Mr Mohamed Nurmohamed, who is the respondent to this appeal, but to whom it is convenient to refer as the Claimant, is an estate agent. From January 2008 until his dismissal on 17 October 2013 he was employed by Chesterton Global Ltd (“Chestertons”), latterly as Director of its Mayfair office. Following his dismissal he brought proceedings in the employment tribunal against Chestertons alleging that he was unfairly dismissed. He claimed that his dismissal was because he had made protected disclosures within the meaning of the Employment Rights Act 1996 – i.e., in the usual shorthand, for being a whistleblower – with the result that it was automatically unfair by reference to section 103A of the Act; but he also claimed for “ordinary” unfair dismissal by reference to section 98. In addition, he claimed to have suffered various detriments, besides his dismissal, because he had made the same disclosures, contrary to section 47B of the Act: the latter claim was made against both Chestertons and its HR Director, Mr Neal Verman. I will refer to Chestertons and Mr Verman as the Respondents, though they are the appellants before us.

Chestertons in its ET3 acknowledged that it was liable for ordinary unfair dismissal, but both it and Mr Verman disputed the claims based on the whistleblower provisions. By a judgment sent to the parties on 4 June 2014 an employment tribunal chaired by Employment Judge Walker upheld both claims. By a judgment handed down on 8 April 2015 the Employment Appeal Tribunal (Supperstone J sitting alone) dismissed the Respondents’ appeal. This is an appeal against that decision.

There were issues before the ET as to whether the Claimant made the disclosures that he alleged; whether, if so, they were protected within the meaning of the Act; and whether in any event they constituted the principal reason for his dismissal or a ground for the detriments of which he complained. But only the second of those issues is live before us. More particularly, the issue is whether the ET was entitled to find that the Claimant had made the disclosures in question in the reasonable belief that they were “in the public interest”, which is one of the elements in the definition of a protected disclosure. That requirement was introduced by amendment in 2013 and has not previously been the subject of any consideration in this Court. Accordingly the charity Public Concern at Work (“PCaW”) has been given permission to intervene.

The Claimant has been represented before us by Mr James Laddie QC and Ms Alice Mayhew; the Respondents by Mr David Reade QC; and PCaW by Mr Thomas Linden QC and Mr Thomas Kibling. In the ET and the EAT the Claimant was represented by Ms Mayhew and the Respondents by Mr Martin Palmer. The quality of the submissions before us was very high.

The scheme of whistleblower protection was created by the Public Interest Disclosure Act 1998. This introduced into the 1996 Act, as Part IVA, a new concept of a “protected disclosure”. Section 47B, introduced into Part V of the Act, gave workers the right to complain to the employment tribunal of being subjected to a detriment, other than dismissal, on the ground that they had made such a disclosure. So far as dismissal is concerned, the 1998 Act introduced into Part X, at section 103A, a new form of unfair dismissal in cases where the reason, or principal reason, for the dismissal of an employee was that they had made a protected disclosure. Establishing a claim of unfair dismissal under this head has three potential advantages: the protection is available from the beginning of the employment, instead of only after a qualifying period; the cap on compensation at section 124 of the Act is disapplied; and in an appropriate case a claim can be made for interim relief.

The term “protected disclosure” is defined in sections 43A-43H of the 1996 Act. The basic structure of those provisions is as follows:

Section 43A defines a protected disclosure as a “qualifying disclosure” which is made by a worker in accordance with any of sections 43C to 43H.

Section 43B defines a qualifying disclosure essentially by reference to the subject-matter of the disclosure: I set it out in full below.

Sections 43C to 43H prescribe six kinds of circumstances in which a qualifying disclosure will be protected, essentially by reference to the class of person to whom the disclosure is made.

As originally enacted, section 43B (1) read as follows:

Those provisions were subject to some exegesis by this Court in Babula v Waltham Forest College[2007] EWCA Civ 174, [2007] ICR 1026. Two points in particular are emphasised in that case, though in truth both are clear from the terms of the section itself:

The definition has both a subjective and an objective element: see in particular paras. 81-82 of the judgment of Wall LJ (pp. 1045-6). The subjective element is that the worker must believe that the information disclosed tends to show one of the six matters listed in sub-section (1). The objective element is that that belief must be reasonable.

