Société Générale, London Branch v Geys

The appellant, Raphael Geys, is a Belgian national. He is in dispute with his former employer, Société Générale, London Branch (“the Bank”), about the amount due to him following his summary dismissal from his employment. His case is that he was dismissed on 6 January 2008, and that he is entitled to a sum contractually due to him in the form of a termination payment amounting to more than €12.5m and to damages for breach of contract. The Bank’s case is that the appellant is entitled to a termination payment of no more than €7m, as he was dismissed on 29 November 2007 or at the latest 18 December 2007. It also maintains that, having regard among other things to the terms of his employment contract, it is not open to the appellant to claim damages. The case went to trial before Mr George Leggatt QC sitting as a Deputy High Court Judge. On 25 March 2010 he found that the appellant was dismissed on 6 January 2008. He gave judgment for the appellant in a sum to be assessed, with a payment on account by 1 April 2010 of €11m, less tax and national insurance contributions, together with interest on all sums due at 1% above base rate from 3 February 2008: [2010] EWHC 648 (Ch), [2010] IRLR 950.

The Bank appealed against that decision on various grounds. The first two related to the date of the appellant’s dismissal. For reasons that will be explained later, the choice of date has significant consequences for the amount that is contractually due to the appellant as a termination payment. The first ground raised the question as to whether, in the context of employment law, a repudiatory dismissal of an employee will by itself terminate the contract even if its repudiation is not accepted. The second raised the question as to when, having regard to the relevant provision in the contract, the right to terminate was validly exercised. The effect of its submissions on these issues, if sound, is that the appellant’s employment was terminated at the latest on 18 December 2007. The third ground related to the construction of a paragraph in the appellant’s employment contract which obliged the Bank to ensure that any bonus award made to the appellant was made in as tax efficient a manner as was possible. The fourth and fifth grounds related to the construction of provisions in the employment contract for the entering into by the appellant of a termination agreement in the event of his employment with the Bank being terminated. The Bank’s case is that the contract provided for a clean break when the employment was terminated and excluded the possibility of claiming damages.

The Court of Appeal (Arden, Rimer and Pitchford LJJ) dismissed the Bank’s appeal on the first and fifth grounds, but allowed its appeal on the second, third and fourth grounds and found that the appellant was dismissed on 18 December 2007. It dismissed a cross-appeal by the appellant as to the date of his dismissal: [2011] EWCA Civ 307, [2011] IRLR 482. The appellant now appeals to this court on the issues raised by the second and third grounds. The Bank cross-appeals against the dismissal of its appeal on the first ground.

On 9 February 2005 the appellant commenced employment with the Bank as the managing director of its European Fixed Income Sales, Financial Institutions Division. He was provided with a written contract of employment. It was offered to him by a letter dated 28 January 2005 with which there were enclosed, among other things, two copies of a contract of employment (“the Contract”) and a copy of the Staff Handbook of the SGUK Group (“the Handbook”). He indicated his acceptance of the offer in the way that the letter required of him and, having done so, commenced his employment.

The Contract, in which the Bank was referred to as “the Company”, contained the provisions that a contract of this kind would be expected to set out as to commencement, job title, remuneration, working hours and duties, overtime, holiday, notice, restrictions upon and after termination of employment, disciplinary rules, choice of law and confidentiality. There was also, in paragraph 5, an elaborate section which extended to more than eleven pages dealing with the employee’s entitlement to participate in bonuses under the Bank’s Fixed Income Sales Scheme (FISS) referred to in paragraph 5.2. It included provision for the making of a termination payment in the event of the termination of the employment, in consideration for which the employee was to enter into a termination agreement in the terms set out in a schedule. Various events that might give rise to a termination were provided for in paragraph 5. Paragraph 13, under the heading of “Notice”, was in these terms:

“Your employment can be terminated on the expiry of 3 months’ written notice of termination given by you to the Company or by the Company to you.”

In paragraph 5.14 it was provided that, if the Company were to terminate his employment in circumstances other than those contained in sub-paragraph 5.6(b)(i)-(iv) (which did not apply in this case):

“the Company will, within 28 days after such termination of your employment, make a payment to you (the ‘Termination Payment’) as specified in paragraph 5.15.”

By paragraph 5.15 it was provided that the Termination Payment was to be equal to the aggregate of (a) the value of the proportion of any award by way of bonus that had been made to the employee but retained by the Company and not yet released, and (b) a Compensation Payment calculated by reference to the date of the termination of his employment. The relevant sub-paragraphs are as follows:

“(iii)

if your employment terminates after 31 December 2006 but before 1 January 2008, the Compensation Payment shall be 0.65 x (S divided by 2) where S is the aggregate of any award or awards made to you under the FISS and the Scheme (whether or not subject to the Deferral under paragraph 5.7) in respect of the calendar years ending 31 December 2005 and 31 December 2006;

(iv)

if your employment terminates after 31 December 2007 but before 1 January 2009, the Compensation Payment shall be 0.65 x (T divided by 2) where T is the aggregate of any award or awards made to you under the FISS and the Scheme (whether or not subject to the Deferral under paragraph 5.7) in respect of the calendar years ending 31 December 2006 [and] 31 December 2007.”

The difference between the payments that would be due to the appellant under sub-paragraphs (iii) and (iv) respectively, depending on the date of his dismissal, has not been precisely identified in these proceedings. But it is common ground that it is substantial. So the answer to the question as to the date when the appellant’s employment was terminated will have a significant bearing on the amount to which he is entitled by way of a termination payment under the contract.

Section 1 of the Handbook, in which the Bank was referred to as “SG”, contained a number of additional terms and conditions of employment. Among them was the following paragraph:

“8.

Notice Periods

8.1

Your Right to Notice

Your entitlement to written notice of termination from SG is the longer of:

•

The period set out in your Contract; or

•

1 week for each complete year of service up to a maximum of 12 weeks’ notice after 12 years’ continuous service.

No notice or payment in lieu of notice will be given where SG is entitled to dismiss you immediately without notice or payment in lieu of notice …

Notice given by SG in writing shall be deemed to have been given by SG upon either being handed to you or sent to your home address (as last notified by you to HR). If such notice is sent by post, it shall be deemed to have been received by you on the second day after posting.

8.2

Giving Notice

You are required to give SG the period of written notice set out in your Contract.

Without prejudice to any other contractual rights and duties relating to your employment, if you fail to give the correct period of notice, SG may require you to give the correct period of notice as required by your Contract.

If SG does, at its absolute discretion, accept less than full notice from you:

•

You shall not be entitled to payment in respect of salary or to receive contractual benefits for the period of notice not worked;

•

You may only be entitled to accrued but untaken holiday pay in respect of that holiday year at SG’s discretion; and

•

You will remain subject all contractual and legal restrictions and obligations.

8.3

Termination by SG and Payment in Lieu of Notice

SG reserves the right to terminate your employment at any time with immediate effect by making a payment to you in lieu of notice (or, if notice has already been given, the balance of your notice period) based upon the value of your:

•

Basic annual salary; and

•

Flexible benefits allowance;

for your notice period (or, if notice has already been given, the balance of your notice period).”

Paragraph 17 of the Contract, under the heading “General Information”, was in these terms:

“This contract is in conjunction with the offer letter, the Staff Handbook of the SGUK Group (as amended from time to time) and the SGUK Compliance Manual which, together with this letter, form the written particulars of employment as required by law. However, in the event of any conflict of any terms set out in this Contract and those contained in the Handbook the terms of this contract will prevail.”

On 29 November 2007 the appellant was called to a meeting at which he was handed a letter which had been written on the Bank’s behalf and was in these terms:

“Termination of Employment

I am writing to notify you that Société Générale, London (‘SG’) has decided to terminate your employment with immediate effect.

In accordance with the terms of your employment contract with SG dated 28 January 2005, SG will arrange for the appropriate termination documentation to be provided to you and your legal adviser.”

The appellant was escorted from the building and did not return to it. But he did not let the matter rest there. He consulted his solicitors after his summary dismissal. They wrote to the Bank on 7 December 2007 asking for further information about the sums that it was offering to pay following the termination of his employment, but also saying that the appellant reserved all his rights. On 10 December 2007 the Bank’s legal department sent the appellant a severance agreement which was said to have been prepared in line with the relevant provisions of his employment contract, together with another letter of the same date which contained a list of the payments that it was proposed should be made to him in consideration of his entering into that agreement. He was asked to agree the terms that were set out in that letter by returning a signed copy, but he declined to do so.

On 18 December 2007 the Bank paid £31,899.29 into the appellant’s bank account. This was the equivalent of his basic salary and flexible benefits allowance for three months. It is agreed that this was a payment that satisfied the monetary requirements of paragraph 8.3 of the Handbook as it was the amount which the appellant would have received had he been given three months’ notice. The appellant became aware of this payment at some point before 2 January 2008 which has not been precisely identified but which the judge found was probably before the end of December 2007. The Bank then sent the appellant a payslip, accompanied by a P45, which set out the various elements of the payment of 18 December 2007 including in lieu pay amounting to £37,500 before deductions. The appellant first saw it on 7 or 8 January 2008 when he returned to London from Belgium where he had spent most of his Christmas and New Year holiday. He said in his evidence that, while he could not be sure what the payment was for, the best guess he could have was that it was intended to be a payment in lieu of notice.

Meantime, on 21 December 2007 the appellant’s solicitors wrote in reply to the Bank’s letter of 10 December 2007 asking for further information, in particular about how the proposed payments had been calculated. They again stated that the appellant’s rights in relation to his employment contract remained reserved. On 2 January 2008 they wrote to the Bank saying that the appellant had decided to affirm his contract of employment. Referring to the payment of 18 December 2007, they said that they reserved his position in relation to those monies until they understood what they constituted.

On 4 January 2008 the Bank’s Human Resources Director wrote to the appellant with regard to his employment with the Bank. The first four paragraphs of that letter were as follows:

“I write further to your meeting with Fred Desclaux and Nigel Holmes on 29 November 2007 to confirm the details of the termination of your employment. Please accept my apologies for the delay in sending these details to you.

1.

Notice Entitlement

Under your terms and conditions of employment, you are entitled to 3 months’ notice of termination of your employment. Société Générale gave you notice to terminate your employment with immediate effect on 29 November 2007 (your Termination Date) and will pay you in lieu of your notice period. This payment will be calculated in accordance with section 1/8.3 of the Société Générale CIB Staff Handbook.

2.

Final Salary Payment

Your notice payment was credited to your bank account on 18 December and your final salary slip and P45 was sent to your home address. This amount was paid to you with deduction of income tax or employee NICs.

3.

Pension Benefits

Your active membership of the SG International Pension Plan (IPP) will cease on 29 November 2007.”

He was also told that the outstanding balance in respect of his annual travel insurance policy would be deducted from his final salary payment and that the policy would continue until 31 March 2008.

Having regard to paragraph 8.1 of the Handbook the appellant must be deemed to have received the Bank’s letter of 4 January 2008 on 6 January 2008. The judge held that this was the first occasion when the Bank notified the appellant that it had exercised its right to terminate the contract under paragraph 8.3. The Court of Appeal held that the contract was terminated on 18 December 2007 when the amount of the payment in lieu of notice was paid into the appellant’s bank account. The Bank’s primary argument was that the contract of employment was terminated on 29 November 2007 when the appellant was summarily dismissed. This was rejected by the Court of Appeal, which held that it was bound by the decisions of the Court of Appeal in Gunton v Richmond-upon-Thames London Borough Council [1981] Ch 448 and Boyo v Lambeth London Borough Council [1994] ICR 727, in which the principle that a repudiatory breach must be accepted was applied to contracts of personal service. Rimer LJ said in para 18 that permission to appeal had been given on this issue solely to keep open the possibility of an appeal to the Supreme Court so that this area of the law could be reconsidered.

Four issues are before the court in this appeal. The first two, which are of general public importance, bear directly on the question as to the date when the appellant’s employment was terminated. The third and fourth are directed solely to the proper construction of provisions in the contract. They can be summarised as follows:

For the reasons given by Lord Wilson, I too would hold that the elective theory is to be preferred – that a party’s repudiation terminates a contract of employment only if and when the other party elects to accept the repudiation. I am persuaded by his careful analysis of the authorities that provide support for his conclusion that the view that repudiation of a contract of employment terminates the contract without the necessity of acceptance by the other party was not as authoritative or as consistent as Lord Sumption indicates in para 128 below. I also think that there are cases, of which this case is a good example, where it really does matter which of the two theories is adopted. The automatic theory can operate to the disadvantage of the injured party in a way that enables the wrongdoer to benefit from his own wrong. The law should seek to avoid such an obvious injustice. Where there is a real choice as to the direction of travel, the common law should favour the direction that is least likely to do harm to the injured party. I agree that we should be very cautious before reaching a conclusion whose result is that a breach is rewarded rather than its adverse consequences for the innocent party negatived: see para 66.

Was Sir John Donaldson clearly right when he declared in Sanders v Ernest A Neale Ltd [1974] ICR 565 at p 571 that an unaccepted repudiation brought a contract of employment to an end? Lord Sumption says that this was an accurate summary of the position as it then stood: paras 128 and 139, below. But I find it hard to disagree with Buckley LJ’s observation in Gunton v Richmond-upon-Thames London Borough Council [1981] Ch 448, 466 that Sanders v Ernest A Neale Ltd was the first case in which the automatic theory was part of the basis for the decision in an employment case. In Thomas Marshall (Exports) Ltd v Guinle [1979] Ch 227 Sir Robert Megarry V-C in his review of the authorities also took that case as his starting point. He described it as the high-water-mark of the doctrine of automatic determination, but said that the authorities on the point were in a state that was far from satisfactory. Shaw LJ, in his dissenting judgment in Gunton, referred to the field that Buckley LJ had covered in his review of the authorities as dubious. He said that, as a result of the ebb and flow of the tide of judicial opinion, the court was left in the slack water of first principles. Only a few months later, in London Transport Executive v Clarke [1981] ICR 355, the majority view in the Court of Appeal was in favour of the position that Sir Robert Megarry V-C adopted in Marshall.

The fact has to be faced that there is still a degree of oscillation between the two theories: David Cabrelli and Rebecca Zahn, The Elective and Automatic Theories of Termination at Common Law: Resolving the Conundrum? (2012) 41 Industrial Law Journal 346, 349. In any case, the question which of the two theories should be adopted is an open question at our level. Which result is, in principle, the most desirable? One must be careful not to assume that, just because in practice the employee may have little choice but to accept the repudiation, he has in law no alternative but to do so. I would endorse Ralph Gibson LJ’s criticism in Boyo v Lambeth London Borough Council [1994] ICR 727, 743 of Buckley LJ’s observation in the Gunton case that in a case of wrongful dismissal the court should easily infer that the innocent party has accepted the guilty party’s repudiation of the contract. If the law requires acceptance of the repudiation, the requirement is for a real acceptance – a conscious intention to bring the contract to an end, or the doing of something that is inconsistent with its continuation. So the question is whether there are sound reasons of principle for holding that the general rule of law that requires acceptance of a repudiation does not apply.

The fact that an application of the automatic theory may produce an injustice is, for me, the crucial point. The question that Sir John Donaldson asked himself in Sanders v Ernest A Neale Ltd [1974] ICR 565, 571 is at the heart of the issue: why should the employee not sue for wages if it is the act of the employer which has prevented his performing the condition precedent of rendering his services? There may be grounds for thinking that the court is less reluctant than it once was to give injunctive relief in such cases, but I would not rest my decision on that point. It is the objection that the party who is in the wrong should not be permitted to benefit from his own wrong that is determinative. The timing of the repudiation may be crucial, and if the automatic theory were to prevail an employer may well be tempted to play this to his advantage – by getting in first before a rise in pay or pension entitlement takes place or, as in this case, a rise in the entitlement to bonuses. I note too that, as Professor Douglas Brodie has pointed out, it is not always true that work is the counterpart of the entitlement to wages. In some contracts wages are given to employees for holding themselves available for work: The Contract of Employment (2008), para 18-09.

