Dunnachie v Kingston Upon Hull City Council

Ever since the introduction by the Industrial Relations Act 1971 of a right not to be unfairly dismissed, compensation for unfair dismissal has been required by law to be “such amount as the tribunal considers just and equitable in all the circumstances having regard to the loss sustained by the complainant in consequence of the dismissal in so far as that loss is attributable to action taken by the employer” (see now the Employment Rights Act 1996, s.123(1)).

In the early days of the new legislation the National Industrial Relations Court, under its first (and only) president Sir John Donaldson, decided in Norton Tool Co Ltd v Tewson [1972] ICR 501 that this formula embraced only quantifiable pecuniary losses. This remained the almost unquestioned orthodoxy until in the speeches of the House of Lords in Johnson v Unisys Ltd [2001] UKHL 13 it was - putting it equivocally for the moment - indicated that, although at common law the House's decision in Addis v Gramophone Co Ltd [1909] AC 488 continued to confine damages for breach of contract to pecuniary losses, the statutory formula was large enough to embrace damages for non-economic harm in unfair dismissal cases.

The decision in Johnson has attracted a good deal of academic comment, not all of it favourable, partly because of its uncovenanted impact on the law of constructive dismissal. But its apparent enlargement of the accepted ambit of compensation led one learned commentator, Professor Hugh Collins (2001) 30 ILJ 305, to remark that a phoenix of truly just and equitable compensation might now rise from the ashes of the hoped-for evolution of the common law of wrongful dismissal.

This appeal is about the phoenix. If the Employment Appeal Tribunal is right and Norton Tool remains good law, the phoenix was an illusion. Whether this is the case depends on two things: first, whether the availability of full compensation was an integral part of their Lordships' reasoning; secondly, if it was not, whether this court should nevertheless now hold it to be the law.

It is not necessary to say a great deal about the history of the case. Mr Dunnachie, an environmental health officer, had begun work with Hull City Council at the age of 19 in 1986, had qualified two years later and by the time he was forced to resign was an acting principal EHO in the council's food section. His resignation on a month's notice in March 2001 was brought about, the Employment Tribunal found, by a prolonged campaign of harassment and undermining on the part of his colleague and sometime line manager Gary Kitching. The tribunal summed it up like this:

“In that connection, we found that Mr Kitching did, for whatever reason, have a low opinion of the applicant’s capabilities. That opinion was misplaced. Nevertheless, he acted upon it by seeking to undermine the applicant whenever the opportunity presented itself. A particularly bad example was his irrational refusal to allow the prosecution of Skelton’s Bakery to proceed. When the applicant challenged that decision by going to their manager, Mrs Cottis, we are satisfied that Mr Kitching retaliated by conducting an in-depth investigation into the management of the applicant’s files, without telling him that he was doing so. He then threatened the applicant with disciplinary proceedings and left the matter hanging in the air. Mr Kitching’s conduct was compounded by that of Mrs Cottis, who failed to alleviate the applicant’s anxieties about the prospect of being suspended. Both she and Mr Duxbury [her line manager] either failed or refused to recognise that the applicant had been a victim of bullying by Mr Kitching. Mr Duxbury deliberately sought to deflect the applicant from making a formal complaint under the respondent’s personal harassment policy. The respondent’s treatment of the applicant by those officers caused his ill health. We are satisfied that there was the clearest evidence of a breach of the implied term of mutual trust and confidence.”

It followed that Mr Dunnachie had been constructively and unfairly dismissed. Because he had a family to support, he had hung on until he found another job to go to and had only then given notice. The new job, with Doncaster City Council, was less well paid, of lower status and at a much greater distance from his home.

This was a bad case of workplace bullying, compounded by an equally serious refusal by management to deal with it. The blow to a conscientious employee's self-esteem which such treatment delivers may well be the unkindest cut of all, worse in many ways than the monetary loss. There was no professional evidence that the distress and its effects had amounted to a recognised psychiatric condition but Mr Dunnachie had been reduced by his treatment to a state of overt despair.

The tribunal gave their decision on 15 May 2002 and adjourned the question of remedies, failing agreement, to 24 May. On 11 June 2002 they handed down extended reasons for their award. The total was £54,940, together with £2752 costs. The breakdown was as follows:

Basic award 3,240.00

Compensatory award

Loss of earnings to date 6,148.16

Loss of future earnings 74,175.35

Additional cost of travel to work 29,514.77

Loss of statutory industrial rights 250.00

Compensation under Johnson v Unisys 10,000.00

The total compensatory award, £123,328.28, had to be reduced to the amount of the statutory cap of £51,700. To the total produced by adding the basic award to the capped compensatory award, the tribunal added an award of costs because they considered that the respondent council had conducted the proceedings high-handedly, in particular by unreasonably threatening Mr Dunnachie with a costs order should he lose.

The council appealed to the EAT against the calculation of the compensatory award - on which they succeeded to the extent of having it remitted to a fresh employment tribunal - and separately against the Johnson award of £10,000. It is the latter to which the present very full and careful judgment of the EAT, delivered by Burton P, is directed. Its conclusions are, first, that the material part of the speech of Lord Hoffmann in Johnson is not integral to the House's reasoning, and secondly that Norton Tool was rightly decided and should continue to be followed. Recognising the importance of the issues, the EAT itself gave permission to appeal.

Although it will be necessary to refer to other provisions later in this judgment, the key provision is now found in Part X, s.123, of the Employment Rights Act 1996 as amended:

The remainder of s.123 and the other provisions referred to in subsections (1) and (2) do not matter to the present argument. The provisions which I have quoted do not differ in any significant way from the formula originally enacted in s.116 of the Industrial Relations Act 1971 and reproduced at intervals in amending or consolidating Acts. It will be relevant, however, that s.116 of the 1971 Act governed not only awards of compensation for unfair dismissal but all awards of compensation under the Act for unfair industrial practices. These included (under s.101) the inducement of victimisation by trade unions; (under s.102) failure by employers to implement agency shop agreements or provide information relevant to collective bargaining; (under ss.103, 107 and 108) unfair exclusion from trade unions and employers’ associations; and (under s.106) victimisation by employers of trade union members (prohibited by s.5), as well as unfair dismissal.

The issue before their Lordships' House in Johnson was whether in an action at common law for wrongful dismissal an employee can recover damages for consequential psychiatric harm. Their Lordships held that he could not, the majority on the ground that a claim for compensation arising out of the manner of a dismissal (which was what was sought) did not lie within the doctrine of Addis v Gramophone Co Ltd [1909] AC 488 and must now be made under the statutory unfair dismissal scheme or not at all. They consequently declined to enlarge the limit placed by Addis on the ambit of damages for breach of contract.

The holding, as reported in the Appeal Cases headnote [2003] 1 AC 518, was:

“… that (per Lord Bingham of Cornhill, Lord Nicholls of Birkenhead, Lord Hoffmann and Lord Millettt) under Part X of the Employment Rights Act 1996, Parliament had provided the employee with a limited remedy for the conduct of which he complained; that, although it was possible to conceive of an implied term which the common law could develop to allow an employee to recover damages for loss arising from the manner of his dismissal, it would be an improper exercise of the judicial function for the House to take such a step in the light of the evident intention of Parliament that such claims should be heard by specialist tribunals and the remedy restricted in application and extent … ”

Lord Steyn concurred in the result, but on grounds of remoteness rather than of the non-existence of a cause of action.

The reference in the headnote to the restricted statutory remedies reflects Lord Nicholls’ reliance (para.2 ) on “matters such as the classes of employees who have the benefit of the statutory right, the amount of compensation payable and the short time limits for making claims”. None of these touch the present question, which is whether there is also a restriction on the kinds of consequence for which compensation may be awarded. As to this, the critical passage comes in the speech of Lord Hoffmann:

“54.

My Lords, this statutory system for dealing with unfair dismissals was set up by Parliament to deal with the recognised deficiencies of the law as it stood at the time of Malloch v Aberdeen Corpn [1971] 1 WLR 1581. The remedy adopted by Parliament was not to build upon the common law by creating a statutory implied term that the power of dismissal should be exercised fairly or in good faith, leaving the courts to give a remedy on general principles of contractual damages. Instead, it set up an entirely new system outside the ordinary courts, with tribunals staffed by a majority of lay members, applying new statutory concepts and offering statutory remedies. Many of the new rules, such as the exclusion of certain classes of employees and the limit on the amount of the compensatory award, were not based upon any principle which it would have been open to the courts to apply. They were based upon policy and represented an attempt to balance fairness to employees against the general economic interests of the community. And I should imagine that Parliament also had in mind the practical difficulties I have mentioned about causation and proportionality which would arise if the remedy was unlimited. So Parliament adopted the practical solution of giving tribunals a very broad jurisdiction to award what they considered just and equitable but subject to a limit on the amount.