A belief may be reasonable even if it is wrong. That is well illustrated by the facts of Babula, where an employee disclosed information about what he believed to be an act of criminal incitement to religious hatred, which would fall within head (a) of section 43B (1). There was in fact at the time no such offence, but it was held that the disclosure nonetheless qualified because it was reasonable for the employee to believe that there was.

Section 43B was amended by section 17 of the 2013 Act, with effect from 25 June 2013, by the insertion of the words which I have italicised into the first part of sub-section (1) as follows:

It was common ground before us, and is in any event clear from the Parliamentary materials to which I refer below, that the object of the amendment was to reverse the effect of the decision of the EAT in Parkins v Sodexho [2001] UKEAT 1239/00, [2002] IRLR 109. In that case an employee was dismissed because he had complained of having to operate a particular machine without supervision, which he said was both a breach of his contract of employment and “a matter of health and safety”. One issue was, as Judge Altman put it at para. 14 of his judgment:

It was widely believed that Parkins v Sodexho extended the scope of whistleblower protection beyond what had been intended by Parliament when enacting the 1998 Act. Paras. 102-103 of the Explanatory Notes to the 2013 Act read as follows:

The inclusion in section 43B of a reference to the public interest was intended to restore the original intention of the Act. The thinking was explained by the responsible Minister, Mr Norman Lamb (the Parliamentary Under-Secretary for Business, Innovation and Skills), while the bill was in the committee stage. An amendment had been proposed to head (b) of section 43B (1), excluding from the legal obligations there referred to “a private contractual obligation which is owed solely to that worker”. The Minister opposed that amendment. I should quote the following passages from his speech in the Public Bill Committee on 3 July 2012 (cols. 385-388), italicising certain key phrases:

It will be noted that the effect of Parkins v Sodexho which it was intended to reverse was repeatedly stated by the Minister as being the according of protection to disclosures made to pursue the worker’s “private” or “personal” interest as opposed to the public interest. It was common ground that it was permissible for us to take note of those passages as confirming the mischief at which the amendment of section 43B was directed.

Sections 43C-43H are less central to the issue before us, but I need to say something about them. As already noted, they set out a range of circumstances in which a qualifying disclosure will be protected, depending principally on the identity of the person to whom disclosure is made. Section 43C, which is the relevant section in this case, is concerned with disclosures to the worker’s employer.

As originally enacted, it was an element in each of sections 43C-43H (except section 43D, which covers disclosures for the purpose of obtaining legal advice) that the disclosure in question be made “in good faith”. The meaning of that phrase in this context was considered in Street v Derbyshire Unemployed Workers Centre[2004] EWCA Civ 964, [2005] ICR 97. In summary, it was held that a disclosure was not made in good faith if the predominant purpose of making it was something other than what Auld LJ referred to, at para. 56 of his judgment (p. 114 C-D), as “the declared public interest purpose of this legislation”, such as out of personal antagonism. As Wall LJ put it at para. 73 (p. 118 A-B), a worker will not be making a disclosure in good faith if:

The requirement of good faith was removed by section 18 of the 2013 Act, also with effect from 25 June 2013. However a new sub-section (6A) was introduced into both section 49 and section 123 of the 1996 Act giving the employment tribunal power to reduce any compensatory award for unlawful detriment or unfair dismissal by up to 25% if it found that the disclosure in question was not made in good faith. In other words, the question of good faith is no longer relevant to liability in a whistleblowing case but it remains relevant to remedy.

The purpose behind the changes effected by section 18 is not apparent from the Explanatory Notes or any other material that we were shown. It might at first sight be thought that the draftsman regarded the introduction of the public interest requirement as rendering the good faith requirement redundant, but counsel were unaware of any explicit indication to that effect (and Mr Linden’s clients at least might be expected to know if there were). It is, however, clear that the draftsman contemplated that a disclosure which was made in the reasonable belief that it was in the public interest might nevertheless be made in bad faith, since otherwise the new sections 49 (6A) and 123 (6A) would never bite: I return to this below.

Given the limited nature of the issue before us, I need only set out the facts (as found by the ET) relating to the disclosures which were the basis of the findings of liability. I can do so fairly shortly, though there was a good deal more factual material relevant to the other issues with which we are not concerned.