The essential difference between the two theories may be said to be that under the automatic theory the decision as to whether the contract is at an end is made beyond the control of the innocent party in all circumstances, whereas under the elective theory it is for the innocent party to judge whether it is in his interests to keep the contract alive. Manifest justice favours preferring the interests of the innocent party to those of the wrongdoer. If there exists a good reason and an opportunity for the innocent party to affirm the contract, he should be allowed to do so: London Transport Executive v Clarke [1981] ICR 355, 367, per Templeman LJ.

I need not elaborate on these and the other points that favour the elective theory, as they have been dealt with so thoroughly by Lord Wilson. I respectfully agree with the conclusion that he reaches in para 97.

For the reasons given by Lady Hale, I too would hold that it was not until 6 January 2008, when the appellant must be deemed to have received the Bank’s letter of 4 January 2008, that the contractual right to terminate under paragraph 8.3 of the Handbook by the PILON (payment in lieu of notice) method was validly exercised and his employment with the Bank was terminated.

This issue directs attention to the provision in paragraph 13 of the Contract which provided that the employment “can” be terminated on the expiry of three months’ written notice of termination given by either side, and to paragraph 8.3 of the Handbook (the PILON provision) under which the Bank reserved “the right” to terminate the employment at any time with immediate effect by making a payment to the employee in lieu of notice. The judge held that there was no conflict between these provisions when the contract was construed as a whole. Paragraph 13 of the Contract could not be read as giving the appellant an unqualified right to three months’ notice of termination because other provisions in the contract such as paragraph 5.8 contained express rights to terminate it with immediate effect. So paragraph 8.3 of the Handbook qualified paragraph 13 of the Contract but was not in conflict with it. Rimer LJ said in para 29 that in his judgment the judge’s answer to this question was obviously correct.

Mr Cavender QC for the appellant said that his primary case was that there was no conflict between these two provisions. He described his argument that there was a conflict as a fall-back position. He said that there did not have to be a complete conflict to bring paragraph 17 into effect. Furthermore the way the Bank dealt with this case suggested that it was not its intention initially to rely on paragraph 8.3 of the Handbook. It only did so retrospectively. The termination of the appellant’s employment should be seen as having been on the basis that he was being given three months’ written notice of termination as provided for by paragraph 13.

It is not obvious that these two provisions are inconsistent with each other. Paragraph 13 of the Contract set out one way of terminating the contract, but it did not say that it is the only way. It used the word “can”, which suggests that it is a course of action that the Bank might take if it wants to. But the Bank reserved the right, as paragraph 8.3 of the Handbook put it, to use the PILON method. The provision in the Handbook can be read as qualifying the provision which is set out in the Contract. In any event the court’s duty, when confronted with two provisions in a contract that seem to be inconsistent with each other, is plain. It must do its best to reconcile them if that can conscientiously and fairly be done: Pagnan SpA v Tradax Ocean Transportation SA [1986] 2 Lloyd’s Rep 646, 653 per Steyn J. That approach, which was endorsed by Bingham LJ in the Court of Appeal [1987] 2 Lloyd’s Rep 342, 350, does not seem to me to give rise in this case to any difficulty.

I would therefore hold that this case must be approached on the basis that it was open to the Bank to use the PILON method which it had reserved to itself by paragraph 8.3 of the Handbook, and that this was what it was seeking to do when the appellant was called to the meeting on 29 November 2007 at which he was handed a letter which had been written on the Bank’s behalf.

The question to which this issue is directed is whether it is open to the appellant to maintain a claim or claims of damages against the Bank in view of the provisions of paragraph 5.16 of and Schedules 1 and 2 to the Contract by which, in consideration of the termination payment provided for by paragraph 5.15, the employee was to waive all contractual and statutory claims against the Bank.

The relevant provisions in paragraph 5.16 are as follows:

“In consideration for the Company making the Termination Payment … you will enter into a termination agreement with the Company (in the form of the draft termination agreement in Schedule 1 of this letter but amended to take account of any payments due to you under this letter and to take account of relevant legislative developments) under which you will waive all contractual and statutory claims against the Company and any Group Company (save for any pension rights accrued to the date of termination of your employment, any personal injury claims that you may have against the Company or any Group Company and save for any accrued rights you may have under the Deferral scheme and any share incentive scheme which will be dealt with subject to and in accordance with the rules of any such scheme) arising out of your employment with the Company and its termination … If the Company and you wish to amend the form of draft termination agreement further than as set out above, such amendments must be agreed within 28 days after the date on which your employment terminates …, failing which you and the Company will enter into the termination agreement in the form of the draft termination agreement in Schedule 1 of this letter only amended to take account of any payments due to you under this agreement and to take account of relevant legislative developments.” [emphasis added]

Schedule 1 is a draft letter addressed to the appellant which sets out the terms of the termination agreement referred to in paragraph 5.16. Paragraph 1 provides that he will receive his normal salary and benefits up to the termination date. Paragraph 2 provides that, subject to the other provisions of the letter, SG will make various payments to him. His entitlement to pay in lieu of notice, if appropriate, is preserved by paragraph 2(i). Paragraph 2(ii) is in these terms:

“[SG shall] pay you an amount of £ (less such deductions as SG is required by law to make) as [compensation for the termination of your employment REWORD AS APPROPRIATE TO INCLUDE SUCH OF THE PAYMENTS REFERRED TO IN SCHEDULE 2 OF THE LETTER AGREEMENT BETWEEN YOU AND THE COMPANY DATED [INSERT DATE] JANUARY 2005 TO WHICH YOU ARE ENTITLED IN ACCORDANCE WITH THE TERMS OF YOUR EMPLOYMENT DEPENDING ON THE CIRCUMSTANCES IN WHICH YOUR EMPLOYMENT TERMINATES] (this includes any entitlement you may have to a statutory redundancy payment)”

Schedule 2 sets out the payments that the Bank would make to him in the event of his employment being terminated by the Company in four alternative circumstances. In paragraph 3, which applies to the circumstances of this case, five sums which the appellant would have earned or to which he would have been entitled under the Contract on its termination are listed, including a compensation payment calculated in accordance with paragraph 5.15(b) of the Contract and a replacement bonus calculated in accordance with paragraph 5.24. The total amount, when computed, is to be inserted in paragraph 2(ii) of Schedule 1.

Paragraph 7(a) of Schedule 1 sets out a number of matters that the appellant is to be taken to have represented and warranted, including that he may have statutory claims for unfair dismissal and a redundancy payment, referred to as the “Alleged Claims”. Paragraphs 7(b) and (c) state:

“(b)

You hereby unconditionally and irrevocably waive the Alleged Claims, and neither you nor anyone else on your behalf will repeat, refer to or pursue the Alleged Claims.

(c)

You accept the payment to be given to you pursuant to this letter in full and final settlement of:-

(i)

the Alleged Claims; and

(ii)

all other claims and rights of action howsoever arising, which you (or anyone on your behalf) have or may have against SG, and/or any Group Company arising from or connected with your employment by SG and/or any Group Company or its termination,

with the exception that this paragraph 7(c) will not apply to any pension rights or pension benefits which have accrued to you up to the Termination Date or to any personal injury claim you may have. You represent and warrant that you are not aware of any personal injury claim subsisting at the date of this letter not [sic] aware of any basis on which you could bring any personal injury claim.”

Paragraph 7(e) sets out, as a fundamental term of the letter, that the payments to be given to him will at all times be conditional on his refraining from pursuing claims against SG or a Group Company and that, if he subsequently pursues such claims in breach of the letter, the payments made to him under the letter will be repayable to SG forthwith on demand. This is to be without prejudice to SG’s right to seek damages from him for the breach referred to and any other breach of the letter.

Mr Jeans QC for the Bank submitted that the purpose of these provisions was to achieve a clean break in the event of termination. It provided for a full and final settlement, the scope of which was defined by the draft agreement set out in Schedule 1. The appellant had the option not to comply with paragraph 5.16, if he thought that he would be better off by not doing so. In that event his claims against the Bank would not have been waived. The words “under this letter” in paragraph 5.16 (which I emphasised when setting out that paragraph in para 27, above) were to be read as referring to the Schedule 1 letter. That was the sense in which the words “pursuant to this letter” in paragraph 7(c) of Schedule 1 were to be read, so the words in paragraph 5.16 of the Contract should be read in the same way. The words in capital letters in paragraph 2(ii), read together with paragraph 3 of Schedule 2, set out all the sums to which the appellant was entitled by way of an amendment to the letter. All other claims, save for those specifically referred to in paragraph 5.16, were waived.

The judge disagreed with the Bank’s interpretation of these provisions. He said that he saw nothing self-evidently logical about an arrangement whereby the appellant could not be entitled both to accept the termination payments and to sue the Bank for damages for breach of contract: [2010] IRLR 950, para 98. It was not obvious why the appellant should be required to abandon a claim for breach of the tax efficiency obligation in paragraph 5.5 of the Contract in order to be entitled to a termination payment which he would equally have been entitled to receive if the Bank had performed its contractual obligation. To require him to give up the claim seemed to the judge to produce a windfall for the Bank, and the implications of its argument were even more unmeritorious in relation to a claim for damages for wrongful dismissal. The consequence of its argument was that the appellant could not pursue his claim for the losses he has suffered without losing his right to the termination payment to which he would equally have been entitled if the contract had been terminated lawfully. That would allow the Bank to profit from its own wrong – a result that seemed to him wholly unreasonable.

The Court of Appeal said that the answer to this issue depended on the correct interpretation of the Contract and its Schedules 1 and 2, and in particular on the relationship between paragraph 5.16 and paragraphs 7(c) and (e) of Schedule 1: [2011] IRLR 482, para 74. On its approach, the words “any payments due … under this letter” in paragraph 5.16 referred to the payments referred to in Schedules 1 and 2. On this reading, it was no part of the scheme of paragraph 5.16 that the termination agreement should include damages as part of the severance package. Paragraph 5.16 was to be interpreted as imposing a mutual obligation on the parties to enter into a termination agreement in the form of the Schedule 1 draft as appropriately amended by reference to Schedule 2. Once any disputes as to the amounts due under it are resolved, the parties are under an obligation to sign the termination agreement. When it is executed the paragraph 7(e) guillotine will fall, with the effect that the appellant will have to cease the pursuit of any pending claims for breach of contract against the Bank whether for wrongful dismissal or otherwise, or else forfeit the termination payments and face a claim for their repayment. The appellant will issue and pursue any new claims at his peril: para 89.

I agree with the Court of Appeal that paragraph 5.16 is to be interpreted as imposing a mutual obligation on the parties to enter into a termination agreement in the form of the draft set out in Schedule 1. As I read that paragraph, the appellant does not have an option not to comply with it as I understood Mr Jeans to have suggested. Paragraph 5.14 provides that, within 28 days after termination of the appellant’s employment, the Company “will … make a payment to you (the ‘Termination Payment’) as specified in paragraph 5.15”. The opening words of paragraph 5.16 tell the appellant what he must do in return:

“In consideration for the Company making the termination payment …. you will enter into a termination agreement with the Company (in the form of the draft termination agreement in Schedule 1 of this letter but amended to take account of any payments due to you under this letter and to take account of relevant legislative developments).”

It seems to me to be plain that these are mutual obligations binding on both parties to the agreement. The Bank is under an obligation to make the termination payment referred to in paragraphs 5.14 and 5.15. The appellant, for his part, is under an obligation to enter into the termination agreement. There is no provision on which he can rely which would entitle him to waive that obligation. If he fails to enter into the termination agreement, he will be in breach of contract and liable to the Bank in damages.

But I cannot agree with the Court of Appeal’s construction of paragraph 5.16. The crucial question is whether the words “under this letter” refer to the draft letter in Schedule 1 or to the entire agreement to which the appellant was invited by the letter of 28 January 2005 to indicate his acceptance. Mr Cavender said that those words should be read in the broader sense, with the result that the draft termination agreement in Schedule 1 was to be amended to take account of all payments due under and in consequence of the agreement, including claims for damages for wrongful dismissal and for a breach of paragraph 5.5. Mr Jeans, on the other hand, supported the meaning attached to those words by the Court of Appeal. He said that the words “under this letter” were to be read as referring to the draft letter in Schedule 1 without amendment, which made it plain that such claims were to be taken as waived.

Two phrases that appear in paragraph 5.16 tend to support Mr Cavender’s argument. The first is to be found in the words which immediately precede the words “under this letter” which we have to construe: “the draft termination agreement in Schedule 1 of this letter.” In that phrase the words “of this letter” must mean of the letter of 28 January 2005 and the contract enclosed with it, to which Schedule 1 is attached. It would be odd if the same words which follow so closely afterwards were to mean something different. The use of the words “the letter” in the second sentence of paragraph 17, which in that context must mean the letter of 28 January 2005, supports this interpretation. Then there is the phrase “amended to take account of any payments due to you under this agreement” which appears at the end of paragraph 5.16. The phrase in which the words I have emphasised appear contains a restatement of the amendment provision at the start of the paragraph where the word “letter” is used. The use of the words “under this agreement” at the end of the paragraph suggests that these words mean the same thing as the words “under this letter” were meant to convey. This does not, to say the least, fit easily with the submission that where the word “letter” is used it means the draft letter in Schedule 1.

For these reasons I am inclined to read the words “this letter” in the sense contended for by Mr Cavender. A desire for finality appears to have been the reason for the provisions in paragraph 5.16 on either of the two competing constructions. The termination letter would serve equally well as a definitive record of all the outstanding financial issues on the construction which I favour, although some of the more difficult issues will no doubt take longer to finalise. I am reinforced in taking this view by two other points which, taken together, seem to me to put the matter beyond doubt. The first is the unreasonable nature of the arrangement, if the Bank is right, for the reasons that the judge identified which I need not repeat but would respectfully endorse: see para 31, above. The second, which is closely linked to the first, raises an issue of principle.

The effect of paragraph 5.16, on the Bank’s interpretation, is to exclude any liability it may have to the appellant in damages for wrongful dismissal and for breach of the tax efficiency obligation in paragraph 5.5 as a consequence of his entering into the termination agreement, which he is bound to do. The approach that ought to be taken to the construction of clauses of this kind is well-established. In Canada Steamship Lines Ltd v The King [1952] AC 192, 208 Lord Morton of Henryton quoted with approval the principles applicable to clauses which purport to exempt one party to a contract from liability for negligence which were stated by Lord Greene MR in Alderslade v Hendon Laundry Ltd [1945] KB 189, 192. In summary, these principles are (1) that if the clause expressly exempts the party in whose favour it is made (the proferens) from liability for negligence, effect must be given to it; (2) if there is no express reference to negligence, the court must consider whether the words used are wide enough to cover it; and (3) if a doubt arises on this point it must be resolved in favour of the other party and against the proferens.

As Lord Dunedin said in W & S Pollock & Co v Macrae 1922 SC (HL) 192, 199, in order to be effective such clauses must be “most clearly and unambiguously expressed.” In Ailsa Craig Fishing Co Ltd v Malvern Fishing Co Ltd [1983] 1 WLR 964, 969H Lord Fraser of Tullybelton said that it was an ordinary principle that such conditions must be construed strictly against the proferens. The principle is commonly applied in cases where the contract which the other party has entered into with the proferens is in a standard form or in terms set out by the proferens which were not negotiable. The more improbable it is that the other party would agree to excluding the liability of the proferens, the more exacting the application of the principle will be.

