Secretary of State for Business, Enterprise & Regulatory Reform v Neufeld & Anor

This is the judgment of the court to which all members have contributed. Two appeals are before us. Both are from the Employment Appeal Tribunal and both raise similar questions. The appellant in both is the Secretary of State for Business, Enterprise and Regulatory Reform, who was represented by Mr Adam Tolley.

The respondent to the first appeal is Mr Richard Neufeld, who was represented by Mr Clive Jones. Permission for that appeal was given by Maurice Kay LJ, whose view was that there was “an element of uncertainty bordering on disarray in the recent EAT decisions and in their relationship with [the decision of the Court of Appeal in Secretary of State for Trade and Industry v. Bottrill [1999] ICR 592]. The issue clearly requires clarification by the Court of Appeal.” The judgment of the Employment Appeal Tribunal under appeal in Mr Neufeld’s case was given by His Honour Judge McMullen QC, sitting alone. He reversed the decision of an employment judge who had found in favour of the Secretary of State.

The respondent to the second appeal is Mr Keith Howe, who has not been represented before us. In his case there was no substantive hearing before the Employment Appeal Tribunal. What happened there was that, at the Secretary of State’s request, Bean J simply dismissed the Secretary of State’s appeal against the decision of the employment judge in favour of Mr Howe and gave permission to appeal to this court. That was done so that the appeal could catch up with the appeal in Mr Neufeld’s case and be heard at the same time.

Both cases arose out of the insolvency of companies in whose businesses Mr Neufeld and Mr Howe were respectively engaged. The only issue in the cases was whether each had been an “employee” of the failed company. If they had, they enjoyed the protection given by section 182 of the Employment Rights Act 1996 (“the ERA”) to employees whose employer has become insolvent. Section 182, in Part XII (“Insolvency of Employers”), provides:

“182.

Employee’s rights on insolvency of employer

If, on an application made to him in writing by an employee, the Secretary of State is satisfied that –

(a)

the employee’s employer has become insolvent,

(b)

the employee’s employment has terminated, and

(c)

on the appropriate date the employee was entitled to be paid the whole or part of any debt to which this Part applies,

the Secretary of State shall, subject to section 186, pay the employee out of the National Insurance Fund the amount to which, in the opinion of the Secretary of State, the employee is entitled in respect of the debt.”

It is unnecessary to cite the other provisions of Part XII. Section 183 describes when an employer has become relevantly insolvent. Section 184 describes the debts to which Part XII applies: they include arrears of pay for up to eight weeks, notice pay to which the employee was entitled, unpaid holiday pay for up to six weeks during a defined 12-month period and any basic award for compensation for unfair dismissal. Section 186 imposes a limit on the total amount per week payable to an employee. Section 188 entitles a claimant who has applied for a payment from the Secretary of State, and who is not paid what he claims he should have been, to make a complaint to an employment tribunal. That is what happened in both cases.

The Part XII rights apply only to someone who was an “employee” of the insolvent employer; and the question whether he was has to be answered as at the date when the company became insolvent (see the decision of the Employment Appeal Tribunal in Rajah v. Secretary of State for Employment, unreported, 7 July 1995, EAT/125/95, Mummery J, the President, presiding). Section 230 of the ERA contains the relevant definitions:

“230.

Employees, workers etc

(1)

In this Act ‘employee’ means an individual who has entered into or works under (or, where the employment has ceased, worked under) a contract of employment.

(2)

In this Act ‘contract of employment’ means a contract of service or apprenticeship, whether express or implied, and (if it is express) whether oral or in writing. …

(4)

In this Act ‘employer’, in relation to an employee or worker, means the person by whom the employee or worker is (or, where the employment has ceased, was) employed.

(5)

In this Act ‘employment’ –

(a)

in relation to an employee, means (except for the purposes of section 171) employment under a contract of employment, and

(b)

in relation to a worker, means employment under his contract;

and ‘employment’ shall be construed accordingly.”

The special feature of each case under appeal is that the claimant was the controlling shareholder and a director of the company. The narrow questions raised by the appeals are whether the decisions of the employment judges were correct. More broadly, the Secretary of State asks the court to clarify the approach to be adopted by employment tribunals when faced with cases involving such a feature: indeed, Mr Tolley made it clear at the outset of his argument that that is the real purpose of these appeals. Consistently with that stance, the Secretary of State has in both cases undertaken not to seek costs from either respondent should he succeed in either appeal. The matter is of considerable importance to him. We were told that in 2008 there were some 12,000 claims by directors on the National Insurance Fund, of which some 600 had gone or were expected to go to employment tribunals. We were not told, because the information was not readily available, how many of these were by directors who were also controlling shareholders, but no doubt the number is considerable. The Secretary of State’s position is that, such is the conflicting guidance in the authorities as to how to resolve the type of question that has arisen in the cases under appeal, it is essential that this court should now give clear guidance on the subject.

In this judgment we will: (i) explain the facts found by, and conclusion of, the employment judge in Mr Neufeld’s case and then summarise the decision of the appeal tribunal; (ii) explain the facts found by, and conclusion of, the employment judge in Mr Howe’s case (there is no judgment of the appeal tribunal in his case); (iii) review the authorities; (iv) summarise what we regard as the principles applicable to cases such as these; and (v) give our decision on each appeal.

The decision of the employment judge

Mr Neufeld’s claim under section 188 of the ERA was decided by an employment judge (Mr J.R. Hardwick, sitting alone) by a judgment promulgated on 1 February 2007. The respondents were: (i) A & N Communications in Print Ltd (“A & N”), the putative employer, which was in insolvent liquidation; and (ii) the Secretary of State for Trade and Industry. Only the Secretary of State actively opposed the claim, which was for a redundancy payment, notice pay and holiday pay amounting to about £10,000. The single issue before the judge was whether Mr Neufeld was an employee of A & N in October 2005 when it became insolvent. His conclusion was that Mr Neufeld was not such an employee and so the claim was dismissed.

The judge found the following facts. Mr Neufeld commenced employment with Newfeld Press Limited (“NPL”) in 1982. He was a member of its sales team. It was common ground by the time the case came before the appeal tribunal that Mr Neufeld was an employee of NPL under a service contract, although the employment judge did not say so expressly. In 1988 Mr Neufeld became a shareholder in and director of NPL. In July 2001 NPL’s undertaking was transferred to A & N. The consequence of that was that the employment contracts of NPL’s employees became treated as if made with A & N. If Mr Neufeld was still an employee of NPL, he too became an employee of A & N.

The judge found that Mr Neufeld held 90% of A & N’s issued shares and was a director. He had two co-directors, Mr Munns and Mr Faulkner, who each had 5% shareholdings. On an unspecified date, which we infer to have been in about July 2001, the three directors agreed that Mr Neufeld would be A & N’s managing director and would also be employed as part of the sales team managed by Mr Faulkner, the sales director. Mr Neufeld’s evidence was that he worked an average of 60 hours a week carrying out sales, management duties and helping with production and that he was a successful sales person who achieved the highest volume of business.

The judge said in paragraph 3.3:

“There were about 18 employees in the business of [A & N] and all the manual employees had written contracts of employment. The three directors did not have contracts of employment.”

We read that as saying that there were 18 employees in addition to the three directors. The second sentence was unfortunately expressed. The sole question before the judge was whether Mr Neufeld was an “employee” of A & N; and, if he was, he did have a contract of employment. The judge was at this point setting out the facts, not making his decision. We interpret him as saying merely that the three directors did not have written contracts of employment: he has not yet decided whether or not they were employees. In paragraph 3.4 he explained that they had worked together for over 20 years (they had all been engaged in NPL’s business). He also found that Mr Neufeld had given personal guarantees in respect of A & N’s business.

In paragraph 4 the judge summarised the opposing submissions. The case for the Secretary of State was that whilst Mr Neufeld paid tax and National Insurance, he was a 90% shareholder of A & N and had given personal guarantees in respect of its liabilities. It was said that he did not take all his holiday entitlement and that, as he owned 90% of A & N, it was not possible for him to have been dismissed by it. We infer that the point there being made was that it followed that he could not have been an employee. That was a bad point that should not have been made (we make clear that Mr Tolley did not appear before the employment judge). Mr Neufeld’s case was that his status as controlling shareholder did not rule out such a contract and he was paid a weekly wage subject to PAYE and National Insurance. His employment contract could not be described as a sham nor was that alleged.

The judge expressed his conclusions in paragraph 5. He started, in paragraph 5.1, by saying that Mr Neufeld “certainly had a contract of employment albeit oral” and noted that it had not been suggested that it was a sham or was otherwise unusual save that it was not in writing. The judge was plainly there referring to the contract that Mr Neufeld was asserting to have existed as at the relevant date and so one might think that that finding marked the point of victory for Mr Neufeld. But no, the judge, having there apparently decided the only point he was required to decide, went on to make a finding to the reverse effect. He referred first to the tests developed by the courts for the purpose of distinguishing between contracts of service and for services, referring to the “economic reality” and “multiple” tests. He said the case law required him:

“… to look at the overall characteristics of the relationship and generally speaking no single factor on its own is generally conclusive. The Tribunal must weigh up each factor against the overall background picture and come to a conclusion.”

He said the one essential ingredient in an employment contract was an obligation upon the employer to provide work and a corresponding obligation upon the employee to accept and perform it. Such mutuality of obligation was not in issue in Mr Neufeld’s case.