The background is that in 2011 a new group of investors acquired a shareholding in Chestertons, which is not a publicly quoted company. Their involvement prompted a review of the system for payment of commission to the sales staff, in the light of concerns that they were calculated only by reference to revenue and without regard to profitability. A new commission system was introduced at the beginning of 2013: this related commission to a substantial extent to the achievement of budgeted profits for the relevant department – in the Claimant’s case the Mayfair office. The Claimant believed that the new system would have a serious adverse impact on his earnings. He objected, but in February 2013 he agreed to the new system subject to some modifications.

Against that background, the Claimant monitored Chestertons’ internal accounts over the following months. At a meeting on 14 August 2013 with Ms Patricia Farley, the director responsible for the London area, he demonstrated a number of what he said were discrepancies in the monthly accounts which appeared to show that the profitability of the Mayfair office was being artificially suppressed so as to reduce the level of commission payable. Two examples to which the Tribunal referred (though the Claimant apparently gave others) were that a depreciation charge had been made which was higher than that budgeted for; and that a figure was included for “staff bonus” when none had been paid. The Claimant described this to Ms Farley as “manipulating the accounts to the benefit of the shareholders”. The Tribunal found that he genuinely believed that such manipulation was occurring and that his belief was reasonable – though it made it clear that it was making no finding that it was correct.

The Claimant repeated what appear to have been essentially the same allegations to Mr Verman on 24 September 2013 and to Ms Farley again on 8 October. As regards the latter meeting, the Tribunal at para. 144 of its Reasons recorded the Claimant’s evidence as follows:

The ET’s Reasons are thorough and carefully structured. The only part with which we are concerned – apart from the findings of fact, which I have sufficiently summarised above – comes under the heading “Did the Claimant make the disclosures in the reasonable belief that they were in the public interest ?”. Paras. 146-148 read as follows:

Slightly confusingly, the Tribunal then goes on, at paras. 149-150, to refer to the second of the two groups identified in para. 148, i.e. “anybody who relied on the accounts”. It acknowledges that it was possible that in the event of a sale of Chestertons’ business potential purchasers might be misled by mis-statements in the accounts, but it discounts this group at the end of para. 149 because “we have no evidence that Mr Nurmohamed had that issue in mind at the time”.

At para. 151 the Tribunal reverts to the first of the two groups affected, the “100 managers”. The paragraph reads:

The ultimate issue for us, as it was for the EAT, is whether the reasoning of the ET was bad in law. That being so, and meaning no disrespect to Supperstone J, I will not lengthen this judgment by setting out his analysis.

The issue in this appeal turns on the meaning, and the proper application to the facts, of the phrase “in the public interest”. But before I get to that question I would like to make four points about the nature of the exercise required by section 43B (1).

First, and at the risk of stating the obvious, the words added by the 2013 Act fit into the structure of section 43B as expounded in Babula (see para. 8 above). The tribunal thus has to ask (a) whether the worker believed, at the time that he was making it, that the disclosure was in the public interest and (b) whether, if so, that belief was reasonable.

Second, and hardly moving much further from the obvious, element (b) in that exercise requires the tribunal to recognise, as in the case of any other reasonableness review, that there may be more than one reasonable view as to whether a particular disclosure was in the public interest; and that is perhaps particularly so given that that question is of its nature so broad-textured. The parties in their oral submissions referred both to the “range of reasonable responses” approach applied in considering whether a dismissal is unfair under Part X of the 1996 Act and to “the Wednesbury approach” employed in (some) public law cases. Of course we are in essentially the same territory, but I do not believe that resort to tests formulated in different contexts is helpful. All that matters is that the Tribunal should be careful not to substitute its own view of whether the disclosure was in the public interest for that of the worker. That does not mean that it is illegitimate for the tribunal to form its own view on that question, as part of its thinking – that is indeed often difficult to avoid – but only that that view is not as such determinative.