55.

In my opinion, all the matters of which Mr Johnson complains in these proceedings were within the jurisdiction of the industrial tribunal. His most substantial complaint is of financial loss flowing from his psychiatric injury which he says was a consequence of the unfair manner of his dismissal. Such loss is a consequence of the dismissal which may form the subject matter of a compensatory award. The only doubtful question is whether it would have been open to the tribunal to include any sum by way of compensation for his distress, damage to family life and similar matters. As the award, even reduced by 25% exceeded the statutory maximum and had to be reduced to £11,000, the point would have been academic. But perhaps I may be allowed a comment all the same. I know that in the early days of the National Industrial Relations Court it was laid down that only financial loss could be compensated: see Norton Tool Co Ltd v Tewson [1973] 1 WLR 45; Wellman Alloys Ltd v Russell [1973] ICR 616. It was said that the word ‘loss’ can only mean financial loss. But I think that is too narrow a construction. The emphasis is upon the tribunal awarding such compensation as it thinks just and equitable. So I see no reason why in an appropriate case it should not include compensation for distress, humiliation, damage to reputation in the community or to family life.

56.

Part X of the Employment Rights Act 1996 therefore gives a remedy for exactly the conduct of which Mr Johnson complains. But Parliament had restricted that remedy to a maximum of £11,000, whereas Mr Johnson wants to claim a great deal more. The question is whether the courts should develop the common law to give a parallel remedy which is not subject to any such limit.”

Lord Bingham (para 1) and Lord Millettt (para 68) both expressed their agreement with the speech of Lord Hoffmann. If the passage of Lord Hoffmann's speech which I have cited was integral to his reasoning, their assent will - for reasons I will come to - have given it the force of binding authority. Antony White QC for the appellant submits that this is the case; John Bowers QC for the respondent submits that it is not. The excellence of their respective arguments has made this and the other questions both easier to grasp and harder to resolve.

I have put the paradigm of judicial authority as sharply as I have done in the preceding paragraph because, as we have correctly been reminded, it is not every remark in a judgment to which the author's fellow-judges can be taken to be assenting when they express agreement with it. As Lord Reid said in Cassell & Co Ltd v Broome [1972] AC 1027, 1087, “concurrence with the speech of a colleague does not mean acceptance of every word which he has said. If it did there would be far fewer concurrences than there are”. The distinction which he made, and which I respectfully adopt, was between views which did and did not form “an essential step in [the judge’s] argument”.

This formulation is not as simple as it sounds. It is often possible to proceed to a conclusion X either by the progression A -> B -> X or by the progression A -> B -> C -> X. Here step C is by definition not objectively essential; but the judge who takes it has elected to integrate it into his or her reasoning. Does an assenting judgment in such a case include assent to C or not? In Jacobs v London County Council [1950] AC 361, 369, Lord Simonds said: “… there is in my opinion no justification for regarding as obiter dictum a reason given by a judge for his decision because he has given another reason also. If it were a proper test to ask whether the decision would have been the same apart from the proposition alleged to be obiter, then a case which ex facie decided two things would decide nothing”. (From the illustrations which he went on to cite it is clear that he had in mind decisions founded on two alternative grounds.) This makes clear sense.

I would conclude therefore that an additional reason which a judge gives for reaching his conclusion is for the purposes of establishing authority an integral part of his reasoning. Once that point is reached, I think the other questions raised by Mr Bowers, as to the reading of judicial minds and the imputation to them of others’ ideas, become mercifully immaterial. Giving full weight to Lord Reid’s cautionary remark, one still has to take expressed assents by other members of the court or committee as adopting at least as much of the leading judgment or speech as its author has chosen to make part of his or her reasons for reaching a particular conclusion. It is always possible, after all, for an assenting judgment to reserve or even to reject (reverting to my algebra) the correctness of proposition C without disturbing the result. So everything turns on whether the view expressed by Lord Hoffmann in the material part of his speech was of this character or was, as Mr Bowers submits it was, a personal reflection on a side-issue, having manifest forensic weight but no judicial authority.

This inevitably depends to a very large extent on how the judge has chosen to express his or her particular view. If it is prefaced by words such as “This is enough to dispose of the appeal, but I would nevertheless observe that …” there will be little difficulty in segregating what follows from the judge’s essential reasoning. If it is prefaced by words such as “This in itself would not be sufficient; but …”, the converse will equally probably follow. But the narrative form which in the eyes of other European judges makes United Kingdom judgments remarkable and, at least in our own eyes, makes them interesting, eschews such formulaic approaches; so that those who come after have sometimes to undertake what Bacon called divination (and what Lord Hoffmann in HIH Casualty v Chase Manhattan Bank [2003] UKHL 6, para. 80, described as having “more in common with reading tea-leaves than with legal reasoning”).

The EAT expressed itself “entirely satisfied” that the passage in question was no more than “the expression of opinion by a very experienced and influential Law Lord as to what employment tribunals ought, notwithstanding Norton Tool, to do in respect of recoverability of non-economic loss in unfair dismissal claims”. Mr Bowers supports this conclusion essentially on two grounds. One is that Lord Bingham and Lord Millettt should not be taken to have assented to dicta on a topic which had not featured at all in the printed cases. I will come in a moment to how the issue entered the argument in their Lordships’ House; but for the reasons I have given, I consider that expressed assents have to be presumed to relate - and to relate only - to whatever is integral to the leading judgment's reasoning, however it may have got there.

Mr Bowers’ second essential ground is that nowhere does Lord Hoffmann or any other member of the Appellate Committee hold in terms that Norton Tool was wrongly decided. This is a much weightier argument. In the ordinary way one would expect a departure at appellate level from a first-instance decision which has enjoyed unquestioned authority for thirty years to be squarely explained. Why then was this not done here?

Mr White has gone some way to providing an answer; although it cannot by itself answer the more fundamental question of the authority of Lord Hoffmann's remarks. The note of argument in the Appeal Cases report of the case, which was not available to the EAT, shows that Norton Tool came without objection into the argument when counsel for the employee, Edward Faulks QC, citing the decision, submitted that to force his client back upon the statutory remedy would deny him compensation for any but economic loss (p.523). David Pannick QC for the employer countered this (p.524) by submitting that there was no good reason, Norton Tool notwithstanding, why tribunals “should not be able to make an award for the psychiatric injury itself as well as the financial consequences. Tribunals should be empowered to make such awards”. Mr Faulks, continuing the role-reversal, urged in reply that the common law should not be developed on the premise that a case such as Norton Tool had been wrongly decided.

Pausing here, it does not look as though Mr Pannick was seeking the reversal of Norton Tool - hence Mr Faulks' guarded reply. He seems to have been arguing that Parliament should reverse it by legislation. Hence too, I think, the less than definite tone of paragraph 55 of Lord Hoffmann's speech, to which Mr Bowers has drawn attention. But the fact remains that Lord Hoffmann’s remarks, in contrast to counsel's submission, are not about the law as it could or should be: they are about the law as it is - specifically about the breadth of the true construction of s.123. They are not hedged or qualified like, for example, his recent remarks in Wainwright v Home Office [2003] 4 All E.R. 969, paras 51-2, on the availability of an Article 8 action for an invasion of privacy: “Speaking for myself , I am not so sure…. Be that as it may…” This very clearly, and in contrast to the material passage in Johnson, is the vocabulary of a judicial aside.

What seems to me, moreover, to indicate that paragraph 55 is not a judicial aside but an integral step in Lord Hoffmann's reasoning is the way he begins paragraph 56: “Part X of the Employment Rights Act 1996 therefore gives a remedy for exactly the conduct of which Mr Johnson complains”. With all respect to the EAT, the preceding passage cannot have had the character that they ascribe to it of simply expressing a view as to what tribunals should do “nowithstanding Norton Tool”. Either Norton Tool continues to bind tribunals, or it is wrong in law and does not. There is no way in which, short of being overruled, it can be stripped of its authority and ignored by tribunals. So whatever he was doing, Lord Hoffmann cannot have been making the suggestion ascribed to him by the EAT.

It seems to me, in the end, that Lord Hoffmann in paragraph 55 was grasping a nettle which, although it had sprung up only in oral argument, he recognised as having a sting: that to deny a remedy at common law on the ground that a parallel statutory scheme existed, when the statutory scheme apparently denied the very remedy that was being sought at common law, was intellectually unsatisfactory and in practice would leave a black hole. It was to fill that jurisprudential space that, as it seems to me, he said what he did in the critical passage.

