Judgments - Macdonald (AP) (Appellant) v. Advocate General for Scotland (Respondent) (Scotland), Pearce (Appellant) v. Governing Body of Mayfield School (Respondents)

(back to preceding text)

    85. The employment tribunal said that the incident of 1 March 1995, when several boys called out "pussy", there were comments about the smell of fish and cat food and Ms Pearce later found an opened can of cat food and half its contents in her coat pocket, was quite different from those which had concentrated on her sexuality as it was directed to a woman and a woman's anatomy and not to woman who was lesbian. But this incident has to be seen in its context. Ms Pearce had just returned to work after a period of sick leave after becoming ill with stress. This followed various incidents in September and October 1994 which were all homophobic in nature. There then followed a further series of incidents from March to May 1995, all of which were of the same character. The incident of 1 March 1995 was, as Miss Booth QC submitted, gender-specific but none the less homophobic in character. The Employment Appeal Tribunal were persuaded that there were no grounds for treating this incident differently from the others [2000] ICR 920, 932, para 12, and I think that they were right.

Sexual harassment

    86. Section 4(1) of the Sex Discrimination Act 1975 identifies victimisation as a distinct category of discrimination. The Act does not accord the same treatment to sexual harassment. If this occurs in the field of employment it is a "detriment" within the meaning of section 6(2), as Lord President Emslie said in Strathclyde Regional Council v Porcelli 1986 SC 137, 144-145; [1986] ICR 564, 568. It is not made unlawful by the Act unless the employer discriminates against the person whom he subjects to it by treating her less favourably than he would treat a man or, if the victim is a man, than he would treat a woman. But it has been suggested that sexual harassment is so plainly discriminatory that there is no need to look for a comparator. This is because the essence of this form of humiliating or degrading treatment is that it selects as its target the victim's gender. The acts or words are gender-specific. So it speaks for itself, it is said, and the search for a comparator is unnecessary. This approach has the obvious advantage of avoiding the difficulty with which I have just been dealing when making the comparison which confronts homosexuals. That is the difficulty which is inherent in using the principle of equal treatment to combat harassment where the conduct in question is equally offensive to members of both sexes: Deakin and Morris, Labour Law 3rd ed, (2001) p 598. But is the approach compatible with the test for discrimination on the ground of sex which has been laid down by the Act?

    87. The point arises in Mr Macdonald's case with respect to the interview to which Wing Commander Leeds subjected him. His complaint is that he was interrogated in unnecessarily explicit detail about his past sexual activities and that its effect was to subject him to unwanted, unreasonable and offensive conduct of a sexual nature such as to constitute sexual harassment. The employment tribunal said that they had no doubt that if he had been faced with a homosexual female officer Wing Commander Leeds would have conducted the interview on very much the same lines, as to do so was what he perceived to be his responsibility as a vetting officer. As the complaint fell to be dismissed because it failed in the application of the comparator test they did not have to decide whether the conduct was of the quality that was needed to support a finding of sexual harassment, were such a finding competent. The Employment Appeal Tribunal disagreed. The principal ground on which they reversed the decision of the employment tribunal was their interpretation of the word "sex" in section 1(1)(a) of the 1975 Act, which they said should be interpreted to include sexual orientation. But Lord Johnston had this to say about the interview [2001] ICR 1, 13, para 32:

    "… given the tribunal's findings of fact as to the nature of the interview, the test laid down in Porcelli v Strathclyde Regional Council [1986] ICR 564 must apply which confirms that if the nature of the conduct is both sexually related and blatantly unacceptable there is no need for a comparator."

    88. The Employment Appeal Tribunal found that, on the facts found generally proved by the tribunal, Mr Macdonald was subjected to sexual harassment and that he was also entitled to compensation in that respect. Although Mr O'Neill no longer seeks to support the decision of the Employment Appeal Tribunal on the interpretation of the word "sex" in section 1(1)(a), he submitted that this finding was sound in law and that Mr Macdonald was entitled to an award of compensation on the ground of sexual harassment even if he was not entitled to compensation for his dismissal.