The position in this case was that the terms of the employment contract were the product of negotiation between the parties. Nevertheless the exclusion clause was conceived in favour of the Bank. The provisions under which the appellant was required to waive all contractual and statutory claims against it, and thus to exempt the Bank from any liability in damages for breach of contract, are at first sight all embracing. But they are not without qualification. The critical words are those that indicate that the draft termination agreement in Schedule 1 may be amended to take account of payments “due to you under this letter.” In order to be effective to achieve what the Bank says it was meant to achieve the agreement had to be clearly expressed. At the very least for the appellant, for the reasons given above, the wording that was chosen was ambiguous. In this situation the ordinary principle must be applied. Any doubt that the wording gives rise to must be construed in favour of the appellant and against the Bank.

I would therefore hold for these reasons that, on a proper construction of paragraph 5.16 and Schedules 1 and 2 of the Contract, the appellant is entitled to maintain against the Bank a claim for damages for wrongful dismissal and a claim for an alleged breach of the tax efficiency provision in paragraph 5.5 and that, if he were to do so, he would not be in breach of the terms on which he is entitled to payment of the termination payment specified in paragraph 5.15.

I would allow the appeal, dismiss the cross-appeal and restore the order that was made by the Deputy High Court judge.

Lord Hope has identified the four issues in this appeal at paragraph 14 of his judgment. On the first issue, the repudiation issue, which is much the most important point in the appeal, I agree with everything which Lord Wilson says in support of the “elective” rather than the “automatic” theory of the termination of an employment contract for repudiatory breach. I also agree with the additional reasons given by Lord Hope for supporting that view. The automatic theory simply cannot work in cases of repudiatory breach which do not amount to express dismissal or resignation. Distinguishing between the two types of repudiation is both impracticable and unprincipled.

On the third and fourth issues, the conflict and paragraph 5.16 issues, I agree with the conclusions reached by Lord Hope. Paragraph 5.16 is not easy to construe, as demonstrated by the different constructions favoured in this court and in the courts below. I therefore share Lord Hope’s view that it was for the Bank, as author of the document which the appellant had to accept if he was to accept the job, to make the position crystal clear.

I turn, therefore, to the second issue, the termination issue. When was the contract terminated in accordance with its terms? In particular, having unsuccessfully attempted to dismiss the appellant summarily on 29 November 2007, when did the Bank succeed in operating the provision for payment in lieu of notice (the PILON clause)?

Amid the welter of case law and academic commentary upon the subjects of both wrongful and unfair dismissal, there appears to be remarkably little discussion of the requirements for a lawful dismissal under the terms of the employment contract. Ever since indefinite terms of employment became the norm, the courts have implied a term that either party may bring it to an end by giving notice (see S Deakin and GS Morris, Labour Law, 6th ed, 2012, paras 5.13, 5.14). In 1963, statute intervened to lay down minimum periods of notice to which the employee is entitled and a lesser period to which the employer is entitled (see now, Employment Rights Act 1996, sections 86 et seq). But the parties are, of course, free to provide expressly in their contracts for longer periods of notice. Statute also permits either party to waive his right to notice on any occasion or to accept a payment in lieu of notice (1996 Act, section 86(3)).

Statute is, however, silent as to the manner in which such notice is to be given. “Notice” is, of course, an ambiguous term. It can refer to the period between the time when an employer or employee is notified that the contract is to be terminated and the expiry of the specified period. Or it can refer to the notification itself. Or both. The statutory provisions focus upon the period of notice required. This is clear from section 86(6), which provides that the section does not affect the right of either party “to treat the contract as terminable without notice by reason of the conduct of the other party”.

Clause 13 of the Contract of employment between the Bank and the appellant, as is usual, dealt with both the manner of notification and the period of notice required: “Your employment can be terminated on the expiry of 3 months’ written notice of termination given by you to the Company or by the Company to you.” The Contract itself contained no provision for payment in lieu of notice (a PILON clause). Clause 17 stated that the contract “is in conjunction with the offer letter, the Staff Handbook of the SGUK Group (as amended from time to time) and the SGUK Compliance Manual which, together with this letter, form the written particulars of employment as required by law”. Clause 18 stated that the Contract, Part 1 of the Staff Handbook of the SGUK Group and the SGUK Compliance Manual “contain the entire understanding between you and the Company”. The wording of these two clauses leaves open the possibility that the Staff Handbook is not, in fact, a contractual document, but rather part of the employer’s rules by which the employee has agreed to abide. This is an interesting question of academic debate, but the point has wisely not been taken on either side in this case. We have proceeded on the basis that the Handbook does indeed form part of the contract between them.

Paragraph 8 of section 1 of the Handbook is set out in full at paragraph 7 of Lord Hope’s judgment. Paragraph 8.1 deals with the employee’s right to notice. In relation to the period of notice, it adds nothing to what would otherwise be the position: the employee is entitled to whichever is the longer of the period specified in his contract or the statutory minimum (the Handbook does not state this in exactly the same terms as section 86(1) of the 1996 Act, but it comes to the same result). In relation to the manner of notification, however, it does add something. It refers to “your entitlement to written notice” and provides for when such notice is deemed to have been given. Even if there were no entitlement to notice in writing in the Contract, therefore, there would clearly be an entitlement to notice in writing under the Handbook.

Paragraph 8.2 deals with the employee’s obligation to give notice. Unlike paragraph 8.1, it is drafted on the assumption that the Contract will provide for the period of written notice to be given by the employee. It does not set out the statutory position as a default. But in both cases the notification given has to be in writing.

Paragraph 8.3 deals with termination by the Bank and payment in lieu of notice. Such PILON clauses are very common in contracts of employment and no doubt this clause is in a form which is also common. Its object is to dispense with the period of notice. The employer “reserves the right to terminate your employment with immediate effect by making a payment to you in lieu of notice”. It says nothing about whether and how the employee is to be notified that his employment is at an end. Is it enough that the payment in lieu is actually made? Or is something more than that required? And if so, what? The resolution of these questions is of great importance to the large numbers of employees and employers who are party to PILON clauses in this form.

Mr Cavender, for the appellant, argues that paragraph 8.3 is dealing only with the period of notice. It allows the Bank to cut this short. It does not deal with the manner of notification. It cannot operate in isolation from clause 13 of the Contract and must be construed alongside that clause. It does nothing to detract from the requirement in clause 13 (and in every other clause of the Contract and Handbook dealing with notification of termination of employment) for notification in writing. Payment into the bank account was not enough, because it was not accompanied by notification in writing that the Bank was terminating his employment by making a payment in lieu of the notice period. The letter of 29 November (set out at paragraph 9 of Lord Hope’s judgment) was not enough to cure that omission, because it did not notify the appellant that that was what the Bank intended to do (indeed, it is not clear that that is what it intended to do on that day). In any event, it could not put the burden upon him of checking whether and when the money had reached his bank account. It had a duty to notify him at that time. The first proper notification which the Bank gave him was the letter of 4 January 2008, set out at paragraph 12 of Lord Hope’s judgment. This was the first time that he was told, clearly and unambiguously and in writing, that the Bank had exercised its right to terminate his employment with immediate effect by making him a payment in lieu of notice. He accepts, therefore, that his contract was validly terminated on 6 January when he was deemed to have received that letter.

In support of his argument, Mr Cavender relies on the general principle that notices to determine contracts should be unambiguous and unequivocal and leave the recipient in no doubt as to the contractual right being invoked. He relies in particular upon the well-known passage in the opinion of Lord Steyn in Mannai Investment Co Ltd v Eagle Star Life Assurance Co Ltd [1997] AC 749, 768:

“Making due allowances for contextual differences, such notices [under a break clause in a lease] belong to the general class of unilateral notices served under contractual rights reserved, eg notices to quit, notices to determine licences and notices to complete: Delta Vale Properties Ltd v Mills [1990] 1 WLR 445, 454E-G. To those examples may be added notices under charter parties, contracts of affreightment, and so forth. Even if such notices under contractual rights reserved contain errors they may be valid if they are ‘sufficiently clear and unambiguous to leave a reasonable recipient in no reasonable doubt as to how and when they are intended to operate:’ the Delta case at p 454E-G, per Slade LJ and adopted by Stocker and Bingham LJJ; see also Carradine Properties Ltd v Aslam [1976] 1 WLR 442, 444. That test postulates that the reasonable recipient is left in no doubt that the right reserved is being exercised. It acknowledges the importance of such notices. The application of that test is principled and cannot cause any injustice to a recipient of the notice.”

Although that case was concerned with the effect of a mistake in an otherwise clear and unambiguous notice, the principle is clear. The reasonable recipient has to be told that the right is being exercised, how and when it is intended to operate. This was not done in this case.

Mr Jeans, on behalf of the Bank, argues that paragraph 8.3 is simplicity itself. The act of making payment brings the employment to an end. There is no requirement of notification. But in any event, Mr Geys knew from the letter of 29 November and later correspondence that the Bank was sacking him, although this did not spell out the basis upon which it was doing so. He knew of the payment into his account before the end of December. The trial judge found that he had probably guessed that the most likely explanation for the credit was that it was a payment in lieu of notice. So even if there is some requirement of notification, this was enough. So his employment ended before the end of 2007, which is the crucial date for the calculation of his termination payment.

In my view, it is quite clear that paragraph 8.3 is not dispensing with whatever requirement there is that the employee be notified of the termination of his employment. The words in brackets (“or, if notice has already been given, the balance of your notice period”) draw a clear distinction between the notice period and notification of the termination of employment and thus strongly suggest that the word “notice” which precedes them also refers to the notice period. The question therefore becomes, to what notification was the employee entitled under the express or implied terms of his contract of employment?

In this connection, it is important to distinguish between two different kinds of implied terms. First, there are those terms which are implied into a particular contract because, on its proper construction, the parties must have intended to include them: see Attorney General of Belize v Belize Telecom Ltd [2009] UKPC 10, [2009] 1 WLR 1988. Such terms are only implied where it is necessary to give business efficacy to the particular contract in question. Second, there are those terms which are implied into a class of contractual relationship, such as that between landlord and tenant or between employer and employee, where the parties may have left a good deal unsaid, but the courts have implied the term as a necessary incident of the relationship concerned, unless the parties have expressly excluded it: see Lister v Romford Ice & Cold Storage Co Ltd [1957] AC 555, Liverpool City Council v Irwin [1977] AC 239.

A great deal of the contractual relationship between employer and employee is governed by implied terms of the latter kind. Some are of long-standing, such as the employer’s duty to provide a safe system of work. Some are of more recent discovery, such as the mutual obligations of trust and confidence. This was referred to by Dyson LJ in Crossley v Faithful and Gould Holdings Ltd [2004] IRLR 377 as an “evolutionary process”. He also described the “necessity” involved in implying such terms as “somewhat protean”, pointing out that some well-established terms could scarcely be said to be essential to the functioning of the relationship. At para 36, he said this:

“It seems to me that, rather than focus upon the elusive concept of necessity, it is better to recognise that, to some extent at least, the existence and scope of standardised implied terms raise questions of reasonableness, fairness and the balancing of competing policy considerations.”

There is much to be said for that approach, given the way in which those terms have developed over the years.

Whatever the test to be applied, it seems to me to be an obviously necessary incident of the employment relationship that the other party is notified in clear and unambiguous terms that the right to bring the contract to an end is being exercised, and how and when it is intended to operate. These are the general requirements applicable to notices of all kinds, and there is every reason why they should also be applicable to employment contracts. Both employer and employee need to know where they stand. They both need to know the exact date upon which the employee ceases to be an employee. In a lucrative contract such as this one, a good deal of money may depend upon it. But even without that, there may be rights such as life and permanent health insurance, which depend upon continuing to be in employment. In some contracts there may also be private health insurance. A person such as Mr Geys, going on holiday over Christmas and the New Year, needs to know whether he should be arranging these for himself. At the other end of the scale, an employee who has been sacked needs to know when he will become eligible for state benefits.

It is necessary, therefore, that the employee not only receive his payment in lieu of notice, but that he receive notification from the employer, in clear and unambiguous terms, that such a payment has been made and that it is made in the exercise of the contractual right to terminate the employment with immediate effect. He should not be required to check his bank account regularly in order to discover whether he is still employed. If he does learn of a payment, he should not be left to guess what it is for and what it is meant to do.

This is not an unreasonable requirement to place upon an employer (or indeed upon an employee giving notice). When an employer sacks an employee it ought to know what it is doing: is it with immediate effect or on notice? If it is with immediate effect, is it because of some misconduct on the part of the employee or in the exercise of a PILON clause? It is not good enough to purport summarily to dismiss the employee without stating a cause and without making a payment, then to realise that there is no right to do that, but that there is the right to terminate under a PILON clause, and so decide to exercise that right without telling the employee that the right is being exercised and the payment has been made.

Given that such a notice is a necessary incident of the relationship, a wise employer would take care to give it in writing. But if the contract does not require writing, it would be possible for an employer to hand over the correct money and clearly state at the same time that this brings the employment to an immediate end, in place of the notice period to which the employee would otherwise be entitled. In the days when wages were normally paid in cash, this would have been a common practice. But if, as is now common, payment is made direct to the employee’s bank account, the employee’s bank is his agent for the receipt of payment, but it is not without more his agent for the receipt of notification of what the payment is for. That notification has to be given to the employee.

On any view, such clear and unambiguous notification was not given in this case. The Bank could easily have done things properly. But for whatever reason they did not do so. Subject, therefore, to the repudiation issue, it was not until 6 January 2008, when Mr Geys must be deemed to have received the Bank’s letter of 4 January 2008, that the contractual right to terminate under the PILON method provided for by paragraph 8.3 of the Handbook was validly exercised and his employment with the Bank came to an end.

In para 14 above Lord Hope helpfully identifies the four issues before the court. I agree with his proposed resolution of the third issue (the conflict issue) and the fourth issue (the para 5.16 issue). I also agree with the resolution of the second issue (the termination issue) proposed by Lady Hale. I address the first issue (the repudiation issue).

In the absence of any direct authority of real weight at this level, the court is required to make a difficult and important choice between a conclusion that a party’s repudiation (albeit perhaps only an immediate and express repudiation) of a contract of employment automatically terminates the contract (“the automatic theory”) and a conclusion that his repudiation terminates the contract of employment only if and when the other party elects to accept the repudiation (“the elective theory”). It is common ground that, whichever theory be chosen, it should apply equally to wrongful repudiations by employers (i.e. wrongful dismissals) and wrongful repudiations by employees (i.e. wrongful resignations); and it is only for convenience, and because it is reflective of the facts of the present case, that I will, at times, refer to the wrongful repudiator as the employer and to the innocent party as the employee.

In light of the fact that a central incident of the automatic theory is that, upon the automatic termination of the contract, the innocent party has a right to damages, the first question must be whether it matters that the contract is terminated forthwith upon repudiation or, instead, survives until some further, terminating, event? The answer is that sometimes it does matter. It depends on the terms of the contract. The date of termination fixes the end of some contractual obligations and, sometimes, the beginning of others. An increase in salary may depend on the survival of the contract until a particular date. The amount of a pension may be calculated by reference to the final salary paid throughout a completed year of service or to an aggregate of salaries including the final completed year. An entitlement to holiday pay may similarly depend on the contract’s survival to a particular date. In some cases an award of damages will compensate the employee for any such loss. But often it will fail to do so. Such failure flows from application of the “least burdensome” principle, namely that damages should reflect only the losses sustained by the employer’s decision to repudiate the contract unlawfully rather than by his having hypothetically proceeded, in the manner “least profitable to the plaintiff, and the least burthensome to the defendant”, to terminate the contract lawfully: see Cockburn v Alexander (1848) 6 CB 791, 136 ER 1459, at pp 814 and 1468, (Maule J), and McGregor on Damages, 18th ed (2009) para 8-093. So, where under the terms of the contract it had been open to the wrongfully repudiating employer to have taken a course which would have terminated the contract quickly as well as lawfully, the damages will be small.