Given what therefore seems to me the true choice, I find myself driven to the conclusion that paragraph 55 is an essential step in Lord Hoffman’s reasoning. If so - indeed because it is so - the assent expressed by Lord Bingham and Lord Millettt gives it the force of binding authority in this court. Its inexorable meaning and effect - acknowledging, as I do, the force of Mr Bowers' point that one would have expected it to be expressly said - are that Norton Tool is not good law.

In case I am wrong - and (for reasons expressed with some muscularity by Lord Justice Brooke) I may well be - about the authority of paragraph 55 of Johnson v Unisys Ltd, it is necessary to consider independently whether Norton Tool was rightly decided. Norton Tool has never been frontally questioned at this level, and we have of course heard much fuller argument than was available to their Lordships on it. In the intervening years the decision has been repeatedly followed in the NIRC and its successor the EAT, and so by industrial and employment tribunals throughout the country: see Robert Normansell (Birmingham)Ltd v Barfied (1973) 8 ITR 171, 174; Vaughan v Weighpack [1974] ICR 261, 265; Wellman Alloys v Russell [1973] ICR 616, 619; Fougere v Phoenix Motor Co Ltd [1976] IRLR 259. In this court and in their Lordships’ House it has been more than once taken as given law: see O'Donoghue v Redcar and Cleveland BC [2001] IRLR 615; Mahmud v BCCI [1998] AC 20, paras. 39, 52. But it is without question open to challenge in this court, and in my judgment Mr White’s challenge to it is well-founded.

Norton Tool was to be one of the seminal decisions of the National Industrial Relations Court, and it has continued to govern the operation of the unfair dismissal jurisdiction during its successive metamorphoses. In it, in relation to what was then s.116(1) of the Industrial Relations Act 1971, the President held:

“The court or tribunal is enjoined to assess compensation in an amount which is just and equitable in all the circumstances and there is neither justice nor equity in a failure to act in accordance with principle. The principles to be adopted emerge from section 116 of the Act of 1971. First, the object is to compensate, and compensate fully, but not to award a bonus; save possibly in the special case of a refusal by an employer to make an offer of employment in accordance with the recommendation of the court or a tribunal. Secondly, the amount to awarded is that which is just and equitable in all the circumstances, having regard to the loss sustained by the complainant. ‘Loss’, in the context of section 116 does not include injury to pride or feelings. In its natural meaning the word is not to be so construed, and that this meaning is intended seems to us to be clear from the elaboration contained in section 116 (2). The discretionary element is introduced by the words ‘having regard to the loss”. This does not mean that the court or tribunal can have regard to other matters, but rather that the amount of the compensation is not precisely and arithmetically related to the proved loss. Such a provision will be seen to be natural and possibly essential, when it is remembered that the claims with which the court and tribunals are concerned are more often than not presented by claimants in person and in conditions of informality. It is not, therefore, to be expected that precise and detailed proof of every item of loss will be presented, although, after making due allowance for the skills of the persons presenting the claims, the statutory requirement for informality of procedure and the undesirability of burdening the parties with the expense of adducing evidence of an elaboration which is disproportionate to the sums in issue, the burden of proof lies squarely upon the complainant.”

The weak link in this reasoning, in my very respectful view, is the proposition that “having regard to the loss” “does not mean that the court or tribunal can have regard to other matters”. One asks why not. If, as Mr Bowers rightly accepts, the governing phrase in the formula for compensation is “such amount as the tribunal considers just and equitable in all the circumstances”, the natural implication of adding “having regard to…” is that the matters which follow must be compensated for but that compensation is to include whatever else justice and equity require.

This highlights another problem with the Norton Tool reading of the statutory formula: it leaves the governing concept - compensation which is just and equitable - without a role. Mr Bowers argues that this is not so. He cannot of course introduce contributory fault on the employee's part by this route, since it has always been separately provided for in the legislation. But he submits that it allows an uplift for something akin to a Smith v Manchester award for loss of employability, and a reduction of the kind established in Polkey v AE Dayton Services Ltd [1988] AC 344, where an unfairly dismissed employee could in the circumstances have been dismissed without unfairness. The difficulty with these examples is that the first (as Mr Bowers recognised) is an orthodox form of economic loss, while the second is not allocated anywhere in the speeches in their Lordships’ House to the ‘just and equitable’ element of the statutory formula. Only in the Court of Session's decision in Campbell v Dunoon HA [1993] IRLR 496, para. 3, is this rationale for Polkey v Dayton advanced.

Mr Bowers has also drawn our attention in this regard to the decision of the House in W Devis & Sons v Atkins [1977] AC 931, where it was the 'just and equitable' test that was held to warrant the reduction or extinction of compensation for an employee who has been unfairly dismissed and then found to have been liable to summary dismissal. This certainly gives reality to the formula in the context of economic loss. But Mr Bowers has been able to give no example, either from authority or hypothetically, of its use to enhance an award of compensation for economic damage. It is, he has to submit, a purely negative factor. The decision of the House, however, goes further than this. Viscount Dilhorne, with whose speech the other members of the Committee agreed, said of the legislative provision with which we too are concerned :

“The paragraph does not, nor did s. 116 of the Act of 1971, provide that regard should be had only to the loss resulting from the dismissal being unfair. Regard must be had to that, but the award must be just and equitable in all the circumstances, and it cannot be just and equitable that a sum should be awarded in compensation when in fact the employee has suffered no injustice by being dismissed”. (p.955)

Lord Simon of Glaisdale (at 959-960) said that although his first impression had been that “having regard to” governed and therefore limited what was “just and equitable in all the circumstances”, the converse reading preferred by Viscount Dilhorne as the natural one did no great violence to language and obviated injustice. Lord Diplock (at 948) seems to have considered that the ordinary grammatical meaning of the provision had to give way in order to achieve this result. But Lord Edmund-Davies and Lord Fraser of Tullybelton concurred with Viscount Dilhorne without any such reservation.

The reasoning of their Lordships in Devis v Atkins seems to me centrally relevant to the question before this court. It endorses the view, which I would independently have reached, that the governing provision for compensation is that it is to be a sum that is just and equitable in all the circumstances. It establishes in terms that resultant loss is not the only element to which regard is to be had. If that reasoning operates, as it did in Devis v Atkins, to diminish an award below the amount of the economic loss actually suffered, there is in my judgment no reason why in an appropriate case it should not operate to elevate it above the amount of purely economic loss.

This is Mr White's principal argument. His alternative argument is that the word “loss” itself is apt to include more than monetary loss. He needs this argument only if he fails in his principal submission that in s.123(1) loss is not the defining category but a sub-set of the larger category of just and equitable compensation.

For my part I find the meaning of “loss” in the present statutory context problematical. Although, as Mr White suggests, loss can and often does include far more things than money, I agree with Mr Bowers that its more natural meaning in s.123 is pecuniary loss. Yet, accepting this, subsection (2)(b) also requires compensation to cover lost benefits, which are not always monetary - they may consist of in-service training or sabbaticals, for example - and which therefore have to have a round figure put on them like other forms of general damage. It ceases, however, to be problematical if one treats loss, as it seems to me Parliament has treated it, as one constant element in a range of things that it may be just and equitable to award compensation for. And that is Mr White's principal argument.

One initial attraction of Mr Bowers’ contrary submission on the principal argument was that the constraining words which follow – “in consequence of … attributable to…” appear to govern only “loss”, leaving wider aspects of compensation hazardously at large. But the answer to his submission is that it will never be just or equitable to award compensation under s.123(1) for harm which is not caused by or is too remote from the unfair dismissal. For the rest, Mr Bowers has candidly accepted that his construction of s.123 involves reading the words “having regard to the loss sustained” as meaning “having regard solely to the loss sustained”. I can find no warrant for doing this, and very clear syntactical and semantic indicators to the contrary - first and foremost that it is not what the statute says. I respectfully resist Lord Justice Brooke's suggestion that this is mere logic-chopping.

Mr Bowers has argued forcefully that if, nevertheless, Parliament has left the impact of Norton Tool untouched for thirty years, during which time it has amended the employment protection legislation in numerous other respects, it should be taken to have abstained deliberately and by so doing to have endorsed the NIRC's decision. He invites us to adopt the approach to be found in EWP Ltd v Moore [1992] 1 QB 460, 474-5, where Bingham LJ, in a short assenting judgment, noted that the construction favoured, despite the anomaly it produced, by all three members of the court was one which had been the subject of a longstanding decision at first instance and that “despite abundant opportunities Parliament has not acted to cure the anomaly, which cannot have escaped the attention of departmental lawyers and administrators”. This is without doubt a frequently relevant consideration. But it is not a canon of construction. One reason is found in what Bingham LJ said next: “The inference must be either that this apparent anomaly is not regarded as such or that it is regarded as a desirable or tolerable anomaly”.

