Allen & Ors v GMB

Prior to 1997, the terms and conditions of employment applicable to local authority employees were set out in different documents which were referable to different categories of employees. Manual workers were governed by the White Book whilst administrative, professional, technical and clerical (APT&C) workers were governed by the Purple Book. A third category of craft workers came under the Red Book but this case does not concern that category. As regards the White Book and Purple Book employees, it was recognised that some gender-based pay inequalities had been allowed to develop. In 1997 a national collective agreement – the Green Book – was negotiated between the relevant trade unions and the local authority employers. The intention was to bring the White Book and the Purple Book employees under a new system with a common pay and grading structure. It was to be known as “single status”. Although the overarching structure of single status was the result of a national agreement, it was envisaged that actual pay scales and pay rates would be devolved to local level and that, in order to eradicate historical inequalities, local agreements would be preceded by local job evaluation studies. Each job would be assessed and placed on the appropriate Green Book scale. This proved to be a complex exercise. In Middlesborough, a job evaluation study was carried out and, in due course, new terms and conditions reflecting it came into effect on 1 April 2005. GMB (the Union) was one of the unions which negotiated the new terms and conditions with Middlesborough Borough Council (the Council).

The complexity of these circumstances is plain to see. From the Council’s point of view, the funding of any pay deal is heavily dependant upon the flow of money from central government and takes place against the backdrop of “capping”. On the other side, the Union has to represent members in different categories whose interests can and do conflict. Put very simply, the Council sought an outcome that was affordable. The Union wanted one that somehow compensated the victims of past inequality but at the same time provided a measure of pay protection for those who were disadvantaged by the job evaluation study and maximised the amount available for future pay across the board. In addition, the Union was constrained by the natural perception that, if it pushed too hard, the consequences might include job losses and contracting out, neither of which would be in the interests of its members.

It is beyond dispute that, faced with these conflicting pressures, the Union decided to give priority to those who needed pay protection and to achieving equality and better pay for the future rather than to maximising claims for past unequal pay. The deal done between the Union and the Council provided the White Book women with some compensation for the historical inequalities (in the region of 25% of the full value of successful equal pay claims) but did not provide the Purple Book women with any such compensation, the Council having apparently taken the view that their equal pay claims were without merit.

The present dispute concerns the claims of White Book and Purple Book women against the Union. They claimed that the Union’s prioritisation of pay protection and future pay over compensation for past inequalities was itself discriminatory – either directly or indirectly or by way of victimisation. Following a lengthy hearing, the Employment Tribunal (ET) rejected the claim of direct discrimination, finding the evidence of the Union officials to be “completely and utterly convincing” in establishing that the reason why the deal with the Council took the form that it did was not gender-based but was “an attempt to keep the peace and to appear to as many people as possible to have done a good job as a union in representing the interests of all members” (paragraph 7.32). It added (at paragraph 7.33):

Moreover, the Union’s negotiators did not perceive pay protection to be a “male issue”, not least because many female clerical staff were among its beneficiaries. The rejection by the ET of the direct discrimination claim has not been challenged on appeal.

On the other hand, the indirect discrimination and victimisation claims succeeded before the ET. The Union successfully appealed those aspects of the case to the Employment Appeal Tribunal (EAT) but the EAT (Elias J, Mr P Jacques CBE and Mr S Yeboah) granted the claimants permission to appeal to this Court on the issue of indirect discrimination. Before turning to the present appeal, it is necessary to set out the relevant statutory provisions and the different reasoning of the ET and the EAT on indirect discrimination.

Section 12(3) of the Sex Discrimination Act 1975 provides:

Indirect discrimination against a woman occurs if a person

This is the wording of section 1(2)(b) as amended by the Sex Discrimination (Indirect Discrimination and Burden of Proof) Regulations 2001. (A further amendment by the Employment Equality (Sex Discrimination) Regulations 2005 came too late to apply to this case and would probably not make a significant difference to the outcome in any event).

The conclusion of the ET was that, by agreeing to a low back pay settlement in order to release more money for pay protection and the future pay line, the Union had engaged in a potentially discriminatory practice. The disadvantaged group were predominantly women. The Union had failed to justify the discriminatory practice. This was because of certain aspects of the Union’s approach of which the ET was critical. There were four principal matters of concern. First, the Union had failed to protect the interests of the claimants by not pursuing proceedings at an early stage so as to establish an early date for the calculation of back pay. Even if the Union had preferred not to litigate, it should have protected the claimants in this way. Secondly, the Union had deliberately omitted to give advice about back pay and had refused to support litigation in order not to antagonise the Council or to delay or impede the progression to single status. Thirdly, the Union had “rushed headlong” into an ill-considered back pay deal. It had accepted too readily the Council’s plea of poverty. Finally, and, it seems, crucially, the Union had failed to give the claimants a fully informed choice about the options available to them. They had not been informed that what they were being offered was substantially less than they might receive following successful litigation and there was no assessment of the litigation risk which the ET considered to be relatively small, at least for some of the claimants. The ET considered that, if the Union was going to require the claimants to make some sacrifice in the interests of other members, then that should have been made plain to them. There had been not only a failure to provide full information but also positive manipulation of “relatively unsophisticated claimants” by suggesting that the offer from the Council was acceptable and placing them in a position where they were in fear that, if they pressed for more, it might lead to job losses and to their being seen as traitors by their colleagues. The ET considered this to be “the worst aspect of the case”.