Third, the necessary belief is simply that the disclosure is in the public interest. The particular reasons why the worker believes that to be so are not of the essence. That means that a disclosure does not cease to qualify simply because the worker seeks, as not uncommonly happens, to justify it after the event by reference to specific matters which the tribunal finds were not in his head at the time he made it. Of course, if he cannot give credible reasons for why he thought at the time that the disclosure was in the public interest, that may cast doubt on whether he really thought so at all; but the significance is evidential not substantive. Likewise, in principle a tribunal might find that the particular reasons why the worker believed the disclosure to be in the public interest did not reasonably justify his belief, but nevertheless find it to have been reasonable for different reasons which he had not articulated to himself at the time: all that matters is that his (subjective) belief was (objectively) reasonable. (Footnote: ⁴)

Fourth, while the worker must have a genuine (and reasonable) belief that the disclosure is in the public interest, that does not have to be his or her predominant motive in making it: otherwise, as pointed out at para. 17 above, the new sections 49 (6A) and 103 (6A) would have no role. I am inclined to think that the belief does not in fact have to form any part of the worker’s motivation – the phrase “in the belief” is not the same as “motivated by the belief”; but it is hard to see that the point will arise in practice, since where a worker believes that a disclosure is in the public interest it would be odd if that did not form at least some part of their motivation in making it.

Finally by way of preliminary, although this appeal gives rise to a particular question which I address below, I do not think there is much value in trying to provide any general gloss on the phrase “in the public interest”. Parliament has chosen not to define it, and the intention must have been to leave it to employment tribunals to apply it as a matter of educated impression. Although Mr Reade in his skeleton argument referred to authority on the Reynolds defence in defamation and to the Charity Commission’s guidance on the meaning of the term “public benefits” in the Charities Act 2011, the contexts there are completely different. The relevant context here is the legislative history explained at paras. 10-13 above. That clearly establishes that the essential distinction is between disclosures which serve the private or personal interest of the worker making the disclosure and those that serve a wider interest. This seems to have been essentially the approach taken by the Tribunal at para. 147 of its Reasons.

The particular issue that arises in this appeal is whether a disclosure which is in the private interest of the worker making it becomes in the public interest simply because it serves the (private) interests of other workers as well. Mr Reade, for the Respondent, submits that that was the approach taken by the Tribunal in this case: the Claimant’s disclosure was found to be in the public interest (or, more accurately, his belief that it was was found to be reasonable) essentially only because 99 other employees were in the same boat. In his submission, that approach was wrong: mere multiplicity of workers sharing the same interest is not enough. In order for a disclosure to be in the public interest, the interests served have to “extend outside the workplace”, in the sense that the disclosure furthers the interests of person other than the workers themselves qua workers. To take the example of holiday pay given by the Minister when discussing the 2013 amendment, a disclosure that a large employer was systematically miscalculating the amounts of holiday pay due to its employees could not reasonably be regarded as a disclosure made in the public interest simply because thousands of workers were affected by it: the interest was inherently personal in nature. Mr Reade emphasised that it was not his position that a disclosure affecting the personal rights of a worker could not also, on the facts of a particular case, be in the public interest, as indeed the Minister made clear. A disclosure, say, which tended to show that hospital doctors were being required to work excessive hours might well be in the public interest, as well as in the personal interests of the doctors themselves, because of the risk to patients; but that would be because of the nature of the disclosure, not because of the number of doctors affected.

For PCaW Mr Linden took a starkly opposite position. Parkins v Sodexho had been concerned with a case of an individual employee whose grievance was unique to him and was not understood to have any wider ramifications. It was unnecessary to treat the amendments introduced by the 2013 Act as excluding from protection any other kind of disclosure. It was highly desirable that there be a bright-line rule enabling workers and those advising them to know with reasonable certainty when a disclosure would be protected. That could only be achieved by treating any disclosure as being “in the public interest” if it is in the interests of anyone else besides the worker making the disclosure.

Mr Laddie, for the Claimant, took a position between those two extremes. He accepted that the mere fact that the disclosure was in the interest of other workers besides the worker making it was not in itself enough to bring it within section 43B (1); but he did not accept that numbers were irrelevant, nor that the disclosure need always be in the interests of persons “outside the workplace” in Mr Reade’s sense. He contended that a disclosure of pay irregularities affecting the entirety of the NHS workforce (over a million employees) would plainly be in the public interest; or, if that case were sought to be distinguished on the basis that the NHS is a public authority, that the same would be the case for Royal Mail (a plc) or indeed the John Lewis Partnership (a private company). The disclosure in such a case would be in the public interest simply because of the number of employees affected. He said that in any case the tribunal in deciding whether a disclosure was in the public interest would have to consider all the circumstances, but he suggested that the following factors would normally be relevant (I have paraphrased them slightly):