    89. In Ms Pearce's case this argument extends to the whole of the conduct to which she claims to have been subjected. It is said that the words that the pupils used when they were harassing her could not have been used of a man, that they were gender-specific and that it follows that this conduct amounted to sexual harassment falling within what was described by Hale LJ in the Court of Appeal as the Porcelli principle: [2002] ICR 198, 204, para 11. Mr Emmerson referred to the way the argument was put by Morison J in British Telecommunications Plc v Williams [1997] IRLR 668, 669, where he said:

    "To affect a person's dignity on the grounds of sex will, as with other forms of sexual harassment, cause a detriment to that person. Thus, proof of sexual harassment, of whatever form, will satisfy the criterion. Because the conduct which constitutes sexual harassment is itself gender-specific, there is no necessity to look for a male comparator. Indeed it would be no defence to a complaint of sexual harassment that a person of the other sex would have been similarly so treated: see Porcelli v Strathclyde Regional Council [1986] ICR 564."

It would, said Mr Emmerson, be quite illogical if what would plainly be sexual harassment in the case of a heterosexual woman were to be held not to be in the case of a woman who was a lesbian because her case had to be compared with a man who was a homosexual.

    90. The source of this argument seems to be the decision of the Court of Session in Strathclyde Regional Council v Porcelli 1986 SC 137; [1986] ICR 564. Mrs Porcelli was employed as a science laboratory technician at a school in Glasgow. Two technicians in the same department named Coles and Reid pursued a vindictive campaign against her for the deliberate purpose of making her apply for a transfer to another school. This conduct was both unpleasant and intimidating. It included various obscene acts which had a sexual innuendo, and the industrial tribunal held that it undoubtedly amounted to sexual harassment. There was no suggestion in that case that Mrs Porcelli was, or was suspected of being, a homosexual. The purpose of the conduct was simply to drive her out of the school, which in the event it did. It was designed to hurt her, and the method that was chosen was to humiliate her by using a particular kind of weapon against her because she was a woman. The tribunal held, with some regret, that had Mrs Porcelli been a man whom the other technicians disliked as much as they disliked her they would have treated him just as unfavourably as they treated her. So, applying the comparator test, they dismissed her application. The Employment Appeal Tribunal reversed this decision on the ground that the tribunal did not approach the case in the right way. Lord McDonald said at p 140 that what they should have done was to ask themselves (1) was there sexual harassment, and (2) if so, was it to the claimant's detriment? The reason which he gave for not mentioning the comparator test was that the aspects of the technicians' behaviour towards Mrs Porcelli which had sexual overtones could have no relevance in their conduct towards a man.

    91. The Court of Session did not overlook the need for a comparison. Lord President Emslie referred to section 5(3) as one of the provisions in the Act which were relevant to Mrs Porcelli's case. The way in which he dealt with it can be seen from how he summed the matter up in the final paragraph of his opinion at p 146:

    "The industrial tribunal reached their decision by finding that Coles' and Reid's treatment of an equally disliked male colleague would have been just as unpleasant. Where they went wrong, however, was in failing to notice that a material part of the campaign against Mrs Porcelli consisted of 'sexual harassment', a particularly degrading and unacceptable form of treatment which it must be taken to have been the intention of Parliament to restrain. From their 'Reasons' it is to be understood that they were satisfied that this form of treatment - sexual harassment in any form - would not have figured in a campaign by Coles and Reid directed against a man. In this situation the treatment of Mrs Porcelli fell to be seen as very different in a material respect from that which would have been inflicted on a male colleague, regardless of equality of overall unpleasantness, and that being so it appears to me that upon a proper application of section 1(1)(a) the industrial tribunal ought to have asked themselves whether in that respect Mrs Porcelli had been treated by Coles (on the ground of her sex) 'less favourably' than he would have treated a man with whom her position fell to be compared. Had they asked themselves that question it is impossible to believe that they would not have answered it in the affirmative."

    92. The key to a proper understanding of this passage is to be found in the fact that the Lord President has identified sexual harassment as a particular form of treatment which was to be distinguished from other forms of unpleasantness. Having done so the application of the comparator test became a formality, because the tribunal's findings indicated that this form of treatment would not have featured in a campaign against a man. But I do not think that he is to be taken as laying down any principle. What he was doing, as he explained at p 144, was examining the question whether the tribunal had correctly applied the provisions of section 1(1)(a) to the facts which it found proved. Lord Brand's observation at p 152 that, if a form of unfavourable treatment is meted out to a woman to which a man would not have been vulnerable, she has been discriminated against within the meaning of section 1(1)(a) reads as if he understood this to be a point of principle. A similar observation is to be found in the opinion of Lord Grieve at p 150, where he suggest that treatment meted out to a woman on the ground of her sex, although less cruel than that accorded to the man, would fall to be regarded as less favourable treatment simply because it was sexually orientated. But that is not how the case appears to have been decided by the Lord President.