The reasons why possible, even desirable, legislative measures do not reach the statute book are legion. They include shortage of time in the government's legislative programme, the want of a sponsor who has come high enough in the ballot for private members’ Bills, a lack of departmental time or interest, a political preference for letting sleeping dogs lie, and so on. Desirable amendments may be laid before Parliament and then lost for procedural or political reasons which have nothing to do with their merits. And these hazards do not beset only minor measures. Important legislative proposals which have the weight of the Law Commission behind them not infrequently fail to reach the statute book for one or more of these reasons. Very often the only inference one can draw from Parliament's failure to revise a statute in response to a particular judicial interpretation of it is that it was not thought to be important enough to earn the attention of ministers or whips. Parliament may be omnipotent, but it is not omnicompetent. Many anomalies highlighted by the judicial interpretation of legislation, as Bingham LJ suggests, are just tolerated.

Here, moreover, there has been a positive disincentive to intervene because the continued presence of a financial cap on the amount of the compensatory award has meant that to enlarge compensation beyond the Norton Tool limit would add little to what most claimants recovered. The cap, as Mr Bowers has helpfully shown us, was still only £8,000 in 1985. By 1998 it had risen to £12,000. If pressure was going to be exerted in or through Parliament, it was on the amount of the cap, rather than in the first instance on Norton Tool, that it was likely to be exerted. It was in 1999 (just after Johnson had been decided in this court) that the cap had to be raised in a single jump to £50,000; and it was arguably only then that the recoverability of compensation for more than economic losses became a serious issue. Although we do not have comprehensive figures, Mr Bowers has shown us the table reproduced in Elias, Napier and Wallington Labour Law: cases and materials (1980), p.630: it indicates that in 1977, when the cap was set at £4,300, more than 95 per cent of awards were under £2000. Only about 4.5 per cent of awards reached or approached the limit. This does indicate a shortfall in awards which might have been made good by a successful challenge to Norton Tool; but it also suggests that the necessary investment in such litigation (which would almost certainly have had to come from one or more trade unions) was unlikely to be justified by the potential return.

However, Mr Bowers points out that ready-made vehicles for the reversal of Norton Tool have actually passed through Parliament in the intervening years: any of the Employment Bills which became law in 1980, 1982, 1993, 1998 and 1999 (or at least those which were not pure consolidating measures) could have been used to reverse the NIRC's decision. This gives Mr White the opportunity to point out that the Bill which became the Employment Act 2002 could just as readily have been used to overset (or at least lay the ghost of) Johnson v Unisys, which by then was being followed by some tribunals but not by others.

For my part, I would be prepared to give some interpretative weight to Parliament's non-intervention if I considered that the meaning of the words enacted in 1971 and at intervals thereafter was doubtful. But I do not consider it doubtful; and I do not consider that Parliament, by letting an erroneous interpretation of its words stand uncorrected, can give such error the force of law.

A number of the arguments addressed to us have been arguments from consequences. One is that employment tribunals are unequipped to adjudicate on contests of psychiatric evidence. Another is that they lack the procedural equipment to deal with conflicting expert testimony. These can at best be secondary arguments, for if Parliament has given tribunals jurisdiction to entertain particular matters, it is their obligation in appropriate cases to do so. But I do not consider the arguments to have substance in any case.

As to the evaluation of expert evidence, a striking feature of this case is that the parties agree about two things: first, that if s.123 takes tribunals beyond economic loss, it necessarily takes them both into aspects of distress falling short of clinical damage and into clinical forms of mental disturbance; and secondly, that even if s.123 has the limited ambit for which the respondent contends, it still includes the proximate economic consequences of psychological harm - for instance the cost of counselling to enable a badly distressed employee to get back to work. The latter being so, the evaluation of conflicting psychiatric or similar evidence is on any view within the jurisdiction of employment tribunals. If so, the terrors implicit in the former begin to fade. They fade further when one calls to mind that tribunals, which of course have always had legally qualified chairmen, have for some time now possessed full case management powers: see the Employment Tribunals (Constitution and Rules of Procedure) Regulations 2001. There is no more reason why they should be overwhelmed by the occasional conflict of expert witnesses than circuit or district judges are.

In considering what Parliament intended in 1971, it is useful to turn to the officious bystander's Parliamentary analogue, the industrious backbencher. If that not wholly imaginary individual had asked the minister in committee what clause 116 would do to help an employee who had been appallingly treated and finally driven out of his job with his self-confidence in tatters, but who had speedily secured himself an equally well-paid job, I doubt whether the minister would have said “Nothing”. I think it much likelier that he would have said that the industrial tribunal would be able to award the employee a round sum which was just and equitable in all the circumstances. He might have added that this would be of particular value to some victims of unfair industrial practices, to which the section 116 formula (as Lord Justice Brooke points out) was also to apply.

If the backbencher had followed up by asking whether victims of other unfair industrial practices (see paragraph 11 ante) were going to have to prove pecuniary loss in order to recover compensation, I have little doubt that he would have received the same answer. Both the backbencher and the minister would have been well aware of the then recent case of Edwards v SOGAT [1971] Ch. 354, which had been heard and decided by this court in July 1970 and had contributed to the pressure for legislation. Mr Edwards, who was black, was a skilled worker in a 100-per-cent union printshop. His employers were compelled to dismiss him after his dues had been allowed to fall into arrear through a union official’s neglect. He had to sue the union in contract. He won his action before Buckley J, but the union appealed to this court on the quantum of damages. The union had at a late stage agreed to readmit him, but it was too late to put him back in his original job, and he had meanwhile found and then lost another job. Lord Denning MR concluded (at 378):

“I feel that damages in a case such as this are so difficult to assess that I would be inclined to view them somewhat broadly. I would start with the loss of earnings which he might reasonably be expected to have suffered over two years from his expulsion. That is what was suggested by Lord Donovan’s Committee. I would then work upwards or downwards from that figure, according to the circumstances of the case.”

Sachs LJ said (at 378-9):

“The union’s liability in damages being clear, this appeal is concerned with their measure – an important matter in the particular circumstances. These damages, of course, sound in contract and not in tort. It is, however, as well to record at the outset … that certain rules laid down in Addis v Gramophone Co Ltd [1909] AC 488 touching damages for wrongful dismissal have no application to the present type of case. In other words, whereas in the former class of cases the damages can contain no element for the difficulty the dismissal causes to a plaintiff in getting fresh employment, the essence of the measure in the present case is an assessment of the financial consequences of that very difficulty.”

Megaw LJ, who preferred to segregate past loss from future loss, said (at 387) of the element of future loss:

“Where there are so many incalculables, it would not be right to seek to give an aura of scientific respectability to the assessment of future damages by purporting to apply arithmetical or actuarial formulae to the assessment, or to any individual factor on which the assessment partly depends. One must try to assess. One cannot calculate.”

Of course future loss of this kind is still financial loss and not therefore the kind of damage with which this appeal is concerned. But the importance of Edwards v SOGAT (which was not cited to their Lordships in Johnson v Unisys Ltd) is that it enlarged the Addis doctrine in relation to what was shortly to become by statute an unfair industrial practice, and that it forms part of the background to the 1971 Act.

In supplementary written submissions invited by the court on related statutory provisions such as those I have just been considering, Mr Bowers has pointed out that the Employment Protection Act 1975, while it effectively reproduced the old s.116 formula in relation to unfair dismissal, replaced the old s.5 rights of trade union members with a separate right (s.53) and a remedy expressed in s.56 in this way:

“… such amount as the tribunal considers just and equitable in all the circumstances having regard to the infringement of the complainant’s right under section 53 above by the employer’s action complained of and to any loss sustained by the complainant which is attributable to that action.”

The words which I have italicised, it is submitted, demonstrate Parliament’s own view, in 1975, that its enactment in 1971 of the remaining words had been restricted pecuniary loss.

There is a fundamental problem about this submission. It is not Parliament but the courts who have the constitutional function of interpreting the words on the statute book. The courts do so, of course, with full and careful regard to all the indicia provided by the legislation itself as to what is meant by the words Parliament has used. But it is Parliament which, by strong constitutional convention, takes from the courts what its legislation means and – if it wishes – amends its legislation accordingly. It may very well be, therefore, that s.56 was enacted in response to Norton Tool, consciously enlarging what had been held to be a general limit on the computation of awards. But such a legislative response does not operate as a Parliamentary adoption or endorsement of the NIRC’s original decision, and Mr Bowers has not argued that it does. He argues that it tells us what Parliament meant when it enacted s.116 in 1971. That is an argument which, on grounds of constitutional principle, I consider to be unsustainable.

“Since interpretation of legislation is a matter for the courts, they are not bound by an expression of Parliament’s opinion as to what the law is, whether expressed in or to be inferred from a statute, as distinct from a positive enactment manifesting an intention to declare the law: 44(1) Halsbury Laws of England (4^th ed. Reissue), para. 1236, fin.”