the numbers in the group whose interests the disclosure served – see above;

the nature of the interests affected and the extent to which they are affected by the wrongdoing disclosed – a disclosure of wrongdoing directly affecting a very important interest is more likely to be in the public interest than a disclosure of trivial wrongdoing affecting the same number of people, and all the more so if the effect is marginal or indirect;

the nature of the wrongdoing disclosed – disclosure of deliberate wrongdoing is more likely to be in the public interest than the disclosure of inadvertent wrongdoing affecting the same number of people;

the identity of the alleged wrongdoer – as Mr Laddie put it in his skeleton argument, “the larger or more prominent the wrongdoer (in terms of the size of its relevant community, i.e. staff, suppliers and clients), the more obviously should a disclosure about its activities engage the public interest” – though he goes on to say that this should not be taken too far.

I am satisfied that Mr Linden’s submission on behalf of PCaW is wrong. An approach to the concept of “public interest” which depended purely on whether more than one person’s interest was served by the disclosure would be mechanistic and require the making of artificial distinctions. It would be extremely unsatisfactory if liability depended on the happenstance of the circumstances of other employees. If Mr Linden were right a disclosure by a worker about the calculation of his holiday pay would be unprotected if he were the only employee affected but would have attracted protection the following day if another employee had been recruited who was affected in the same way. It is in my view clear that the question whether a disclosure is in the public interest depends on the character of the interest served by it rather than simply on the numbers of people sharing that interest. That is in my view the ordinary sense of the phrase “in the public interest”; but if there were any doubt about the matter the position is clear from the legislative history. The essence of the “Parkins v Sodexho error” which the 2013 Act was intended to correct was that a worker could take advantage of “whistleblower protection” where the interest involved was personal in character. Such an interest does not change its character simply because it is shared by another person. The advantage of achieving a bright line cannot be obtained by distorting the natural meaning of the statutory language.

It might be thought to follow from my rejection of Mr Linden’s argument that I should accept Mr Reade’s opposite submission that mere multiplicity of persons whose interests are served by the disclosure of a breach of the contract of employment can never, by itself, convert a personal interest into a public interest: if the essential question is the character of the interest served, why should that character be changed by the fact that the number of individuals whose interests is engaged is 200 or 2,000 any more than when it is two ? I see the logical attraction of that argument, and I was initially minded to accept it. However, on reflection, I do not think it would be right to take that position. The statutory criterion of what is “in the public interest” does not lend itself to absolute rules, still less when the decisive question is not what is in fact in the public interest but what could reasonably be believed to be. I am not prepared to rule out the possibility that the disclosure of a breach of a worker’s contract of the Parkins v Sodexho kind may nevertheless be in the public interest, or reasonably be so regarded, if a sufficiently large number of other employees share the same interest. I would certainly expect employment tribunals to be cautious about reaching such a conclusion, because the broad intent behind the amendment of section 43B (1) is that workers making disclosures in the context of private workplace disputes should not attract the enhanced statutory protection accorded to whistleblowers – even, as I have held, where more than one worker is involved. But I am not prepared to say never. In practice, however, the question may not often arise in that stark form. The larger the number of persons whose interests are engaged by a breach of the contract of employment, the more likely it is that there will be other features of the situation which will engage the public interest.

Against that background, in my view the correct approach is as follows. In a whistleblower case where the disclosure relates to a breach of the worker’s own contract of employment (or some other matter under section 43B (1) where the interest in question is personal in character (Footnote: ⁵)), there may nevertheless be features of the case that make it reasonable to regard disclosure as being in the public interest as well as in the personal interest of the worker. Mr Reade’s example of doctors’ hours is particularly obvious, but there may be many other kinds of case where it may reasonably be thought that such a disclosure was in the public interest. The question is one to be answered by the Tribunal on a consideration of all the circumstances of the particular case, but Mr Laddie’s fourfold classification of relevant factors which I have reproduced at para. 34 above may be a useful tool. As he says, the number of employees whose interests the matter disclosed affects may be relevant, but that is subject to the strong note of caution which I have sounded in the previous paragraph.

