MR S ESSOP & OTHERS v HOME OFFICE (UK BORDER AGENCY) [2014] UKEAT 0480_13_1605

EMPLOYMENT APPEAL TRIBUNAL

BAILII case number: [2014] UKEAT 0480_13_1605

Case No Appeal No. UKEAT/0480/13

Between

MR S ESSOP & OTHERSAPPELLANT

HOME OFFICE (UK BORDER AGENCY)RESPONDENT

Before

Mr Justice Langstaff

Date 16 May 2014

[1]APPEARANCES For the Appellant MS KARON MONAGHAN QC MS NICOLA BRAGANZA (of Counsel) Instructed by: Thompsons Solicitors City Gate East Tollhouse Hill Nottingham NG1 5FS For the Respondent MR JOHN-PAUL WAITE (of Counsel) Instructed by: Treasury Solicitors Department One Kemble Street London WC2 B 4TS SUMMARY RACE DISCRIMINATION: INDIRECT In a test case, it was assumed that BME candidates disproportionately failed the CSA test, passing which was necessary to progress to higher grades in the Civil Service. The Respondent argued successfully that claims of indirect discrimination could not succeed unless the individual Claimants could prove the reason they failed the test. Unless they could do so they could not show that they were at "that disadvantage" as required by statute: showing that the group of them who were BME were disadvantaged was not enough. Held There was no evidence to suggest the individual Claimants were not at "that disadvantage" in a situation in which it was known that the group of which they were part suffered it, where the precise reason why an apparently neutral criterion had that effect was unknown. THE HONOURABLE MR JUSTICE LANGSTAFF (PRESIDENT) The appeal is against a decision made by Employment Judge Baron at a pre-hearing review at London South Employment Tribunal, in test cases in which it was asserted that black and minority ethnic (BME) candidates over the age of 35 were systematically less likely than non-BME and younger candidates to pass a Core Skills Assessment ("CSA") which it was necessary to pass in order to achieve promotion to the post of HEO or above in the Civil Service.

[2]For the purposes of the hearing it was assumed:-(a) There was a statistically significant difference between the success of BME/older candidates and younger non BME candidates sitting the CSA test.

(b) There was no particular personal factor specific to any individual Claimant that might explain this.

[3]The Judge held that a Claimant had to show that not only was the systematic disadvantage true of the group of which the Claimant (as an older BME candidate) was part, but also why the Claimant had failed the assessment. Thus in paragraph 40 the Judge held:

"It will be necessary for each of the Claimants to prove the reason for his/her failing of the CSA test."

The Statutory Background

[4]The issue is purely one of the proper interpretation and application of statute. The central statutory provisions are those contained in the Equality Act 2010 . Section 39(2) provides that:- "An employer(a) (A) must not discriminate against an employee of A's

(b) (B) – (a) as to B's terms of employment; (b) in the way A affords B access, or by not affording B access, to opportunities for promotion, transfer or training or for receiving any other benefit, facility or service…"

[5]The definition of that which is discrimination is contained in Part 4 of the Act. Section 19, headed "Indirect Discrimination" provides (so far as material):- "(1) A person(a) (A) discriminates against another

(b) (B) if A applies to B a provision, criterion or practice which is discriminatory in relation to a relevant protected characteristic of B's (2) For the purposes of sub-section (1), a provision, criterion or practice is discriminatory in relation to a relevant protected characteristic of B's if – (a) A applies, or would apply, it to persons with whom B does not share the characteristic (b) it puts, or would put, persons with whom B shares the characteristic at a particular disadvantage when compared with persons with whom B does not share it,

(d) A cannot show it to be a proportionate means of achieving a legitimate aim. (3) The relevant protected characteristics are – age;…. …….. race….." Although Section 19 looks at the disadvantage suffered by those who share a protected characteristic of one sort when compared to those who do not (Section 19(2)(b)) the impact of Section 23 must also be taken into account. In a section of the Act entitled "Discrimination: Supplementary", Section 23 provides relevantly that:

"Comparison by reference to circumstances (1) On a comparison of cases for the purposes of section… 19 there must be no material difference between the circumstances relating to each case."

