Gibb v Maidstone & Tunbridge Wells NHS Trust

This is an appeal with permission granted by Sir Richard Buxton (on one point) and by Arden LJ (on the appellant’s remaining points) against the judgment of Treacy J given in the Queen’s Bench Division on 28 April 2009 ([2009] EWHC QB 862) by which he dismissed the appellant’s claim to enforce the terms of a severance agreement (“the compromise agreement”) and rejected her alternative claims in unjust enrichment and breach of contract.

The core facts of the case are crisply and accurately summarised by Treacy J at paragraphs 1 – 15 of his judgment. I see no point in replicating this basic history in different words, though I shall have to enter into greater detail on some aspects of the facts in confronting particular issues. This is what Treacy J said:

The sum of £75,000 odd paid to the appellant (with the blessing of the Department of Health, thus the Secretary of State) was provided for in the compromise agreement to be paid in lieu of due notice to terminate the appellant’s contract of employment. Her claim in these proceedings is for the balance of the whole sum of £250,000 agreed: that is, the sum of £175,000 odd stated to be by way of a compensation payment. That is resisted on the basis, upheld by the judge, that it was ultra vires the Trust to agree the payment: it was “irrationally generous”. I shall call this the “ultra vires issue”. The appellant’s riposte is to assert that if the compromise agreement was indeed ultra vires the Trust, then she had a remedy in restitution on the footing that in the events which had happened the Trust had been unjustly enriched at her expense. I shall call this “the restitution issue”. In the alternative the appellant asserts that the Trust was in breach of its obligation of trust and confidence imposed by her contract of employment. I shall call this “the contract issue”. Sir Richard Buxton granted the appellant permission to appeal against the judge’s decision on the ultra vires issue, and Arden LJ on the two further heads of claim raised by the appellant.

I turn to the ultra vires issue. The Trust’s case is that its undertakings in the compromise agreement were “irrationally generous” to the appellant and therefore beyond its power. Though some qualifications were voiced in the course of argument, counsel were broadly agreed that a public body such as the Trust, in entering into a contractual obligation to make payments to an employee or ex-employee as was done here, is constrained by a general duty arising in public law not to enter into undertakings which may be described as “irrationally generous”. This is a term of art, since the reference to irrationality is to unreasonableness in the Wednesbury ([1948] 1 KB 223) sense – a decision so unreasonable that no reasonable decision-maker could have arrived at it.

It is of course elementary that the discretionary decisions of every public body are constrained by the Wednesbury rule. What distinguishes this case is that the decision-maker itself, the Trust, asserts its own irrationality: it seeks to escape the coils of a contractual obligation it has entered into by suggesting that it could not rationally have signed up to it. I agree with Mr White QC for the appellant that the nearest case in the books, perhaps the only other in which the decision-maker has pleaded his own alleged unreasonableness by way of defence to a claim, is Newbold v Leicesterhire CC [1999] ICR 1182. It is convenient to examine this authority before going any further.

In Newbold a contract had been entered into in 1989 between the respondent council and their street cleansing drivers, the appellants, by which the drivers would forego existing rights to extra pay for stand-by and emergency call-out duties in return for lump sums equivalent to four times the annual difference between their former and new earnings, subject to a cap of one year’s earnings. It was agreed that this new scheme should be applied to the appellants in 1991. However the council was advised (I summarise) that the multiplier of four years would be vulnerable to challenge as excessively generous. The council re-calculated downwards the payments that would be due. The appellants sued for breach of contract. The council defended the claim by asserting that the unamended scheme was, as they had been advised, irrationally generous and therefore ultra vires. The judge at first instance accepted that contention. This court allowed the appellants’ appeal. Simon Brown LJ as he then was said:

All the considerations there described by Simon Brown LJ are present in this case, in which, for the reasons given by Simon Brown LJ, the Trust has in my judgment a very steep hill to climb. I should indicate that in my view it is unhelpful to view the Newbold case as merely an instance of the elementary principle that a public body must not exercise its powers for an improper purpose: see Roberts v Hopwood [1925] 1 AC 578. Newbold is an especially acute case because it involves a public body’s reliance on its own (putative) public law wrong. Moreover cases such as Hinckley [2000] LGR 9 and Foster [2000] AER (D) 2407, relied on by Miss McNeill QC for the Trust, are to be understood as instances where a local authority unlawfully sought to set in place arrangements which would allow it to make payments above a permitted statutory maximum. Allsop v North Tyneside BC [1992] ICR 639, also relied on by Miss McNeill, was specifically distinguished in Newbold.

