243 men all of whom can comply. If we include those who, to our knowledge, comply with difficulty it is 3 women out of 31 against 1 man out of 243. Similar figures appear if we restrict the pool to 24 hour officers in Somerset West, the figures then being 243 men who can all comply against 1 or 3 women out of 19 who cannot comply.

In percentage terms those three analyses give a differences of between approximately 3.2% and 15.8%. Also two of them contain the feature that existed in Edwards (No 2) that all the men could comply. We add that these figures also provide some support for the inference drawn in the third sentence of paragraph 16 of the Extended Reasons (see paragraph 38 above).

48. As we have mentioned earlier the argument advanced on behalf of the Chief Constable on the finding of the Employment Tribunal is put in the alternative namely that they erred in law in applying the wrong approach, test or criteria, or alternatively that the decision was perverse. The arguments are however both essentially that the difference of 2.26% which the Employment Tribunal found in respect of the pool they identified (i) cannot be a considerably smaller percentage, and (ii) leads inevitably to the conclusion that the first stage (disparate effect) is not satisfied
49. The first argument is that in applying s. 1(1)(b)(i) SDA the Employment Tribunal (i) should have had regard only to that percentage and its basis not had regard to other matters that they took into account, or (ii) should have given that percentage and its basis far greater weight and the other factors they took into account less weight. The second (and effectively undisputed argument) is that applying the approach, test or criteria advanced on behalf of the Chief Constable the conclusion reached by the Employment Tribunal is outside the range of permissible options and is thus perverse.
50. In rounds terms the argument was the argument that was advanced and rejected in Edwards (No 2) but with the additional points that (i) the Seymour-Smith case has now been decided by the ECJ and the House of Lords, and (ii) the percentage difference of 2.26% found by the Employment Tribunal is lower than the percentage under consideration in Edwards (No 2).
51. As appears above we accept that there is considerable force in the argument that a percentage difference of 2.26% is inherently unlikely to found the establishment of the first stage (see in particular paragraph 36(10)). Indeed we accept that taken on its own such a percentage difference does not found the first stage (see in particular paragraphs 36(7), (8) and (10)). It follows that if there are other factors to be taken into account they need to be persuasive.
52. For the reasons we have given we are of the view that with the qualification and confirmation referred to in paragraph 36(10) Edwards (No 2) remains good law and that:

(a) a flexible approach that has regard to factors other than the identified percentage difference can, and in this case should, be taken in deciding whether the first stage (disparate effect) is established, and

(b) it does not inevitably, or necessarily, follow from the percentage difference between the proportions of men and women who can comply with the requirement identified by the Employment Tribunal in respect of the pool they have chosen that the first stage is not established.

Therefore and although as the Court of Appeal recognised in Edwards (No 2) (and the Employment Tribunal would have known from its obvious reliance on thereon) that case, and this one, are difficult cases in our judgment the Employment Tribunal were correct in law to have regard to factors other than the percentage difference identified in respect of the pool they chose. Part of the difficulty relates to the weight to be given to factors other than the identified percentage.

53. Although we agree with leading counsel for the Chief Constable that paragraph 18 of the Extended Reasons is an important paragraph in our judgment it is not the only, or the crucial, paragraph that identifies the factors taken into account by the Employment Tribunal. As we have already indicated in our judgment paragraphs 14 to 21 of the Extended Reasons have to be read together and with paragraphs 22 to 25, 27, 28 and 32 thereof to identify the factors taken into account by the Employment Tribunal. This has the consequence that in our judgment the Employment Tribunal correctly had regard to (i) the findings in paragraph 15 as to application of the requirement, and this paragraph should be read in particular with paragraphs 22 to 25, (ii) the three inferences in paragraph 16 (the last of which relating to the percentage identified being a bare minimum is repeated in paragraph 18), (iii) the acceptance in paragraph 17 of the point that the requirement could affect both men and women and thus to its inherently likely effects, (iv) the point in paragraph 19 that the difference is explicable in terms of sex discrimination and thus again its inherently likely effect which in this paragraph is linked to the points made on childcare which should in our judgment be read with paragraphs 28 and 32, and (v) the point that the percentage difference was significant and not fortuitous.
54. In our judgment point (iv) accords with the point we have made in paragraph 23 above and the points made in Edwards (No 2) in respect of childcare (see in particular paragraphs 25 and 27).
55. Having regard to paragraph 14 we are not entirely clear how the Employment Tribunal took into account the points made in paragraph 20 relating to Somerset West. As we read paragraph 20 they are taking them into account by way of confirmation of, or as support for, their view focusing on the pool they selected. As we have pointed out in our judgment Somerset West (or that District together with others applying the requirement in the same or a similar manner) were possible pools and the Employment Tribunal could have regard to the position in those Districts and compare it with the force as a whole. Further in our judgment the lack of any findings as to whether five of the other Districts applied the same or a similar approach to the application of the requirement as that taken in Somerset West:

(a) does not invalidate consideration of the position in Somerset West alone. It only means that another comparison (or other comparisons) could have been made which might have lead to some modification of the relevance, or impact, of Somerset West when taken alone, and

(b) does not mean that we should leave the findings in paragraph 20 out of account or remit so that findings in respect of the application of the requirement in other Districts can be made.

56. It follows that in our judgment by taking into account the factors they identified in the Extended Reasons (and are referred to in paragraphs 53 to 55 above) in addition to the percentage difference they identified the Employment Tribunal did not err in law in the test, criteria or approach they applied.
57. Further, in our judgment, the conclusion of the Employment Tribunal that the combination of those factors had the result that the first stage (disparate effect) was established was within the range of conclusions open to them. Indeed, and although it is unnecessary for us to do so, having particular regard to the points made in (i) paragraph 19 of the Extended Reasons and paragraphs 53(iv) and 54 above, and (ii) paragraph 20 of the Extended Reasons and thus paragraphs 47 and 55 above, we agree with their conclusion.

The cross appeal

58. In view of the above conclusion it is not necessary for us to further consider the cross appeal. However we would repeat that in our view counsel for Ms Chew was correct not to pursue his argument, based on Meer, that the force as a whole was not a permissible choice as the pool. Further, in our judgment, if the Employment Tribunal had chosen Somerset West (or that District together with others with the same, or a similar, approach to the application of the requirement) as the pool a comparison with the position demonstrated in the force as whole would have been permissible and appropriate (see again paragraph 37 above).

General comment

59. In view of the conclusion we have reached it is not necessary for us to consider the arguments described in paragraph 14(c) above. Further the conclusion we have reached and the reasoning in support of it (including our conclusion that the Employment Tribunal were entitled to conclude that the percentage difference they found was based on the minimum of people who cannot comply) mean that we do not have to remit (i) for the determination of the points of fact raised before us which were not the subject of findings by the Employment Tribunal and which we cannot resolve, or (ii) because of the matters referred to in paragraphs 10 and 11 above.

Overall conclusion

60. The Chief Constable's appeal is dismissed and no order is made on the cross appeal.