These propositions are well demonstrated by the facts of the present case. Lord Hope explains in para 6 above why the appellant’s termination payment would be substantially increased if his contract of employment were to have terminated after 31 December 2007. Had the effect of the Bank’s wrongful repudiation been to terminate it on or prior to that date, his damages would not cover his loss of the increase in payment. For, as Lady Hale observes in para 61 above, it would have been easy for the Bank lawfully to have operated the PILON clause in para 8.3 of the Handbook. Indeed it could, by proper operation of that clause, lawfully have dismissed the appellant on 29 November 2007 itself. So his damages for the Bank’s unlawful repudiation of the contract on that date would, by application of the least onerous principle, be no more than nominal. Superficially, however, it may be said to be paradoxical that the principle should demand a hypothesis that the Bank would have operated the PILON clause immediately and validly in circumstances in which in fact it delayed its attempted operation of the clause until 18 December 2007 and thereafter, until 6 January 2008, it operated it invalidly.

The central task in this part of the appeal is therefore to identify the date when the appellant’s contract terminated; and, in my respectful view, it is not, as Lord Sumption suggests in para 120 below, to analyse the enforceability of what he calls the core obligations. He proceeds to suggest in para 140 below that the application of the elective theory, of which the result, of course, would be to exclude a conclusion that the contract terminated on 29 November 2007, would “give rise to significant injustice in this case”. There, with respect, I part company with Lord Sumption. Before I consider the detail of the authorities, I find it helpful to stand back and to remind myself of the overall effect of the automatic theory. It is to reward the wrongful repudiator of a contract of employment with a date of termination which he has chosen, no doubt as being, in the light of the terms of the contract, most beneficial to him and, correspondingly, most detrimental to the other, innocent, party to it. We must, I suggest, be very cautious before turning basic principles of the law of contract upon their head so that, in this context, breach is thus to be rewarded rather than its adverse consequences for the innocent party negatived. It is, says Professor Freedland in The Personal Employment Contract, 2003, at p 390 “a matter of concern if the common law of wrongful dismissal functions so as to invite opportunistic breach of contract”. My view of the location of the justice of the case is opposite to that of Lord Sumption: it is that, in that the Bank failed to operate its own PILON clause lawfully until after 31 December 2007, it should not be able to revert to its unlawful act on 29 November as the reason why the contract did not survive for the final 32 days of the year.

In the jurisprudence of England and Wales prior to the decision in Gunton v Richmond-upon-Thames London Borough Council [1981] Ch 448, the fullest analysis of the rival advantages of the automatic and the elective theories, in the light of such earlier relevant authority as existed, was conducted by Sir Robert Megarry VC in Thomas Marshall (Exports) Ltd v Guinle [1979] Ch 227. As Warner J observed in Irani v Southampton and South West Hampshire Health Authority [1985] ICR 590, 598, the Vice-Chancellor’s analysis was powerfully reasoned. There it was the employee who, following various breaches of contract on his part, wrongfully repudiated it by purported resignation half way through its fixed term. The employer sought interlocutory injunctions that he should neither solicit its customers nor use information confidential to it. Towards the end of his analysis, which begins on p 236, the Vice-Chancellor said, at p 243:

“Above all, I think the courts must be astute to prevent a wrongdoer from profiting too greatly from his wrong... [W]hy should the court’s inability to make a servant work for his employer mean that as soon as the servant refuses to do so the court is forthwith disabled from restraining him from committing any breach, however flagrant, of his other obligations during the period of his contract? I would wholly reject the doctrine of automatic determination …”

The Vice-Chancellor thereupon proceeded, at p 247, to make both of the requested injunctions on the basis that they were in support of the employee’s implied duty of fidelity and good faith which, as the Vice-Chancellor had expressly noted at p 243, bound the employee only for as long as the contract subsisted.

Contracts of employment often include provisions which are expressed to bind the parties following the termination of the contract: Rhys-Harper v Relaxion Group plc [2003] UKHL 33, [2003] ICR 867, para 36 (Lord Nicholls). For example, they may oblige the employee not to compete with the employer for a specified period nor to use information which he has obtained in confidence during the period of his employment. Or, as in the present case, they may oblige the employer, within a specified period following termination of the contract, to make to the employee a termination payment, to be calculated in accordance with terms specified in it, and may oblige the employee, in consideration of the payment, to enter into a termination agreement on terms also therein specified. Such provisions of the contract are, by their terms, enforceable following its termination. The enforceability of, for example, a restrictive covenant by the repudiator against the innocent party is now the subject of some debate: Rock Refrigeration Ltd v Jones [1997] ICR 938. There is no problem about the enforceability of such provisions against the repudiator. But authorities to that effect shed no light on the issue between the elective and the automatic theories because the provisions do not depend on the survival of the contract.

By contrast, however, authorities in which, following an unaccepted wrongful repudiation, provisions which do notsurvive the termination of the contract have been enforced againstthe repudiator must, in my view, be taken to be examples of the operation of the elective theory. Adoption by this court of the automatic theory would leave them unjustifiable. For example, the Thomas Marshall case was far from being the first example of the enforcement of a covenant against competition during the contract and following its wrongful repudiation.

Thus, in Lumley v Wagner (1852) 1 De GM & G 604, 42 Eng Rep 687, Miss Wagner agreed to sing operatic roles for Mr Lumley for the months of April, May and June 1852, at Her Majesty’s Theatre and not to sing elsewhere during that period. She wrongfully repudiated the contract and proposed, instead, to sing for Mr Gye at the Royal Italian Opera, Covent Garden. The Lord Chancellor, Lord St Leonards, acknowledged, at pp 619 and 693, that he could not order Miss Wagner to sing for Mr Lumley. But he held that he could, and should, order her not to sing for Mr Gye; and it is clear from Mr Lumley’s pleading, set out at pp 607 and 689, that the injunction was to endure only “during the existence of the agreement”, i.e. until 30 June 1852.

In Whitwood Chemical Co v Hardman [1891] 2 Ch 416, Lindley LJ observed, at p 428, that he regarded the decision in Lumley v Wagner “as an anomaly to be followed in cases like it, but an anomaly which it would be very dangerous to extend”. He made clear that the danger was that its extension might represent a movement towards the specific performance of a contract of employment. He did not suggest that there was any anomaly in the analysis that Miss Wagner’s contract had continued notwithstanding her repudiation of the contract. Indeed in William Robinson and Co Ltd v Heuer [1898] 2 Ch 451 the Court of Appeal, in a constitution over which the same judge, then Master of the Rolls, presided, made an injunction, analogous to that made against Miss Wagner, against an employee who had wrongfully resigned after three of his five contractual years of service and who was breaking his covenant not, “during this engagement”, to work for any rival business. The court expressly, held, at p 458, that the employee’s engagement continued; and it made an injunction against his working for a rival business for the remaining two years.

In Warner Brothers Pictures Inc v Nelson [1937] 1 KB 209 Bette Davis had, in 1934, entered into a contract of employment with Warner Brothers which, at their option, could continue until 1942. In 1936 she repudiated the contract and proposed to break her covenant not, during its currency, to participate in any other film for any other company. Branson J, at p 222, enjoined her from doing so “during the continuance of the contract or for three years from now, whichever period is the shorter”.

Into a different, yet equally significant, category fall cases in which an employer wrongfully repudiates a contract of employment in circumstances in which its terms require him to have implemented a disciplinary procedure. The law is clear that an injunction may issue so as to enforce the requirement; and the absence of a right to claim damages for breach of a duty to follow a disciplinary procedure (see Edwards v Chesterfield Royal Hospital NHS Foundation Trust [2011] UKSC 58, [2012] 2 AC 22) makes the availability of the injunction particularly precious. But it is self-evident that, had the wrongful repudiation already automatically terminated the contract, an injunction would not issue so as to require observance of a procedure designed to determine whether the employer was entitled to terminate it.

Thus in Jones v Lee [1980] ICR 310 the managers of a Roman Catholic school wrongfully dismissed its headmaster following his divorce and remarriage to a former teacher at the school. The dismissal was wrongful because it was in breach of a term of his contract of employment which gave him the right to a hearing before the local education authority prior to his dismissal. The Court of Appeal enjoined the managers from dismissing him or purporting to dismiss him prior to any such hearing. In the Irani case, cited above, Warner J made an analogous injunction. In Robb v Hammersmithand Fulham London Borough Council [1991] ICR 514 the claimant, who was Hammersmith’s director of finance, had been responsible for speculative, indeed unlawful, investments of its funds. Hammersmith invoked a contractual disciplinary procedure with a view to dismissing him for lack of capability but it abandoned the procedure and wrongfully dismissed him with immediate effect. Morland J granted him an injunction so as to restrain Hammersmith from giving effect to its purported dismissal of him and, as the judge explained at p 523, so as to restore his entitlement to the ventilation of his defence through the disciplinary procedure.

In my view the proponents of the automatic theory fail to explain how the competition and the disciplinary cases are consistent with it. To describe them as examples of the enforcement only of collateral obligations would, I believe, be to fail to engage sufficiently with their significance.

How and when did the automatic theory take hold? To what extent has it taken hold? To what extent should it take hold?

Equity took the view that the remedy of specific performance, or analogous injunction, should not be available so as to require an employee who had wrongfully resigned to go back to work or to require an employer who had wrongfully dismissed the employee to take him back. “[T]he courts,” said Sir George Jessel, Master of the Rolls, in Rigby v Connol (1880) 14 ChD 482, 487, “have never dreamt of enforcing agreements strictly personal in their nature, whether they are agreements of hiring and service, being the common relation of master and servant, or ...”. In Chappell v Times Newspapers Ltd [1975] ICR 145 Geoffrey Lane LJ explained, at p 178, that “if one party has no faith in the honesty or integrity or the loyalty of the other, to force him to serve or to employ that other is a plain recipe for disaster”. This has made a contract of employment into a special case – but only in terms of remedies. Indeed where, notwithstanding an employer’s wrongful repudiation, trust and confidence between the parties have not been forfeit, an injunction, analogous to specific performance, may be granted to restrain implementation of its purported notice: Hill vCA Parsons & Co Ltd [1972] Ch 305. The big question whether nowadays the more impersonal, less hierarchical, relationship of many employers with their employees requires review of the usual unavailability of specific performance has been raised, for example by Stephenson LJ in the Chappell case, at p 176, but is beyond the scope of this appeal.

Where did the unavailability of specific performance leave the wrongly dismissed employee? Specifically, could he sue for his wages on the basis that at any rate he had remained ready, able and willing to resume his work for the employer? The Victorian work ethic helped to provide a negative answer. In Goodman v Pocock (1850) 15 QB 576, 117 ER 577, Erle J said at pp 583-584 and 580:

“I think that the servant cannot wait till the expiration of the period for which he was hired, and then sue for his whole wages on the ground of a constructive service after dismissal. I think the true measure of damages is the loss sustained at the time of the dismissal. The servant, after dismissal, may and ought to make the best of his time; and he may have an opportunity of turning it to advantage.”

Ever since then the law has been clear that, save when, unusually, a contract of employment specifies otherwise, the mere readiness of an employee to resume work, following a wrongful dismissal which he has declined to accept, does not entitle him to sue for his salary or wages. “He cannot”, as Salmon LJ said in Denmark Productions Ltd v BoscobelProductions Ltd [1969] 1 QB 699, 726, “sit in the sun...”. The law takes the view that it is better for the employee (as well, of course, as for the employer) that his claim for loss of wages or salary should be confined to a claim for damages and therefore be subject to his duty to mitigate them by taking all reasonable steps to find other work. This principle is not without its critics. In Boyo v Lambeth London Borough Council [1994] ICR 727, 747 Staughton LJ observed that, unconstrained by authority, he would not have accepted it; and, in his dissenting judgment in Cerberus Software Ltd v Rowley [2001] ICR 376, Sedley LJ suggested, at p 386, that it was “one of the great unresolved questions of employment law”. But, even if the question can be said to be unresolved, this court is not invited to resolve it. The facts of this appeal leave no room for an attack on the principle. It has added to the making of a contract of employment into a special case – but, again, only in terms of remedies.

Until 1955 there was no suggestion in the jurisprudence of England and Wales – or elsewhere in the world of the common law – that a wrongful repudiation of a contract of employment automatically brought it to an end. The need for the innocent party to elect whether to accept the repudiation, in accordance with general principles of the law of contract, was taken as read: see, for example, Boston Deep Sea Fishing and Ice Co v Ansell (1888) 39 Ch D 339, 365 (Bowen LJ) and General Billposting Co Ltd v Atkinson [1909] AC 118, 122 (Lord Collins).

Then came the important decision of the High Court of Australia in Automatic Fire Sprinklers Proprietary Ltd v Watson (1946) 72 CLR 435. Its clear indorsement of the elective theory still holds sway in Australia: Byrne v Australian Airlines Ltd (1995) 131 ALR 422. In the Automatic Fire Sprinklers case the employer’s purported dismissal of the employee was wrongful for two reasons. First, it was in breach of contract. Second, it was in breach of a war-time regulation which prohibited his dismissal without the consent of the Director-General of Man Power. The employee did not accept the wrongful repudiation and sued for his salary for the year which followed it. In relation to the first issue the court was unanimous: it was that, although at common law the contract was not at an end, the employee was nevertheless not entitled to sue for his salary. But the way in which it expressed that conclusion may, in retrospect, have been significant: for it said that, although the contract was not at an end, the relationship was at an end. Thus Latham CJ, who was in the minority only on the second issue, said, at p 451:

“Thus the wrongful dismissal determines the relationship of master and servant created by the contract, even though the servant may not have accepted his dismissal as entitling him to regard the contract as discharged. Any other view would in effect grant specific performance of a contract of personal service, a remedy which the courts have always refused in such a case.”

The second issue related to the effect of the breach of the regulation; the majority held that its effect was to preclude the termination even of the relationship of master and servant, with the result that the employee was entitled to recover his salary.

The High Court’s reference to the termination of the relationship of master and servant, as distinct from the terminationof their contract, was no more than its convenient short-hand for the common law’s long rejection of a claim for wages or salary. Some subscribers to the elective theory have considered the distinction useful. Thus in the Gunton case,[1981] Ch 448, Brightman LJ explained, at pp 474-475, that although a wrongful dismissal, if not accepted, left the contract in being, the status, or relationship, of the parties to it no longer existed and that obligations not necessarily dependent on the existence of the relationship might alone survive. But other subscribers to the elective theory have criticised the distinction. In Dietman v Brent London Borough Council [1987] ICR 737, Hodgson J referred to it, at p 753, as “a little difficult to understand”. In their article entitled Theories of Termination in Contracts of Employment: the Scylla and Charybdis, (2003) 19 JCL 134, Hough and Spowart-Taylor described it, at p 144, as “deeply problematic”. I myself regard the distinction as unhelpful, indeed confusing. It has offered easy pickings for proponents of the automatic theory, whom it enables to argue, with superficial force, that, if the wrongful repudiation terminates the relationship, it must also then terminate the contract.