In addition to the learning I have mentioned Miss McNeill referred also (skeleton argument paragraph 5) to s.26 of the National Health Service Act 2006, which provides “An NHS Trust must exercise its functions effectively, efficiently and economically”. The Trust’s power to make arrangements to compensate employees is conferred by Schedule 4. HM Treasury Guidance, set out in a letter of 18 August 2005 referred to as DAO(GEN) 11/05, states (I summarise) that the level of such compensation should only exceptionally exceed statutory entitlements (that is, under the employment legislation) and where that is proposed Treasury approval should be obtained. However these matters, though certainly part of this case’s context, in my judgment have no critical bearing on the result, because the Wednesbury rule (of which “irrational generosity” is as I have stated an instance) arises by force of the common law, and so constrains the conduct of the Trust irrespective of s.26 of the 2006 Act, and certainly irrespective of Treasury guidance.

The core of the Trust’s case is crisply expressed by the learned judge as follows:

It is necessary to look a little more closely at the facts. We have seen that Capsticks’ advice on 21 September 2007 was that the Trust’s financial exposure to the appellant’s claim was likely to be in the range from about £90,000 to £250,000. Mr Edwards of Capsticks attended the meeting of the Remuneration Committee on 28 September 2007, when it was decided that the appellant’s contract of employment should be terminated well in advance of the HCC report (which was to be published on 10 October 2007) and that that should be achieved by means of a negotiated settlement. Capsticks had prepared a draft compromise agreement. After discussions with her trade union and solicitors the appellant entered into the contract whose essential terms I have described.

In the course of argument Mr White relied on certain written evidence from the Chairman of the Remuneration Committee which was not addressed by the judge below, although it was deployed before him. This material was provided for the purpose of a review by the Department of Health commissioned in October 2007 and completed in November 2007. While it might be said to offer no more than an ex post facto justification for the terms of the compromise agreement, it refers to a series of objective facts whose force, as perceived by the Chairman, I see no reason to doubt. This is what he said:

The judge for his part acknowledged that the Trust sought by the compromise agreement to obtain certain material advantages. He appreciated (paragraph 58), as was plainly the fact, that the Trust was anxious to terminate the appellant’s employment with a “clean break” and a confidentiality clause, obviating the “potential risk of a contest before the Employment Tribunal, which could serve to prolong the issues arising from the outbreak in the public domain, and distract the management from working to recover the position”. The judge referred also to “the cost of proceedings before the Employment Tribunal, both in legal and management terms”. These considerations, as it seems to me, sit well alongside the observations of the Chairman of the Remuneration Committee; indeed to an extent they develop the logic of what he had to say.

There was also the “Business Case” prepared by Mr Coode, the Trust’s Human Resources Director, on 26 October 2007. He stated that the likely costs of resisting an unfair dismissal claim in the Employment Tribunal would include legal costs of up to £80,000, Executive preparation time, twenty “person days”: £10,000, and Director time at a five day tribunal hearing: £2,500. Mr Coode also referred to “inestimable costs” which would arise in dealing with further media coverage and “reputational damage” which might be suffered by the Trust.

The judge was unimpressed by Mr Coode’s document as having been created after the event (paragraph 67): it was “not of great weight, since [it] is plainly an attempt to justify ex post facto the actions of the Trust”. As it happens Mr Coode gave advice to the Remuneration Committee on 28 September 2007, when he referred to a possible settlement package of up to £300,000, though the figure was not worked out. In any event Mr Coode’s document of 26 October 2007 sits perfectly well both with the observations of the Chairman of the Remuneration Committee and the summary of advantages which, as I have shown, the judge himself acknowledged were anticipated by the Trust to flow from the compromise agreement. It seems to me to be plain that the Trust was moved by the prospect of the costs, direct and indirect, overt and hidden – some perhaps unquantifiable – of meeting a tribunal claim brought by the respondent. That proposition was not so much rejected by the judge as a matter of fact; rather he attached no weight to it for the reasons given in paragraphs 67 and 68 of the judgment, to which I have already referred.