Thus the comparison to be made is not simply a crude comparison between all those who share the same protected characteristic and all those who do not. The comparison must be between those who are in circumstances which are not materially different, one from the other, save that in the case of one of the two comparison groups they share the same protected characteristic whereas the members of the other do not. In relation to the anti-discrimination provision in Section 39, the members of the comparator groups must obviously both be employed by, or be potential candidates for employment by, the same employer. There may be other material differences. Ultimately, the identification of those differences which are material for the purposes of the comparison will be for a Tribunal itself to determine as a matter of fact and assessment, though always subject to guarding against reducing the numbers in the comparison groups by over definition so as to lose sight of the essential comparison to be made (see University of Manchester v Jones

[1993] ICR 474

[6]If a complaint is of a failure to observe the requirement of section 39 not to discriminate in the way the employer affords an employee access to opportunities for promotion, transfer or training, an Employment Tribunal may by virtue of Section 120 determine the complaint. If it finds that there has been discrimination by virtue of Section 19 (i.e. indirect discrimination) then Sections 124(4) and (5) come into play. They provide:- "(4) Sub-Section (5) applies if the Tribunal –(a) finds that a contravention is established by virtue of Section 19, but

(b) is satisfied that the provision, criterion or practice was not applied with the intention of discriminating against the complainant (5) It must not make an order under Sub-Section (2)(b) unless it first considers whether to act under Sub-Section (2)(a) or (c)." Sub-section (2), there referred to, reads as follows:- "(2) The Tribunal may - (a) make a declaration as to the rights of the complainant and the respondent in relation to the matters to which the proceedings relate; (b) order the respondent to pay compensation to the complainant;

(c) make an appropriate recommendation." If it awards compensation, it must award an amount which might be awarded by a County Court as if the proceedings had been in tort (Sections 124(6) and Section 119).

[7]The effect of sub-sections 124(4) and (5) is thus to open up the possibility of a Tribunal merely making a declaration of rights, or a recommendation in an appropriate case, but not making an award of compensation. Since sub-sections 124(2)(a) - (c) are not mutually exclusive, it may be more likely that compensation and a declaration are made than merely the latter alone: but it is certainly contemplated by the section that there may be cases in which although there is no loss to be compensated it is nonetheless appropriate for a declaration to be made. Just as the provisions are not mutually exclusive, nor do all three have to be honoured – the operative word in the opening line of sub-section 2 is "may".

[8]The focus of the appeal was on the precise force of the words "at that disadvantage" in the expression in section 19(2)(c) "it puts… B at that disadvantage…".

[9]Both parties accept that a requirement of the Civil Service for applicants for a post of HEO and above to pass a CSA test is a PCP within the meaning of section 19(2). There is no obvious reason to conclude that it is racially biased, or that it favours the younger over the older. If that were so, the provision would almost certainly have constituted direct discrimination. Cases of indirect discrimination arise where an apparently neutral criterion actually has a disparate effect as between persons who have different protected characteristics. Examples are notorious in the folk lore of civil rights: such as the requirement there once was in some Southern States of the Union for voters to have to demonstrate an ability to read or write - this apparently racially neutral criterion excluded most of those who, being black and former slaves, had had no education. The overt discriminatory intent behind such a provision may easily be recognised, despite the cloak of disguise in its ostensible neutrality – but it is not always so obvious. The bias created by or inherent in a particular measure may be much more subtle.

[10]It is well-recognised in discrimination cases that discrimination is rarely overt. It need not be intentional. Yet a subconscious covert bias may affect an outcome. It will not be easy to detect, not least because those holding the bias may with all honesty argue that what they have thought, done and said is not at all to discriminate but rather to hold all people equal. Thus a practice which is adopted may be intended to be fair, but in its effects may produce a discriminatory result. The reasons why despite the good intentions it does so may be far from clear – the practice (or provision or criterion) may simply lack transparency.