The judge accepted that Mr Neufeld was a working director, was on the payroll and paid his tax and National Insurance. He accepted that “a large shareholding in itself does not debar that person from being an employee it is just one of the factors to be taken into account in the overall picture.” He did not, however, explain how it was or might be relevant in such a picture. He rejected the absence of a written contract as counting against Mr Neufeld’s claim. He then considered the question of the extent of A & N’s effective control over Mr Neufeld. Whilst the evidence was that all three directors had worked together over 20 years in a collegiate atmosphere and had together resolved to put A & N into liquidation, nevertheless as a 90% shareholder Mr Neufeld held the ultimate control and, in the face of disagreement with his co-directors, could have removed them and obstructed any efforts by them to remove or discipline him. He said that Mr Neufeld had said that “three quarters of the time when he was a director he did not take his full holiday entitlement” although we were told that his evidence was that he only failed to take one quarter of his holiday entitlement. The judge noted that employees generally take their full holiday allowance and said this was a consideration pointing against employee status “and more to one of operating one’s own business.”

In paragraph 5.7 the judge said that what pointed particularly against Mr Neufeld’s claimed employee status was the giving by him of the guarantees: one for £10,000 in relation to a machine and another for up to some £25,000 in relation to sales financing. A & N’s bank manager had asked him to give them. He had also made a personal loan of £20,000 to A & N. The judge regarded these as significant factors to put into the balance. He expressed his conclusions as follows:

“5.8

In my view [Mr Neufeld] has endeavoured to put a gloss on these issues. If a factoring company has advanced money and the sales do not materialise because of insolvency, it can have recourse to any guarantee. [A & N] itself may not be worth pursuing. There is no doubt in the Tribunal’s mind that in arriving at these arrangements [Mr Neufeld] was seeking to give an advantage to [A & N]. However at the same time in the Tribunal’s view he was involving himself in potential losses and liability. Very few employees would enter into such an arrangement where their own capital is at risk. In my view it points to [Mr Neufeld] running his own business as a manager and major shareholder of that business seeking commendably to secure ongoing finances through its bankers.

5.9

I have also taken into account against the overall background the very significant shareholding of [Mr Neufeld]. I am clear that the preponderance of characteristics of the relationship between [A & N] and [Mr Neufeld] point very much away from one of employer/employee particularly against the dimension of the personal guarantees.

5.10

It follows that as [Mr Neufeld] has been adjudged not to be an employee of [A & N] his claim against the Secretary of State must fail.”

The judge’s judgment shows his approach to have been as follows. The question required a consideration of the overall characteristics of the working relationship between Mr Neufeld and A & N. No single factor on its own was likely to be conclusive either way. Ignoring the guarantees, loan and Mr Neufeld’s control of A & N, there was apparently no doubt (apart only from the point about the amount of holiday he took) that he worked under what appeared to be a genuine employment contract with A & N. But the fact that he gave personal guarantees on A & N’s behalf, lent money to it and controlled it showed that in reality he was not an employee. They showed he was running his own business as a manager and major shareholder. The “preponderance of the characteristics in the relationship” showed that he was not an employee.

The decision of the Employment Appeal Tribunal

As a preliminary to the disposal of the appeal, Judge McMullen clarified the facts found by the employment judge. In paragraphs 16 and 17 Judge McMullen held that he had found Mr Neufeld to have been employed by NPL as a salesman from 1982 and that the same employment contract continued in and following July 2001, when the three directors agreed that Mr Neufeld – who held 90% of A & N’s shares -- was to be the managing director. There was, Judge McMullen rightly said, no reason why a director and shareholder of a company cannot also be an employee of the company. He also said that if, as the judge appeared to have found, Mr Neufeld’s 1982 employment contract (whether in its original form or as varied) continued after 2001, a simple solution to the case was that it would still have been in place in October 2005. There is, if we may say so, much to be said for that: the employment judge made no express finding to the effect that the original employment contract was ever discharged, although it is implicit in his conclusion that at some unidentified point in A & N’s history it was.

In paragraph 20 of his judgment, Judge McMullen indicated that the real focus of the appeal was on a consideration of what each party had done under the contract, which we read as meaning under the purported contract. In a case such as Mr Neufeld’s, where there was no written contract, that was obviously the correct approach: it was not enough for Mr Neufeld merely to assert that he had an employment contract, it was necessary for him to prove that what he actually did showed that he had such a contract. The employment judge had found that, on the face of it, Mr Neufeld did have such a contract, but had ultimately held that he did not. Judge McMullen’s view was that, in arriving at that conclusion, the employment judge had taken account of irrelevant matters. He summarised as follows why in his view the judge had erred in law:

“That the parties had other relationships such as shareholder, managing director, creditor or guarantor is relevant only to the extent that it reflects upon the conduct of the parties in carrying out the contract of employment. To adopt an approach which involves consideration of matters outside the conduct of the parties in performing the contract of employment casts the net too wide and is wrong in law. With respect, the Employment Judge in our case who considered ‘the preponderance of factors’ was wrong to include factors which did not reflect upon the conduct of the parties in the performance of the contract of employment.”

That conclusion meant that Mr Neufeld’s appeal had to be allowed. The ordinary consequence would be a remission of the case to the employment tribunal for its reconsideration on the correct basis but the parties agreed that Judge McMullen should make the decision himself rather than remit it. Judge McMullen did so by reference to the most recent decision on the type of issue that was before the employment judge, that of the Employment Appeal Tribunal in Clark v. Clark Construction Initiative Ltd [2008] IRLR 364 (Elias J, the President, presiding), a decision post-dating that of the employment judge. Elias J had there reviewed the authorities and outlined eight non-exhaustive factors for a tribunal’s consideration. Judge McMullen applied the factors, concluded that Mr Neufeld was an employee of A & N within the meaning of section 230 of the ERA and reversed the decision of the employment judge. There is no need to refer to the factors at this stage. We will come to them in the course of our review of the authorities.

The decision of the employment judge

Mr Howe’s claim under section 188 came before Employment Judge Burton on 26 November 2007, his reasons being promulgated on 18 January 2008. Mr Howe’s claim against the Secretary of State, the only respondent, was for a statutory redundancy payment of £765 following the insolvency of Track Records Music Limited (“TRM”). The issue was whether Mr Howe was an employee of TRM.

The judge’s findings were these. In 1979 Mr Howe started the business of Track Records, which was at that stage no more than his trading name. He was originally the sole proprietor of the business although between 1990 and 1995 he traded in partnership with his father. In December 2004 Mr Howe incorporated TRM. Track Record’s assets were transferred to it and it began to trade. Mr Howe held 100% of its shares and was the sole director. He arranged to pay himself a weekly salary of £255 and paid tax and National Insurance as an employed person. TRM’s profits were invested in its business. It had several employees, none of whom had written contracts of employment and nor did Mr Howe. He gave guarantees for TRM’s liabilities to suppliers. In August 2005 TRM needed a cash injection and Mr Howe arranged for it to borrow £50,000 from the bank, for which he gave a guarantee secured on his house. He also gave a guarantee to the landlords of TRM’s shop. By Christmas 2006 TRM was in financial difficulty. Mr Howe sought advice from insolvency practitioners, he ran the business down and by 4 August 2007 had disposed of TRM’s stock and the fixtures and fittings in its shop. He put TRM into voluntary liquidation and his claimed employment was brought to an end.

The judge said that had the relevant guidance to his decision been that of this court in Bottrill (see paragraph [2]above), he would have held that Mr Howe’s status as the owner of all TRM’s shares prevented him from being an employee of TRM. But he regarded the then recent decision of the Employment Appeal Tribunal in Nesbitt v. Secretary of State for Trade and Industry [2007] IRLR 853 (Underhill J presiding) as requiring him to find, as he did, that Mr Howe was an employee of TRM. He read Nesbitt as directing him that, even though TRM was wholly owned by Mr Howe, it was for him to find that Mr Howe was such an employee unless only he was satisfied that the purported contract of employment between TRM and Mr Howe was a sham. There was no evidence that it was a sham. The judge found that it was clear that there was a contract of employment between TRM and Mr Howe. He held that the Secretary of State was bound to pay Mr Howe the claimed redundancy payment of £765.

The decision of the Employment Appeal Tribunal

As we have said, Bean J simply dismissed the Secretary of State’s appeal and gave permission to appeal. The modest size of Mr Howe’s award probably explains why he was neither present nor represented before us. He was not disadvantaged by that because we had the benefit of a full argument on the issues from Mr Jones, for Mr Neufeld.

The general questions raised by both appeals are: (i) can a controlling shareholder and director of a trading company become an employee of that company under a contract of employment; (ii) if yes, are there any guidelines which may assist tribunals in deciding whether in any particular case such a shareholder/director has become an employee? The first question is not a novel one and has been considered in many authorities. It comes before us afresh because of what is regarded as confusion as to the principles which are applicable in answering it.

The logical starting point is a reminder of the essentials of a contract of employment. MacKenna J provided a well-known summary in Ready Mixed Concrete (South East) Ltd. v. Minister of Pensions and National Insurance [1968] 2 QB 497, at 515:

“A contract of service exists if these three conditions are fulfilled. (i) The servant agrees that, in consideration of a wage or other remuneration, he will provide his own work and skill in the performance of some service for his master. (ii) He agrees, expressly or impliedly, that in the performance of that service he will be subject to the other’s control in a sufficient degree to make that other master. (iii) The other provisions of the contract are consistent with its being a contract of service.”

MacKenna J’s judgment continued with an elaboration of each condition. It is unnecessary to refer to all of what he said, or to the more recent authorities which have considered the essentials of a service agreement: the appeals do not require it. We do, however, at least refer to MacKenna J’s elaboration of the “control” condition, which is of present relevance. He referred to “control” as including:

“… the power of deciding the thing to be done, the way in which it shall be done, the means to be employed in doing it, the time when and the place where it shall be done.”