The first reason was that “there was no proper financial analysis done”. But the settlement advice given by Mr Edwards of Capsticks was, so far as it went, in the nature of a business case; I do not with respect consider that the judge was entitled to dismiss, as abruptly as he did, the contents of Mr Coode’s document of 26 October 2007; and the remarks of the Chairman of the Remuneration Committee (to which as I have said the judge made no reference) seem to me to carry considerable weight. It is to be noted, moreover, that the Trust called none of the actual decision-makers to give evidence.

More deeply the importance attached by the judge to what he saw as a want of financial rigour in the Trust’s decision-making suggests, I think, an error of approach on his part. He was concerned to decide a Wednesbury question, not to reach his own conclusions as to what financial prudence might require. That question was “for the council’s own assessment and not for the court”: Newbold at 1187D – E. To the extent (at least) that there was any scope for a range of opinions on the cost of not settling with the appellant, the Trust had to consider a series of matters not all of which were clear-cut or implied financially precise outcomes.

This brings me directly to the judge’s other two reasons for dismissing the basis on which the Trust entered into the compromise agreement. The second reason was that “the Trust was working on the basis that there would in the absence of a Compromise Agreement be a situation of unfair dismissal with a maximum award” (paragraph 67, last sentence). This is somewhat Delphic. However from earlier passages in the judgment (see in particular paragraph 59) it seems tolerably clear that the judge contemplated that absent a compromise agreement the Trust would swiftly have dismissed the appellant and no less swiftly settled with her at the maximum value of the tribunal claim which would then have been clearly open to her, thus limiting their cost exposure to a broadly foreseeable amount well below their commitment under the compromise agreement. Hence, in the judge’s view, the anticipated costs saving delivered by the compromise agreement was largely illusory. But on the evidence the position was by no means so straightforward. The Trust acted on the instructions of the Department of Health. On 11 October 2007 (see the judge’s paragraph 14, cited above) the Department instructed the Trust to withhold any payment to the appellant, and subsequently (paragraph 15) declined to authorise any payment beyond the sum due in respect of her six-month notice period. In a letter of 20 March 2008 to the Strategic Health Authority Mr Flory, Director-General of NHS Finance, Performance and Operations (who gave evidence for the Trust), stated:

In all these circumstances I do not think it can be assumed that absent the compromise agreement the Trust would have simply conceded then and there and offered to pay the maximum statutory compensation. As with most counterfactuals, or ‘what-ifs’, it is not possible to determine with confidence what precisely would have happened; but it is to my mind clear that as at 28 September 2007 when the Remuneration Committee met, the alternative to a compromise such as was proposed by Capsticks was by no means as clear-cut as a more or less straightforward calculation of the cost of an unfair dismissal claim at once conceded at full value. Miss McNeill fairly makes the point (skeleton argument paragraph 22) that the Department’s instructions to the Trust to withhold payment came after the compromise agreement had been entered into; but this by no means demonstrates that as at 28 September 2007, absent the compromise agreement, the Trust would without more have settled the appellant’s statutory claim.

There is a further point. An unfair dismissal claim is not in all respects to be equated with a common law action which a defendant can simply choose to settle by a monetary offer. Here the decision of the Employment Appeal Tribunal (presided over by Tucker J) in Telephone Information Services Ltd v Wilkinson [1991] IRLR 148 is instructive. It is enough to cite this passage from the headnote:

In the circumstances the judge’s second reason for dismissing the basis on which the Trust entered into the compromise agreement does not in my judgment withstand scrutiny.

The third reason (paragraph 68) was that in fixing the terms of the compromise agreement the Trust had regard to the appellant’s many earlier years of good service, and the time it might take her to find other employment; and these were legally irrelevant considerations. In my judgment these matters do not by any means fall to be regarded as legally irrelevant. I do not see why an employer such as the Trust, faced in difficult and perhaps controversial circumstances with the need to terminate a long-standing employee’s contract, should be obliged in settling terms of severance to disregard past service and the employee’s future likely difficulties. In such a case a reasonable employer is not limited to the replication of the statutory maximum available to the employee through legal redress. He will not show undue favours; but the constraint of rationality will not close the door on some degree of generosity for the sake of good relations and mutual respect between employer and employee: not only for the sake of the employee in question, but it may be also for the employer’s standing and reputation as such. This position is unaffected by the terms of guidance or instructions from the Treasury, neither of which is a source of law.