    93. In my opinion Morison J read too much into Porcelli when he said in British Telecommunications Plc v Williams [1997] IRLR 668, 669 that it would be no defence to a complaint of sexual harassment that a person of the other sex would have been similarly so treated. It was precisely because the tribunal's reasons showed that they were satisfied that sexual harassment in any form would not have been used in a campaign which Coles and Reid directed against a man that Lord President Emslie felt able to say that if they had asked themselves whether Mrs Porcelli had been treated less favourably than a man they would have been bound to answer the question in the affirmative. In so far as Lord Grieve and Lord Brand may be taken to have been suggesting that this was not a relevant question where the harassment is sexually orientated, I would disapprove of their observations. I respectfully agree with the way Ward LJ dealt with Porcelli in Smith v Gardner Merchant Ltd [1999] ICR 137, 147-148. As he said, the conclusions in that case were conclusions of fact and Morison J was wrong to elevate them into a principle of law.

    94. There is no escape, then, from the need to resort to a comparison. The words "less favourable treatment" in section 1(1)(a) render this inevitable. It may be that the conduct complained of is so specific to the claimant's gender that there is no need to do more than to ask the question, to which the answer may well be, as Ward LJ put in Smith v Gardner Merchant Ltd [1999] ICR 137, 148, res ipsa loquitur; Deakin and Morris, Labour Law 3rd ed, (2001) p 598, note 6. But that conclusion may be more easily drawn in cases of sexual harassment which do not involve any homophobic element than in cases such as those of Mr Macdonald and Ms Pearce where the context for the abuse is the abuser's belief that the victim is a homosexual. This is because those who abuse homosexuals tend to pick on them not because of their gender but because they are homosexual. The form which the abuse takes may well be specific to the gender of the person who is being abused, but this is because the terminology which is used to describe homosexuals and the acts which they can perform with each other tend to vary according to the gender of those who are involved in this relationship. That is not to exclude the possibility that an abuser may treat a woman who is a homosexual less favourably than he would treat a male homosexual. That may indeed happen, and an employment tribunal must always be alert to this possibility. But whether this is so will be a question of fact in each case.

    95. The facts as found by the employment tribunal in the cases of Mr Macdonald and Ms Pearce leave no room for doubt on this issue. In Mr Macdonald's case the finding is that the interview would have been no less intrusive if it had been a female homosexual officer who was being interviewed. In Ms Pearce's case the finding is that the pupils would not have treated a hypothetical homosexual male teacher more favourably. On these findings they were not discriminated against on the ground of their sex.

The employer's liability for sexual harassment

    96. This was an issue in Ms Pearce's case only. This is because the sexual harassment was perpetrated not by staff at the school where she was working but by pupils. They were third parties for whose acts the school was not vicariously liable. It was contended that the school's liability depended on whether it could be said that it had "subjected" her to this treatment within the meaning of section 6(2)(b) of the 1975 Act. If, as I would hold, she was not discriminated against on the ground of her sex, this question is not one that need be answered. But I should like to say something about it, as it seems to me that a misunderstanding has arisen as to the test that ought to be applied. The issue is an anxious one, as the test in question - the "control" test - has been consistently applied in employment law since it was first formulated. It is particularly important for schools and other educational establishments, in view of the measures which are now being taken to extend protection to homosexuals against discrimination on the ground of their orientation and against sexual harassment. But it affects all cases where the harassment is perpetrated by third parties for whose acts the employer is not vicariously liable under section 41 of the 1975 Act or, in race relations cases, section 32 of the Race Relations Act 1976.

    97. The employment tribunal based their decision on this part of the case on the control test which was formulated by the Employment Appeal Tribunal in its reasons for its decision in Burton v De Vere Hotels Ltd [1997] ICR 1. The claimants in that case, who were both black, were held to have been subjected to racial harassment by their employers when the speaker at a dinner in the hotel where they were employed made racially offensive remarks about them as they were clearing tables during his performance. The employment tribunal held that their employers had not discriminated against them contrary to the Race Relations Act 1976. Its finding were that, although the hotel manager was aware of the offensive nature of the speaker's act, he simply did not address his mind to what might happen vis-à-vis the applicants that night and that had he addressed his mind to the risk before the event he would have taken steps to protect them. The implication of these findings was that, if he had addressed his mind to this risk, he would have treated white waitresses who were exposed to the risk of offensive conduct in the same way. In other words, he did not discriminate against the applicants on racial grounds. The tribunal had no doubt that they suffered a detriment that night within the meaning of the Act. But it was not their employers who subjected them to it: [1997] ICR 1, 5D-6D.