In considering justification, the ET noted that it was common ground that the appropriate test was to ask whether the means adopted by the Union were proportionate to the attainment of a legitimate aim. The following passages are particularly relevant:

In paragraph 7.58 the ET recorded that what troubled it most was “the way in which the back pay offer (or lack of one in the case of the APT&C workers) had been ‘sold’ to the membership”. Various factors which ought to have been explained to the women were not so explained. It was not explained that they were being offered substantially less than they were likely to receive if they succeeded in litigation. The chances of success for the White Book claimants were considered by the ET to be “very good” and, for the Purple Book claimants, “fairly good”. However, the Union “did not even assess – let alone inform the members of – litigation risk”. It was not explained that part of the reason for their being asked to settle for a lower figure was so that pay protection could be maximised. “There is nothing wrong with asking members to make a sacrifice if they are informed of what the amount of their sacrifice will be and to what it will be applied.” Nor were they informed that another part of the reason was to provide a better eventual pay line for everybody for the future. The ET added:

Turning finally to victimisation, the ET concluded that the claimants had been less favourably treated by reason of having done the protected act of instructing their solicitor, Mr Cross, to take their cases to the ET. It is plain that the Union has a low opinion of Mr Cross and “no win, no fee” litigation. The ET said (at paragraph 7.77):

Upon the Union’s appeal to the EAT, there was some debate about the appropriateness or otherwise of the language used by the ET in paragraph 7.69 of its decision. However, in paragraph 78 of its decision, the EAT stated:

That defined the parameters of the main issue before the EAT which has now become the main issue in this Court. The EAT allowed the Union’s appeal on justification, concluding that the ET had fallen into a “fundamental error”. The approach of the EAT is apparent from the following passages in the judgment:

The EAT considered that, even if the Council had been pressed more firmly, and had found further funds to meet the Union’s demands, any extra monies would still have been distributed according to the balance that was in fact struck by the Union. Turning to the central issue of misleading information, the EAT stated:

In addition to allowing the Union’s appeal on indirect discrimination, the EAT also allowed the appeal against the finding of victimisation. However, nothing now turns on that.

Indirect discrimination arises when a person applies to a woman a provision, criterion or practice (PCP) which he applies or would apply equally to a man but which is such that it would be to the detriment of a considerably larger proportion of women than of men. The woman’s claim then succeeds if the person cannot show the PCP to be justifiable irrespective of the sex of the person to whom it is applied and she establishes that it is to her detriment. For the purposes of this case it is important to identify the PCP because it is the PCP itself which the person must show to be justifiable if he is to avoid a finding of discrimination.

What was the PCP in this case? The clearest statement by the employment tribunal is in paragraph 7.40 of its decision. As a manifestation of its policy to achieve single status with the minimum of losers, the Union

On this basis, I would characterise the PCP as the deal that was done with Middlesborough as a result of that policy. It fell to the Union to justify it.

It was common ground before the ET, and remains so, that the appropriate test was to ask whether the means adopted by the Union were proportionate to the attainment of a legitimate aim.

The crucial findings of the ET are to be found in these passages:

In other words, the Union succeeded in establishing that the PCP had a legitimate aim. However, the ET went on to conclude:

I have referred to the factors which informed that conclusion: denying the White Book members a fully informed choice (the advice given to Mrs Spayne was “the antithesis of that”); refusing to support litigation against Middlesborough; the way in which the deal was “sold” to the membership; “rushing headlong into accepting an ill-considered back pay deal, accepting the Council’s plea of poverty without question”; “the way in which relatively unsophisticated Union members were manipulated into either accepting the offer or doing nothing in the case of the APT & C workers by such alarmist information”; and finally (at paragraph 7.69):

I have emphasised that final sentence because, essentially, it is by reference to it that Mr Stafford QC seeks to uphold the conclusion of the ET that the Union had failed to justify the PCP. Before the EAT and in this Court, his stance has been that the Union had been entitled to adopt the priorities that it did, but that the means adopted to achieve those priorities were disproportionate.