For all these reasons it is not in my judgment shown that the compensation package provided for in the compromise agreement amounted to “irrational generosity” on the part of the Trust. I would accordingly allow the appeal on the ultra vires issue.

If my Lords agree, my conclusion on the ultra vires issue disposes of the whole case in the appellant’s favour. But out of respect for the parties’ submissions and the judge’s views I should address the restitution issue.

This aspect of the appellant’s claim proceeds on the premise that (contrary to her primary case on the ultra vires issue) the compromise agreement is indeed void as being “irrationally generous”. On that footing the claim is formulated as restitution of the value of benefits conferred on the Trust under a void contract. In paragraph 11A of the Re-amended Particulars of Claim it is alleged that the appellant’s entering into the compromise agreement secured in the Trust’s hands the following benefits: (1) the value of the statutory unfair dismissal claim which she forewent and the costs of contesting that claim which the Trust saved, (2) the confidentiality requirement imposed on the appellant, (3) the immediate cessation of the appellant’s employment, and (4) the avoidance of any internal grievance procedure of which the appellant might have contested to the detriment of the Trust in terms of “significant human and/or financial resources”. The appellant seeks to recover (paragraph 11A(3)) “compensation or damages in the maximum sum intra vires of the [Trust], to be assessed by the Court”.

Miss McNeill submits that the law relating to unjust enrichment does not recognise a category of case in which the defendant is held to have been enriched by the claimant’s having foregone a claim against him. She observes, correctly, that the law in this field has been developed incrementally. She cites Mann J’s judgment in Charles Uren v First International Finance Ltd [2005] EWHC Ch 2529 which collects a number of germane dicta,including this from Lord Diplock in Orakpo v Manson Investments [1978] AC 95, 104C:

There is, I think, something of a tension underneath this reasoning. It is between these two propositions. (1) The categories of unjust enrichment claims cannot be closed, for if they were this branch of the law would be condemned to ossify for no apparent reason; and nothing could be further from the common law’s incremental method. But (2) such a claim must fall “within one of the hitherto established categories of unjust enrichment” which suggests (at least) that the categories rather than any overriding principle are paramount. The authorities’ reluctance to assert first principles may be ascribed to the justified fear of the palm tree: if the principle of unjust enrichment does no more than to invite one judge after another, case by case, to declare that this or that enrichment is inherently just or unjust, it is not much of a principle. That is why, with all due deference, I wonder whether Lord Hoffmann’s formulation in Banque Financière de la Cité v Parc (Battersea) Limited [1999] 1 AC 221 at 234C – D has not too much of a broad-brush or legislative flavour. He states there are four components to a claim in unjust enrichment. First, the defendant must be enriched by receipt of a benefit. Secondly, the benefit must be at the claimant’s expense. Thirdly, the defendant’s retention of the benefit must be shown to be unjust. Lastly, there must be no policy reasons for denying a remedy. The third requirement seems to be quite unqualified.

If one looks at the matter from what is perhaps a more modest standpoint, we may see at once that clear reasoning is at least required for the elaboration of any extension of unjust enrichment. Clear reasoning, if it allows a claim in seemingly new circumstances, will provide clear analogues with other cases. No doubt this is what Mann J had in mind when he qualified his reference to established categories by the phrase “or some justifiable extension thereof”.

I make these points only to show, with respect, that Miss McNeill’s forceful plea that this case lies outside the established categories of unjust enrichment may do less than justice to the subtleties of the way the law develops. For his part, Mr White submits that as matters stand a claim of this kind is well accommodated by authority. He cites Eastbourne BC v Foster [2002] ICR 234, in which the council had entered into a void and unenforceable agreement with an employee designed to enable him to qualify for enhanced pension benefits on reaching his 50^th birthday notwithstanding that his post was shortly to be terminated on grounds of redundancy. At paragraph 32 of the judgment Rix LJ stated:

It is true, as Miss McNeill submitted, that in that case payments had actually been made over in pursuance of an ultra vires contract. But Mr White would submit that the present case does no more than build on well established ideas of justice plainly expressed in Eastbourne BC v Foster. He refers moreover to the equitable principle that giving up a valid claim may supply the “value” which has to be shown by a bona fide purchaser for value without notice: Thorndike v Hunt (1859) 3 De Gex & Jones 563, 570, and Taylor v Blakelock (1886) LR 32 Ch D 560, 568, 570. Mr White submits that such cases provide at least indirect authority for the proposition that the foregoing of a valuable claim may confer a benefit – an enrichment – on the party not so pursued.