The automatic theory made its appearance in the jurisprudence of England and Wales in 1955 almost in parenthesis. The case of Vine v National Dock Labour Board [1956] 1 QB 658 (CA) and [1957] AC 488 (HL), concerned a registered dock worker employed by the Board on terms set by a statutory scheme. The Board wrongfully dismissed him and the House of Lords, reversing the majority decision of the Court of Appeal, held, by reference to the terms of the scheme, that the trial judge had been entitled to declare that his dismissal had been invalid. In his dissenting judgment in the Court of Appeal, Jenkins LJ, at p 674, contrasted the effect of the scheme with the “ordinary case of master and servant” in which – so he proposed - “the repudiation or the wrongful dismissal puts an end to the contract, and the contract having been wrongfully put an end to a claim for damages arises”. In the House of Lords Viscount Kilmuir LC, said, at p 500:

“This is an entirely different situation from the ordinary master and servant case; there, if the master wrongfully dismisses the servant, either summarily or by giving insufficient notice, the employment is effectively terminated, albeit in breach of contract. Here, the removal of the plaintiff’s name from the register being, in law, a nullity, he continued to have the right to be treated as a registered dock worker with all the benefits which, by statute, that status conferred on him. It is therefore right that, with the background of this scheme, the court should declare his rights.”

Although there may be some ambiguity in his use of the word “effectively”, the Lord Chancellor is generally there taken to have indorsed the proposition of Jenkins LJ in support of the automatic theory. The basis of the proposition – which, as will already be clear, played no part in the reasoning of the decision of the House - had been, and remained, unexplained. In its recent affirmation of the elective theory in Paper Reclaim Ltd v Aotearoa International Ltd [2007] 3 NZLR 169, the Supreme Court of New Zealand referred, at para 18, to “the difficulty of the proposition as a statement of law, as opposed to a statement of practical consequence for the employee”.

Two months after its decision in the Vine case the appellate committee heard McClelland v Northern Ireland General Health Services Board [1957] 1 WLR 594. The respondent had given six months’ notice of termination of the appellant’s employment as a clerk. But, in the absence of her gross misconduct, incapacity or ill-health, there was no express provision in the contract for the respondent to terminate it, whether on six months’ notice or reasonable notice or otherwise. The majority of the committee held that such a provision could not be implied; and accordingly it declared that her contract had not been validly terminated. No reference was made to the decision in the Vine case. The elective theory was applied without argument to the contrary.

Mr Jeans presents the judgment of the Privy Council in Francis v The Municipal Councillors of Kuala Lumpur [1962] 1 WLR 1411 as an example of the application of the automatic theory. There is no doubt that the employee, wrongfully dismissed, was confined to a claim for damages. But part of the Board’s analysis was inconsistent with the theory. Lord Morris of Borth-y-Gest said, at p 1417-1418:

“In their Lordships’ view, when there has been a purported termination of a contract of service a declaration to the effect that the contract of service still subsists will rarely be made. This is a consequence of the general principle of law that the courts will not grant specific performance of contracts of service. Special circumstances will be required before such a declaration is made... there are no circumstances in the present case which would make it either just or proper to make such a declaration.”

In Decro-Wall International SA v Practitionersin Marketing Ltd [1971] 1 WLR 361, the defendant had contracted to be the exclusive marketeer in the UK of tiles manufactured by the claimant in France. So it was a contract for the provision of services. One of the questions before the Court of Appeal was whether, as the claimant contended, its own wrongful repudiation of the contract had automatically brought it to an end. The claimant relied on dicta of Lord Reid in White and Carter (Councils) Ltd v McGregor [1962] AC 413, 428 and 429, which (so the claimant said) suggested that, where, following a wrongful repudiation of a contract for the provision of services, the completion of the contract by the innocent party would have required the repudiator’s cooperation, the repudiation automatically brought the contract to an end. The court explained that Lord Reid’s remarks could not bear the weight thus sought to be placed on them and that, in the passage quoted by Lord Sumption at para 114 below, Lord Hodson, with whom Lord Tucker had agreed, had expressly reaffirmed the survival of the contract beyond repudiation until acceptance and irrespective of the availability of specific performance: see p 370 (Salmon LJ), p 375 (Sachs LJ) and p 381 (Buckley LJ). “It is”, said Sachs LJ at p 375, “the range of remedies that is limited, not the right to elect”. Salmon LJ, at p 370, and Sachs LJ, at p 376, also, in passing, expressed their provisional rejection of the application of the automatic theory to a contract of employment but Buckley LJ, at p 381, left that point open.

In paragraphs 114 and 115 below, albeit under the rubric only of “The general law”, Lord Sumption lays stress on Lord Reid’s dicta in the White and Carter case. I agree with the Court of Appeal’s treatment of them in the Decro-Wall case. In particular Lord Reid was not addressing the enforceability of terms of a contract of employment which are not dependent on mutual cooperation and thus, in that context, the wider question of the proper treatment of a wrongful repudiation.

In Sanders v Ernest A Neale Ltd [1974] ICR 565 the National Industrial Relations Court, of which the President was Sir John Donaldson, dismissed an appeal by employees against the conclusion of an industrial tribunal that their dismissals had not been attributable to redundancy. The first question was when their dismissals had occurred. The court assumed that the employer had wrongfully repudiated their contracts. It held that it had thereby automatically terminated them; and it proceeded to conclude that the tribunal had been right to hold that the terminations had not been attributable to redundancy. At pp 570-571 Sir John addressed the validity of the proposition that “a servant cannot sue for wages if he has not rendered services, and the wrongful dismissal prevents him rendering services”. He proceeded as follows:

“It being admitted that a wrongful dismissal does prevent a servant from so suing, there must be some other explanation. The obvious, and indeed the only, explanation is that the repudiation of a contract of employment is an exception to the general rule. It terminates the contract without the necessity for acceptance by the injured party.”

Six years later, in the Gunton case, Buckley LJ was to observe, at p 466, that, to the best of his knowledge, the Sanders case was the only ordinary employment case in which the automatic theory was part of the basis of the decision. But if, as also appears to me, it was in that sense the high-water mark of the automatic theory, it was scarcely the result of a flood tide. Sir John Donaldson’s reasoning was to jump from the absence of some remedies to the absence of all rights, heedless in particular of contractual rights other than to payment of wages or salary. As Deakin and Morris state in Labour Law, 2012, 6th ed, para 5.38, application of the automatic theory is “a case of the tail wagging the dog”. I am a late convert to the cliché as an effective means of explaining a point; and another, apt to the context, would involve babies and bath-water. In his article entitled Remedies for Breach of the Contract of Employment [1993] CLJ 405, Professor Ewing wrote, at pp 410-411:

“So the rights of the parties are to be driven and determined by the availability of remedies; the contract is automatically terminated by the unilateral repudiation of either party, simply because it is not capable of specific performance. As such the argument is hopelessly circular.”

The circularity is that there is no remedy so there is no right so there is no remedy. The professor proceeded, at p 415, to describe the automatic theory as “a bastard doctrine, which is difficult to reconcile with the general principles of contract law”. In Treitel: The Law of Contract, 13th ed, 2011, at para 18-006, Professor Peel identifies other types of contract, such as a sale of goods or a charter of a ship, in which, following a wrongful repudiation, the innocent party may be unable to require full payment under the contract yet in which no doubt is raised about the continuation of the contract pending his election.

In the Gunton case [1981]Ch448 the employer wrongfully repudiated the employee’s contract of employment by dismissing him for disciplinary reasons without complying with the contractual disciplinary procedure. The Court of Appeal held that, if (which Shaw LJ doubted) the termination of the contract depended upon his having accepted the wrongful repudiation, the employee had nevertheless done so. Therefore the question of his obtaining an injunction analogous to that in Jones v Lee [1980] ICR 310 did not arise; and the decision related to the appropriate measure of his damages. But there was a discussion about the automatic theory, which Shaw LJ favoured, and the elective theory, which Buckley and Brightman LJJ favoured. Shaw LJ referred to the basic principle of the common law, which afforded to the innocent party a right to elect whether to accept a wrongful repudiation and claim damages or to call for performance in accordance with the contract. He proceeded, at p 459:

“This practical basis for according an election to the injured party has no reality in relation to a contract of service where the repudiation takes the form of an express and direct termination of the contract in contravention of its terms... There may conceivably be a different legal result where the repudiation is oblique and arises indirectly as, for example, where the employer seeks to change the nature of the work required to be done or the times of employment; but I cannot see how the undertaking to employ on the one hand, and the undertaking to serve on the other can survive an out-and-out dismissal by the employer or a complete and intended withdrawal of his service by the employee. It has long been recognised that an order for specific performance will not be made in relation to a contract of service... If the only real redress is damages, how can its measure or scope be affected according to whether the contract is regarded as still subsisting or as at an end? To preserve the bare contractual relationship is an empty formality.”

But Buckley LJ said, at pp 468-469:

“Why should the doctrine operate differently in the case of contracts of personal service from the way in which it operates in respect of other contracts? I for my part can discover no reason why it should do so in principle. It cannot be because the court will not decree specific performance of a contract of personal service, for there are innumerable kinds of contract which the court would not order to be specifically enforced, to which the doctrine would undoubtedly apply... [But] in a case of wrongful dismissal in the absence of special circumstances the damages recoverable on the footing of an accepted repudiation must, I think, be as great as, and most probably greater than, any damages which could be recovered on the footing of an affirmation of the contract by the innocent party and of the contract consequently remaining in operation. So... a wrongfully dismissed servant really has, in the absence of special circumstances, no option but to accept the master’s repudiation of the contract.

It consequently seems to me that, in the absence of special circumstances, in a case of wrongful dismissal the court should easily infer that the innocent party has accepted the guilty party’s repudiation of the contract. I do not think, however, that it is impossible that in some cases incidental or collateral terms might cause the injured party to want to keep the contract on foot.”

In the course of the affirmation of the elective theory by the Saskatchewan Court of Appeal in Smart v Board of Governors of South Saskatchewan Hospital Centre (1989) 60 DLR (4th) 8, Bayda CJS commented on the observations of Buckley LJ, at p 17:

“This position of ‘being better off to accept the repudiation’ in which the innocent employee so often finds himself in practice and the courts’ commensurate readiness to find acceptance have, in my respectful view, tended to seduce some legal analysts into concluding that the innocent employee is obliged in law to accept the repudiation, or, alternatively, does not have the option in law to treat the contract as continuing. But, as... Buckley [LJ] explicitly pointed out, that conclusion is erroneous. It is important to remember that there are times when it is in the innocent employee’s practical interest to continue the contract in law.”

But Buckley LJ’s suggestion that acceptance of a wrongful repudiation should easily be inferred – and his consequent dilution of the effect of the theory which he himself was commending – has attracted powerful criticism, not least by Professor Brodie in The Contract of Employment (2008), para 18.10, and by Ralph Gibson LJ in the Boyo case [1994] ICR 727, 743. There is certainly no point in conferring upon a party an election to which some other principle of law is applied so as to deprive it of real value; and in my view Buckley LJ’s suggestion should be treated cautiously. Ralph Gibson LJ proceeded to accept that, following a wrongful repudiation, contractual obligations which did not depend on the existence of the relationship of master and servant, such as terms as to disciplinary procedures and competition, continued to exist. But, “subject”, so he said, “to that qualification”, he would, in the absence of the binding authority of the Gunton case, have preferred the automatic theory. I do not understand how a theory can be preferred “subject to” a qualification which is entirely inconsistent with it.

Apart from the decision in 1994 in the Boyo case, cited above, in which the employee represented himself and the court felt reluctantly obliged to apply the elective theory in accordance with the decision in the Gunton case, the most recent domestic decision of significance is London Transport Executive v Clarke [1981] ICR 355. Its date demonstrates that, for an entire generation, the issue between the two theories has been substantially quiescent. The employee went to Jamaica for seven weeks contrary to the terms of the contract and to the employer’s express instructions. So it was a repudiatory breach falling short of purported resignation. On the contrary, the employee wished to resume his employment upon his return. While he was away, however, the employer told him, by letter, that his employment was at an end. The first question posed by his complaint of unfair dismissal to the industrial tribunal related to the identity of the party who had terminated the contract. Lord Denning MR, evidently prepared to apply the automatic theory even to a repudiatory breach falling short of purported resignation, held, at p 366, that, upon his departure, the employee had himself terminated the contract. But Templeman and Dunn LJJ held that the termination had occurred only when, by its letter, the employer had accepted his repudiatory breach. So he had indeed been dismissed, albeit (so they proceeded to hold) not unfairly. Templeman LJ, with whose reasoning Dunn LJ agreed, said at pp 366-367:

“The general rule is that a repudiated contract is not terminated unless and until the repudiation is accepted by the innocent party...

[C]ontracts of employment cannot provide a general exemption to that rule because it would be manifestly unjust to allow a wrongdoer to determine a contract by repudiatory breach if the innocent party wished to affirm the contract for good reason. Thus in Thomas Marshall (Exports) Ltd v Guinle [1978] ICR 905, which contains a full discussion of principles and of the conflicting authorities, a contract of employment was repudiated by the employee. The court could not enforce specific performance of the contract for personal services, but Sir Robert Megarry VC enforced against the wrongdoing employee at the behest of the innocent employer who had not accepted the repudiation a confidentiality and non-competition obligation which was only effective during the continuance of the contract. Repudiation cannot determine a contract of service or any other contract while there exists a reason and an opportunity for the innocent party to affirm the contract.”

Templeman LJ added, at p 368, that the suggested exception was “contrary to principle, unsupported by authority binding on this court and undesirable in practice”.

Such might have been good quotations with which to conclude my judgment. For I entirely agree with them and cannot improve on them; and they seem particularly apt to the present case, in which the appellant had an obvious reason – and in my view a good reason – for not accepting the Bank’s wrongful attempt to terminate his contract until after 2007.

But another big question remains: how far would any application of the automatic theory extend? Mr Jeans suggests that the theory should be applied only to wrongful dismissals and resignations which are “express and immediate” or “outright”. The suggestion is somewhat analogous to that made by Shaw LJ in the Gunton case, in the passage quoted in para 90 above, in which he would have limited the application of the theory to an “express and direct” or “out-and-out” wrongful termination, as opposed – “conceivably” – to an oblique and indirect repudiation.

Any proponent of the automatic theory needs to be able to draw the contours of its application and to justify them logically. The following questions arise:

In proposing that the court should indorse the automatic theory, the Bank invites it to cause the law of England and Wales in relation to contracts of employment to set sail, unaccompanied, upon a journey for which I can discern no just purpose and can identify no final destination. I consider, on the contrary, that we should keep the contract of employment firmly within the harbour which the common law has solidly constructed for the entire fleet of contracts in order to protect the innocent party, as far as practicable, from the consequences of the other’s breach.

I agree that the appeal should be allowed and the order of the Deputy Judge restored as proposed by Lord Hope. I add a few words of my own in recognition of the main points of difficulty. I have nothing to add on the conflict issue, on which I agree entirely with Lord Hope’s analysis.

The most significant issue, which has divided the court, is the repudiation issue. Lord Sumption’s historical analysis of the development of the law in this area is powerful and of great interest. However, I am not in the end persuaded that it should provide the answer to this case. That review, like Lord Wilson’s equally powerful response, shows how both courts and academics have grappled with, and sought to reconcile, the apparently conflicting rules and remedies which judicial pragmatism has devised to meet the special features of employment contracts.