[11]This is part of a more general problem of evidence and analysis. It is a fact of life that though it is acknowledged that certain actions are more likely than others to produce a given result, it may not always be possible to say precisely how they do so. It may be sufficient for some purposes merely to identify the effect (an example canvassed in argument was that of smoking, which it is acknowledged causes lung cancer, though it remains scientifically unknown precisely how it does so, and which part of the tobacco smoke has that effect: yet the known effect is sufficient to justify the legislature in restricting the marketing of the product.) For other purposes - such as those of ameliorating the adverse effects caused by the action in question - it may be necessary to know more about the precise causal mechanism (on the crude analogy of smoking, if for instance the purpose in question is the production of a "safe" or at least "safer" cigarette it might be necessary to know more precisely what the harmful part of cigarette smoke is). In short, the degree of particularity with which a cause of a given effect needs to be identified depends upon the nature of the question being addressed.

[12]EJ Baron found that the "particular disadvantage" within Section 19 which had been suffered in the present case was that there was an increased likelihood of an older BME candidate failing the CSA test. He accepted the Respondent's case that it was not simply sufficient for the purposes of section 19(2)(b) and/or (c) to identify that this was the effect: he held the statute required the Claimants to prove on the balance of probabilities what was the reason for the lower pass rate. The Claimants disputed that they needed to show this. The Tribunal Decision

[13]The central reasoning of the Employment Judge was that the clear wording of the statutory provisions should be followed: "First it is for the parties to agree, or for the Tribunal to ascertain, what was the particular disadvantage caused by the provision, criterion or practice to the group who do not share the characteristic in question." (Though the Judge observed, rightly in my view, that that may well involve a determination of whom the relevant group consists, he made no such particular determination in the present case). He found there to be a particular disadvantage. Then, at paragraph 39, he accepted the arguments of Mr Waite who appeared for the Respondent. He summarised these as being to the effect that section 19(2)(c) required each Claimant to show that as a fact s/he was less likely to be able to pass the test – it was not sufficient simply to show the fact of failure. If the Claimants' contentions were correct, it would mean that an individual could make a claim based upon statistics wholly irrespective of whether the cause or the adverse differential impact suffered by the group of which the individual was part actually affected that individual. He suggested that claims of indirect discrimination should start by identifying the reason for the adverse impact on the individual Claimant, and then move on to see if that impact was shared by the relevant group. As an example, he supposed a job requiring that a successful candidate had a high level of spoken English. If such a requirement would put BME candidates generally at a particular disadvantage within Section 19(2)(b), but the case being considered was one of a particular candidate with excellent spoken English who failed to secure appointment, the correct way of approaching it would not be to rely upon his membership of the group, but to ascertain the reason for the failure. If it were otherwise, the employer would be faced with having to justify a PCP which had not in fact had any adverse impact on the Claimant in question. Individuals would be able to benefit from "a statistical fluke".

[14]In accepting these arguments,

the Judge observed:

"Once the particular disadvantage caused by the PCP has been found then it would be necessary for the tribunal to determine that the relevant Claimant was actually put at that disadvantage. The mere fact of failure of the CSA test in any particular case is not determinative of whether the Claimant has been put at that disadvantage. If the Tribunal finds that the Claimant was him/herself put at the group disadvantage applicable to Section 19(2)(b), then the definition of discrimination is satisfied, subject to the defence of justification. 39. It is then necessary for the tribunal to go further and consider whether the indirect discrimination was unlawful under Section 39(2)(b). Did the discrimination as found under Section 19 result in the relevant Claimant being denied access to promotion? It is quite possible that absent any discrimination the relevant Claimant would not have passed the CSA test. It is my conclusion that in such circumstances the discrimination would not be unlawful under Section 39. 40. Therefore on the issue before me as agreed between Counsel my conclusion is that it would be necessary for each of the Claimants to prove the reason for his/her failing of the CSA test."