Notwithstanding Miss McNeill’s tenacity and ingenuity it seems to me that the facts of this case are readily aligned to established categories of unjust enrichment. If everything else is equal I can see no principled distinction between a benefit consisting in money paid and a benefit consisting in a claim foregone. For the purpose of this branch of the law the material benefit may take many forms. In Way v Latilla [1937] 3 AER 759 it consisted in providing information about gold mines and effecting introductions.

I conclude that the facts here are in principle amenable to an unjust enrichment claim. However the judge concluded that on the facts the appellant had not been enriched or received any benefit by reason of the compromise agreement’s being void: in his view any advantage obtained by the Trust accrued “because the [appellant] and her advisors with full knowledge of the stance which the [Trust] was taking as to the validity of the agreement [had] not issued proceedings in time before the Employment Tribunal”(paragraph 86). Time had expired, as I understand it, on 4 January 2008. The judge held that before that date the appellant was aware of the Trust’s reasons for withholding payment under the compromise agreement – namely that it was ultra vires – and had access to legal advice on the matter.

The judge’s conclusion looks like a straightforward finding on causation. But it masks an important distinction. The focus in this case is on the benefit, the enrichment, enjoyed by the Trust at the appellant’s expense; not on the question whether the appellant has suffered a loss which she could have avoided. The contrast is critical. In the conventional case of contractual or tortious damage, the policy of the law is of course to compensate the claimant for loss occasioned by another’s fault. In a case of unjust enrichment, the policy of the law is to strip the defendant of a benefit he has wrongly received at the claimant’s expense. In the former instance the defendant may have received no benefit whatever. In the latter, his benefit is centre stage. It is therefore no surprise that in the unjust enrichment case the law takes a much looser view of causation where the question is, was the benefit received at the claimant’s expense, than in the conventional damages case where the question is, was the claimant’s loss caused by the defendant’s fault. It is now all but a commonplace to acknowledge that a forensic assessment of causation in any given case will be heavily influenced by the nature of the interest which as a matter of justice the law seeks to protect in the particular case. Often the issue tested by causation is, where should responsibility lie? And in that case, of course, the assessment of causation is by no means a value-free exercise.

There are many judicial pronouncements in this field, but I may confine myself to what has been said specifically with reference to claims of unjust enrichment. Thus in Niru Battery Manufacturing Co (No 2) [2003] 2 AER (Comm) 365 Moore-Bick J as he then was observed at paragraph 40 that

I conclude that the judge’s finding based on a straightforward model of causation was, with respect, over-simplistic. It left out of account the vital fact of the Trust’s enrichment. But it is also in my view fragile (to say the least) on the evidence, which I ought therefore to address. On 17 December 2007 the appellant’s solicitors sought performance of the compromise agreement and threatened proceedings. By its reply two days later the Trust did not take the ultra vires point; they sought further time in which to consider their reply. On 21 December 2007 the appellant agreed to postpone the commencement of proceedings for one month “in order to enable your client to meet with the Department of Health”. On 17 January 2008 the Trust sought “the indulgence of [the appellant] to a further extension of at least one week”. But the date (as I have said, 4 January 2008) by which tribunal proceedings should have been commenced had already passed. On 29 January 2008 the Trust expressly took the position that they would not pay the compensation stipulated in the compensation agreement because the latter had been entered into ultra vires their lawful powers.

It was not pleaded in the Trust’s defence, originally or by amendment, that any sum alleged to be recoverable by the appellant was lost to her by reason of her own failure to issue a protective claim in the tribunal. Nor was any such argument advanced for the Trust in counsel’s written opening skeleton or written closing submissions. The appellant was not cross-examined about it.