In choosing between them, I attach particular weight to the fact that, in spite of the force of the criticisms directed at the election theory, and at some of the reasoning of the majority in Gunton [1981], the law as there stated has stood for 30 years, apparently without evidence of practical difficulty or injustice. That in turn drew on the characteristically comprehensive review of the subject by Sir Robert Megarry VC in Thomas Marshall (Exports) Ltd v Guinle [1979] Ch 227. It also followed settled authority in the High Court of Australia dating back to 1946 (Automatic Fire Sprinklers), which has since been reaffirmed at that level (Byrne v Australian Airlines Ltd (1995) 131 ALR 422) and, as we were told, followed consistently elsewhere in the common law world: see, most recently Paper Reclaim v Aotearoa [2007] 3 NZLR 169 (New Zealand Supreme Court).

That approach seems apt also to the particular context of paragraph 5.15, under which the termination payment arises. I am not persuaded that a general distinction can be drawn, as Lord Sumption suggests, between the existential (“obligations which go to the continued existence of the employment relationship”); and the collateral. Nor do I find it helpful (as in some of the submissions before us) to talk of the continuation of a mere “shell” or “husk” contract. As in any other case, the nature and extent of the contractual remedies at any time must depend on the context, the terms of the contract, and the circumstances of the breach.

In the present case, the contract provided a detailed code for what was to happen during and after the period of service. The elaborate provisions for termination were an important part of the contractual rights provided to the employee. Paragraph 5.15 fixed the amount of the termination payment by reference to when “your employment terminates”. I see no reason why, for the purposes of that clause, the employer should not be held to the date of termination in accordance with the contract, rather than permitted to advance that date by repudiatory breach.

On the termination issue, after some hesitation, I have come to the conclusion, for the reasons given by Lady Hale, that the payment on 18 December 2007 did not effect a lawful termination. It is true that on the facts of this case, that may seem somewhat formalistic, and the consequences disproportionate. The employee can have been in no doubt by that stage that his employment was at an end, and could no doubt readily infer the purpose of the payment once he became aware of it. However, as she says, it is not unreasonable to expect an employer relying on a PILON clause to make the position clear. Although no formal written notice was required, it was necessary for the employer to ensure that the payment was unequivocally identifiable as an exercise of the power under para 8.3. That was not done. Accordingly, I agree that the contract was not lawfully terminated until 6January 2008.

Turning finally to the paragraph 5.16 issue, I have seen more force than my colleagues in the respondent’s case. I find Rimer LJ’s reasoning on the construction of the termination agreement (para 77) persuasive. Arguably, the clearest thing about paragraph 5.16 is the contrast between the “payments due to you under this letter (or under this agreement)” to which (subject to agreed amendments) the employee is entitled, and “all contractual and statutory claims… arising out of your employment… and its termination…”, which he is required to give up. On ordinary principles of contractual interpretation, the former would not be read as including the latter.

I accept that, if one starts from the premise (following the Court of Appeal) that the termination agreement was mandatory, in the sense that the employee was compelled to enter the agreement and take the payment, the result could be said to be unreasonable. On that view, I agree with Lord Hope that there is a strong case for applying the principle that an agreement purporting to exclude liability for breaches of contract should be narrowly construed contra proferentem.

However, it can be looked at the other way round. The company’s obligation to make the termination payment, and that of the employee to enter the termination agreement, are not expressed as mutual, concurrent obligations. The first obligation is that of the employer to make the payment. The employee’s obligation to enter the agreement is expressed as one undertaken “in consideration” for the making of the termination payment. Arguably that could be construed as leaving the employee free to waive the payment, and thus avoid the obligation to enter the agreement. Such a construction would also avoid an “unreasonable” result, and might be thought to strain the language less than that proposed by Lord Hope. It is also consistent with the last sentence of the Schedule 1 letter, which appears to assume that the “offer” is one which can be accepted or rejected.

However, in view of the unanimity of my colleagues on this issue, and since it does not appear to be a point of any more general significance, I see no purpose in carrying my doubts to the point of dissent.

Mr Geys is a lucky man. He had a responsible and highly paid job with an entitlement to participate in a profit-sharing bonus scheme dependent on the performance of his division, in addition to discretionary bonuses. The other side of the coin was that he had no contractual job security. Under his contract of employment, his employers, Société Générale (“SG”), were entitled to dismiss him at any time without cause either upon three months’ notice or “with immediate effect by making a payment to you in lieu of notice.” This is what happened to Mr Geys. He was called to a meeting on 29 November 2007 and given a letter informing him that SG had decided to terminate his employment with immediate effect and that the “appropriate termination documentation” would follow. In accordance with the time-honoured ritual, he was then taken to clear his desk and escorted from the building by security staff. There could not have been the slightest doubt that his employment relationship with SG was at an end. He cannot have supposed that he had been dismissed for cause, for no cause was stated. The only reasonable inference was that SG was purporting to dismiss him summarily without cause, as they were entitled in principle to do.

Fortunately for Mr Geys, SG did not understand their own contract. It is common ground that if they had handed him a cheque for his payment in lieu of notice at the meeting on 29 November, his dismissal would have taken effect according to his contract at once. Because the right to terminate with immediate effect is exercisable “by making a payment in lieu of notice”, it is common ground that the purported dismissal with immediate effect on 29 November was a repudiatory breach of contract by SG. They were not entitled to dismiss him with immediate effect from 29 November, but only with effect from the payment in lieu. It was, however, a repudiation of the most technical kind. There was no doubt about SG’s right to dismiss him with immediate effect if they set about it in the right way. For this reason, as I understand the majority to accept, SG’s mistake in itself caused him no loss. It made a practical difference of only three weeks and a legal difference of just over five. It made a practical difference of three weeks because the payment in lieu was in fact received on 18 December by Mr Geys’ bank on his behalf. If knowledge of the payment by Mr Geys himself was required (which I doubt), he had it by his own admission when he consulted his account on-line some time in late December. As he accepted in cross-examination, he saw the payment from SG and realised that “it had to be – thought would probably be, yes, compensation pay of – in lieu. That is the best guess one could have.” In the circumstances, it could not have been anything else. SG’s mistake made a legal difference of just over five weeks because the majority of this Court is of the opinion that the payment, although received by Mr Geys’ bank on his behalf on 18 December, was by a term to be implied into the contract ineffective to bring it to an end until 6 January, when SG unequivocally told him what he had already appreciated in late December, namely that the payment was in lieu of notice.

The result is that although the employment relationship was dead for all practical purposes from 29 November, and Mr Geys contributed nothing to SG’s fortunes after that date, he is in a position to argue that technically the contract limped on as a formal “shell” or “husk” (to use the terms deployed in argument) into January 2008. The financial consequences of this, if it is right, are considerable. The effect of paragraph 5.15(b)(iii) and (iv) of the contract is that if Mr Geys’ “employment terminates” after 31 December 2007, he is entitled to a “Compensation Payment” assessed by reference to the aggregate of his bonus awards for the calendar years 2006 and 2007, whereas if it terminates on or before that date, it will be assessed by reference to his awards in 2005 and 2006, which were substantially lower. The figures are disputed, but the result is likely to be that SG’s breach, although it has caused Mr Geys no substantial loss, will have brought him a windfall amounting to several million euros. Rarely can form have triumphed so completely over substance.

Accordingly, the main question on this appeal can be shortly stated. If an employer repudiates a contract of employment, does it end forthwith, leaving the employee to claim damages so far as the repudiation has caused him any? Or does it end only if and when the employee elects to accept the repudiation as bringing the contract to an end?

The law of employment is based partly on contract and partly on statute. The interface between the two can sometimes give rise to difficulty. But not in this case. It is common ground that the present issue depends entirely on the common law. It follows that the starting point is to examine the relevant general principles of the law of contract.

The general rule is that the repudiation of a contract does not necessarily bring the contract to an end. The innocent party has a right to choose either (i) to accept the repudiation, thus bringing the primary obligations in the contract to an end but leaving him with a right to enforce the secondary obligation to pay damages for the loss of the bargain; or (ii) to treat the contract as subsisting and claim any sums falling due under it as and when they fall due, together with any damages for the repudiating party’s failure to perform as and when performance should have occurred. These principles had been applied for many years by the time that they were first articulated in Hochster v De la Tour (1853) 2 E & B 678 in England and Howie v Anderson (1848) 10 D 355 in Scotland, as the citations in the former case show. Their most recent and authoritative restatement is to be found in the speech of Lord Diplock in Photo Production Ltd v Securicor Transport Ltd [1980] AC 827. The concept was memorably expressed by Asquith LJ in Howard v Pickford Tool Co Ltd [1951] 1 KB 417, 421, when he described an unaccepted repudiation as “a thing writ in water.”

This is sometimes called the “elective theory” of repudiation. The expression is, however, misleading because it suggests that the innocent party’s right to treat the contract as subsisting necessarily follows from the unilateral character of the other party’s repudiation. In fact, the right to treat the contract as subsisting has never been absolute. It is subject to important exceptions and qualifications. These can be illustrated from older cases, but were first coherently articulated by Lord Reid, delivering the leading judgment for the majority in White & Carter (Councils) Ltd v McGregor [1962] AC 413. The facts of this case are well known. White & Carter contracted with the Respondent to put advertisements for his garage on litterbins. The Respondent purported to cancel the contract without any right to do so, but the company chose to ignore the cancellation, continued to perform as if nothing had happened and sued for the agreed price of their services, which was much greater than the damages that they would have suffered had they accepted the repudiation. The Appellant succeeded because of what Lord Reid called the “peculiarity” that the contract could be performed without any co-operation from the Respondent. Lord Reid said at p 429:

“Of course, if it had been necessary for the defender to do or accept anything before the contract could be completed by the pursuers, the pursuers could not and the court would not have compelled the defender to act, the contract would not have been completed and the pursuers' only remedy would have been damages.”

Lord Hodson (with whom Lord Tucker agreed) appears to have agreed with this. At p 445, he observed:

“The true position is that the contract survives and does so not only where specific implement is available. When the assistance of the court is not required the innocent party can choose whether he will accept repudiation and sue for damages for anticipatory breach or await the date of performance by the guilty party.” [Emphasis added].

Lord Reid’s qualification about co-operative agreements has subsequently been accepted and applied. The most significant decisions are Hounslow London Borough Council v Twickenham Garden Developments Ltd [1971] Ch 233, and Attica Sea Carriers Corporation v Ferrostaal Poseidon Bulk Reederei GmbH [1976] 1 Lloyd’s Rep 250. It has also been treated as good law in other cases where it nevertheless was found not to apply on the facts, because properly analysed the contract could be performed without the co-operation of the repudiating party: see Isabella Shipowner SA v Shagang Shipping Co Ltd [2012] EWHC 1077 (Comm), paras 37-41.

These decisions are authority for a general rule that the innocent party to a repudiated contract cannot treat it as subsisting if (i) performance on his part requires the co-operation of the repudiating party, and (ii) the contract is incapable of specific performance, with the result that that co-operation cannot be compelled. The purpose of the right to treat a repudiated contract as subsisting is to enable it to be performed at the option of the innocent party. It is difficult to see why the law should recognise such a right in a case where the contract cannot be either performed or specifically enforced.

The rationale for all this is closely connected with the reasons for the law’s reluctance to grant specific performance of certain kinds of contract. Specific performance, like any equitable remedy, is discretionary, but in the present context the discretion is largely determined by well established principles. These have always been influenced by a strong pragmatic aversion to the specific enforcement of contractual obligations in circumstances where they sterilise productive resources or lead to their wasteful allocation. In his dissenting judgment in the Court of Appeal in Co-operative Insurance Society Ltd v Argyll Stores (Holdings)Ltd [1996] Ch. 286, 304, Millett LJ put the point in this way:

“The competing arguments in the present case, and the difference in the views of the members of this court, reflect a controversy which has persisted since the dispute between Sir Edward Coke and Lord Ellesmere LC. Sir Edward Coke resented the existence of an equitable jurisdiction which deprived the defendant of what he regarded as a fundamental freedom to elect whether to carry out his promise or to pay damages for the breach. Modern economic theory supports Sir Edward Coke; an award of damages reflects normal commercial expectations and ensures a more efficient allocation of scarce economic resources. The defendant will break his contract only if it pays him to do so after taking the payment of damages into account; the plaintiff will be fully compensated in damages; and both parties will be free to allocate their resources elsewhere. Against this there is the repugnance felt by those who share the view of Fuller CJ in Union Pacific Railway Co v Chicago, Rock Island and Pacific Railway Co (1896) 163 US 564, 600 that it is an intolerable travesty of justice that a party should be allowed to break his contract at pleasure by electing to pay damages for the breach. English law has adopted a pragmatic approach in resolving this dispute. Equitable relief is discretionary and exceptional. Courts of equity have never enforced the performance of all contracts, whatever their nature. Over the centuries rules of practice have evolved so that the parties can know in advance which contractual obligations will be specifically enforced and which sound in damages only. The leading principle is usually said to be that equitable relief is not available where damages are an adequate remedy. In my view, it would be more accurate to say that equitable relief will be granted where it is appropriate and not otherwise; and that where damages are an adequate remedy it is inappropriate to grant equitable relief.”

Millett LJ’s dissent was subsequently upheld in the House of Lords [1998] AC 1, where Lord Hoffmann observed at pp. 15-16:

“From a wider perspective, it cannot be in the public interest for the courts to require someone to carry on business at a loss if there is any plausible alternative by which the other party can be given compensation. It is not only a waste of resources but yokes the parties together in a continuing hostile relationship.”

Subject to the intervention of statute, contracts of employment are governed by the same principles as other contracts, except in those cases where their subject-matter gives rise to compelling policy considerations calling for a different approach. But the relationship of employer and employee is especially liable to give rise to policy considerations of this kind, because its incidents have significant social and economic implications. They affect a high proportion of the adult population and have a profound impact both on their personal lives and on their relationships with others. When it comes to enforcing an unwanted relationship of employer and employee, there are altogether more sensitive considerations involved than those governing most other more contractual bargains. As Fry LJ put it in De Francesco v Barnum (1890) 45 Ch D 430, 438, the courts are “very unwilling to extend decisions the effect of which is to compel persons who are not desirous of maintaining continuous personal relations with one another to continue those personal relations.”

Historically, there have been three main reasons for this. The first is that the relationship of employer and employee was traditionally regarded as a highly personal one. In an age of large corporate enterprises many of whose employees perform routine jobs, the personal character which was once typical of employment relationships has lost much of its former importance. But employment is nonetheless a relationship based on mutual trust and confidence, a factor which has assumed growing importance in the way that the law has developed over the past thirty years. Second, the difficult and litigious history of industrial relations in the United Kingdom in the late nineteenth and early twentieth centuries reinforced the sensitivity which the common law had always had about any intervention by a court which might force the parties to continue in a relationship which has been described as “at once interdependent and oppositional”: The Oxford History of the Laws of England, vol xiii (2010), p 623. This is why the common law rule against injunctions requiring an employee to work has for many years been statutory: see, currently, section 236 of the Trade Union and Labour Relations (Consolidation) Act 1992. This makes it more difficult to justify intervening in a way that forces an employer to employ someone if the law is to maintain the ordinary principle that remedies should operate mutually or not at all. Third, legal thinking in this area has always been influenced by a concern for the productive use of resources, including labour. This is evident in the development of the common law relating, for example, to restrictive covenants and, at a more macro-economic level, to the economic torts of interference with contractual relations and procuring a breach of contract and aspects of the law of conspiracy. It is reflected in the abiding concern of the common law to ensure the terminability of contracts of employment, without prejudice to the subsequent regulation of the financial consequences by an award of damages. The harsher consequences of this approach for individuals have been mitigated in the last half century by a parallel scheme of statutory protection of employment, operating within defined limits and administered by specialised statutory tribunals with limited jurisdiction over purely contractual disputes. But the statutory protection of employment overlays the common law without necessarily altering it. Indeed, it makes the development of a more stringent standard of employment protection at common law unnecessary and perhaps inappropriate. That much is apparent from the decision of the House of Lords and this court that the employer’s obligation to maintain mutual trust and confidence does not apply to or survive a wrongful dismissal: Johnson v Unisys Ltd [2003] 1 AC 518, Edwards v Chesterfield Royal Hospital NHS Foundation Trust [2012] 2 AC 22.