What the legislative history does illustrate, however, is the extent to which Norton Tool has been regarded as settled law. In this respect I entirely share Lord Justice Brooke’s concerns. But, accepting his view and that of Mr Justice Evans-Lombe that paragraph 55 of Johnson v Unisys Ltd does not bindingly overrule Norton Tool, it is in my opinion not open to this court, when called upon to decide for itself whether Norton Tool is good law, to defer to the decision of a lower court which, though venerable and respected, it considers to be wrong. We can neither refuse to decide the point nor decide it on grounds of convenience.

To hold this is not to hold that every upset caused by an unfair dismissal carries a compensatory award. The power is there to permit tribunals to compensate an employee for a real injury to his or her self-respect. It is likely to become material principally in cases of constructive dismissal where the employee has been driven from his or her job. For the ordinary case of unfair dismissal, assuming that there is no reinstatement or re-engagement, it is the basic award which is there to compensate for the unfairness.

For these reasons, in my judgment, the statutory formula now found in s.123(1) and (2) embraces non-pecuniary losses caused (in the sense associated with the ordinary principles of remoteness) by an unfair dismissal. To the extent that Norton Tool holds otherwise I consider it to have been wrongly decided.

The EAT considered that the employment tribunal's award for the distress suffered by Mr Dunnachie in consequence of the treatment which constituted his constructive dismissal was insufficiently explained. They did not hold that it was in itself excessive. Nevertheless Mr Bowers has, without objection, sought to sustain the EAT's decision on the latter ground.

The employment tribunal gave a separate set of extended reasons for their award. They summarised the basis of the award for non-economic loss which, pursuant to their understanding of Johnson v Unisys, they proposed to make:

4.

As appears from our decision on the merits, the conduct of the respondent’s employee Mr Kitching and the poor management of his superiors effectively drove the applicant from his employment with the respondent. In order to protect his economic interests, the applicant obtained alternative work with Doncaster Metropolitan Borough Council. He began work with that authority on 10 April 2001.

5.

The conduct of the respondent’s employees seriously undermined the applicant’s health so that he was off work with stress for about three weeks in November 2000. He had, immediately before that date, been driven by the circumstances in which the respondent had put him to lose his self control in front of his boss.

6.

As a result of losing his job with the respondent, the applicant has lost the congeniality of working in the city of his birth where his family live. He now has to endure an extra 64 miles per day in travel to work. That impacts on time with his family.

7.

Furthermore, the treatment of the applicant by the respondent seriously undermined his self confidence and self esteem so that he has been receiving professional help from a counsellor. He is now better.

It should be added that the loss of self-control mentioned in this passage was not a passing tantrum. It was a humiliating collapse marking the culmination of a prolonged campaign of denigration and harassment which the respondent's management had compounded by inertia. The full fact findings of the employment tribunal had been set out in 34 carefully written paragraphs of their decision on liability, and their decision on compensation must be read together with this.

I am satisfied for my part that employment tribunal's award of compensation for non-economic loss sits at an appropriate point on the scale which has now been given the authority of this court in Vento v Chief Constable of West Yorkshire Police [2003] ICR 318. £10,000 is the centre point of the middle band allocated by Mummery LJ to cases which are serious but not the most serious. That, it seems to me, fairly represents what the tribunal found to have happened, not least because it was aggravated by the respondent's persistence in continuing aggressively to blame the applicant for his own misfortune in the tribunal proceedings.

I am also satisfied that the employment tribunal gave an adequate and proper explanation of its award. The EAT considered that it had not done so. Burton P said in paragraph 52 that there was “no, or no sufficient, consideration of causation and no explanation or reasoning as to how the sum is arrived at”.

The tribunal reminded itself that Lord Hoffmann in Johnson had said that he saw no reason why in an appropriate case compensation should not reflect distress, humiliation, damage to reputation in the community or to family life and psychiatric damage caused by the manner of dismissal. They reminded themselves of the scale proposed in Tchoula v ICTS (UK) Ltd [2000] IRLR 643, the precursor of Vento, which set £10,000 as the top of the lower bracket of discrimination awards. This was the amount for which the applicant had asked and was in their view a reasonable sum:

“For the applicant has, as a result of the respondent's actions, suffered some loss of professional status. He was humiliated and distressed by the manner of his dismissal. That process lasted over several months. The consequences have clearly affected his family life as we have described earlier. If anything, the applicant's claim is modest in this regard and we have no difficulty in identifying it as being properly at the upper end of the lower scale in Tchoula” (para.18).

I accept readily that in most legal decisions it is necessary to go from the facts to the conclusion by way of an explanation of how the former leads to the latter. But when damages or compensation at large are being assessed, a recital of or a reference to the facts which justify the award, accompanied by any necessary indication of the scale of quantification which is being used, is ordinarily a sufficient explanation of the award itself. It enables the parties to know, so far as it can be expressed, why the award is what it is; and it enables an appellate tribunal to say whether it was a permissible award or not.

Compensation for non-economic loss brought about by the manner of an unfair dismissal (we have not been asked to decide whether it also embraces the fact of such a dismissal) is on authority and on principle recoverable. The award of such compensation by the employment tribunal in the present case was not excessive and was adequately explained. I would accordingly allow the appeal and restore the award.

Given the division of opinion both within this court as to the effect of Johnson v Unisys Ltd and between this court and the EAT as to the sustainability of Norton Tool, as well as the impact of these vagaries of authority upon the work of employment tribunals, litigants and advisers, we propose, if asked and if not dissuaded, to grant Hull City Council leave to appeal to the House of Lords in order that a definitive answer may be given to what is ultimately a short question of statutory construction. In the meantime employment tribunals should manage, list and decide cases in the knowledge that the last word has not been said, but is going to be said in the foreseeable future, on this topic.

In this appeal I have had the advantage of reading in draft the judgments of Lord Justice Brooke and Lord Justice Sedley who disagree on the main point for decision, namely, whether the provisions of section 123 of the Employment Rights Act 1996 which substantially re-enacts of the provisions of section 116(1) of the Industrial Relations Act 1971 in which they first appeared, contain a power to award compensation for harm, in the section referred to as “loss”, sounding in general, as opposed to special, i.e. pecuniary damage. A good example of such general damage would be compensation for psychiatric illness resulting from the circumstances of the dismissal for which the employer was responsible but in respect of which no identifiable pecuniary loss had resulted. Thus the “loss” would be such harm as susceptibility to depressive illness resulting in injury to future employment prospects. The facts of the present case are another example.

In Norton Tool Company Ltd v Tewson [1972] ICR 501, Sir John Donaldson, sitting in the National Industrial Relations Court, decided that such “general” damage, not quantifiable as pecuniary loss was not recoverable in the employment tribunals under section 116(1) of the 1971 Act. That decision, though not binding on this court, has passed unchallenged until the decision of the House of Lords in Johnson v Unisys Ltd [2003] 1 AC 518. In that case an employee, who had a past history of mental illness, sought to recover in proceedings at common law damages for wrongful dismissal including claims for such general damage. His claim was struck out at first instance as disclosing no reasonable cause of action and that decision was upheld in both the Court of Appeal and the House of Lords on the basis that a claim for damages for wrongful dismissal at common law was inconsistent with the provision by Parliament of a statutory remedy covering the same ground. At paragraph 55 of the report in the record of the speech of Lord Hoffmann, the following passage appears:-

“55

In my opinion, all the matters of which Mr Johnson complains in these proceedings were within the jurisdiction of the Industrial Tribunal. His most substantial complaint is of financial loss flowing from his psychiatric injury which he said was a consequence of the unfair manner of his dismissal. Such loss is a consequence of the dismissal which may form the subject matter of a compensatory award. The only doubtful question is whether it would have been open to the tribunal to include a sum by way of compensation for his distress, damage to family life and similar matters. As the award, even reduced by 25% exceeded the statutory maximum and had to be reduced to £11,000, the point would have been academic. But perhaps I may be allowed to comment all the same. I know that in the early days of the National Industrial Relations Court it was laid down that only financial loss could be compensated: see Norton Tool…. It was said that the word “loss” can only mean financial loss. But I think that is too narrow a construction. The emphasis is upon the tribunal awarding such compensation as it thinks just and equitable. So I see no reason why in an appropriate case it should not include compensation for distress, humiliation, damage to reputation in the community or to family life.”

In the present case the Employment Tribunal awarded £10,000 to Mr Dunnachie to compensate him for such general damage arising from the circumstances of his dismissal. That award was disallowed by the Employment Appeal Tribunal.