I do not consider that the judge’s factual conclusion on causation (paragraph 86 of his judgment) can be supported; and I conclude that the appellant’s alternative case in unjust enrichment, if I am wrong on the ultra vires issue, is well made out. In sum, I would accept the submission made at paragraph 39 of Mr White’s principal skeleton argument:

Moreover I see no reason why the Trust’s enrichment should not in principle extend to all the benefits pleaded in paragraph 11A of the Re-amended Particulars of Claim (see paragraph 24 above). Mr White (principal skeleton argument paragraph 33) quotes Professor Birks’ observation (Unjust Enrichment, 2^nd edn. 2005) that “[w]hat works for money must work for value received in other forms”. He refers also to County of Carleton v City of Ottawa (1965) 52 DLR (2d) 220, a decision of the Supreme Court of Canada which exemplifies the recovery of a benefit not obtained by the provision of money, goods or services to the defendant (the claimant had mistakenly discharged the defendant’s statutory duty). However given the judge’s strong reservations at paragraphs 87 – 91 of the judgment, the non-monetary elements would I think fall to be valued very modestly. The Trust’s benefit obtained by the appellant’s foregoing her claim for compensation for unfair dismissal is worth £69,590: see paragraph 33 of the judgment.

I propose merely to express a view on this final issue, and to do so very shortly. It is common ground, and the judge below found, that the Trust was in breach of its contractual duty of trust and confidence owed to the appellant (paragraph 112). The breach consisted in the provision of reckless assurances that the Trust had obtained all necessary approvals for the compromise agreement. But the judge held (paragraph 117) that the claim was defeated by what has become known as the “Johnson exclusion area”. This is a reference to Johnson v Unisys [2001] ICR 480, whose effect the judge described as follows (paragraph 113):

Mr White submits that the appellant’s claimed loss does not flow from her dismissal – if it did, it would indeed fall within the Johnson exclusion area, whose purpose is to bar from the ordinary courts claims which by statute should be brought by way of unfair dismissal proceedings. By contrast Mr White’s case is that the breach of contract in question (the reckless assurances) caused the appellant to enter into the compromise agreement, and thus sign away her potential unfair dismissal claim. She entered into the compromise agreement before the termination of her contract of employment; the fact that the loss fell in after it does not show that the termination caused the loss. This seems to me to be an entirely plausible analysis of what took place, and if it were necessary to decide the issue, I would uphold it.

The judge also rejected this head of claim on the basis that

For all the reasons I have given, I would allow the appeal. If my Lords agree, counsel will no doubt assist us as to the appropriate orders to be made.

As a bystander at the execution of Admiral Byng explained to Candide:

The trustees of the hospital group of which Ms Gibb held the post of chief executive officer learnt that a damning report on its standards of patient care was coming from the Healthcare Commission. The draft report included this:

The trustees took the view that one answer to such criticism, when the storm broke, would be that they had already taken remedial measures, and that the sacrifice most likely to propitiate the deities of Whitehall and the media was their chief executive officer, Ms Gibb. The fact that she personally had done nothing to merit dismissal was a problem, of course, but not an insuperable one. Provided the Trust was willing to pay the necessary price, she could be dismissed both unlawfully (that is to say, without notice) and unfairly (that is to say without any good cause). Both of these were quantifiable in money.

The trustees, through their remuneration committee, also wanted to show some measure of acknowledgment that Ms Gibb had given years of blameless service as CEO and that she was losing her job principally because it was she who held it at a time when serious shortcomings had been found in the hospital’s standards. Moreover, by accepting a compensation package that included a vow of silence she would be sparing the Trust a public controversy about where responsibility actually lay, the near-certainty of an adverse tribunal finding, a serious drain on management time and resources, further potential costs and – importantly – the damaging effect of all these things on staff morale and performance at a time when the Trust needed a new start.

For evidence of what the board were thereby escaping in terms of media intrusion and hostility, one need only look at Ms Gibb’s account of what happened to her:

These would all have been potentially relevant considerations for an employment tribunal. The range within which they might be accorded a monetary value was the subject of expert legal and management advice given to the remuneration committee. The committee’s decision to go to the top of that range is explicable by the regard they chose to have to the kinds of consideration I have mentioned. While some may be regarded as self-interested, the interests were those of the Trust at least as much as of its board members personally. Their powers, moreover, by virtue of §26 of Sch. 4 to the National Health Service Act 2006, included a power to pay not simply contractual sums and damages but gratuities “by way of compensation to … any of the NHS trust’s employees who suffer loss of office or employment”.