The question of principle raised by the appeals is tested best by the consideration of the extreme position arising in the case of a so-called “one man” company. By that we mean a company in which one individual owns all its issued shares and is its sole director and thus has total dominion over its affairs. If in principle such an individual can have a contract of employment with his company, so in principle can an individual whose control is less than total but which is nevertheless still sufficient to give him control of the company in general meeting and of the constitution of its board of directors.

Approaching the matter apart from authority, it might be thought that the main conceptual barrier in the way of such a company granting a valid employment contract to its “one man” would be that there could in practice be no relevant control of the putative employee so as to meet that particular condition of an employment contract. In theory, the control would be exercisable by the company, a legal person separate from the individual. In practice, it would be exercisable by the putative employee himself since he controls the company. It is easy to conclude that that cannot be real control, with the consequence that a central condition of a contract of employment is missing.

It is, however, too late in the development of our jurisprudence, at any rate at the level of this court, to regard that particular control issue as providing a threshold obstacle to the creation of a valid contract of employment between a company and the one man who wholly controls it. The decision in Lee v. Lee’s Air Farming Ltd [1961] AC 12 shows that it is not. Whilst, as a decision of the Privy Council, it is not strictly binding on us, its correctness as an authority has not, so far as we are aware, been challenged, nor did Mr Tolley challenge it. He recognised that this court regarded it as sound law in Bottrill.

Mr Lee formed Lee’s Air Farming Limited for the purpose of conducting an aerial top-dressing business. He held 2,999 of its 3,000 issued shares, the other share being held by a solicitor. Article 32 of the articles of association appointed him governing director and conferred upon him “the full government and control of the company”. Article 33, as amended, provided that the company was to employ him as its chief pilot at an annual salary to be arranged by him and that “in respect of such employment the rules of law applicable to the relationship of master and servant shall apply as between the company and [Mr Lee]”. Mr Lee, as governing director and controlling shareholder, exercised unrestricted control over the company’s affairs and made all decisions relating to contracts for aerial top-dressing, contract prices, the manner in which the company’s aircraft was to be employed and the methods employed in carrying out the company’s work. He exercised unfettered control over the company’s operations at all times.

Mr Lee was killed in a crash whilst piloting the company’s aircraft on an aerial top-dressing operation. His widow sued the company on the basis that he had been employed under a “contract of service” with the company and that, under the Workers’ Compensation Act 1922 of New Zealand, his death in service as a result of an accident in the course of his employment entitled her to compensation. The contest was, in reality, between the widow and the company’s insurers. The only issue was whether or not Mr Lee was an employee of the company. The Court of Appeal of New Zealand held that, as the governing director of the company, he could not be its employee. Its view was that “the duty both of giving orders and obeying them” lay upon Mr Lee in his dual capacity as (in effect) employer and worker and the two offices were incompatible. It followed that “there could exist no power of control and therefore the relationship of master-servant was not created.” On the widow’s appeal to the Privy Council, her argument was that it was the company, not Mr Lee, that was the employer and it was the company that:

“… was entitled under the contract to control both what Lee did and how he did it, and this control remained the control of the company whoever might be the agent empowered by the company to exercise it”. ([1961] AC 12, at 21)

The argument for the company was that “[c]ontrol means a real living person to control another real living person” ([1961] AC 12, at 22); and that where “one gets the controller and the controlled person residing in the same person there is no control … it is wholly illusory.” ([1961] AC 12, at 23).

The Privy Council allowed the appeal and held that Mr Lee was an employee under a contract of service with the company. The judgment of Lord Morris of Borth-y-Guest is instructive for the purposes of the present appeals and plays an important role in the later authorities. Despite their length, we will therefore cite two passages ([1961] AC 12, at 24 to 27):

“The Court of Appeal thought that [Mr Lee’s] special position as governing director precluded him from being a servant of the company. On this view it is difficult to know what his status and position was when he was performing the arduous and skilful duties of piloting an aeroplane which belonged to the company and when he was carrying out the operation of top-dressing farm lands from the air. He was paid wages for so doing. The company kept a wages book in which these were recorded. The work that was being done was being done at the request of farmers whose contractual rights and obligations were with the company alone. It cannot be suggested that when engaged in the activities above referred to the deceased was discharging his duties as governing director. Their Lordships find it impossible to resist the conclusion that the active aerial operations were performed because the deceased was in some contractual relationship with the company. That relationship came about because the deceased as one legal person was willing to work for and to make a contract with the company which was another legal entity. A contractual relationship could only exist on the basis that there was consensus between two contracting parties. It was never suggested (nor in their Lordships’ view could it reasonably have been suggested) that the company was a sham or a mere simulacrum. It is well established that the mere fact that someone is a director of a company is no impediment to his entering into a contract to serve the company. If, then, it be accepted that the respondent company was a legal entity their Lordships see no reason to challenge the validity of any contractual obligations which were created between the company and the deceased. …

Nor in their Lordships’ view were any contractual obligations invalidated by the circumstance that the deceased was sole governing director in whom was vested the full government and control of the company. Always assuming that the company was not a sham then the capacity of the company to make a contract with the deceased could not be impugned merely because the deceased was the agent of the company in its negotiation. The deceased might have made a firm contract to serve the company for a fixed period of years. If within such period he had retired from the office of governing director and other directors had been appointed his contract would not have been affected. The circumstance that in his capacity as a shareholder he could control the course of events would not in itself affect the validity of his contractual relationship with the company. When, therefore, it is said that ‘one of his first acts was to appoint himself the only pilot of the company,’ it must be recognised that the appointment was made by the company, and that it was none the less a valid appointment because it was the deceased himself who acted as the agent of the company in arranging it. In their Lordships’ view it is a logical consequence of the decision in [Salomon v. Salomon & Co [1897] AC 22] that one person may function in dual capacities. There is no reason, therefore, to deny the possibility of a contractual relationship being created as between the deceased and the company. If this stage is reached then their lordships see no reason why the range of possible contractual relationships should not include a contract for services, and if the deceased as agent for the company could negotiate a contract for services as between the company and himself there is no reason why a contract of service could not be negotiated. It is said that therein lies the difficulty, because it is said that the deceased could not both be under the duty of giving orders and also be under the duty of obeying them. But this approach does not give effect to the circumstance that it would be the company and not the deceased that would be giving the orders. Control would remain with the company whoever might be the agent of the company to exercise it. The fact that so long as the deceased continued to be governing director, with amplitude of powers, it would be for him to act as the agent of the company to give the orders does not alter the fact that the company and the deceased were two separate and distinct legal persons. If the deceased had a contract of service with the company then the company had a right of control. The manner of its exercise would not affect or diminish the right to its exercise. But the existence of a right to control cannot be denied if once the reality of the legal existence of the company is recognised. Just as the company and the deceased were separate legal entities so as to permit of contractual relations being established between them, so also were they separate legal entities so as to enable the company to give an order to the deceased.” (Emphasis added)

In a later passage of his judgment, at [1961] AC 12, at 29, Lord Morris explained why, on the facts, the relevant contract was a contract of service, and not a contract for services.

Lee’s case therefore establishes two propositions of direct present relevance. First, that an individual who owns all the shares in, and is the sole director of, a company – and so has total dominion over it – can also be employed by that company under a contract of service. Secondly, that it is no answer to the claimed creation of such a contract that the “control” condition that is essential to it is not satisfied. The answer to that point, even in relation to a “one man company” case, is that the company and the one man are not the same person; and it is the company that exercises the relevant control. In Lee’s case the employer was the company and the employee was Mr Lee. The control necessary for the purposes of the claimed contract of service was exercisable by the company and it made no difference that in practice, so long as Mr Lee remained the sole governing director, that control would be and was exercised by him as the company’s agent. The close identity that in reality existed between the company and Mr Lee did not prevent a contract for service being created.

We cannot leave Lee’s case without a comment on Lord Morris’s observation that it had not been, nor could be, suggested that “the company was a sham or a mere simulacrum” (our emphasis), an observation that has been discussed in the later authorities to which we shall come. It is a difficult expression because a registered company has legal personality. However, we consider that Lord Morris was probably there using the words “sham” and “mere simulacrum” as synonyms for essentially the same idea and had in mind the limited types of case in which an individual (as in Lee’s case) owns all the shares in a company and the courts have considered it right for policy reasons to “pierce the veil” of incorporation and treat the company as the alter ego of the controlling shareholder, that is to treat them as one. In such a case, any suggestion that the individual had a service contract with the company would not succeed.

It appears to us that such circumstances, at least in a case in which the company is a genuine trading company, would be exceptional. No such question arises in these appeals, nor did it arise in the authorities to which we were referred. We propose therefore to say no more about it. On the other hand, as Bottrill in this court makes clear, a preliminary question which may more commonly arise in a case in which a controlling shareholder claims to have a service contract with his own company will be whether the putative contract (rather than the company) is genuine or a sham. That is because the reality in such cases is that the controlling shareholder will have been the directing mind and will behind the purported creation of his own contract. That factor will be likely in many cases to require a careful scrutiny of the claim that a valid employment contract has been created.

As to the circumstances in which the contract might be regarded as a sham, the classic case will be that described by Diplock LJ in Snook v. London and West Riding Investments [1967] 2 QB 786, at 802:

“As regards the contention of the plaintiff that the transactions … were a ‘sham,’ it is, I think, necessary to consider what, if any, legal concept is involved in the use of the popular and pejorative word. I apprehend that if it has any meaning in law, it means acts done or documents executed by parties to the ‘sham’ which are intended by them to give to third parties or to the court the appearance of creating between the parties legal rights and obligations different from the actual legal rights and obligations (if any) which the parties intend to create. … But one thing, I think, is clear in legal principle, morality and the authorities (see Yorkshire Railway Wagon Co. v. Maclure (1882) 21 Ch. D. 309 and Stoneleigh Finance Ltd v. Phillips [1965] 2 QB 537), that for acts or documents to be a ‘sham’ with whatever legal consequences follow from this, all the parties thereto must have a common intention that the acts or documents are not to create legal rights and obligations which they give the appearance of creating. No unexpressed intentions of a ‘shammer’ affect the rights of a party whom he deceived.”