The traditional insistence of the courts that contracts of employment are not specifically enforceable has not, at least in the last half century, been dogmatically applied to every obligation under such contracts. It is important to distinguish between the core obligations which provided the original rationale of the rule, and what for want of a better word I will call collateral obligations. In my respectful opinion the difficulties which the majority find with the “automatic” theory of termination as applied to contracts of employment are largely attributable to their failure to make this distinction. The core obligations are those which are fundamental to the continued existence of the employment relationship, essentially the obligation of the employee to work and the concomitant obligation of the employer to continue to employ and pay him. When courts say, as they always have, that a contract of employment is not specifically enforceable, they are almost always talking about obligations of this kind. The present appeal is about the core obligations under Mr Geys’ contract of employment. We are concerned with the question whether, in any legally meaningful sense, it can be said that Mr Geys had an obligation to work after 29 November 2007 or SG an obligation to pay him in respect of the period after that date. If the answer to these questions is No, it must be difficult to suggest that there was any subsisting contract of employment between them. What follows is directed only to those obligations under a contract of employment which determine whether the relationship created by it is to subsist. I shall return to the question of collateral obligations later.

The law on these core obligations dates back to the early nineteenth century. In a previous age, Lord Mansfield had held that a wrongfully dismissed employee was entitled to his wages accruing after termination, on the principle that the employer should not be allowed to take advantage of his own wrong: Temple v Prescott (1773), cited in The Oxford History of the Laws of England, vol xiii (2010), p 645. But this view was decisively rejected in all the subsequent case-law. The rule that the innocent party to a repudiated contract of employment was not entitled to treat it as subsisting or recover wages accruing after dismissal was established after a difference on the point had arisen between Lord Ellenborough and Lord Tenterden. In Gandell v Pontigny (1816) 4 Camp 375, 171 ER 119, a merchant’s clerk was unlawfully dismissed and declined to accept the repudiation, notifying his employer that he held himself available to work for him. Lord Ellenborough awarded him his full salary, for practical purposes a decree of specific performance. But in Archard v Hornor (1828) 3 Carr & P 349, 172 ER 451, Lord Tenterden limited the award in a comparable case to damages representing the dismissed employee’s wages up to the time of his unlawful dismissal. His view was consistently accepted thereafter in preference to Lord Ellenborough’s: see Snelling v Lord Huntingfield (1834) 1 CM & R 20, 149 ER 976, Fewings v Tisdal (1847) 1 Exch 295, 154 ER 125 (where the history is reviewed in the successful argument of Greenwood).

In French v Brookes (1830) 6 Bing 354, 130 ER 1316, the law was said to have been settled in this sense for many years. The facts were sufficiently close to the present case to repay attention. John Oliver French was employed for three years as the manager of a mine in South America on terms that he might be dismissed either on a year’s notice or on payment of a year’s salary and the cost of his passage home. Half way through the term the local agents of the company decided to make economies by suppressing Mr French’s post and dismissed him without either notice or the year’s salary in lieu. He declined to accept the validity of his dismissal, declared his intention of carrying on and sued for a sum which although described as damages was in fact the total amount that he would have received had the contract subsisted. The jury awarded him only his actual loss. Dismissing his claim for the balance, Tindall CJ said, at pp 360-361:

“[Sergeant] Wilde’s motion stands on the construction of the agreement: he argues, that the contract between the parties not having been determined in the mode pointed out by the agreement, it must be considered as subsisting for the whole time originally contemplated. But this action, like others of the same sort, is brought because the contract has been violated; and the case has been correctly dealt with if the jury have given damages for the breach... The jury, therefore, have not erred if they have put the plaintiff in the same situation as if the directors, upon dismissing him, had paid at the time twelve months' salary, and a reasonable sum towards defraying his expenses from South America to England... If any special damage had been alleged and proved, as resulting from the directors not having paid the year's salary at the time of the dismissal, the jury might have found for that...”

The modern law starts with a trio of cases in which the Plaintiff was dismissed by a public authority or an organ of a public authority without the power to do so: Vine v National Dock Labour Board [1957] AC 488, Francis v Municipal Councillors of Kuala Lumpur [1962] 1 WLR 1411, and Ridge v Baldwin [1964] AC 40. In each of these cases, there was either a contract of employment or a relationship regarded as legally analogous. But all three cases had the distinctive feature that the decision to dismiss was not only repudiatory in the contractual sense but was, as a matter of public law, a nullity. In each of them, however, the position in the ordinary contractual context was considered, whether by way of either contrast or analogy. Since these are decisions of high authority and it is implicit in the majority’s view that they were wrongly decided, or at least wrongly reasoned, it is I think worth examining them.

In Vine v National Dock Labour Board [1957] AC 488 the House of Lords held, overruling the Court of Appeal, that a docker was entitled to a declaration that he had been unlawfully dismissed by the Board. He had been dismissed by a committee which had no power to do so under the relevant regulations. The decisive consideration was that his dismissal was a nullity as a matter of public law. Viscount Kilmuir LC (p 500), adopting the reasoning of the dissenting judgment of Jenkins LJ in the Court of Appeal, observed:

“This is an entirely different situation from the ordinary master and servant case; there, if the master wrongfully dismisses the servant, either summarily or by giving insufficient notice, the employment is effectively terminated, albeit in breach of contract.”

Lord Keith of Avonholm (p 507) said:

“This is not a straightforward relationship of master and servant. Normally, and apart from the intervention of statute, there would never be a nullity in terminating an ordinary contract of master and servant. Dismissal might be in breach of contract and so unlawful but could only sound in damages.”

Lord Morton (p 504) and Lord Cohen (p 507) both adopted the judgment of Jenkins LJ, the former expressing himself “content to adopt, without qualification, everything that he said on the point.” Lord Somervell (p 513) agreed on this point with Lord Morton. These remarks were obiter. But they were clearly considered statements of principle, which formed an integral part of the reasoning.

In Francis v Municipal Councillors of Kuala Lumpur [1962] 1 WLR 1411, Mr Francis was dismissed by the Kuala Lumpur Council on 1 October 1957 from his position as a clerk. The Council had no power to do this, because regulations conferred the power on the president of the Council alone. Mr Francis’s case was that the decision was a nullity, and that accordingly he remained in the Council’s employment, just as Mr Vine remained an employee of the National Dock Labour Board. Lord Morris, delivering the advice of the Privy Council, distinguished Vine’s case on grounds which are unclear but for present purposes do not matter. The relevant point is that they proceeded by analogy with an ordinary contract of employment and held that the dismissal, although wrongful, had been immediately effective to terminate Mr Francis’s employment. Lord Morris expressed the Board’s reasons (p 1417) as follows:

“Their Lordships consider that it is beyond doubt that on October 1, 1957, there was de facto a dismissal of the appellant by his employers the respondents. On that date he was excluded from the council's premises. Since then he has not done any work for the council. In all these circumstances it seems to their Lordships that the appellant must be treated as having been wrongly dismissed on October 1, 1957, and that his remedy lies in a claim for damages. It would be wholly unreal to accede to the contention that since October 1, 1957, he had continued to be, and that he still continues to be, in the employment of the respondents.”

Ridge v Baldwin [1964] AC 40 concerned the dismissal for misconduct of a chief constable, not technically an employee but a public officer. The dismissal was a nullity as a matter of public law. Lord Reid (at p 65) contrasted the position under a contract of employment, where it would not have been a nullity:

“The law regarding master and servant is not in doubt. There cannot be specific performance of a contract of service, and the master can terminate the contract with his servant at any time and for any reason or for none. But if he does so in a manner not warranted by the contract he must pay damages for breach of contract.”

These statements of principle were accepted in a succession of cases which arose in a purely contractual context, without the public law element: see Barber v Manchester Regional Hospital Board [1958] 1 WLR 181, Cranleigh Precision Engineering Ltd v Bryant [1965] 1 WLR 1293, 1304-1305; Denmark Productions Ltd v Boscobel Productions Ltd [1969] 1 QB 699; Ivory v Palmer [1975] ICR 340, 354 (Browne LJ). The most significant of them is Denmark Productions. The case arose from the repudiation by a pop group of its contract with its manager. The issue was whether the manager was entitled to claim an account of profits on the footing that his contract had never lawfully been terminated, or was limited to a claim for damages for loss of the bargain. The contract was not a contract of employment but a contract for services. However, the Court of Appeal held by analogy with the law relating to employment contracts that the manager could not claim his remuneration on the footing that the contract subsisted. Salmon LJ said (p 726):

“It has long been well settled that, if a man employed under a contract of personal service is wrongfully dismissed, he has no claim for remuneration due under the contract after the repudiation. His only money claim is for damages for having been prevented from earning his remuneration: Goodman v Pocock; French v Brookes; Fewings v Tisdal.A managing director, for example, engaged at £10,000 a year, who has ten years of his service agreement to run, is dismissed without cause. He cannot sit in the sun for ten years drawing his salary on the basis that he is ready, able and willing to serve as managing director if only the company would allow him to do so. His sole money claim is for damages and he must do everything he reasonably can to mitigate them.”

Harman LJ in the same case said, at p 737:

“I am, therefore, of opinion, and in this I concur with my brother Salmon, that the true cause of action of the plaintiffs was for damages for wrongful dismissal and that the action as framed for an account is misconceived. An employee dismissed in breach of his contract of employment cannot choose to treat the contract as subsisting and sue for an account of profits which he would have earned to the end of the contractual period: he must sue for damages for the wrongful dismissal and must, of course, mitigate those damages so far as he reasonably can.”

In 1974, Sir John Donaldson P reviewed the case-law in the National Industrial Relations Court in Sanders v Ernest A Neale Ltd [1974] ICR 565 and concluded (p 571) that the repudiation of a contract of employment “terminates the contract without the necessity for acceptance by the injured party”. As a summary of the position as it then stood, this was clearly right. There was a long, authoritative, and broadly consistent consensus in favour of the principle that an unaccepted repudiation of a contract of employment which terminated the relationship also brought the contract to an end, in law as well as in fact.

Even the apparent exceptions were consistent with the underlying principle. Leaving collateral obligations aside for the moment, most of them are cases in which, unusually, the repudiation did not bring an end to the relationship of employer and employee. Such cases are quite different from the case of a dismissal or resignation, actual or constructive, which bring the relationship to an end. Thus in Cranleigh Precision Engineering Ltd v Bryant [1965] 1 WLR 1293, Roskill J acknowledged in terms the general rule which treated the contract as at an end upon an unaccepted repudiation, but said that he would not have applied it in the case before him because both parties ignored the event said to constitute the repudiation. The relationship continued, not just as a legal construct but in fact. The employee went on working and the employer continued to pay him. In Hill v CA Parsons & Co Ltd [1972] Ch 305, the Court of Appeal declined to treat the contract as at an end in the “exceptional” circumstances of that case. These were that the dismissal notice was invalid and the employee retained the confidence of the employer. It had only dismissed him at the insistence of a trade union, which had imposed a closed shop agreement of a kind which was about to become unlawful. As Lord Denning pointed out at p 314B, “In the ordinary course of things, the relationship of master and servant thereupon comes to an end: for it is inconsistent with the confidential nature of the relationship that it should continue contrary to the will of one of the parties thereto.” The position was very similar in Rigby v Ferodo Ltd [1988] ICR 29, where the repudiation consisted in the employer’s unilateral imposition of a reduction in wage rates but the relationship did not end. The employer wished to go on employing the whole workforce, and had indeed imposed the reduction in order to make that possible. The employee for his part continued to work and receive wages, albeit reserving his rights and protesting about their reduced amount. Lord Oliver, delivering the sole reasoned speech in the House of Lords, expressly reserved his opinion on what the position would have been if there had been an “outright dismissal or walk-out”: see pp 33D, 34E-F, 35B-C. In my view, this is the true rationale of the cases in which the courts have specifically enforced contractual disciplinary procedures, As Ralph Gibson LJ pointed out in Boyo v Lambeth London Borough Council [1994] ICR 727, 743H, such procedures do not depend on the continued existence of the relationship of employer and employee. They are, in the terminology that I have been using, collateral. The courts, developing a principle originally derived from public law, have been willing to enforce them even if the effect is to prolong the period of employment. This does not impinge on the traditional objections of the common law to the specific enforcement of the employment relationship, because of the collateral character of the disciplinary procedures and because the possibility that an internal disciplinary procedure may result in the employee’s reinstatement makes it premature to regard that relationship as at an end.

Sir John Donaldson also observed in Sanders that the principle which he regarded as well established in the field of employment represented an exception to the general rule of the law of contract allowing the innocent party to elect whether to accept the repudiation or affirm the contract. In my view this was a fundamental misunderstanding of the position which, although often repeated, has had an unfortunate effect on more recent developments in this area of law. In fact, the rule which Sir John Donaldson applied, far from being an exception to the ordinary principles of the law of contract, exemplified the ordinary operation of those principles. The general principle is that the innocent party to a repudiated contract cannot treat it as subsisting unless he can either perform it without the co-operation of the other party or compel that co-operation. In the case of a contract of employment, neither condition is satisfied.

All of the cases which I have cited, as well as those to which I shall come, are agreed that the employer’s refusal to allow the employee to earn his wages by excluding him from work does not give rise to a right to recover the wages, but only to a claim for damages. Moreover, the courts have never applied to contracts of employment the doctrine of deemed performance endorsed by the House of Lords in Mackay v Dick (1881) 6 App Cas 251, according to which a party who is prevented by the non-co-operation of the counterparty from satisfying a condition precedent to his right to receive remuneration may be deemed to have earned it notwithstanding the condition. Why have the courts been so absolute in their refusal to contemplate a claim by a wrongfully dismissed employee for his wages? The reason is sometimes said to be that he has not earned it because under a contract of employment the obligation of the employee is to do the work, not just to hold himself available to do it. But this is certainly not a general principle of employment law, as the old cases on sick pay (before it became statutory) and the more recent ones on “go-slows” and other forms of partial industrial action tend to show: see Cuckson v Stones (1858) 1 E & E 248, 256 (Lord Campbell CJ), Miles v Wakefield Metropolitan District Council [1987] AC 539, 561B-C (Lord Templeman) and the discussion in Freedland, The Personal Employment Contract (2003), 212-223. Another possible explanation is that to allow the employee to recover his wages after a wrongful dismissal would be a form of specific enforcement of the contract, and that the problem is the unavailability of that particular remedy. That, however, hardly seems more satisfying. After all, if the contract subsists, the wages are a debt. It is hard to see why any of the objections to making the remedy of specific performance available to enforce a contract of employment should apply to the recovery of an accrued debt. If there were such an objection, it would apply equally to an action for wages accrued under a contract which had not been repudiated, but it is clear that it does not. The only rational explanation of the rule that a wrongfully dismissed employee cannot sue for his wages is that once the employee has been dismissed, albeit wrongfully, there is no longer a contractual obligation to pay the wages, and therefore no debt on which to sue. This can only be because the contract terminated upon the dismissal. It terminated because the contract is a co-operative agreement whose performance requires the engagement and mutual confidence of both sides. It is therefore not possible for the employee to treat it as subsisting once the employer has repudiated it and brought their de facto relationship to an end. The consequence, as the editors of Chitty on Contracts, 30th ed (2008), vol 1, para 24-032, point out with reference to co-operative contracts generally, is that “the party not in default may be compelled to treat the prevention of performance as a repudiation of the contract and to sue for damages for the breach.”