Submissions

[15]Ms Monaghan QC (who did not appear below) and Ms Braganza (who did) argued that the effect of the Tribunal's decision is that it would be impossible for any of the Claimants, who neither assessed nor scored their test papers, to prove why they failed the test. Only the Respondent would know: and as pointed out above, even they might not appreciate why or how the CSA had the sustained and significant adverse effects on BMA candidates which it did. The fact that it did was all that was necessary to answer the question whether there should be liability. To require a Claimant to prove the reason for failure in the test was more akin to the requirements of a direct discrimination claim, which (see Section 13 of the Equality Act 2010 ) refers to less favourable treatment having been suffered "because of" a protected characteristic. Moreover, indirect discrimination does not concern inequality of treatment, to which the Judge's remarks appeared to be directed, but of outcomes. As was said in R (on the application of E) v Governing Body of JFS and others [2009] UKSC15, per Baroness Hale at 56-57:- "The basic difference between direct and indirect discrimination is plain: see Mummery LJ in R (Elias) v Secretary of State for Defence [2006] EWCA1293, paragraph 119. The rule against direct discrimination aims to achieve formal equality of treatment : there must be no less favourable treatment between otherwise similarly situated people on grounds of colour, race, nationality, or ethnic or national origins. Indirect discrimination looks beyond formal equality towards a more substantive equality of results: criteria which appear neutral on their face may have a disproportionally adverse impact upon people of a particular colour, race, nationality or ethnic or national origin." Further, the Tribunal's judgment was contrary to the statutory code of practice, duly approved by Parliament, (the Equality and Human Rights Commission Statutory Code of Practice 2011), to which Tribunals must have regard in accordance with Section 15 of the Equality Act 2006. The Code relevantly states:

"4.11 In some situations, the link between the protected characteristic and the disadvantage might be obvious; for example, dress codes create a disadvantage for some workers with particular religious beliefs. In other situations it would be less obvious how people sharing a protected characteristic are put (or would be put) at a disadvantage, in which case statistics or personal testimony may help to demonstrate that a disadvantage exists. … Example: a consultancy firm reviews the use of psychometric tests in their recruitment procedures and discovers that men tend to score lower than women. If a man complains that the test is indirectly discriminatory, he would not need to explain the reason for the lower scores or how the lower scores are connected to his sex to show that men had been put at a disadvantage; it is sufficient for him to rely on the statistical information."

[16]This passage shows that the Code contemplated there was no need further to require a reason to be established.

[17]The Tribunal had erred in its approach to establishing indirect discrimination: all that was needed was for the Claimants to show group and particular disadvantage, which they did by reference to the pass rates generally and by the individual Claimants and those sharing their protected characteristics also failing the CSA and not being promoted.

[18]Mr Waite argued, on behalf of the Respondent argued, as he had below, that the reason for the words "that disadvantage" was to ensure that Claimants do not succeed without themselves being victims of the discriminatory conduct which impacts on some members of the group. For instance, a requirement that all employees work full time may disadvantage those women with child-care responsibilities, but is not discriminatory against a woman with no such responsibilities who wishes to work part-time in order to pursue a hobby.

[19]Contrary to the way his arguments had been represented by the Claimants, he had never suggested that Claimants needed to show that the conduct of which they complained resulted from a protected characteristic they happened to share – but was merely that a cause of group disadvantage must be identified which also operated to the disadvantage of the individual Claimant concerned. The implications of the Claimants' arguments were too wide – if successful they would mean that any person who failed a test and was part of a protected group which performed proportionately less well in that test would be able to demonstrate that they were the subject of indirect discrimination, thereby requiring the assessor to prove that the conduct of which complaint was made was objectively justified.

[20]In Eweida v British Airways

[2009] IRLR 78

(when before the Appeal Tribunal) Elias P. giving judgment said (at paragraph 45):

" 45. There is some merit in the argument that the change in wording permits a court to find a particular disadvantage even with respect to those who can and do comply with the provision. An example might be a woman who wishes for child care reasons to work part time but feels compelled to work full time, which is a job requirement, because her employer will not consider the possibility of part time work and she cannot afford to lose her job. It may well be that the current definition would permit a claim of that nature. Equally, when determining whether there is a group disadvantage, such a person could be considered to be part of the disadvantaged group notwithstanding a reluctant willingness to comply with the requirement, although we suspect that examples of people prepared to compromise strongly held religious beliefs in that way would be rare. (The issue whether the employer knew of the objection could arguably become relevant in those circumstances.) 46. However, in order to fall within the terms of the legislation, it is still necessary that the particular disadvantage relied upon should stem from the religious beliefs held by the claimant. It is not enough that persons of the same religion and belief are fortuitously affected by the provision. It must be something connected with the religion or belief that causes the adverse effect. That is so however the pool is defined."