In most cases in which there arises a question as to whether the claimed employment contract is a sham, there will be what purports to be a formal written contract, or at least a board minute or memorandum purporting to evidence or record the contract: a “shammer” is hardly likely to rest his case on the claim that his contract was an oral one. An inquiry into whether any purported contract does amount to a sham does not limit the court or tribunal to a consideration of the evidence as at the time of its making. It will usually also be highly material to see what the company on the one hand and the shareholder/director on the other have actually done under the purported contract: that will be likely to shed light on its genuineness or otherwise. In principle, however, a similar problem could arise where the alleged employment contract is an oral one, for it might be said in response that the basis on which such a contract is alleged is a mere pretence and is false. This court has recently considered the concept of a sham in an employment law context: Protectacoat Firthglow Ltd v. Miklos Szilagyi [2009] EWCA Civ 98.

Having made those comments about the possibility of a purported service agreement in favour of a controlling shareholder being a sham, we emphasise that these appeals are not about allegedly sham contracts either. In the cases under appeal there was no question of either contract being a sham. The question in each is whether, accepting that the evidence supported the conclusion that the shareholder/director had worked under what was ostensibly a genuine service agreement, the fact that it was made in favour of an individual who was also the company’s controlling shareholder – who had personally invested in the business and assumed personal liabilities for it -- prevented the contract from amounting to aservice agreement which could be recognised as such. The real question, therefore, is whether the putative employee’s control of the company is fatal to the creation of a service agreement in his favour.

Lee’s case might be thought to have provided a resounding “No” in answer to that question. But it did not. Mummery J, when President of the Employment Appeal Tribunal, made clear his own views on the topic, which can be said to have side-stepped the effect of Lee’s case. The first of his decisions was McQuisten v. Secretary of State for Employment, unreported, 11 June 1996, EAT/1298/95. Following the company’s entry into liquidation, Mrs McQuisten made a claim to a redundancy payment from the Secretary of State under the statutory predecessor of Part XII of the ERA. She was an 85% shareholder and (with her son) one of the company’s two directors. She worked full time for it, her work including two hours a day on the books and seven hours a day labouring in the company’s sawmill under the direction of her son and daughter, the other shareholders. She paid Schedule E tax and National Insurance contributions. She had guaranteed a company loan. The industrial tribunal found that she was not an employee. The Chairman was particularly influenced by the fact that she could secure to herself virtually complete control of the company. His conclusion was that she was, “in effect, carrying on a business under the auspices of a limited company”, a view reinforced by the fact that she was a guarantor of the company.

On Mrs McQuisten’s appeal, the appeal tribunal directed themselves (correctly) that the question before the industrial tribunal was one of fact and that they could only interfere with the decision if it was flawed by an error of law. They rejected the argument that the Chairman’s decision was perverse and upheld it as a decision to which he was entitled to come. He was, they said, entitled to place more weight on the control factor than on other factors and he made no error in that respect. The importance of their decisionlies in Mummery J’s postscript to his judgment, headed “Majority Shareholding”. Its essence was that an individual in control of a company could not normally be dismissed from any contract of employment save with his own concurrence; and, if he could not be dismissed, it was difficult to see how he could be regarded as an employee. There were, Mummery J said:

“… real legal difficulties in regarding a majority shareholder as an employee when he is ultimately able to exercise his shareholder’s powers to determine every question concerning his employment – his rates of pay, his duties, his holidays, his pension arrangements and, most important, whether he continues to work for the company or not. It is unlikely that Parliament intended to include such persons within the scope of the protection of the 1978 Act [the Employment Protection (Consolidation) Act 1978].”

Given that view, it is no surprise that the appeal was dismissed. A question arises as to how such view was reconcilable with Lee’s case, which does not appear to have been cited. Mummery J returned to this theme in the joined appeals of Buchan v. Secretary of State for Employment; and Ivey v. Secretary of State for Employment [1997] IRLR 80 (“Buchan”) and this time he had been referred to Lee’s case. Buchan involved like claims by a majority shareholder and director to have been an employee of his insolvent company. Mr Buchan had a 50% shareholding in his company, worked for it full time as a scanner operator and sales manager, received £35,000 a year subject to PAYE and National Insurance and was entitled to five weeks’ holiday a year. He did not have a written service contract. Mr Ivey had a 99% shareholding in his company and did have a written service contract, one entitling him to a salary of £28,500 and 22 days’ holiday a year. In both cases the industrial tribunal found that the claimants were not employees. The appeal tribunal dismissed both appeals.

In paragraph 8 of his judgment, Mummery J identified four principles which are supported by authority and correct: (1) a limited company is a distinct legal entity from its shareholders and directors; (2) a director, and managing director, of a limited company may have a contract of service with it; (3) a shareholder may likewise have a contract of service with it; and (4):

“a person will usually be regarded as having entered into and as working under a contract of service if he is bound by contract (whether express or implied, whether in writing or orally), (a) if he is bound to devote all his time to the affairs of the company and to do all in his power to develop and extend its business, and (b) if he is engaged on terms which provide for the payment of a regular fixed salary and for determination of that contract.”

The last point is important. If the putative contract was not a sham (in the Snook sense), there is a separate potential question whether it is properly to be characterised as a contract of employment rather than some other form of contract. That involves an analysis of the parties’ respective contractual rights and obligations.

In paragraph 9, Mummery J expressed his view that the application of his four legal principles either will or may not be capable of applying to an individual who, by reason of his controlling shareholding in the company, is able to prevent his dismissal from his position in it. His opinion was that the critical provisions of the Employment Protection (Consolidation) Act 1978 (his point would apply equally to the successor provisions of the ERA) gave legal protection to a person dismissed unfairly (including dismissal for redundancy on the insolvency of an employer) and provided guaranteed State payments for employees in that event. He said it would be inconsistent with the purposes of the 1978 Act to extend such protection to a person who cannot be dismissed without his consent. The argument that Lee’s case showed that such persons can be employees of their company ignored the different context in which Lee’s case was decided, that of a claim for compensation under the New Zealand Workers’ Compensation Act 1922. Mummery J said:

“12.

The issue in this case arises in the context of employment protection legislation. Taking the facts of Mr Lee’s case it is difficult in an employment protection context to conceive of circumstances in which Mr Lee could have been regarded as an employee of the company for the purposes of making a claim against the company for a redundancy payment or for unfair or wrongful dismissal. None of those events could give rise to any remedy by him against the company, because the wrong of which he would complain could not occur without his concurrence. That feature would also affect the position of a guarantor of a liability of the company, such as the Secretary of State in relation to redundancy and other payments. The liability of the Secretary of State is to make a payment which the insolvent employer is liable to make, but cannot make because of lack of funds. If a person has no remedy against the company over which he exercises control, he cannot have any claim against the guarantor of the company’s liability.”

Mr Buchan was in fact dismissed by the company’s administrative receivers, but the appeal tribunal held that as his 50% shareholding had previously enabled him to block any decision to dismiss him, he was never an employee of the company and his purported dismissal did not make him one. Mr Ivey was, a fortiori, similarly not an employee: the label on his purported service contract was not conclusive as to its substantive effect.

As we shall explain, this court in Bottrill later disagreed with and rejected the basis on which Mummery J distinguished Lee’s case. His point that a controlling shareholder of a company cannot also be its employee for the purposes of the employment protection legislation is no longer to be regarded as sound. The Secretary of State’s attempted resurrection of the point before the employment judge in Mr Neufeld’s case was mistaken.

Before the decision in Bottrill’s case, Buchan was considered by the Court of Session in Fleming v. Secretary of State for Trade and Industry [1997] IRLR 682. We must refer to Fleming, since its approach was later endorsed by Bottrill. Mr Fleming held 65% of the shares in the company and worked alongside its employees, working the same 40 hours per week that they did. He had given guarantees to the company’s main suppliers. He had no written contract of employment nor was there a board minute or memorandum recording an agreement to employ him. His remuneration was paid under the PAYE system although when the company ran into financial difficulty he suspended collection of his remuneration until sunnier days. They did not come, the company went into liquidation and he was dismissed by the liquidator. He claimed redundancy and statutory notice payments out of the National Insurance Fund but the Secretary of State refused to accept that he had been an employee. The industrial tribunal dismissed his application, holding that he was not an employee. It was clearly influenced by the fact that he was a majority shareholder, had guaranteed the company’s obligations and was in control of the company: he was not under anyone else’s control. The appeal tribunal dismissed his appeal and endorsed Mummery J’s views in McQuisten that “where the person involved is a majority shareholder in the company …, his status as an employee must be seriously in doubt.”

Mr Fleming appealed to the Court of Session. He challenged the proposition that a majority shareholder could not be an employee of the company. The counter argument was that (i) the question for the industrial tribunal was one of fact, (ii) the fact that his remuneration was taxed under Schedule E by the PAYE system was not significant since a director was an office-holder and taxable on the same basis, (iii) the giving of guarantees was a relevant consideration, and (iv) the shareholder control pointed away from an employment relationship. The court (The Lord President (Lord Rodger), Lord Kirkwood and Lord Coulsfield) dismissed the appeal. The judgment of the court was delivered by Lord Coulsfield, who accepted the arguments for the Secretary of State. He said:

“9.