Before this court it was first argued that the passage in the speech of Lord Hoffmann which I have set out above formed part of the ratio of his decision binding upon us, because two other members of the committee agreed with him, Lord Bingham and Lord Millett, and so the question is decided in favour of the employee in this case. I respectfully agree with the conclusion of Lord Justice Brooke, for the reasons which he sets out in his judgment, that what Lord Hoffmann says in the passage which I have set out above was obiter and is not binding on this court.

It was further argued on behalf of the appellant that, notwithstanding that this court might conclude that it was not bound by what Lord Hoffmann said in Johnson’s case, it should nonetheless conclude, that the Norton Tool case was wrongly decided, notwithstanding that it has stood as authority on the construction of section 123(1) of the 1996 Act and its predecessors down to today.

In my judgment, if the matter was free of authority, the words of sub-section (1) are sufficiently widely drawn so as to permit the tribunal to award compensation for any loss which can be shown to directly flow from an unfair dismissal including compensation for “general damage”. With respect to those who think otherwise there does not seem to me to be any justification for construing the word “loss” as referring only to pecuniary or economic loss. In my view the construction I favour is supported by the use of the word “include” in the first line of sub-section (2) where recoverable loss is expressed to “include” certain types of pecuniary or economic loss set out in sub-section (a) and (b), those words indicating that the extent of the types of cases in which compensation was recoverable were not confined to those described in sub-section (2). In the above quoted passage from the speech of Lord Hoffmann in Johnson’s case he says “it was said that the word “loss” can only mean financial loss. But I think that is too narrow a construction”.

I accept that for the purpose of construing section 123 of the 1996 Act it is necessary to attempt to discern what the intention of Parliament was when it enacted section 116 of the 1971 Act, which created the remedy for “unfair dismissal” thereby breaching the dam on recovery of damage for breach of contract of which the decision of the House of Lords in Addis v GramophoneCo Ltd [1911]AC 488was a principal building block.

In his speech in Johnson’s case Lord Millett at page 548 of the report conducted an overview of the relevant legislation. He said this:-

“The 1971 Act did not expressly provide that the jurisdiction of the Industrial Tribunals was exclusive, but it did not need to. It was clearly predicated on the existing state of the law established in Addis’ case… and confirmed in Mallochs’ case… There would have been no point (for example) in excluding certain categories of employee from obtaining compensation for unfair dismissal if they could obtain a remedy by way of damages at common law… . Prior to 1996 therefore the jurisdiction of the Industrial Tribunals to award compensation for unfair dismissal was exclusive in practice, not because it was made so by statute, but because it was premised on the absence of a corresponding remedy at common law.

But the common law does not stand still. It is in a state of continuous judicial development in order to reflect the changed perceptions of the community. Contracts of employment are no longer regarded as purely commercial contracts entered into between free and equal agents. It is generally recognised today that “work is one of the defining feature of peoples lives”; that “loss of ones job is always a traumatic event”; and that it can be “especially devastating” when dismissal is accompanied by bad faith:… . This change of perception is of course partly due to the creation by Parliament of the statutory right not to be unfairly dismissed.”

Lord Millett continues at page 550:-

“But the creation of the statutory right has made any such development of the common law [the development of a parallel common law cause of action] both unnecessary and undesirable. In the great majority of cases the new common law right would merely replicate the statutory right; and it is obviously unnecessary to imply a term into a contract to give one of the contracting parties a remedy which he already has without it. In other cases, where the common law would be giving a remedy in excess of the statutory limits or to excluded categories of employees, it would be inconsistent with the declared policy of Parliament. In all cases it would allow claims to be entertained by the ordinary courts when it was the policy of Parliament that they should be heard by specialist tribunals with members drawn from both sides of industry.”

For my part I am unable to discern any reason why the legislature, in 1971, should have intended to exclude from recovery under the new remedy they were creating, compensation for any type of loss which could be shown to flow from unfair dismissal. The scheme of the legislation was, in the words of Lord Hoffmann at page 543 of the report of the Johnson case to adopt “the practical solution of giving the tribunals a very broad jurisdiction to award what they consider just and equitable but subject to a limit on amount.” There were also, of course, limits on the sorts of employee who could claim such compensation. With respect it does not seem to me that anything in the contents of the Royal Commission on Trade Unions and Employers Associations summarised in my Lord Justice Brooke’s judgment, points against this conclusion.

The exclusive jurisdiction of the employment tribunals was made express by the introduction of section 205 of the 1996 Act. It is accepted that the effect of that section taken with the decision in Johnson’s case means, if the narrow construction of the word “loss” in section 123(1) contended for by the employers is correct, that there is no remedy in respect of general damage as a result of personal injury if caused by the circumstances of an unfair dismissal, which, if it had been negligently caused would have been recoverable. I cannot believe that that result is one that the legislature in 1971 would have contemplated.

In supplementary written submissions invited by the court our attention has been drawn by counsel to sections 101 – 108 of the 1971 Act and thus to the provisions for compensation in respect of breaches of other requirements of the 1971 Act, apart from section 22, and which were, by that Act, referred for adjudication and assessment to the Industrial Tribunals and the Industrial Court as an aid to the construction of section 22 of the 1971 Act and so to the construction of section 123 of the 1996 Act. It was submitted “that it cannot have been the intention of Parliament to allow the possibility of an award of injury to feelings, not only for unfair dismissal, but where for example collective agreements have been broken, or where an employer has negotiated with a body other than a recognised trades union or joint negotiating panel (section 55), or if an employee induced a breach of contract.” It was pointed out that it was difficult to see how a trades union or employers association can suffer injury to feelings.

It does not seem to me that the fact that compensation in respect of other breaches of the requirements of the 1971 Act, where it was impossible or highly unlikely that breach of those requirements would have resulted in non-pecuniary loss to the complainant, should lead to the conclusion that section 116, and thus section 123 should be construed so as to exclude recovery of such loss where it has actually been suffered and the words used by the relevant section (section 116/123) are wide enough to include it. It is to be observed that a number of the breaches of the requirements of the 1971 Act referred to the Tribunals were capable of resulting in non-pecuniary loss, for example, unfair and unreasonable disciplinary action contrary to section 66 and 70 and trade union victimisation contrary to section 5.

The supplementary written submissions included a further submission on behalf of the respondents arising from the provisions of the Employment Protection Act 1975. My Lord, Lord Justice Sedley, deals with this submission at paragraph 45 and following in his judgment with which I respectfully agree. I would add that if it had been the intention of Parliament in enacting section 56 of the 1975 Act so as deliberately to import into that section a power to award what I have described as general damage for breach of the requirements of section 53 in recognition of a perceived inability to award such compensation for unfair dismissal under section 76 of the same act (an intermediate predecessor of section 123 in substantially similar terms) one would have expected Parliament to do so in much plainer terms than those used in section 56. I am unable to discern what policy considerations can have led to a Parliamentary intention to give a right to award general damage in respect of breaches of the requirements of section 53 while not doing so in respect of awards of compensation for unfair dismissal.

I respectfully agree with the conclusions of Lord Justice Sedley on this part of the case, and for the reasons which he gives, together with those which I have sought to express, I conclude that the Norton Tool case must be treated as wrongly decided and that section 123(1) should be construed as empowering employment tribunals to award compensation for non-pecuniary damage flowing from the circumstances of unfair dismissal.

It was submitted that such a construction of section 123 would open the floodgates to a tide of claims for the recovery of such compensation. I see no reason why this should happen provided that the employment tribunals are firm in requiring proof that an applicant has been occasioned any such loss.

As to the issues in respect of the quantum of the award on which my Lords are agreed I also agree and have nothing to add.

I would allow this appeal.

The work of employment tribunals administering the law relating to compensation for unfair dismissal was plunged into chaos by paragraph 55 of Lord Hoffmann’s speech in Johnson v Unisys. The scale of the chaos is vividly described in paragraphs 1 and 2 of the EAT’s judgment now under appeal. After explaining how compensation claims for what could generically be described as non-economic loss had mushroomed since the Johnson decision, Burton J said:

“This latter head of loss has, in the course of the hearing before us, been exemplified or described by reference to different cases in a number of different ways, and under a number of different headings. These include physical injury, recognised psychiatric illness (including mental illness, neurosis and personality change), 'non-psychiatric illness' and (in alphabetical order) anger, anguish, anxiety, damage to family life, damage to reputation, depression, disappointment, frustration, grief, humiliation, hurt, impact on family life, inconvenience, injury to feelings, loss of congenial employment, loss of self-confidence, loss of esteem, mental distress, mental suffering, stress, upset and worry.

Some tribunals have awarded such compensation, concluding that they are entitled, or obliged, to do so, since Johnson. Some have declined to do so, concluding that they were not bound by Johnson to do so and had no jurisdiction to do so. Examples of awards in respect of such non-economic loss drawn from the cases before us, and from other decisions in the Employment Tribunals known to or discovered by Counsel before us, have ranged between £250 and £10,000.”