[21]Mr. Waite contends that persons of the same ethnicity would be "fortuitously affected" if the interpretation for which Ms. Monaghan argues were correct. Claimants would be permitted to "ride on the back" of an unidentified disadvantage which affected only some of the group of which they were part, where they were not themselves at that disadvantage at all.

[22]Although there was no authority which bore directly on the point in issue, in Homer v Chief Constable

[2012] UKSC 15

it was the identification of the cause of the group disadvantage which permitted the Claimant in that case to succeed – he could not achieve the promotion dependent upon having a law degree because, being 62, he did not have sufficient time before the normal retirement age of 65 to gain one.

[23]As for the Claimants' reliance on the Code, he submitted that parts were capable of supporting his own argument. Thus, he submits materially, the Code also provides:

"It is not enough that the provision, criterion or practice puts (or would put) at a particular disadvantage a group of people who share a protected characteristic. It must also have that effect (or be capable of having it) on the individual worker concerned. So it is not enough for a worker merely to establish that they are a member of the relevant group. They must also show they have personally suffered (or could suffer) the particular disadvantage as an individual. Example : An airline operates a dress code which forbids workers in customer-facing roles from displaying any item of jewellery. A Sikh cabin steward complains that this policy indirectly discriminates against Sikhs by preventing them from wearing the Kara bracelet, However, because he no longer observes the Sikh articles of faith, the steward is not put at a particular disadvantage by this policy and could not bring a claim for indirect discrimination."

Discussion and Conclusions

[24]The starting point for decision in a jurisdiction which is statutory must always be the words of that statute. The first observation to be made is that the wording of section 19 of the Equality Act does not in terms require members of a disadvantaged group to show why they have suffered the disadvantage, in addition to the fact that they have done so. This is in itself a sufficient answer to the Respondents' case, without the need for further consideration. The Judge was asking that a further matter be established by a Claimant, for which the law does not specifically provide.

[25]On this approach, the Judge's assertion in paragraph 39 that "the mere fact of failure of the CSA test in any particular case is not determinative of whether the Claimant has been put at that disadvantage" has no logical basis. That is because, in this case, the disadvantage to which the group was disproportionately subject was that of the greater risk of, or actual, failure of, the test. The "mere fact of failure of the test" summarised the potential disadvantage. The Claimants suffered exactly that disadvantage.

[26]The second observation is that where, as here, domestic provisions implement requirements of EU law, that law too is relevant and may be decisive: for any domestic legislation must be interpreted so far as possible to conform with EU law. Though the prohibition of discrimination on grounds of race has, by comparison with that on the ground of sex, only recently been adopted by the EU, there is no material distinction in the approach to be taken. Discrimination on ground of race is a social evil, just as it is on the ground of sex, and (as Advocate General Lenz put it in his opinion in the case of Enderby v Frenchay Health Authority

[1994] ICR 112

, ECJ , when considering the different ways in which indirect discrimination could be categorised) that:

"15. The purpose of a conceptual scheme is to comprehend methods by which women are placed at a disadvantage in their working lives and not to create additional obstacles to claims being made before the courts in respect of sex-related pay discrimination. For this reason, a formalistic approach should not be adopted when categorising actual instances where women are placed at a disadvantage at work. In accordance with the result-orientated line taken by the Court of Justice in the past, a pragmatic approach ought to be pursued….. 27. The concept of indirect discrimination is a legal concept which enables cases of unequal treatment, for which there is an objective justification but which in fact result in the woman being disadvantaged, to be included as an instance of unlawful sex discrimination"

The facts of that case are well-known: the decision was that although separate collective bargaining processes, each free of any obvious taint of sex discrimination, had resulted in jobs done predominantly by one sex being less well paid than jobs of equal value, also done for the same employer but predominantly by those of the opposite sex, the employer was required to go further than simply demonstrating that the cause of each was a separate bargaining process and show that the difference was based on objectively justified factors wholly unrelated to any discrimination on grounds of sex. Its importance for present purposes is two-fold: first, as Ms Monaghan submits, an employer's practice may not be transparent, such that discrimination may be disguised, often to the employer itself – and relevant and significant statistics may uncloak it; second, that where such statistics demonstrate disparate outcomes as between two groups, one with a protected characteristic and one without, that is sufficient proof without more to call for the employer to show that what has caused the difference is objectively justified, and is not itself tainted by the relevant discrimination.