… In the present case, it seems to us that, taking all the factors together, there was ample material to entitle the industrial tribunal to reach the conclusion that the appellant was not an employee for the purposes of the employment protection legislation. There were … factors on which the appellant could rely on [sic] as pointing towards an employment relationship, such as the fact that the appellant worked alongside other employees of the company, with the same hours of work, and had no other employment. On the other hand, the fact he was able to decide to draw no salary during the last month of the company’s existence, the fact that he personally guaranteed its obligations and the fact that he held a substantial majority shareholding are all factors which point to the opposite direction, as is the fact that there was no written record of his terms of employment. The fact that the salary which the appellant received was paid through the PAYE system is, in our view, neutral. In all the circumstances, therefore, we have no doubt that there was ample material on which the industrial tribunal was entitled to reach the conclusion that the appellant was not an employee.

10.

In reaching the above conclusion, we have treated the fact that the appellant held a majority shareholding in the company as a relevant factor. We do not see how it could, in common sense, be doubted that the fact that a person is a shareholder is a relevant factor. The significance of that factor will depend on the circumstances, and the weight to be given to it may vary with the size of the shareholding. It is true that, as the appellant pointed out, a change in the size of a shareholding might, on that view, lead to a change in the view of the status of a particular person. The decision as to whether a person is or is not an employee must, however, be taken on all the relevant factors at the material time. The shareholding position at the material time must, in our view, be a relevant factor. It will, however, usually only be one of a number of such factors, and it is not impossible that regard might be had to the way in which the person in question comes to be a shareholder, or to be a majority shareholder. As in any such decision, all the circumstances have to be considered.”

In paragraphs 11 and 12 of the judgment, Lord Coulsfield recorded that counsel for the Secretary of State had come close to submitting that there was a rule of law that a controlling shareholder of a company could not be treated as an employee of the company for the purposes of the employment protection legislation, whatever might be the situation in other contexts. The argument was based on inter alia the decisions in McQuisten and Buchan. Lord Coulsfield regarded Buchan as also coming close to asserting the existence of such a rule. The court expressed its reluctance to see the introduction of any such purported rule of law into the determination of what was well established as being a question of fact and it declined to enter into a consideration of the argument. It contented itself with saying that:

“12.

… It can easily be seen, as we have said, that the fact that the claimant is a majority shareholder is always a relevant factor. We are not, however, convinced that it would be proper to lay down any rule of law to the effect that the fact that a person is a majority shareholder necessarily and in all circumstances implies that that person cannot be regarded as an employee, for the purposes of the employment protection legislation.”

We have found it difficult, with respect, to know what to make of Fleming and express respectful reservations as to the emphasis it places on the relevance of the putative employee’s status as a majority shareholder in determining whether or not he is also an employee of the company. Of course in answering that question, his status as a director/controlling shareholder will or may be one of the factors which will have given rise to such question and will be part of the factual background against which the question falls to be answered. But the answer to it must ultimately depend on what he agreed with the company he would do, whether he did it and whether what he so agreed and did show that he had a contract of employment with the company. If such inquiry shows that he did have such a contract, his status as director/controlling shareholder will not ordinarily require any different conclusion. Nor can we see the relevance of the fact that he gave guarantees to creditors of the company or that he did not draw his salary during the last month of the company’s existence.

As to the giving of guarantees, a controlling shareholder/director may enter into different relationships. He may arrange for the company to borrow money from the bank, for which he may give a personal guarantee. He may also arrange a contract between the company and himself, which may be a contract of service or a contract for services. We cannot see a logical connection between his giving such a guarantee and the ascertainment of the nature of any such contract between the company and himself, which must depend on the nature of the obligations undertaken.

As to his not drawing his salary, we can see that it would point against the existence of a contract of employment if his remuneration had been generally irregular. But we find it difficult to see how it could be concluded that he was not entitled to the payment of a salary if it had been paid with regularity until the final month: and if he was contractually entitled to it, the fact that he did not take it could not retrospectively diminish his right.

We come now to what is currently the leading case in this corner of the law, the decision of this court in Secretary of State for Trade and Industry v. Bottrill [1999] ICR 592. Mr Bottrill acquired a shelf company called Magnatech UK Limited. He acquired the only issued share – and so controlled the company -- and was a director. He had a co-director. He had set the business up under an arrangement with a group of companies in the United States, which was to invest in the company and acquire an 80 per cent shareholding in it. A draft agreement to this effect was, however, never executed. The industrial tribunal held that Mr Bottrill’s shareholding was only intended to be a temporary matter. He had a written service agreement for a term of three years with the company under which he was paid a salary from which tax and National Insurance contributions were deducted. He was not paid director’s fees. The company became insolvent and Mr Bottrill made a claim for a redundancy payment against the Secretary of State. The industrial tribunal held he was an employee of the company and upheld his claim. They held that the factors pointing towards an employment relationship (the service agreement, salary payments, holiday and sick pay arrangements) outweighed that pointing against it (his “theoretical” control of the company by his 100% holding – theoretical because of the arrangement with the United States group).

The appeal tribunal, by a judgment delivered by Morison J, the President, dismissed the Secretary of State’s appeal ([1998] ICR 564). The Secretary of State placed reliance on Buchan, which Morison J said had given rise to concern by its apparent disregard of the separate legal identities of a company and its controlling shareholder. Morison J considered Lee’s case and expressed the view that, context apart, the question in Buchan was indistinguishable from it; and that the different context of Lee’s case did not provide a basis for the distinction that Buchan had drawn. On the other hand, Buchan had been decided as recently as July 1996 and Morison J was reluctant within a mere 18 months to adopt an inconsistent approach. He regarded the solution as lying in the Court of Session’s decision in Fleming. Contrary to the view expressed in Fleming, Morison J considered that Mummery J had purported to establish a principle of law to the effect that a controlling shareholder could not be an employee for the purposes of the employment protection legislation in the ERA. But he regarded Fleming as deciding that there was no such principle and that the question whether someone is or is not an employee is always one of fact.Morison J then said ([1998] ICR 564, at 574):

“The shareholding of a person in the company by which he alleges he was employed is a factor to be taken into account, because it might tend to establish either that the company was a mere simulacrum or that the contract under scrutiny was a sham. In our judgment it would be wrong to say that a controlling shareholder who, as such, ultimately had the power to prevent his own dismissal by voting his shares to replace the board, was outside the class of persons given rights under the Act of 1996 on an insolvency.”

We cite that passage for two reasons. First, as this court would confirm on the appeal against the appeal tribunal’s decision, Morison J was there correct to depart from the principle that Mummery J had espoused in Buchan. Secondly, because whereas Fleming had contented itself with saying merely that the fact of a majority shareholding is a factor to take into account – without saying why – Morison J now did explain why. In our judgment his explanation (drawing on what Lord Morris had said in Lee’s case about the company being a “sham or a mere simulacrum” and adding the important point that the fact of the controlling shareholder’s control would be relevant to whether the purported contract was a sham) was right. Lee’s case disposed of the argument that a controlling shareholder cannot be subject to the control condition necessary for the creation of a valid contract of employment. Once both that argument and the Buchan fallacy are removed from the arena, the onlyrelevance of the putative employee’s control of the company is likely to be in the context of the consideration of whether (a) the contract he is asserting was a sham rather than a genuineemployment contract (for the obvious reason that he will have been pulling the strings that will have purportedly created it) or (b) if intended to be genuine, its terms and the way in which it was intended to be performed (that is, in essence, the services to be rendered and the remuneration to be paid) were compatible with its being classed as a contract of employment. Once it is found that it is both a genuine contract, and apparently also a genuine employment contract – by which we mean that it is just what it purports to be -- it cannot cease to be such merely because the employee has a shareholding ranging from, say, 51% to 100% in the company. By that point in the inquiry, the size of his shareholding and the extent of his control will ordinarily have become irrelevant.

We would also observe as follows. In Fleming the contract in question was not in writing, whereas in Bottrill it was. Where a written contract is concerned, the question of whether it is a contract of employment or not will turn, subject to any evidence that it is a sham, essentially on its terms and matters of general law (see Chitty on Contracts, 30^th ed, 2008, Vol II at para 39-010ff). Where, however, the contract is not in writing, it will of course need to be proved by other means, and, given the greater informality likely to obtain in “one-man” companies, the fact and extent of the alleged employee’s control of the company may be a factor for consideration.

The Secretary of State appealed to this court, which consisted of Lord Woolf MR, Peter Gibson and Mantell L.JJ ([1999] ICR 592). Lord Woolf delivered the judgment of the court dismissing the appeal. He opened the judgment by saying that the issue was whether Mummery J was right in Buchan to hold that a controlling shareholder could not be an employee of his company or the Court of Session was right in Fleming to hold that it was always a question of fact in all the circumstances. Lord Woolf cited extensively from Buchan and Fleming. He then came to Morison J’s judgment and referred to the reference in it to Lee’s case (which Lord Woolf misdescribed at [1999] ICR 592, 600 as a reference to another case, but there can be no doubt that he was intending to refer to Lee’s case). He cited from Morison J’s judgment, including the passage we have cited. He then came to the court’s conclusion.

Lord Woolf said the court had no hesitation in preferring the approach of the appeal tribunal in Bottrill and of the Court of Session in Fleming as compared with that of Mummery J in Buchan. He said that Mummery J’s attempted distinguishing of Lee’s case would result in the concept of an “employee” being given a special meaning for the purposes of employment protection legislation which it does not have in other contexts. There was no justification for that distinction. Lord Woolf then said this at [1999] ICR 592, 603:

“We recognise the attractions of having in relation to the Act of 1996 a simple and clear test which will determine whether a shareholder or a director is an employee for the purposes of the Act or not. However, the Act does not provide such a test and it is far from obvious what Parliament would have intended the test to be. We do not find any justification for departing from the well established position in the law of employment generally. That is whether or not an employer and employee relationship exists can only be decided by having regard to all the relevant facts. If an individual has a controlling shareholding that is certainly a fact which is likely to be significant in all situations and in some cases it may prove to be decisive. However, it is only one of the factors which are relevant and certainly is not to be taken as determinative without considering all the relevant circumstances.”