We were led to understand that in one region the chaos was such that a regional chairman of employment tribunals decided sensibly, but with doubtful authority, to instruct all the tribunals in his region to follow a common line until such time as the uncertainties were straightened out by a higher court. Since legal aid is not available in this class of litigation, costs are generally irrecoverable, and employment law advisers very often operate on restricted financial resources, it is most unfortunate that so much effort has had to be expended throughout the country over the last three years in the task of determining whether this passage of Lord Hoffmann’s speech is binding, and if it is not, whether it should be followed at all.

In the very long paragraphs 44-47 of the EAT’s judgment, which cover no fewer than ten A4 pages, Burton J and his very experienced colleagues described the extent of the massive practical problems thrown up by this paragraph of Lord Hoffmann’s speech. If these matters had been fully ventilated in the courts before the speeches in Johnson were released, and if the House of Lords had showed clearly that they had borne all these considerations in mind and were still determined to give a quite novel interpretation to very familiar statutory language, then of course employment tribunals and the lower courts would have loyally followed their decision, however troublesome the practical consequences. But the point was never argued at all. It would be very odd if the doctrine of stare decisis compelled us to follow Lord Hoffmann unquestioningly in circumstances as peculiar as these.

The problem arose in this way. When Johnson was argued in the Court of Appeal (see the report at [1999] ICR 809) Mr Patrick Elias QC appeared for the employers. Although he argued that a newly invented implied duty not to dismiss an employee unfairly would be inconsistent with section 205 of the Employment Rights Act 1996 (“the 1996 Act”), which required a claim for unfair dismissal to be made by way of complaint to an employment tribunal “and not otherwise”, it does not appear from the judgment of Lord Woolf CJ in that case that the point loomed very large in the Court of Appeal. Norton Tool was not cited, and the idea that that case was wrongly decided formed no part of Mr Elias’s argument.

Mr David Pannick QC appeared for the employers in the House of Lords. The contention that the concurrent existence of a statutory remedy must stifle any thought of creating a possible new duty at common law loomed very much larger in the argument at that level, and this contention ultimately prevailed. But Mr Pannick never suggested that Norton Tool was wrongly decided. As Lord Steyn correctly said (at para 11 of his speech in Johnson), Mr Johnson’s claim was concerned solely with the recovery of special damages for financial loss. The quite separate question whether an employee might recover compensation for anxiety and mental stress arising from the manner of his dismissal was never raised before Judge Ansell or the Court of Appeal and was not an issue before the House of Lords. Indeed, during the course of his argument Mr Pannick suggested that it might be a good idea if tribunals were empowered to make such awards. It does not seem to have occurred even to that very distinguished advocate that in fact tribunals had possessed the requisite power for nearly 30 years, but nobody had ever realised this.

Against this unprepossessing background, three principal matters arise for our decision on this appeal:

(i)

Are we bound by paragraph 55 of Lord Hoffmann’s speech in Johnson?

(ii)

If we are not so bound, should we nevertheless follow it?

(iii)

If the answer to (i) or (ii) is “yes”, should we restore the Employment Tribunal’s award of £10,000 compensation?

My answer to question (i) is unhesitatingly “no”. The point was never argued before the House of Lords and it was not necessary for the decision of the majority of the House. They were convinced by the arguments (i) that the duty of mutual trust and confidence could not be called in aid at a time when the employer wished to end the employment relationship; and (ii) that the existence of a statutory remedy precluded them from importing a quite different implied duty into an employment contract, breach of which would sound in damages at common law. There is no sign in the assenting speeches of Lord Nicholls and Lord Millettt that they were concerned with any issues relating to the adequacy of the statutory remedy as an instrument of complete justice. It was sufficient for them that it existed, and they were unattracted by the idea of identifying a new common law right covering the same ground as the statutory right not to be unfairly dismissed. As Lord Millettt observed (at para 80), all coherence in our law would be lost if it did.

If authority is needed for the proposition that we are not bound by paragraph 55 of Lord Hoffmann’s speech in Johnson, even if the point had ever been argued, I am content to refer to a passage in the judgment of Schreiner JA in the South African case of Pretoria City Council v Levison (1949 (3) SA 305,317):

“As I understand the ordinary usage in this connection, where a single judgment is in question, the reasons given in that judgment, properly interpreted, do constitute the ratio decidendi, originating or following a legal rule provided (a) that they do not appear from the judgment itself to have been merely subsidiary reasons for following the main principle or principles, (b) that they were not merely a course of reasoning on the facts (cf Tidy v Battman (1934) LJKB 158 at p 162) and (c) (this may cover (a)) that they were necessary for the decision, not in the sense that it could not have been reached along other lines, but in the sense that along the lines actually followed the result would have been different but for the reasons.”

Mr White, for his part, referred us to the well-known speech of Lord Simonds in Jacobs v London County Council [1950] AC 361, 369. Lord Simonds was there addressing the problem that arises when several issues are raised in a case and a determination of one of them is decisive in favour of one or other of the parties. What, in these circumstances, is the binding effect of the court’s determination on the other issues? It is not in my judgment necessary to refer any further to this speech since Lord Hoffmann’s observations in paragraph 55 of his speech in Johnson did not determine an issue raised by anyone. In a different context it might have been said that he was on a frolic of his own.

I turn therefore to the second main question we have to answer. If we are not bound by paragraph 55, should we nevertheless follow it? The search for an answer to this question compels us to ask: What do the words of section 123(1) of the 1996 Act (as amended) mean? According to what principles should we construe them?

The 1996 Act was in essence a consolidation Act, following upon the earlier consolidation of this part of our statute law in 1978. It brought together in a convenient way a number of disparate statutory provisions relating to employment rights. For all practical purposes the critical language of section 123(1) is the same as the language of section 116(1) of the Industrial Relations Act 1971 (which, where it differs, I have placed in square brackets as opposed to italics):

“…the amount of the compensatory award [compensation] shall be such amount as the [court or] tribunal considers just and equitable in all the circumstances, having regard to the loss sustained by the complainant [aggrieved party] in consequence of the dismissal [matters to which the complaint relates] in so far as that loss is [was] attributable to action taken by the employer [by or on behalf of the party in default].”

The critical words we have to interpret therefore represent “straight consolidation”, which “consists of reproduction of the original wording without significant change” (FAR Bennion, Statutory Interpretation (4^th Edition, 2002), p 516). Nobody suggested to us that the original wording had changed its meaning since 1971 because of its proximity to the different statutory provisions which now surround it as a consequence of consolidating the law. We must apply the rule, or presumption, that this provision must be construed as if it had remained in the original Act. In Galloway v Galloway [1956] AC 299, 320 Viscount Radcliffe said:

“I must confess that I do not lend a sympathetic ear to this last and almost mystical method of discovering the law, least of all when it depends upon a consolidating Act the function of which is to repeat, but not to amend, existing statute law.”

Because this is a consolidation Act we do not have to explore the ramifications of the principles of construction explained by the House of Lords in R v Chard [1984] AC 279, 294-5 per Lords Scarman, Roskill and Templeman, and in B (A Minor) v DPP [2000] AC 428, 465-6 and 473-4, which are concerned with a different problem of statutory interpretation.

On first principles, therefore, we must go back to 1971 to ascertain what Parliament meant when it created for the first time a statutory right to compensation for unfair dismissal and directed industrial tribunals to award as compensation

“… such amount as the … tribunal considers just and equitable in all the circumstances, having regard to the loss sustained by the aggrieved party in consequence of the matters to which the complaint relates, in so far as that loss was attributable to action taken by or on behalf of the party in default.”

In order to identify the mischief which these statutory provisions were enacted to avoid (but not to interpret the remedy provided by Parliament to defeat that mischief: see Black-Clawson International Ltd v Papierwerke Waldhof-Aschaffenburg AG [1975] AC 591) we can go back to the report of the Royal Commission on Trade Unions and Employers’ Associations (1968), Cmnd 3623.

Chapter IX of this report addressed these matters. The Commission described how under the present law an employer was legally entitled to dismiss an employee whenever he wished and for whatever reason, provided only that he gave due notice. At common law he did not even have to reveal his reason, much less just justify it (para 521). Even if an employee was dismissed without due notice, his remedy was limited to a claim for the net amount of the wages he had lost as a result of the wrongful dismissal. He had no other legal claim at common law, whatever hardship he suffered as a result of the dismissal. In this context the Commission referred (at para 522) to the decision of the House of Lords in Addis v Gramophone Co Ltd [1909] AC 488.