This was the background to the bizarre legal situation now facing the court. What precipitated the claim was that the Trust was directed by the Department of Health to renege on its own agreement. Although there had been a failure to follow Whitehall guidance about clearance of such agreements (the Trust had taken the normal course in the NHS of getting approval from the strategic health authority instead), this was not the reason for the direction. The reason was that ministers wanted to be able to announce that the agreement was being blocked. The result was that the Trust, when sued by Ms Gibb for failing to honour the greater part – not the whole - of it, found itself compelled to deny its own power to enter into it. (The logical oddity of a partial nullity does not seem to have occurred to anyone until Rimer LJ raised it in the course of argument.)

The basis of the Trust’s volte-face is, or purports to be, the Wednesbury doctrine of limited powers - although, as Laws LJ pointed out during argument, the case developed by Ms McNeill QC for the Trust corresponds more nearly with the Padfield doctrine of limited purposes. I think it is necessary, in this situation, to say something about the place of the ultra vires doctrine in modern public law.

The ultra vires doctrine was imported into public law during the later part of the 19th century from its original home, company law, where it was born fully grown, like Pantagruel, in Colman v Eastern Counties Railway Co (1846) 16 L.J.Ch 73. Its aim was to stop the repeated transgression of their own legal powers by joint-stock railway companies. But by the date of the first edition of Seward Brice’s Treatise on the Doctrine of Ultra Vires (1874) it had become apparent that it was an engine of abuse in the hands of those whom it was meant to constrain, because it enabled limited companies to renege on their debts and other contractual obligations by denying their own power to incur or enter into them.

In his preface Brice wrote:

It is serious enough if a private law corporation reneges on its agreements for want of power to make them (something which may, however, be rarer nowadays because of the typical breadth of articles of association). It is even more serious if a body incorporated by statute for public purposes can do so in a case such as the one before the court. This is not only because public bodies, with access to competent legal advice, can be expected not to act on whims and, when accused of doing so, are generally found not to have done so. It is because if a public body can denounce its own commercial agreements as having been excessively generous – in other words can invite the court to recalculate its liability – it will not be only at the authority’s own instance that this can happen. It will be able to happen at the instance of any person or body with a sufficient interest - here, for example, a local patients’ organisation or the Secretary of State or even (since the rule in Foss v Harbottle is not thought to apply to public corporations) a dissident member of the body itself. It does not matter, I readily accept, that this might create an entire new litigation industry: as Holt CJ said in Ashby v White (1703) 2 Ld. Raym. 938, “if men will multiply injuries, actions must be multiplied too”. What matters is that the autonomy of statutory bodies like the Trust will be irrevocably compromised: the enlargement of what counts as a public law wrong will mean that every financial decision of a public body is open to scrutiny by the courts on the motion of anyone with a sufficient interest. Only the legal profession would regard such a development as desirable.

None of this is to cast doubt on the availability of judicial review to correct excesses of power or other public law wrongs on the part of statutory bodies. But it is clear, in my judgment, that Parliament in setting up the present NHS structures intended no such thing in relation to their routine financial decisions. The 1996 Act is generous in its grant of financial autonomy to trusts. It authorises each trust to “employ such staff as it considers appropriate” and to pay them “such remuneration and allowances as it considers appropriate” (Sch.4 §25). Such powers are to be exercised in accordance with ministerial regulations and directions, but there is no suggestion that the agreement with Ms Gibb transgressed any of these. They are also to be exercised within the limits set by public law: hence the present dispute.