Having offered that general statement of principle, essentially in line with the guidance in Fleming, the court explained why the appeal should be dismissed (in fact, counsel for the Secretary of State had conceded that, if the court disagreed with Buchan, the appeal had to be dismissed). The court then, at counsel’s request, gave the following guidance for like cases in the future ([1999] ICR 592, at 603, 604):

“… We are anxious not to lay down rigid guidelines for the factual inquiry which the tribunal of fact must undertake in the particular circumstances of each case; but we hope that the following comments may be of assistance.

The first question which the tribunal is likely to wish to consider is whether there is or has been a genuine contract between the company and the shareholder. In this context how and for what reasons the contract came into existence (for example, whether the contract was made at a time when insolvency loomed) and what each party actually did pursuant to the contract are likely to be relevant considerations.

If the tribunal concludes that the contract is not a sham, it is likely to wish to consider next whether the contract, which may well have been labelled a contract of employment, actually gave rise to an employer/employee relationship. In this context, of the various factors usually regarded as relevant (see, for example, Chitty on Contracts, 27^th ed. (1994), vol. 2, pp. 703-704, para. 37-008), the degree of control exercised by the company over the shareholder employee is always important. This is not the same question as that relating to whether there is a controlling shareholding. The tribunal may think it appropriate to consider whether there are directors other than or in addition to the shareholder employee and whether the constitution of the company gives that shareholder rights such that he is in reality answerable only to himself and incapable of being dismissed. If he is a director, it may be relevant to consider whether he is able under the articles of association to vote on matters in which he is personally interested, such as the termination of his contract of employment. Again, the actual conduct of the parties pursuant to the terms of the contract is likely to be relevant. It is for the tribunal as an industrial jury to take all relevant factors into account in reaching its conclusion, giving such weight to them as it considers appropriate.”

Later cases have shown that the guidance offered in the last paragraph has given rise to difficulties of interpretation and application. The court was addressing the point that where there was a genuine intention to enter into a contract, and a genuine intention that it should be a contract of employment, there could still be a live issue as to whether the contract was properly classed as one of employment. Difficulties have arisen because the ensuing comments embrace two different things: (a) the extent of the power of the shareholder/director to control the company’s decision making (for example, in relation to whether to enter into or to terminate the contract) and (b) the content of the contract. We should set out our understanding of the position.

Factor (b) is critical: what did the company (acting through the shareholder/director) agree with the shareholder/director that he was going to do qua contracting party and for what remuneration? There is in a sense an artificiality about that whole question, but companies are artificial legal entities and the ability of a “one man company” to enter into a contract with that one man was the very point established by Salomon. It follows from that ruling, as Lord Morris said in Lee, that a company controlled by a single person is equally capable of entering into a contract of service or a contract for services with that individual.

What then is the potential relevance of factor (a) (the extent of the individual’s power to control the company’s decision making) to factor (b) (the content of the contract)? Where the contract (and for this purpose it is being assumed that there was a valid contract) was in writing and sufficiently explicit to establish its content, we cannot see the relevance of factor (a). If, however, the contract was not in writing, or was expressed only in short form, so that it is necessary to examine the conduct of the parties in order to deduce the content of the contract, the position of the individual and manner in which the company’s affairs were conducted provide the factual setting for the inquiry. As regards the court’s point earlier made (in the passage we have cited in paragraph [57]) as to the significance of the putative employee being a controlling shareholder, our comment is the same as we have made in paragraph [49] in relation to the like point made by the Court of Session in Fleming.

The next case in the chronology is the decision of this court in Connolly v. Sellers Arenascene Ltd [2000] ICR 760, which we can deal with relatively shortly. The ultimate question arising against a complicated background was whether the claimant had an employment contract in the company he controlled with an 88% shareholding. He had a written contract of employment which the industrial tribunal found was not a sham. But the tribunal (their decision pre-dating this court’s decision in Bottrill)nevertheless found that he was not an employee. They accepted that he behaved as an employee. What, however, counted against his being one was that his status as a shareholder gave him an interest over and above that as an employee and he stood to gain as a shareholder if the company prospered. He was, the tribunal found, at pains to retain absolute control.

That sort of reasoning was, with respect, erroneous for it was essentially to the mistaken effect that the owner of the company cannot also be its employee. If right, it meant that the decision in Lee’s case was wrong. The appeal tribunal, presided over by Morison J, reversed the tribunal’s decision. In the light of Morison J’s approach to the like question in Bottrill, that was no surprise. This court (Pill, Chadwick L.JJ and Wright J) dismissed the appeal. By then this court’s decision in Bottrill had been delivered. Pill LJ delivered the leading judgment, with which the other members of the court agreed. His view was that the industrial tribunal had attached a significance to the controlling shareholding that excluded a proper consideration of the other relevant factors. He pointed out, rightly in our view, that the tribunal had misunderstood the relevance of the claimant’s shareholding. Its decision could be upheld only if the controlling shareholding was a decisive factor, “which, in the circumstances and on the evidence, in my view, it was not.” The essence of this court’s decision appears to have been to the effect that, once the tribunal’s irrelevant considerations were rejected, there was nothing in the fact of Mr Connolly’s control of the company preventing him from being an employee.

It is now necessary to see how the guidance in Fleming and Bottrill has been applied in subsequent cases, in particular by the Employment Appeal Tribunal. The first case is Hauxwell and another v. Secretary of State for Trade & Industry and another 19 June 2002, EAT/386/01, Wall J presiding. The appellants were Mr and Mrs Hauxwell, who each held 50% of the shares in, and were directors of, the respondent company. The employment tribunal had found that terms and conditions of employment had been issued to them but that all decisions in relation to the company were made by them jointly, they having complete control of it. They were joint guarantors of the company. On appeal, Mr Hauxwell amplified the facts (we do not understand how he could have done so on an appeal lying solely on the ground of an alleged error of law) by showing that he had a written contract of employment, worked fixed hours and was paid a salary fixed in advance. The employment tribunal had rejected the claims of both Mr and Mrs Hauxwell to be employees. The decisive consideration was their control of the company (they relied on Buchan but made no reference to Bottrill). Their view was that “Mr and Mrs Hauxwell were the company, they were not employees of it.”

The appeal tribunal reviewed Bottrill and Fleming. It upheld the employment tribunal’s decision as having been based on a finding of fact that the tribunal was entitled to make and which involved no error of law. In his judgment, Wall J accepted the submission that the tribunal had undertaken a balancing exercise between the facts and circumstances pointing to and away from an employment relationship; the significance attached to the Hauxwells’ controlling shareholding did not exclude a proper consideration of other factors (notably that they were paid and worked on the same basis as other employees); and the appeal tribunal considered there were further factors supporting the tribunal’s decision, including that neither appellant could dismiss the other and each was subject to the control of the other rather than of the company.

Hauxwell is therefore a case in which the appeal tribunal endorsed the view that the function of an employment tribunal is to look at all the circumstances, balance them in some way and make a finding as to whether the claimant is or is not an employee. That approach reflects one reading of the guidance in Fleming and Bottrill. Our difficulty with it is that it appears to us, with respect, to permit or require a balancing exercise without explaining the supposed relevance of the various factors required to be put into the scales.

The next authority is the Employment Appeal Tribunal’s decision in Gladwell v. Secretary of State for Trade and Industry [2007] ICR 264 (Elias J, the President, presiding). The claimant was a 50% shareholder and director of the company. The remaining shares were held by his co-director. The claimant worked full-time for the company, was on its payroll and his salary was paid after deduction of PAYE and National Insurance contributions. He had produced a purported written contract of employment, which did not include the signature page, although the evidence was that the claimant would have signed it twice “because I was employer and employee” (evidence given in the context that, had it been formalised, he would have signed both for himself and the company). The purpose of the contract was to record an agreed reduction in salary. The claimant had also put substantial capital into the company by way of a subscription for an issue of preference shares. The employment tribunal regarded it as a borderline case. There were obvious factors pointing towards an employment contract, although the purported written contract, as an incomplete document, deserved little weight; pointing the other way were the injection of capital, the question of the claimant’s control and his statement that he was both employer and employee, which the tribunal said was an impossibility. The tribunal found he was not an employee.

On the appeal to the Employment Appeal Tribunal, Elias J referred to the guidance in Bottrill, including its reference to Fleming, Elias J noting that in that case Lord Coulsfield had considered that it was material that the director had drawn no salary during the last months of the company’s life and had personally guaranteed its obligations. Elias J also referred to Connolly’s case. He recognised that the authorities required the tribunal to look at the various competing factors but said that the tribunal had erred in regarding the question of control as decisive. The tribunal regarded the question of control (in this case joint control) as likely to be inconsistent with the status of an employee. He said that was inconsistent with the authorities and came close to reinstating Buchan, which Bottrill had declined to follow. Elias J said that ([2007] ICR 264, at 272):

“28.

… A majority shareholder will in practice act as the employer, making decisions on behalf of the company in which he has shares, but that does not prevent him being an employee, as [Bottrill] and subsequent cases show.

29.

Particularly material here is [Connolly] in which the Court of Appeal concluded that the employment tribunal had given the fact that the employee had a controlling shareholding ‘a significance which excludes a proper consideration of other relevant factors’: para 16. In essence I think that accurately summarises the error which the tribunal made here, notwithstanding that certain passages in the judgment would suggest otherwise.”