[27]The approach to be derived from EU law is purposive. If, therefore, a domestic statutory provision may be read either as requiring not just that Claimants prove that they have suffered a disadvantage similar to that which the group of which they are a member has suffered as a whole, but also an additional factor (for which the legislation does not specifically provide); or alternatively as not requiring this additional factor to be established, the initial question will be whether to construe the legislation in this way advances the broad purpose of the legislation or impedes it.

[28]The purpose of the provision – eliminating the adverse effects of "disguised" discrimination – is not advanced, but hindered, by requiring the additional proof to which the Employment Judge referred at paragraph 40. If it is clear from reliable and significant statistical or other evidence that a process adopted by an employer has results which disadvantage a particular racial or cultural group in comparison to others, but neither the employer nor its employees can point to a particular feature of the process which has that result, or explain why it does, to require either to show the reason for the disadvantage in any individual case is to ask them to do that which they cannot do. To make liability conditional upon their being able to do so is thus to remove any legal constraint upon it, and to permit the disproportionate effect to continue. If it is said that a person must first be subject to the process, to see what the result is, then this sets him up disproportionately to fail. Nor where the process is the administration of a test does it help an employer to argue that the results of that test in the case of a particular individual are so poor that whatever the unknown feature of that test which is responsible for its disproportionate effects he would not have succeeded – for this could not properly be assessed without isolating the particular feature itself which caused a disadvantage only apparent on a survey comparing one with others.

[29]The answer to Mr. Waite's complaints that the unmeritorious could succeed by tail-coating the meritorious is three-fold. First, showing that a Claimant has suffered a particular disadvantage as required by Section 19(2)(b) is essential to but not sufficient for his success in the claim. It remains open to the employer to show that the PCP producing this effect is objectively justified (that is, it is a proportionate means of achieving a legitimate aim). Justification affects application of the PCP to which it relates to the group as a whole, so it may not fully answer the theoretical case of the individual "tail-coating" Claimant: but the less is the disadvantage suffered generally by the group, the more likely it is that the application of the PCP will be justified. Since what is proportionate is a proper balance between the importance of the object to be achieved by applying the PCP, the means of achieving it and its discriminatory effect, taken as a whole, an argument that the PCP can be objectively justified is more likely to succeed as the degree of risk of disadvantage reduces. If the hypothetical "tail coater" is disadvantaged for reasons likely to be true of members of the group only exceptionally, then holding the PCP not to have been justified causes little by way of practical problem, whereas conversely if the reasons for the "tail coater" suffering the disadvantage are common to many in the group of which he is part, the PCP is more likely to be upheld as validly justified, since it will cause proportionately less disadvantage in pursuit of the legitimate aim it targets.

[30]The practice, widespread amongst employers, of ethnic monitoring is an example of these principles in action. Such a practice is designed to provide information from which an employer can evaluate whether an apparently neutral practice has results, biased in favour of particular groups, when it would otherwise be unaware of any such potential bias. It allows a chance of identifying the features responsible for it, or to alter the practices which have given rise to the skewed results, so that an equality of result as between different racial groupings can best be achieved. This practice is adopted, despite the fact the employer has done its best to ensure a fair and equal process and if challenged would argue that it was entirely neutral, because it is recognised that any process, however carefully designed and well intended, may inadvertently produce unequal results. The logic of monitoring with a view to critical evaluation of process, which has been important in the improvement of equal opportunities, is consistent with the Claimants' case, and inconsistent with the Employment Judge's decision.

[31]The second answer may be at the level of remedy. If there is material to show that the disadvantage was not suffered for a reason true of the group in general, but for one individual to the Claimant, then the natural inference (that the Claimant has suffered the disadvantage for the same reason as that true of the group when compared to others) may be displaced (that is not the present case, which is one in which there is nothing to suggest this). If there is powerful material to show that a given person would succeed on the test, irrespective of its unidentified bias, then to that extent less compensation may be awarded; just as in the case of the "certain failure" hypothesised by Mr. Waite's argument, the chance that he would not succeed on any appropriate test in achieving the higher salaries to which success is a gateway would significantly moderate and might even eliminate any award of future loss. It may well be that in an appropriate case the Tribunal would think that only a declaration should be made.