The use of these limits by a public authority to renege on agreements which on the face of them lay within their powers gives resonance to the note of caution sounded by Simon Brown LJ in his concurring judgment in Newbold v Leicester City Council[1999] ICR 1182, quoted by Laws LJ in §6 above. It also prompts a wider reflection about the use of what has become the Wednesbury mantra. When, by accident or by design, the voters of Wednesbury returned a majority of sabbatarian councillors who proceeded to make a by-law barring children from the local cinemas on Sundays, it was not unexpected that the owners of the Walsall Street Gaumont, a recently and expensively reconstructed picture palace with over 1,500 seats, would challenge it. (Footnote: ¹ ) Nor was it unexpected, in 1948, that both the Divisional Court and this court were reluctant to intervene; but one needs to have in mind Lord Diplock’s warning in R v IRC, ex p Federation of Self-Employed[1982] AC 617, 649, that “any judicial statements on public law if made before 1950 are likely to be a misleading guide to what the law is today”. As with all Lord Diplock’s pronouncements, the choice of 1950 as a watershed will not have been arbitrary.

But, for all the reasons given by Laws LJ, I agree that there was in the present case nothing irrational, even in classical Wednesbury terms, about the severance payment which the board agreed to, and that no irrelevant considerations entered into the decision to make it. It is also necessary, in my view, to say that the time is past when this schematic and unsubtle approach to public law issues was generally useful. Without dwelling on the reasons, I would draw attention to Lord Cooke of Thorndon’s comments on the genesis of Wednesbury reasonableness in his essay ‘The struggle for simplicity in administrative law’. (Footnote: ² ) Abuses of power, which are what all public law is at root about, are not best detected by tick-list. As Lord Greene had said in his presidential address to the Holdsworth Club ten years before he decided the Wednesbury case:

This is not for a moment to say that profligate expenditure by a public body is beyond the reach of the courts. Even company law has drawn a well-known (and much-transgressed) line requiring that any cakes and ale (corporate hospitality as it is now called) must be consonant with the company’s interests: per Bowen LJ in Hutton v West Cork Railway Co (1883) 23 Ch. 654. And in public law there is at least one example of judicial auditing of civic expenditure by reason of its sheer magnitude. The High Court of Ireland in 1894 had to consider the appeal of the Dublin councillors against a surcharge for their expenditure on the picnic which accompanied their annual inspection of the Vartry waterworks in the Wicklow Hills. In a judgment (R (Bridgeman) v Drury[1894] 2 IR 489) which must rank as one of the great exemplars of judicial restraint, O’Brian CJ said:

But none of this comes close to justifying the retaking by a court of a financial and management decision which lay within the powers and purposes of the Trust, whatever reservations the court itself might have had about the computation and cost of the deal. To start by dissecting the figures is both to assume the very thing that has yet to be established – that the Trust has exceeded its own powers – and to substitute the court’s judgment for that of the Trust. It is only if the figures are inexplicable on their face, or palpably inflated in the light of evidence, that the court will in general be justified in examining their elements, and then not in order to remake the calculation but to see if it has indeed gone beyond the bounds set by law.

In this situation I consider, with respect, that Treacy J erred by letting himself be drawn into acting more nearly as auditor than as judge. On the scale of severance payments not only in the private sector but in parts of the public sector, £240,000 was not on its face outlandish compensation for the arbitrary termination of a career which it was unlikely Ms Gibb would be able to resume or resurrect. If so, the only other question was whether it had been reached on a false basis. As to this, I agree with the reasons given by Laws LJ for holding that it had not.

Given my conclusion on the vires issue, which corresponds with that of the other members of the court and so is dispositive of the appeal, I prefer to express no view on the other two issues.

The effect of unwarranted departmental interference has thus been to trap the Trust between a rock and a hard place and to expose it, in its attempt to escape, to heavy legal costs. Central government (which, it seems, will be picking up the bill) might have done better to recognise that the Trust, in reaching the agreement, had been making the best of a bad job; and perhaps better still to recognise that the bad job had been the decision, which the Department does not appear to have cavilled at, to sacrifice on the altar of public relations a senior official who had done nothing wrong.

Perhaps those responsible will now reflect that, since such blame as the report allocated was subsequently accepted by the Trust’s board; all of whom resigned following publication of the report, there had been no good reason to dismiss the CEO; and that all this money, both compensation and costs, could have been spent on improving hygiene and patient care in the Trust’s hospitals.

For the reasons given by Laws LJ, I agree that Ms Gibb’s appeal should be allowed on the ultra vires issue. For the further reasons he has also given, I also agree that she is entitled in the alternative to succeed on the restitution issue. I would prefer to express no view on the contract issue.