The central question then becomes whether the misselling and manipulation, as found by the ET, can be characterised as the means to the attainment of the admittedly legitimate aim. In this context, it is appropriate to consider first whether the acts found to be reprehensible were in fact part of the PCP. In the course of submissions questions were raised as to whether the “back payers” were part of the ET’s finding of the PCP and, if so, whether the PCP embraced the misselling and the manipulation. It was suggested that the legitimate aim as formulated in paragraph 7.54 of the ET’s decision (set out in paragraph 22, above) was concerned exclusively with future pay issues – pay protection and the pay line – and not at all with back pay. With respect, I do not accept that suggestion. When reference is made to “the best possible pay protection”, that carries with it the implication “if necessary, at the expense of back pay”. Any other meaning would be at variance with the entire thrust of the decision.

The possibility that the PCP may have embraced the misselling and the manipulation was conceived by reference to a sentence in paragraph 7.45 which begins:

It is true that the adoption of the PCP – “all losers must be protected to the greatest extent possible” – must have assumed that steps would have to be taken with a view to persuading the back payers to accept sacrifices to that end. Whenever a union negotiates on behalf of members who may have conflicting interests inter se, the same problems will arise. More for one group often means less for another.

The analysis and taxonomy of aims and means is often difficult. Although the ET and the EAT were in agreement about the identification of the aim or aims, they parted company when identifying the means. The EAT took a narrower view, identifying the means as being, in effect, the balance that was struck in the deal with the Council. If that is right, then the aims and means of achieving them do coalesce in the sense described by Elias J. However, in my judgment that is to accord too narrow an ambit to the concept of “means”. If, as I have suggested, the adoption of the PCP assumed that steps would have to be taken with a view to persuading the back payers to accept sacrifices, including forbearance from litigation to enforce their perceived statutory rights, I would regard such steps as part of the means by which the achievement of the aim was to be pursued. Whilst it would not be helpful to prescribe sub-categories of means - for example, primary/secondary, principal/ancillary – it is pertinent to observe that the concept of “means” is not inherently narrow and that a number of different actions may each be part, and a necessary but not equal part, of the means of achieving the aim.

As I read the decision of the ET, it concluded that the PCP “involved persuading” the claimants and others in a similar position to make the sacrifices to which I have referred. In other words, such persuasion was part of the sine qua non and, as such, it and the form it took were part of the means. In my judgment, that was a permissible, indeed correct approach. Having taken it, I do not consider that the ET fell into any material legal error when, on the basis of the facts found, it concluded that the means were disproportionate to the achievement of the overall aim or aims. I should record that although Elias J considered the speech of Lord Nicholls in Barry v Midland Bank [1999] ICR 842 to be supportive of his analysis, I do not think that it is. In that case there was no question of a need to persuade the disadvantaged women to sacrifice perceived statutory entitlements and no question of manipulation or anything like it.

Ultimately, the approach of the EAT and of Mr Cavanagh in his submissions to this Court, is to marginalise the manipulation by treating it as a causation issue. Thus, the EAT found it

On the basis of the finding of the ET that many of the claimants would have withheld their support had they appreciated the extent of their sacrifices, the EAT considered that the Union would either have had to modify its stance by changing its priorities or continue to adopt the same basic strategy and risk alienating the claimants. It proceeded to the hypothetical analysis which I repeat here for ease of reference (paragraphs 88-89):

There are a number of problems with this analysis. First, it is undermined by the failure to characterise the chosen method of persuasion – the manipulation – as part of the means of achieving the legitimate objective. Secondly, although the objective was a legitimate one, it was not the only possible legitimate one. If it were achievable only by disproportionate means, then it would not be susceptible to justification. To conclude otherwise would be to licence disproportionality. Thirdly, to the extent that the EAT considered that the outcome would have been the same in any event – and that is something not based upon any finding of the ET – that, in my judgment, is potentially relevant to the remedy rather than liability, which was founded upon the incontrovertible detriment of disparate impact. I do not doubt that the unresolved question of remedy will raise difficult issues when it comes to be considered but it is important to ensure that potential remedy issues do not distort the consideration of proportionate means. Fourthly, Mr Cavanagh seeks to rely on the temporal aspect. The strongest evidence of manipulation was contained in a letter which was sent to the White Book claimants only after agreement had been reached with the Council and which was never sent to the Purple Book claimants at all. Thus, he submits, it came too late to count as part of “the means” in the former case and was irrelevant to the latter case. However, that is to ignore the breadth of the findings of the ET against the Union between paragraphs 7.56 and 7.69 of its decision, to which I have referred at length in paragraphs 11 to 12 of this judgment.

For all these reasons, I have come to the conclusion that, in this unusual case, the ET did not err in law in its approach to justification and that the EAT was wrong to allow the Union’s appeal on this central issue.