The result was that the appeal tribunal remitted the claim for a re-hearing by a different tribunal.

That decision can be regarded as moving in the direction of the conclusion that the importance that Fleming and Bottrill attached to the question of the employee’s control of the company was misplaced. The next decision of the Employment Appeal Tribunal, which follows the same path, is an important one: it was decisive of the employment judge’s decision in Mr Howe’s case. It is that of the Employment Appeal Tribunal in Nesbitt and another v. Secretary of State for Trade and Industry [2007] IRLR 847, Underhill J presiding. The picture was a familiar one. Mrs and Mr Nesbitt respectively owned 52.9% and 47% of the shares in the company (one other share – 0.1% -- being owned by Mrs Nesbitt’s mother, Mrs Franklyn) and all three were directors, Mrs Franklyn also having considerable business experience. Mr and Mrs Nesbitt had written contracts of employment from the start and were paid salaries commensurate with their roles as senior managers. They did not receive directors’ fees and no dividends were paid. Following the company’s insolvency, the employment tribunal rejected their claims to be employees.

The appeal tribunal allowed their appeals and held that the employment tribunal should have found them to be employees. Underhill J opened his judgment by saying that the guidance in Bottrill had continued to give rise to difficulty. He said (paragraph 3) that on the face of it Mr and Mrs Nesbitt plainly appeared to have been employees. They had contracts of employment and their relationship with the company had been conducted on a basis consistent with an employment relationship. There was no question of the contracts being a sham. Lee’s case showed that there was no reason why a controlling shareholder cannot be an employee of his company. Underhill J referred to Buchan, Fleming and Bottrill. He cited the same passage in Morison J’s judgment in Bottrill (drawing on Lee’s case) that we have cited, one suggesting that the relevance of the putative employee’s control of the company is that it might go to show either that the company was “a mere simulacrum” or that the employment contract was a sham.

At paragraphs 15 and 16 of his judgment, after citing this court’s guidance in Bottrill, Underhill J referred to the final paragraph of the guidance (that which arises once the tribunal has found that the purported contract is not a sham). He described it as “not entirely straightforward”. He referred to the statement that the question of control will “always [be] important” and asked why. He pointed out that on one view of its guidance the court’s “retreat from [Buchan] is quite limited.” He felt difficulty in interpreting what the guidance amounted to. He moved on to Connolly, and regarded it as indicating that the fact that the putative employer has total control of a company, and is at pains to retain it, is not sufficient to show he is not an employee if he otherwise satisfies the criteria for employment status. He referred to Gladwell, from which he derived the view of Elias J that “it is not the law that the fact that a majority shareholder ‘will in practice act as the employer’ prevents him from being an employee.”

Underhill J’s conclusion was that, by its guidance in Bottrill (in line with the guidance in Fleming), this court attached to the putative employee’s control of the company an importance that was difficult to reconcile with the reasoning for its decision in Bottrill – which accepted that Lee rather than Buchan provided the answer to the case. But if Lee was right, why did the guidance say that – questions of sham apart – control will always be important? Bottrill does not answer that question. Underhill J’s view was as follows:

“27.

For reasons which will appear, I do not ultimately need to decide which approach is correct. I should however say that I believe that the law is that the fact that a claimant under the employment protection legislation is a majority shareholder and a director of the company which employs him does not affect his status as an employee unless the tribunal finds that the company is a ‘mere simulacrum’ in the sense discussed at paragraph 12 (and thus, by the same token, that the contract between it and the putative employee is a sham). That seems to me the right approach in principle.”

Underhill J’s “mere simulacrum” reference derives from Lord Morris in Lee and was picked up by Morison J in the passage in his judgment in Bottrill that we have cited and discussed. Underhill J interpreted “mere simulacrum” cases as confined to those:

“12.

… where it appears that there is no real intention to vest the business in the company in question or, therefore, to distinguish between the two roles of director and employee – in which case, as Morison P’s formulation acknowledges, the contract of employment would necessarily also be a sham.”

We have earlier expressed our own view on what Lord Morris probably had in mind. But we regard it as unnecessary to extend this judgment by discussing this particular point further. A preliminary question, if it arises, is whether there is a genuine contract between a company and its controlling shareholder. That will involve a consideration of all the facts relevant to such an inquiry, the outcome of which will be a finding of sham or no sham. If the outcome is a finding of no sham, the next question is whether the contract that has been proved amounts to one of employment. On these issues we agree with Underhill J’s approach. We should perhaps also say that we do not interpret Underhill J as suggesting that, once the sham hurdle has been overcome, the claimed contract will automatically be regarded as one of employment. The court or tribunal will also have to be satisfied as to that, as we have no doubt Underhill J fully recognised.

The only further matter we need to add about Underhill J’s judgment is that he disposed of the appeal by allowing it. His reason for doing so was that, although “all the indicia were that both Mr and Mrs Nesbitt were employees,” the only reason why the employment tribunal had found against the existence of an employment relationship was because of the putative employees’ control of the company. That was reasoning that this court in Connolly and Elias J in Gladwell had held to be insufficient.

We come finally to the decision of the Employment Appeal Tribunal in Clark v. Clark Construction Initiatives Ltd and another [2008] IRLR 364. This was not a claim against the Secretary of State but, in circumstances we need not explain, the case raised the like question of whether the controlling shareholder of a company was also its employee. The employment tribunal approached the resolution of that question by purporting to apply the guidance of (inter alia) Bottrill. They said:

“[That guidance] would seem to establish that what we have to do is to look at the whole picture which we have done. We should balance out all the factors and make a reasoned conclusion. We do that balancing exercise. It seemed to us that during the first and middle phases the claimant was in business on his own account and not employed.”

In his judgment on the appeal to the appeal tribunal, Elias J reviewed the authorities and cited the guidance given by this court in Bottrill. He referred to Connolly (as a case in which the employment tribunal “had been unduly swayed by the single feature of the controlling shareholding”), Gladwell and Nesbitt. Elias J then suggested three circumstances in which it may be legitimate for a tribunal or court not to recognise what is on its face a legitimate employment contract. First, the exceptional case in which the company is itself a sham (which we understand to mean the case in which it is regarded simply as the alter ego of the individual). Second, where the contract was entered into for an ulterior purpose, for example to secure some statutory payment from the Secretary of State. Third, where the parties do not conduct their relationship in accordance with the contract. This will either be because they never intended to and the purported contract was a sham in the sense of Diplock LJ’s description in Snook v. London & West Riding Investments [1967] 2 QB 786, at 801; or because the relationship has ceased to reflect the contractual terms. The latter type of case is the one that the Bottrill guidance had in mind in twice emphasising the potential relevance of whether the conduct of the parties is consistent with the contract.

We respectfully agree with Elias J’s summary of the types of case in which the court or tribunal may find on the facts that the purported contract is not a genuine contract. But, as we have said, that type of issue does not arise under either appeal before us, we received no argument on it and we were not invited to attempt to provide general guidance on it. We propose, therefore, to say no more about Elias J’s suggested categories save two things. First, we would not wish to be taken as saying that there may never be other factual circumstances in which a conclusion of sham might be made. Second, an investigation of how the parties have conducted themselves under the purported contract may prove different things. We explain what we here have in mind in our summary of the relevant principles that we set out below.

Having dealt with that aspect of the Bottrill guidance, Elias J then said this:

“98.

How should a tribunal approach the task of determining whether the contract of employment should be given effect or not? We would suggest that a consideration of the following factors, whilst not exhaustive, may be of assistance:

(1)

Where there is a contract ostensibly in place, the onus is on the party seeking to deny its effect to satisfy the court that it is not what it appears to be. This is particularly so where the individual has paid tax and national insurance as an employee: he has on the face of it earned the right to take advantage of the benefits which employees may derive from such payments.

(2)

The mere fact that the individual has a controlling shareholding does not of itself prevent a contract of employment arising. Nor does the fact that he is practice able to exercise real or sole control over what the company does (Lee).

(3)

Similarly, the fact that he is an entrepreneur, or has built the company up, or will profit from its success, will not be factors militating against a finding that there is a contract in place. Indeed, any controlling shareholder will inevitably benefit from the company’s success, as will many employees with share option schemes ([Connolly]).

(4)

If the conduct of the parties is in accordance with the contract that would be a strong pointer towards the contract being valid and binding. For example, this would be so if the individual works the hours stipulated or does not take more than the stipulated holidays.

(5)

Conversely, if the conduct of the parties is either inconsistent with the contract (in the sense described in para. 96) or in certain key areas where one might expect it to be governed by the contract is in fact not so governed, that would be a factor, and potentially a very important one, militating against a finding that the controlling shareholder is in reality an employee.

(6)

In that context, the assertion that there is a genuine contract will be undermined if the terms have not been identified or reduced into writing (Fleming). This will be powerful evidence that the contract was not really intended to regulate the relationship in any way.

(7)

The fact that the individual takes loans from the company or guarantees its debts could exceptionally have some relevance in analysing the true nature of the relationship, but in most cases such factors are unlikely to carry any weight. There is nothing intrinsically inconsistent in a person who is an employee doing these things. Indeed, in many small companies it will be necessary for the controlling shareholder personally to give bank guarantees precisely because the company assets are small and no funding will be forthcoming without them. It would wholly undermine the Lee approach if this were to be sufficient to deny the controlling shareholder the right to enter into a contract of employment.

(8)

Although the courts have said that the fact of there being a controlling shareholding is always relevant and may be decisive, that does not mean that that fact alone will ever justify a tribunal in finding that there was no contract in place. That would be to apply the Buchan test which has been decisively rejected. The fact that there is a controlling shareholding is what may raise doubts as to whether that individual is truly an employee, but of itself that fact alone does not resolve these doubts one way or another.”