    98. The Employment Appeal Tribunal held that there was an error of law in this decision. This was on the view that by imposing upon the applicants the burden of showing that there was different treatment on racial grounds the employment tribunal had imposed on them a burden which was additional to that imposed by the Act, namely the burden of showing that the employers were themselves affected by racial bias or animus: [1997] ICR 1, 6H-7C. But the employment tribunal was doing no more than the Act itself requires: see section 1(1)(a) of the 1976 Act. There was no additional burden. The question for the employment tribunal was simply whether, on racial grounds, the applicants were treated by their employers less favourably than they treated, or would have treated, other employees. It appears that the Employment Appeal Tribunal's view that the tribunal imposed an additional burden was based on a concession, which was itself based on a misunderstanding of Strathclyde Regional Council v Porcelli 1986 SC 137; [1986] ICR 564, that where the treatment under consideration is "race specific" there is no need for the employee to show that the employer himself would have treated a white person differently: [1997] ICR 1, 6H. The effect of the concession was to divert the Employment Appeal Tribunal from the question whether the employers treated the applicants less favourably that they would have treated other persons in comparable circumstances. I agree with my noble and learned friend Lord Nicholls of Birkenhead that Burton was wrongly decided.

    99. Smith J said that the question in the Burton case was what was meant by "subjecting" in section 4(2)(c) of the Act, and that an employer "subjects" an employee to the detriment of racial harassment if he causes or permits the racial harassment to occur in circumstances in which he can "control" whether it happens or not [1997] ICR 1, 7C-E. She said that the tribunal should ask themselves whether the event in question was something which was sufficiently under the control of the employer that he could by the application of "good employment practice" have prevented the harassment or reduced the effect of it: pp 9H-10B. But this approach is not based on anything which is to be found in the statute. I agree that it should now be disapproved.

    100. The problems that the Burton "control" test and the concept of "good employment practice" which it involves may give rise to are well illustrated by what happened in Ms Pearce's case.

    101. The employment tribunal said that if they had to make a finding about the abuse which Ms Pearce suffered they would have held that, with the exception of the incident of 1 March 1995 when the school did act properly and promptly and supported Ms Pearce, it did not take appropriate steps and was therefore directly responsible for sex discrimination. But they did not make any findings to the effect that the school was in a position to control or lessen what the pupils were doing to Ms Pearce. Instead they limited their findings to criticism of the school for failing to make it clear that homophobia was unacceptable and for not supporting her with the result that she felt isolated and unsupported. They appear to have misunderstood what Smith J said in the passage which I have quoted. Even if the taking of these steps was desirable, a failure to take them would not make the school liable under the statute if it could not prevent or reduce the extent of what was happening. In the Court of Appeal [2002] ICR 198, 213, para 39 Hale LJ said that the employment tribunal correctly identified the "control" test in Burton and Rhule and that their conclusion that the school could and should have done more to support Ms Pearce "must mean" that had it done so the pupils' treatment of Ms Pearce would have improved. But there were no findings to this effect, as Burton J in the Employment Appeal Tribunal [2000] IRLR 548, 554 para 19 was right to point out.

    102. A finding by the employment tribunal that the steps which could and should have been taken by the school could have prevented or reduced the extent of Ms Pearce's sexual harassment would not have been a mere formality. Burton J in the Employment Appeal Tribunal, p 554, para 20 said that, if the school was to be found liable and other schools were to avoid being held so liable, there must be careful findings of fact, after full investigation, leading to the conclusion that the steps which the school could have taken and failed to take would have prevented or reduced the extent of the discrimination. I agree. And in the Court of Appeal Judge LJ at p 217-218, paras 60-62, with whom Henry LJ agreed at p 224, para 88, drew attention to the very real problems that would arise if one were to apply the test of "control" to educational establishments. As he said, it is difficult to equate the process of education with the sort of "control" that a hotel can exercise over its customers.