That disposes of the central issue on this appeal. However, Mr Cavanagh, by way of a Respondent’s Notice and generally, submits that the decision of the ET was vitiated by other legal errors. The following points merit consideration.

Mr Cavanagh submits that this is a case about direct discrimination, i.e. less favourable treatment, or nothing. The less favourable treatment was established but the direct discrimination case failed because the ET was satisfied that it was not on the ground of the claimants’ sex. This, submits Mr Cavanagh, led to the ET committing “the fundamental error” of equiperating less favourable treatment which was not on the ground of sex with indirect discrimination simply because it affected more women than men. It was wrong to squeeze the less favourable treatment into the catchment of a “practice” because the language of “provision, criterion or practice” covers apparently neutral gateways and barriers which have a disparate impact. It does not extend to everything done by a respondent which adversely affects women more than men.

I do not accept this submission. The fact that something amounts to less favourable treatment (but not unlawfully so) does not in itself exclude the possibility that it is also the application of a PCP. Mr Cavanagh seeks to rely on Rutherford v Secretary of State [2006] UKHL 19, [2006] ICR 785 but the ratio of that difficult case is concerned with the identification of the comparative groups rather than with what is or is not a PCP.

This leads to a related submission by Mr Cavanagh that what is attributed to the Union in the present case is not a PCP and the striking of a negotiated balance or deal is not the “application” of a PCP to the members – it simply affects them. These and cognate submissions were considered but rejected by the EAT in paragraphs 56 – 65 of the judgment. I entirely agree with that part of the judgment. Moreover, it would be inappropriate to adopt a more restrictive approach to the statutory language, given the purpose and policy of the legislation and the Directive which stands behind it.

Put simply, this submission is that the law of indirect discrimination was not intended or designed to cover a case such as this. Mr Cavanagh points to the plurality of interests which exist within the membership of a large trade union, gender imbalances between different parts of the membership, the difficulty of reconciling competing interests and the risk of facing this kind of litigation as a result of a complex negotiation. However, trade unions come within the law on indirect discrimination by reason of sections 12 and 1(2)(b) of the Sex Discrimination Act. When they seek to achieve legitimate aims by proportionate means, they have nothing to fear. I believe that to be the usual, indeed almost invariable, situation. It is the unusually strong and adverse findings of fact that made the present case so unusual. I am not impressed by this appeal to policy.

Mr Cavanagh makes a discrete submission about disparate impact. It takes the form of a criticism of the ET’s identification and analysis of the advantaged and disadvantaged groups. He says that the ET misdirected itself. The relevant passages in the decision are as follows:

Mr Cavanagh submits that this is too simplistic, in particular because some of the group identified as disadvantaged would themselves benefit from pay protection and everybody would benefit from better pay levels in the future. I do not agree with this approach. The EAT also rejected it. Elias J said (at paragraph 69 – 70):

I cannot improve on that analysis. Mr Cavanagh makes a further submission, criticising the ET for a subsequent reference to “the doubly advantaged group”, being those who would benefit from pay protection and the best of the improved pay line. He suggests that that was a “muddying of the waters”. However, in my judgment it adds little to and does not undermine the integrity of the exercise described in paragraphs 7.45 and 7.46.

One of the grounds of appeal to the EAT advanced on behalf of the Union was that the ET substituted its own view of how the Union ought to have approached the negotiations. Although the EAT made a passing reference to this ground of appeal in paragraph 83 of its judgment, it did not give detailed consideration of it, preferring to concentrate on what it considered to be the “more fundamental error” about aims and means which, as I have said, I do not find to have been an error at all. Mr Cavanagh now seeks to resurrect the “substitution” point by way of the Respondent’s Notice. He submits that the ET allowed “their speculative beliefs as to what they would have done in the circumstances [to] affect the issue of objective justification”. I do not accept that this submission succeeds in identifying a material error of law on the part of the ET. In Hardy & Hansons PLC v Lax [2005] EWCA Civ 846, [2005] ICR 1565, Pill LJ said (at paragraph 32),

That rejection makes Mr Cavanagh’s submission more difficult, as he appreciates. Some of his criticism of the ET may be justified – for example, its suggestion that the Council might have been pressed to increase the available funds by, among other things, a sale of capital assets appears to ignore legal constraints upon such a course. However, in the context of the totality of the ET’s reasoning, such criticisms seems to me to be minor and immaterial. They do not detract from the validity and force of the finding of disproportionate means.

It follows from what I have said that I would allow this appeal and restore the decision of the ET on indirect discrimination. The case would then need to be remitted to the ET for a remedies hearing, which may raise even more difficult issues than are customarily found in an indirect discrimination case against an employer.

I agree.

I also agree.