Turning to the Tribunal’s reasoning in the present case, if one takes para. 151 of its judgment by itself, there is force in Mr Reade’s submission that this is a case where it found the disclosure to be in the public interest simply because of the numbers of employees affected. But if one looks at its earlier findings, and specifically at para. 144, there are other features in the situation which might be said to render disclosure in the public interest. Specifically, the disclosure was of what was said to be deliberate wrongdoing, and the alleged wrongdoing took the form of mis-statements in the accounts to the tune of £2m-£3m. (Footnote: ⁶) If the accounts in question were the statutory accounts, even of a private company, the disclosure of such a mis-statement would unquestionably be in the public interest. The fact that the accounts in question were only internal makes the position less black-and-white; but internal accounts feed into the statutory accounts, and we are dealing here with a very substantial and prominent business in the London property market. It is debatable whether the Tribunal, which was navigating uncharted waters, fed those factors into its assessment that it was reasonable to regard disclosure as being in the public interest even though it did not expressly say so. But, even if it did not, I believe that they would only have reinforced the conclusion to which it came based on the numbers alone, so that any error of law in its reasoning was immaterial.

For those reasons I do not believe that there was any error of law in the Tribunal’s decision, and I would dismiss the appeal.

I agree with my Lord, Underhill LJ, that the correct approach to the test for determining whether a disclosure is made in the public interest is the one he sets out at paragraph 37 of his judgment, which I have seen in draft, and this appeal should be dismissed. I add some observations of my own because I did not share his initial attraction to the approach advocated by Mr Reade because I consider that it has similar disadvantages to the approach advocated by Mr Linden.

Mr Reade and Mr Linden argued that the approaches for which they contended would provide certainty. Both approaches are binary. Both, albeit from different ends of the spectrum, suffer from the disadvantages of binary approaches: in their different ways, they can ignore substance and lead to undue formalism. Mr Linden’s does so for the reasons given by my Lord with which I agree: it would be mechanistic and require the making of artificial distinctions. But Mr Reade’s does so as well. He argued that the interests of persons other than the workers themselves qua workers must be furthered. If mere multiplicity of persons whose interests are served by the disclosure of a breach of the contract of employment can never, by itself, convert a personal interest into a public interest and there is a need positively to show the presence of an interest of a different nature, a virtually dispositive effect is given to a factor which, while of major importance, should not have such an effect. I have in mind the example of disclosure that an employer of thousands of workers was systematically miscalculating the amounts of holiday pay due to them. Excluding the factors identified by Mr Laddie (see para. 34 above) would, in my judgment, have an undesirable “chilling effect” on the operation of provisions which, broadly speaking seek to encourage disclosures by employees to employers.

Certainty is undoubtedly important, and some formalism is often a necessary price for the increased certainty that can be obtained from “bright line” rules such as those for which Mr Reade and Mr Linden argue. But the achievement of certainty can be a chimera, and the disadvantages of binary approaches should not be forgotten. “Bright line” rules can deflect attention from the underlying purpose served by legislation which uses open textured terms such as “public interest”. This may be particularly so in a context such as the present where the legislation seeks to protect the reasonable beliefs of the person who makes a disclosure.

The interpretations for which Mr Reade and Mr Linden argue minimise the need for choice in the application of the amendment to section 43B of the Employment Rights Act 1996 enacted by section 17 of the Enterprise and Regulatory Reform Act 2013 to individual cases. This is because they fasten on a feature or features present in a plain case and insist that these are both necessary and sufficient either for another case to fall within the scope of the provision or to prevent it from doing so. As Professor H.L.A. Hart stated in his discussion of formalism and rule-scepticism in The Concept of Law (1961, pp. 126-7):

In my judgment, these considerations are relevant in interpreting section 43B as amended. They explain why I agree with the more nuanced approach that is reflected in Mr Laddie’s submissions and in my Lord’s judgment and why that would neither produce undesirable uncertainty nor undermine the purpose of the amendments made to section 43B by the 2013 Act to address the effect of Parkins v Sodexho.

I too agree with my Lord, Underhill LJ, that the appeal should be dismissed, essentially for the reasons that he gives. I also find myself in sympathy with the observations of my Lord, Beatson LJ, about the disadvantages of the approaches put forward by Mr Linden and Mr Reade.