This consensus was first seriously challenged by way of dictum in Decro-Wall International SA v Practitioners in Marketing Ltd [1971] 1 WLR 361 and Thomas Marshall (Exports) Ltd v Guinle [1979] Ch 227, and finally as part of the ratio in Gunton v Richmond-upon-Thames London Borough Council [1981] Ch 448.

Decro-Wall was not about a contract of employment. It concerned an oral contract between a manufacturer and his exclusive distributor in the United Kingdom. It was held that upon the repudiation of the contract by the manufacturer, the distributor was entitled to treat the contract as subsisting. The contract was essentially a framework agreement which precluded the manufacturer from selling to any one else in the United Kingdom and the distributor from distributing any one else’s competing products, but imposed no obligation on the distributor to buy any goods. The distributor therefore had no obligation which required the manufacturer’s co-operation. He had no more than a right (in effect an option) to buy goods from the manufacturer, which would in principle have been specifically enforceable. The Court of Appeal therefore considered that the case was governed by the actual decision in White & Carter and not by Lord Reid’s qualification to it concerning co-operative agreements. None of the members of the court regarded the arrangement as having any analogies with a contract of employment. For these points, see pp. 369H, 370F-G (Salmon LJ), 376B (Sachs LJ), 381D-F (Buckley LJ). For present purposes, the case is mainly important for a dictum of Salmon LJ at pp 369-370, responding to the citation in argument of his own judgment (see above) in Denmark Productions Ltd v Boscobel Productions Ltd [1969] 1 QB 699. Salmon LJ said that he doubted whether an unaccepted repudiation could bring an end to a contract of employment in law “although no doubt in practice it does”. In law, he thought that the position was (i) that the contract continued in being, (ii) that it would not, however, be specifically enforced because the employee had not worked and had not therefore earned his remuneration; (iii) that the employee’s only remedy was to sue for his lost wages as damages for the employer’s breach in preventing him from earning them (presumably from time to time as they would have fallen due); and (iv) that the only thing that prevented the employee from sitting idle for the rest of the contractual term and collecting damages equal to his lost wages was the condition that he should have taken reasonable steps to mitigate his loss by finding alternative employment. Sachs LJ (at p 375H) appears to have taken the same view, observing that “In such cases it is the range of remedies that is limited, not the right to elect.”

In Thomas Marshall (Exports) Ltd v Guinle [1979] Ch 227, Sir Robert Megarry V-C was concerned with an express covenant in the contract of the Plaintiff’s managing director against using or disclosing its confidential information during or after his employment. The managing director had resigned before the end of the contractual term. It was submitted that this was a repudiation which brought the contract to an end and with it any obligation to observe the restrictive covenants. The Vice-Chancellor’s main concern about this submission was that if correct it meant that the employee could bring an end to his own primary obligations under the information covenant by unilaterally renouncing the contract. He reviewed the case-law and, adopting the dicta in Decro-Wall, rejected the submission on the ground that the employer had elected to treat the contract as subsisting. In his view, therefore, it continued to bind the employee. It seems to me that the result was clearly right for an altogether simpler reason. The covenant in question expressly bound the employee both during and after his employment. It was therefore irrelevant when the employment relationship or the contract embodying it ended.

The statements in these two cases were adopted and expanded as a matter of decision by the Court of Appeal in Gunton v Richmond-upon-Thames London Borough Council [1981] Ch 448. Richmond Council had dismissed Mr Gunton from his employment as a college registrar on disciplinary grounds and excluded him from work, but without properly following the disciplinary procedure incorporated into his contract. This failure was found by the Court of Appeal to have caused him no prejudice, because he had exercised a right of appeal and had received a fair hearing on the appeal, albeit that his appeal failed. Mr Gunton sued on the footing that his employment had not been terminated in law. He claimed damages and a declaration that he was entitled to remain in the council’s employment until he retired or was lawfully dismissed. His original claim was essentially a claim in public law, for a declaration that the decision to dismiss him was a nullity, which came before the Chancery Division in the last period of the integration of public and private law, before the new Order 53 separated the streams. In the Court of Appeal, however, the issue was analysed in private law terms. By a majority (Buckley and Brightman LJJ), the Court of Appeal made the declaration and awarded him damages equal to his losses from the time of his exclusion from work until the expiry of one month’s notice notionally served on the day when a proper disciplinary proceeding could have been concluded. The striking thing about Gunton, however, is that both judges in the majority endorsed the common law’s long-standing recognition that the employment relationship was thereafter at an end. Both of them accepted the traditional refusal of the common law to allow any remedy other than damages. However, both treated the contract as having a continuing vestigial existence.

Buckley LJ accepted that the employee could not sue in debt for his remuneration in respect of any period after his exclusion from work, because “the right to receive remuneration and the obligation to render services are mutually interdependent” p. 468E. Nonetheless, he concluded that the contract must have a continuing existence in order to give effect to the employee’s right to elect whether to accept the repudiation or affirm the contract. This right was part of the general law of contract and there were no principled reasons for applying a different rule to contracts of employment: pp 467-468. He offers no explanation of how a contract can be said to subsist in spite of the absence of any obligation on either side to perform its core obligations.

Like Buckley LJ, Brightman LJ also accepted that there was no right to sue for wages after the employer’s repudiation, although he expressed the reason for this differently at p 473B:

“An employee's remedy, if he is unlawfully dismissed by his employer, is damages. He cannot obtain an order for specific performance because it is not available to compel performance of a contract of service against an unwilling employer.”

Unlike Buckley LJ, he did produce a rationalisation of the continued existence of the contract, by positing a distinction between Mr Gunton’s status as an employee, which was terminated when he was excluded from work, and the contract of employment, which subsisted until it was lawfully terminated. His analysis, at pp 474-475, is sufficiently important to be worth quoting in full:

“It is clear beyond argument that a wrongfully dismissed employee cannot sue for his salary or wages as such, but only for damages. It is also, in my view, equally clear that such an employee cannot assert that he still retains his employment under the contract. If a servant is dismissed and excluded from his employment, it is absurd to suppose that he still occupies the status of a servant. Quite plainly he does not. The relationship of master and servant has been broken, albeit wrongfully by one side alone. The same would apply to a contract for services, such as an agency. If a two year agency contract is made between principal and agent, and the principal wrongfully repudiates the contract of agency after only one year, quite plainly the agent cannot hold himself out as still being the agent of the principal. He is not. The relationship of principal and agent has been broken. I do not think it follows, however, from the rupture of the status of master and servant, or principal and agent, that the contract of service, or the contract of agency, has been terminated by the wrongful act of the master or the principal. What has been determined is only the status or relationship. So in the result the servant cannot sue in debt for his wages, which he is wrongfully deprived of the opportunity to earn; or for his fringe benefit, such as the house which the carpenter in Ivory v Palmer [1975] ICR 340 had the right to occupy as part of his emoluments. As the relationship of master and servant is gone, the servant cannot claim the reward for services no longer rendered. But it does not follow that every right and obligation under the contract is extinguished. An obligation which is not of necessity dependent on the existence of the relationship of master and servant may well survive; such as the right of the master in Thomas Marshall (Exports) Ltd v Guinle [1979] Ch 227 that the servant should not during the term of the contract deal on his own account with customers of the plaintiff company.”

The distinction made by Brightman LJ between the employee’s status and his contractual rights, the one terminating on the employer’s repudiation and the other subsisting, was new to the English case-law, but it was not entirely new to the common law. It had previously been accepted by the High Court of Australia in Automatic Fire Sprinklers Pty Ltd v Watson (1946) 72 CLR 435, 454 (Latham CJ), 469 (Dixon J); cf. Byrne v Australian Airlines Ltd (1995) 131 ALR 422, 432. It is right to point out that if Gunton was rightly decided on this ground, then Mr Geys will not be entitled to recover his profit-related bonus based on the calendar years 2006 and 2007, because his right to such a bonus depends on Clauses 5.15(b)(iii) and (iv) of the contract, which depend on when his “employment terminates”, and not (if it is different) on when the contract terminates.

Shaw LJ dissented, essentially on the ground that the continued vestigial existence of an unperformable contract was an artificial fiction devoid of any connection with the true state of affairs. His reasons are sufficiently summed up in the following passage from p 459 of his judgment:

“... I cannot see how the undertaking to employ on the one hand, and the undertaking to serve on the other can survive an out-and-out dismissal by the employer or a complete and intended withdrawal of his service by the employee. It has long been recognised that an order for specific performance will not be made in relation to a contract of service. Therefore, as it seems to me, there can be no logical justification for the proposition that a contract of service survives a total repudiation by one side or the other. If the only real redress is damages, how can its measure or scope be affected according to whether the contract is regarded as still subsisting or as at an end? To preserve the bare contractual relationship is an empty formality. The servant who is wrongfully dismissed cannot claim his wage for services he is not given the opportunity of rendering; and the master whose servant refuses to serve him cannot compel that servant to perform his contracted duties. In this context remedies and rights are inextricably bound together. It is meaningless to say that the contract of service differs from other contracts only in relation to the availability of remedies in the event of breach. The difference is fundamental, for there is no legal substitute for voluntary performance.”

In my opinion, Shaw LJ’s reasoning is unanswerable. The consensus as it had stood up to the 1970s was correct, and Gunton was wrongly decided. My reasons are as follows:

Much of Mr Cavender QC’s excellent argument on this point was directed not so much to justifying Gunton as to persuading us that whatever doubts we might entertain about its correctness, it had stood for thirty years without apparently creating problems or giving rise to any injustice. The law, he submitted, should be left as it is rather than disturbed for the sake of mere “doctrinal” purity. There are certainly cases where that is the right way of dealing with settled but legally anomalous decisions. But this is not one of them. In the first place, Gunton has always been a controversial decision. It was the decision of a divided court. It was recorded in Rigby v Ferodo Ltd [1988] ICR 29, 34 that the Court of Appeal had given leave to appeal to the House of Lords with a view to its correctness being determined, although in the event the appeal was resolved on another point. Two years later, in Octavius Atkinson & Sons Ltd v Morris [1989] ICR 431, 436B-C, Sir Nicholas Browne-Wilkinson V-C observed that “the correct legal result unhappily remains unresolved”. Gunton was followed with strong and express misgivings by the Court of Appeal in Boyo v Lambeth London Borough Council [1994] ICR 727, some of which foreshadowed the argument before us. Ralph Gibson LJ said that if it had been open to him he would have preferred the analysis of Sir John Donaldson in Sanders v Ernest A Neale Ltd [1974] ICR 565, and Staughton LJ declared a preference for the dissenting judgment of Shaw LJ. Even in 2012, its position has been described as “far from assured”: Cabrelli and Zahn, “The elective and automatic theories of termination at common law: Resolving the conundrum?”, Industrial Law Journal vol 41 (2012), 346, 354-355. Secondly, there is no basis for Mr Cavender’s assertion that the decision in Gunton has given rise to no difficulty or injustice. Its application would give rise to significant injustice in this case, for reasons which I have sought to explain at para 110. It cannot, with respect, be an answer to say, as the majority do, that their approach is required in order to prevent SG from profiting from its own wrong and to “negative” the impact of that wrong on Mr Geys. These are proper functions of an award of damages. Mr Geys’ problem is that the particular feature of SG’s conduct which was wrongful, i.e. the temporal separation of the dismissal and the payment in lieu of notice, has not caused him any significant loss. It is no part of the purpose of the law to reflect moral indignation about SG’s conduct, even assuming that SG’s mistake calls for moral indignation, which I doubt. Third, and more generally, it is always dangerous to allow the law to part company with reality in this way. It leads to unexpected and highly technical results, which businessmen and employees are unlikely to anticipate unless they are particularly well advised. In this case, even a mighty corporation like SG misunderstood the position. How are more modest enterprises to do so? We cannot know what other problems the decision in Gunton has thrown up since it was decided, because it is binding at every level below this one and has therefore had to be borne in silence by any one who lacked the stomach to embark on litigation with a view to taking the issue to the House of Lords or the Supreme Court. Fourth, the law as it was stated in Vine’s case made for certainty in a way which is not true of the law stated in Gunton. If the contract subsists after the employee’s exclusion from work, it will often be extremely difficult to determine with any confidence when it terminates. This will depend on the often ambiguous facts said to constitute an acceptance of the repudiation, or on highly technical questions about the validity of notices and payments such as those which arise in this case. I have already drawn attention to the additional uncertainty associated with the question of mitigation. By comparison, if the contract ends when the employment relationship ends, the position is clear. There is no reason to believe that we would be inviting unforeseeable difficulties by recognising the termination of the relationship. Why should they be any greater now than they were during the very long period before the 1970s when that was thought to be settled law?

A good deal of attention was devoted in the course of argument to the implications for other contractual obligations of concluding that an employee cannot treat the contract of employment as subsisting after a repudiation which terminates the employment relationship de facto. In my opinion, this question has very little bearing on the present issue, once it is appreciated that we are concerned only with those obligations which go to the continued existence of the employment relationship. In Gunton itself, at p 475, Brightman LJ envisaged that the extinction of the relationship of employer and employee might well be survived by any “obligation which is not of necessity dependent on the existence of the relationship.” Echoing this view, Lord Oliver pointed out in Rigby v Ferodo Ltd [1988] ICR 29, 34D that even if Gunton was wrongly decided and the contract terminated with the relationship of employer and employee, that would not necessarily bring an end to those contractual obligations “which do not of necessity depend on the existence of the relationship of master and servant.” I think that this is right, and significant. In many contracts of employment, and perhaps in most modern ones, there is a large number of obligations which do not depend on the existence of the employment relationship. One example is the specific enforcement after a repudiation of express or implied covenants against competition, as in Lumley v Wagner (1852) 1 De G M & G 604, 42 ER 687. In appropriate cases, this may be subject to the proviso that the repudiation was not by the party in whose favour the covenant was included: see General Billposting Co Ltd v Atkinson [1909] AC 118. Another example is a covenant against the disclosure of confidential information, such as the one considered in Thomas Marshall (Exports) Ltd v Guinle [1979] Ch 227. Whether collateral obligations of this kind continue to bind after the termination of the contract or the underlying relationship will normally depend on the construction of the contract, or the exact nature of the implication if the obligation in question is implied. This is not the place for a general review of the kind of obligations which survive termination of the contract and are sufficiently collateral to warrant specific enforcement. What is clear is that it is not necessary to prolong the life of a repudiated contract of employment in order to justify this body of law. It follows that it will not be affected one way or the other by the outcome of this appeal.

I would allow SG’s cross-appeal on the ground that the contract terminated on 29 November 2007, when it was repudiated by SG and Mr Geys was excluded from work.

On that footing the question raised by Mr Geys’ appeal whether, if the contract subsisted after that date, it came to an end upon the crediting of payment in lieu into his bank account or upon his noticing the payment later in December, does not arise. For my part, I would have held that if Mr Geys (contrary to my opinion) was entitled to affirm the contract after the unequivocal notice of dismissal given to him on 29 November, then all that was required to satisfy Clause 8.3 of the Handbook was the making of the payment in lieu. That seems to be more consistent with both the reality of the situation and the approach of the Court of Appeal in Abrahams v Performing Right Society [1995] ICR 1028 and the Employment Appeal Tribunal in Cerberus Software Ltd Rowley [2000] ICR 35.

On the so-called “full and final settlement” issue, which turns on the construction of paragraph 5.16 of the letter agreement, I agree with Lord Hope.