[32]Mr. Waite's argument that the Claimants' case should be rejected because of the adverse consequences of upholding it made by reference to the illustration of the woman who wished to pursue a hobby, rather than child-care, being permitted to succeed on a claim for indirect discrimination is only superficially convincing. On examination it does not address the present case. In his illustration, the reason why the PCP caused relevant disadvantage is known. Whatever the proper analysis of such a case may be, it is not comparable with a case such as the present where, (I assume) despite the best will in the world, the parties remain unclear why precisely the disadvantage is suffered. In such a case there is no basis yet established for distinguishing between the disadvantage to one member of the group as opposed to another. The example would be pertinent if the employer were able to identify the disadvantageous feature of its arrangements with sufficient particularity to show that A, a member of the group which is potentially affected by the arrangements as a whole, was not in fact disadvantaged. In circumstances such as the present he cannot do so.

[33]At paragraph 27 of his reasons, the Judge recorded Mr. Waite as submitting that "claims of indirect discrimination must start by identifying the reason for the adverse impact on the individual Claimant, and then move on to see if that impact is shared by the relevant group." There was an echo of this in his argument on appeal. He pointed out that most claims began in this way, whereas the claims in the present case began with statistical evidence of differential impact. This approach, he argued, was at odds with custom and practice in the field of indirect discrimination claims.

[34]If and insofar as his argument was (or implied) that it was necessary that a claim should proceed in this way, I reject it – and, it follows, hold that the Judge was in error in this respect also – for a number of reasons. First, it introduces a formalism of approach which is not required by the statute – though such an approach may well be adopted in a given case, the issue is not whether it can but whether it has to be, and the legislation simply does not require it. Secondly, the legislation is if anything suggestive of the opposite approach – to identify group disadvantage first and only then proceed to ask if an individual has also suffered it (" that " disadvantage). If it were the other way round, the statute might more appropriately ask if the group had suffered, e.g. "that disadvantage (suffered by the individual)". Thirdly, it requires a reason to be shown for the disadvantage, and (see the discussion above) this may not be possible in many circumstances, even though it can be shown on the evidence that there has been a disadvantage. Fourthly, seen from the perspective of the Employment Appeal Tribunal I simply do not accept that general experience in employment cases is as the argument suggests: rather, as the comments of the Advocate General at paragraph 15 of his Opinion in Enderby show, there are a number of possible routes to a desirable end, which is the identification of discrimination with a view to its elimination. Evidence of disparate impact there must be: but to seek to prescribe the form that evidence should take by reference to what is for the time being said to be prevailing practice would be to permit habit, or preference, to obscure substance and is likely to lead to sterile debate as to what actually is currently prevalent.

[35]Accordingly, I hold that the Employment Judge was in error. The appeal must be allowed. The claims should proceed before the Tribunal in accordance with this judgment. The parties are at odds as to whether I should direct a remission to the same or to a different Tribunal. There is no particular advantage here of the case being remitted to the same Tribunal: indeed, that could cause delay or inconvenience as it is re-assembled. The matters which arose related to the issue with which I have just dealt, so it is unnecessary to revisit it evidentially. Pragmatically, there is nothing much to be gained by remission to the same panel. As to whether I should order a fresh Tribunal, I have had regard to the factor given such emphasis by Mr Justice Burton in Sinclair, Roche & Temperley v Heard

[2004] IRLR 763

: the professionalism of a Tribunal. I see no reason why the Judge and members would not loyally follow the judgment of the Appeal Tribunal. The judgment, though erroneous, is not made in a way which betrays any lack of care or suggests any predisposition to a result in favour of the Home Office on the matters which are yet to be litigated. Taking these considerations into account, in all the circumstances of the case, any Tribunal could properly hear the matter. Accordingly, I shall order merely that the matter be remitted to be heard as the Tribunal considers most convenient especially with regard to expedition. Though I suspect this will be before a freshly constituted Tribunal, whether in whole or part, I do not order that it must be.