    103. This is not to say that Ms Pearce was not entitled to protection against the abuse which she suffered, which was plainly unacceptable. In this respect she was in the same position as any other member of staff. It was the responsibility of the school to face up to the problem of abuse by pupils irrespective of the form in which it might come and against whom it might be directed. Whether the steps which it could have taken to instil the necessary sense of respect and discipline would have been prevented or reduced the extent of the sexual harassment is another matter. I agree with the majority in the Court of Appeal that the employment tribunal failed to address this vital issue, and that their decision on this part of the case cannot be supported. But I also think that the practical difficulties which its approach reveals provides a further reason for departing from the Burton control test.


    104. For these reasons, and for those given by my noble and learned friend, Lord Rodger of Earlsferry with which in all essential respects I agree, I would dismiss these appeals.


    105. Like all of your Lordships, I too would dismiss the appeals in both these cases. My reasons are the same as those expressed in the Opinion of my noble and learned friend Lord Nicholls of Birkenhead. I also agree specifically that Strathclyde v Porcelli [1986] SC 137, British Telecommunications v Williams [1997] IRLR668, Burton v de Vere Hotels Ltd [1996] IRLR 596 and Go Kidz Go v Bourdouane (1996) should be disapproved. Further, I specifically agree with what my noble and learned friend has said on the Human Rights Act point in paragraphs 21 to 24 of his Opinion. In these circumstances I will confine myself to some short observations on points which I believe to be of particular relevance to these and similar cases and are necessary if the same errors of reasoning are to be avoided in future.

    106. The first point is that the statutory law has not stood still during the 27 years since the Sex Discrimination Act 1975 was passed in anticipation of the 1976 Equal Treatment Directive. Confidence in the workability of anti-discrimination laws has increased and Acts have been passed to extend the scope of the legislation and to cover new types of discriminatory conduct. There is now further draft legislation in preparation to give effect to the 2000 and 2002 Directives, where similar developments have occurred. It is a mistake to construe the 1975 Act as if it had been passed within the last few years. It had a limited scope and was so intended. Section 1(1)(a), defining direct discrimination, uses the simple criterion: "A person discriminates against a woman … if, on the ground of her sex, he treats her less favourably than he treats or would treat a man." Section 5 reinforces this by identifying a woman as a female and a man as a male and, in respect of the comparison required by s.1, by stipulating that the circumstances must be "the same, or not materially different". There has been a tendency by courts to treat the 1975 Act as having made greater strides than it did - that it included sexual orientation - that it independently prohibited sexual harassment. The tendency is understandable since the conduct in question may be every bit as objectionable as that covered by the Act, or even more objectionable. But the tendency is erroneous and has led to further errors.

    107. The second point is that proper regard must be had in construing an Act of Parliament to the words actually used. It is they which define the effect of the legislation. The now abandoned argument that the phrase "on the ground of her sex" can mean "on the ground of her sexual orientation" fails to respect the language of the statute. The two things are not the same. The use of different words give the phrases different meanings. To disregard this simple fact is to fall into the error rightly, and memorably, condemned by Lord Atkin as long ago as 1941: see [1942] AC 245.

    108. The third point is one which has affected the validity of the appellants' arguments both in your Lordships' House and in the courts below. An essential part of the criterion is that the discriminator's treatment of the complainant should have been "on the ground of her sex". The question must be answered: was it on that ground that he treated her less favourably? The distinction between sex and sexual orientation demonstrates the necessity of asking the question. Suppose that an employer advertises a vacancy saying - "the job is suitable for either a man or a woman but anyone who is a homosexual will not be considered". Or, suppose that a personnel manager simply receives a letter which does not disclose the sender's gender but does disclose that he or she is a homosexual, and replies refusing employment. In neither case can the discrimination have been on the ground of sex since the person, in the latter example, did not know the inquirer's sex and, in the former, expressly excluded the relevance of sex. In the case of Mr Macdonald, the policy being applied to his disadvantage was expressed in the terms: "Homosexuality, whether male or female, is considered incompatible with service in the Armed Forces." (AP 3392 Vol 5, leaflet 107, emphasis supplied.) For the sake of clarity, the same document also defined homosexuality as being "behaviour characterised by being sexually attracted to members of the same sex". This unequivocally states that the policy is on the grounds of sexual orientation, not on the ground of anyone's sex. Similarly, in the case of Shirley Pearce, it was an agreed fact that the school would have treated a male teacher in exactly the same way as it treated her, in other words, her particular sex was immaterial. This is fatal to their respective cases.

continue previous