After referring to some of the arrangements that had recently been made to mitigate some of these hardships, the Commission reported (at para 525) that there was a very general feeling, shared by employers as well as trade unions, that the present situation was unsatisfactory. The Commission said it shared this belief in full (para 521). It described how people built much of their lives around their jobs, and how for workers in many situations dismissal was a disaster. It ended this section of its report by saying (at para 529) that it believed that it was urgently necessary to give workers better protection against dismissal. In spite of a change of Government and all the political acrimony that accompanied the public discussion of the Commission’s report, the statutory language we have to interpret unquestionably represents Parliament’s choice of remedy for this mischief.

As a matter of history our law recognised two ways of providing financial remedies for wrongs done in the field of employment relations. Both contractual or fiduciary duties might be in question. The common law courts awarded damages as compensation for the breach of a contractual duty. The courts of equity awarded equitable compensation for the breach of a fiduciary duty. It is therefore hardly surprising that we find in this new scheme for statutory compensation, covering not only unfair dismissal but also unfair industrial practices and unfair treatment by a registered trade union or employers’ association (see para 97 below) that the words “just” and “equitable” both govern the tribunal’s consideration of the amount of compensation it should award in “all the circumstances”.

Parliament also directed tribunals about the matter to which it should have regard when they decided the amount of compensation it considered just and equitable. This was not to be palm-tree justice. Tribunals were to have regard to the loss sustained by the aggrieved party in consequence of the matter to which his complaint related. Of course in a context different from that of industrial relations legislation in 1971 the word “loss” might properly be interpreted to include loss of face, loss of confidence, loss of reputation, loss of health, loss of peace of mind, loss of dignity, loss of amenity etc etc, but I find it inconceivable that in this particular context Parliament intended the word to mean anything other than financial loss. And it is striking that nobody ever thought it worth appealing Norton Tool to the Court of Appeal in quest of a wider meaning. While in the context of an unfair dismissal an individual employee might not have thought the game worth the candle in the presence of the statutory cap to compensation, different considerations would apply if one of our great trade unions and their distinguished advisers had ever thought the point worth arguing.

I regard as logic-chopping the suggestion that because the word “solely” does not appear in the sub-section, Parliament intended industrial tribunals to range more widely and have regard to other considerations it did not condescend to mention. It should be remembered that in 1971 these tribunals were in their infancy. This Act marked the first major enlargement of their jurisdiction. In later years, as the EAT’s judgment showed, Parliament was willing to entrust them with steadily more extensive powers. But these were very early days. They were not even being entrusted with the resolution of very low value claims for wrongful dismissal in breach of contract, though the 1971 Act contained empowering provisions which could be triggered off when the time was considered ripe for such an extension: even then they had to keep well clear of personal injury claims, however much bound up with the subject matter of the complaint (see section 113).

It is tempting to interpret an Act passed in 1971 through the eyes of 2003, but the temptation must be resisted. In 1971 the financial limits of a county court judge’s jurisdiction were much more restricted than they are today. Damages for loss of reputation may still be sought only in the High Court 32 years later. Compensation for psychiatric injury still bristles with difficulties: very significant difficulties relating to causation have yet to be resolved. The seed of an implied contractual duty of mutual trust and confidence was not even germinating. In claims based on the breach of a duty of care, damages could only be awarded for psychiatric injury, anxiety or distress if linked with a claim for damages for physical injury (except in the sluggishly developing corner of the law concerned with the “nervous shock” cases). In claims based on breach of contract, Addis v Gramophone Co Ltd dominated the law, although a slim line of authority permitting the award of such damages following the breach of a particular type of contract was starting to emerge. Equity, for its part, concerned itself only with financial loss, and not with any injury to feelings or psychiatric injury consequent on the breach of a fiduciary duty.

In these circumstances, unless impressed by very clear language, I cannot believe that Parliament in 1971 intended to give these new tribunals the sweeping powers suggested by Lord Hoffmann in paragraph 55 of his speech in Johnson v Unisys. In 1971 they would have been unfitted for the task. The results would have been chaotic. In my judgment the EAT in the present case was wholly justified in maintaining the authority of the Norton Tool case for the reasons it gave.

Sedley LJ has asked what answer an industrious backbencher would have received to the question he poses in paragraph 43 of his judgment. He would have received the reply that clause 116 provided a comprehensive compensation code not only for complaints of unfair dismissal, but also for all the other matters for which an industrial tribunal was now being empowered to award compensation on a complaint under Part VI of the 1971 Act. It would have been explained to him that the same code had been devised to provide remedies for:

(i)

A breach of a right to be a member of a trade union, or not to be a member, or if a member, to take part in trade union activities at an appropriate time (s 5);

(ii)

The application of pressure by an employer to anticipate the result of a ballot in respect of a closed shop by a lock out or the threat of a lock out or to induce a trade union to refrain from making an application for a ballot for a closed shop (s 16);

(iii)

A breach by a union or an employer to break a legally enforceable collective agreement (s 36);

(iv)

The instigation or threatened instigation of industrial action pending questions as to the recognition by the employer of a sole bargaining agent (s 54);

(v)

Engaging in collective bargaining with a union other than a recognised union in the context of collective bargaining procedures regarding recognition of a union (s 55);

(vi)

A breach by an organisation of workers of the guiding principles set out in section 65 (s 66);

(vii)

A breach by an organisation of employers of the guiding principles set out in section 69 (s 70);

(viii)

Inducing or threatening to induce a breach of contract in contemplation or furtherance of a trade dispute, except when done by a trade union or employers’ association (s 96);

(ix)

Taking industrial action in support of an unfair industrial practice (s 97);

(x)

Taking industrial action against parties extraneous to a dispute (s 98).

He would have been told that the Government did not intend for the present to give industrial tribunals the duty of deciding difficult questions of law relating to the recoverability of general damages (not attributable to ascertainable financial loss) in relation to this very wide variety of different situations. The law on some of these matters was difficult enough for judges in the mainstream courts, and he would have been told that the Government judged it premature to give the new tribunals even the task of deciding whether damages should be payable for wrongful dismissal (indeed, they were not considered suitable for this task for a further 20 years).

He would also have been told that people aggrieved in all these fields might well suffer in some or more of the ways generically described by the EAT as “non-economic loss” in paragraph 1 of its judgment (see paragraph 75 above), but the Government had decided as a matter of policy that compensation for all these complaints should be limited to provable financial loss and not include something akin to a solatium for hurt feelings or damages for psychiatric illness. This was why the Government had not accepted the recommendation of the Donovan Commission (at para 553) to the effect that a tribunal should be able to compensate an unfairly dismissed employee for injured feelings and reputation. It would have been very easy to draft the Bill to accommodate such claims if that had been the intention. Instead the word “loss” was used.

Like Sedley LJ, I would not attach great significance to the 1975 Act as a guide to the interpretation of the 1971 Act. But given that the 1971 Act was repealed in its entirety in 1974, it is very surprising that when the unfair dismissal provisions reappeared (via Schedule 1 to the Trade Unions and Labour Relations Act 1974) in the Employment Protection Act 1975 Sedley LJ’s industrious backbencher did not take industrious steps to ensure that the language of section 76 (which was virtually identical to section 116 of the 1971 Act) made in future it crystal clear that it was the intention of Parliament that compensation for unfair dismissal should extend beyond mere financial loss. Not only did Norton Tool remain unchallenged, but Parliament did not take this early golden opportunity of rectifying things so that its original intention (if Sedley LJ is right) would be effectively fulfilled.

I would conclude this part of my judgment in this way. When interpreting the meaning of an expression in an Act of Parliament, it is not sufficient merely to analyse the meaning of particular words in isolation. They have to be interpreted in their context. An important part of the 1971 context was Parliament’s decision not to entrust the remedies for all these unfair practices, including unfair dismissal, to the regular courts, but to tribunals containing a majority of lay people. The instructions they gave to these tribunals were to have regard to the loss sustained by the victims of each of these practices and nothing more. The National Industrial Relations Court in Norton Tool provided a very natural interpretation of those instructions and nobody in the next 30 years, so far as I am aware, ever suggested that this interpretation was wrong. In my judgment it was right, and thousands of skilled employment lawyers (and academic lawyers, too?) in all the years that followed 1972 were right not to challenge its correctness.

Finally, although in the light of my earlier conclusions I regard the issue as an academic one, I consider that the award of £10,000 by the employment tribunal was not manifestly too high. They were in a very good position to judge the gravity of the respondent council’s conduct and the effect that it had had on Mr Dunnachie, a grown man of 34. At one time Mr Dunnachie was in such distress at his employers’ unfeeling conduct that he was reduced to crouching with his hands around his head on the floor in the office of the council’s Chief Public Protection Officer shouting “No!” There is nothing I can usefully add to Sedley LJ’s judgment on this issue, with which I agree.