The appeal tribunal held that the employment tribunal’s conclusion that the claimant was not an employee was well open to it on the facts.

The thrust of Mr Tolley’s submissions for the Secretary of State was that we should express our disagreement with the line taken by the Employment Appeal Tribunal in Nesbitt and Clark and re-endorse the guidance given by the Court of Session in Fleming and by this court in Bottrill. In essence, he has urged us to regard Fleming and Bottrill as still endorsing an approach whereby a genuine contract which, on a consideration of all the circumstances, appears to be a contract of employment, should nevertheless be rejected as amounting to such a contract on the ground that the “employee” is himself a director/controlling shareholder of the company and, as such a director/controlling shareholder, acts (e.g. by giving guarantees to a bank) as any such person may act. We have indicated above how the guidance in Fleming and Bottrill has indeed led to difficulties so that the time has come for fresh guidance. We do not accept Mr Tolley’s submissions. We will seek to draw the threads together in a summary way.

There is no reason in principle why someone who is a shareholder and director of a company cannot also be an employee of the company under a contract of employment. There is also no reason in principle why someone whose shareholding in the company gives him control of it – even total control (as in Lee’s case) – cannot be an employee. In short, a person whose economic interest in a company and its business means that he is in practice properly to be regarded as their “owner” can also be an employee of the company. It will, in particular, be no answer to his claim to be such an employee to argue that: (i) the extent of his control of the company means that the control condition of a contract of employment cannot be satisfied; or (ii) that the practical control he has over his own destiny – including that he cannot be dismissed from his employment except with his consent – has the effect in law that he cannot be an employee at all. Point (i) is answered by Lee’s case, which decided that the relevant control is in the company; point (ii) is answered by this court’s rejection in Bottrill of the reasoning in Buchan.

Whether or not such a shareholder/director is an employee of the company is a question of fact for the court or tribunal before which such issue arises. In any such case there may in theory be two such issues, although in practice the evidence relevant to their resolution will be likely to overlap. The first, and logically preliminary one, will be whether the putative contract is a genuine contract or a sham. The second will be whether, assuming it is a genuine contract, it amounts to a contract of employment (it might, for example, instead amount to a contract for services). We make clear that we are not of course suggesting that cases raising the first issue are likely to be common, and we think it probable that they will be relatively exceptional. Despite the repeated references in the authorities to the theoretical possibility of a contract being a sham, no such case has been discovered in the principal authorities to which we have been referred. We make no attempt to give any prescriptive guidance as to the resolution of such issues, but we at least offer the following general observations.

In cases involving an alleged sham, there will, as we have said, almost invariably be what purports to be a formal written employment contract, or at least a board minute or a memorandum purporting to record or evidence the creation of such a contract. The task of the court or tribunal will be to decide whether any such document amounts to a sham in the sense of the Snook guidance (and see also Protectacoat, to which we referred in paragraph [37]). Any such inquiry will usually require not just an investigation into the circumstances of the creation of the document but also into the parties’ purported conduct under it, which will be likely to shed light on the genuineness or otherwise of the claimed contract. The fact that the putative employee has control over the company and the board, and so was instrumental in the creation of the very contract that he is asserting, will obviously be a relevant matter in the court’s consideration of whether the contract is or is not a sham. It will usually be the feature that prompted the inquiry in the first place.

An inquiry into what the parties have done under the purported contract may show a variety of things: (i) that they did not act in accordance with the purported contract at all, which would support the conclusion that it was a sham; or (ii) that they did act in accordance with it, which will support the opposite conclusion; or (iii) that although they acted in a way consistent with a genuine service contract arrangement, what they have done suggests the making of a variation of the terms of the original purported contract; or (iv) that there came a point when the parties ceased to conduct themselves in a way consistent with the purported contract or any variation of it, which may invite the conclusion that, although the contract was originally a genuine one, it has been impliedly discharged. There may obviously also be different outcomes of any investigation into how the parties have conducted themselves under the purported contract. It will be a question of fact as to what conclusions are to be drawn from such investigation.

In a case in which no allegation of sham is raised, or in which the claimant proves that no question of sham arises, the question (or further question) for the court or tribunal will be whether the claimed contract amounts to a true contract of employment. As we have indicated, given that the critical question in cases such as those under appeal is as to whether the putative employee was an employee at the time of the company’s insolvency, it will or may be necessary to inquire into what has been done under the claimed contract: there will or may therefore need to be the like inquiry as in cases in which an allegation of sham is made. In order for the employee to make good his case, it may well be insufficient merely to place reliance on a written contract made, say, five years earlier. The tribunal will want to know that the claimed contract, perhaps as subsequently varied, was still in place at the time of the insolvency. In a case in which the alleged contract is not in writing, or is only in brief form, it is obvious that it will usually be necessary to inquire into how the parties have conducted themselves under it.

In deciding whether a valid contract of employment was in existence, consideration will have to be given to the requisite conditions for the creation of such a contract and the court or tribunal will want to be satisfied that the contract meets them. In Lee’s case the position was ostensibly clear on the documents, with the only contentious issue being in relation to the control condition of a contract of employment. In some cases there will be a formal service agreement. Failing that, there may be a minute of a board meeting or a memorandum dealing with the matter. But in many cases involving small companies, with their control being in the hands of perhaps just one or two director/shareholders, the handling of such matters may have been dealt with informally and it may be a difficult question as to whether or not the correct inference from the facts is that the putative employee was, as claimed, truly an employee. In particular, a director of a company is the holder of an office and will not, merely by virtue of such office, be an employee: the putative employee will have to prove more than his appointment as a director. It will be relevant to consider how he has been paid. Has he been paid a salary, which points towards employment? Or merely by way of director’s fees, which points away from it? In considering what the putative employee was actually doing, it will also be relevant to consider whether he was acting merely in his capacity as a director of the company; or whether he was acting as an employee.

We have referred in the previous paragraph to matters which will typically be directly relevant to the inquiry whether or not (there being no question of a sham) the claimed contract amounts to a contract of employment. What we have not included as a relevant consideration for the purposes of that inquiry is the fact that the putative employee’s shareholding in the company gave him control of the company, even total control. The fact of his control will obviously form a part of the backdrop against which the assessment will be made of what has been done under the putative written or oral employment contract that is being asserted. But it will not ordinarily be of any special relevance in deciding whether or not he has a valid such contract. Nor will the fact that he will have share capital invested in the company; or that he may have made loans to it; or that he has personally guaranteed its obligations; or that his personal investment in the company will stand to prosper in line with the company’s prosperity; or that he has done any of the other things that the “owner” of a business will commonly do on its behalf. These considerations are usual features of the sort of companies giving rise to the type of issue with which these appeals are concerned but they will ordinarily be irrelevant to whether or not a valid contract of employment has been created and so they can and should be ignored. They show an “owner” acting qua “owner”, which is inevitable in such a company. However, they do not show that the “owner” cannot also be an employee.

We have, however, twice -- and deliberately -- used the word “ordinarily” in the last paragraph. We have used the word not because we foresee other circumstances but because “never say never” is a wise judicial maxim.

We respectfully agree with the essence of the factors referred to by Elias J in paragraph 98 of his judgment although we add a comment on four of them. Mr Tolley criticised his first factor as amounting to a suggestion that the mere production of a written contract purporting to be a contract of employment will shift to the opposing party the burden of proving that it was not a genuine such contract. We doubt if Elias J was intending to refer to a legal burden. In cases where the putative employee is asserting the existence of an employment contract, it will be for him to prove it; and, as we have indicated, the mere production of what purports to be a written service agreement may by itself be insufficient to prove the case sought to be made. If the putative employee’s assertion is challenged the court or tribunal will need to be satisfied that the document is a true reflection of the claimed employment relationship, for which purpose it will be relevant to know what the parties have done under it. The putative employee may, therefore, have to do rather more than simply produce the contract itself, or else a board minute or memorandum purporting to record his employment.

We consider that Elias J’s sixth factor may perhaps have put a little too high the potentially negative effect of the terms of the contract not having been reduced into writing. This will obviously be an important consideration but if the parties’ conduct under the claimed contract points convincingly to the conclusion that there was a true contract of employment, we would not wish tribunals to seize too readily on the absence of a written agreement as justifying the rejection of the claim. In both cases under appeal there was no written service agreement, but the employment judges appear to have had no doubt that the parties’ conduct proved a genuine employment relationship.

As for Elias J’s seventh and eighth factors, we say no more than that we regard them as saying essentially what we have said above in our “never say never” paragraph.

We can deal with these shortly. It can fairly be said in the employment judge’s defence in Mr Neufeld’s case that he did his best to apply the Bottrill guidance but the end result was that he allowed his apparently clear finding that Mr Neufeld had a genuine employment contract to be trumped by the irrelevant consideration that he had given personal guarantees for the company, lent money to it and had a controlling shareholding in it. That consideration led him into legal error and in our view the appeal tribunal was right to reverse his decision and hold that Mr Neufeld was an employee. Once the irrelevant considerations were stripped away, there was no alternative conclusion but that Mr Neufeld was an employee. We might add that, apart from the wholly irrelevant considerations, the one factor that the employment judge regarded as pointing away from there being an employment contract was that Mr Neufeld did not take his full holiday entitlement. We would not, with respect, have regarded that as such a pointer. It might have been different if he had taken more than his entitlement: that would perhaps have raised a question as to whether the contract was a genuine one.

In Mr Howe’s case, the Bottrill guidance led the employment judge to the brink of the same error that was made by the employment judge in Mr Neufeld’s case but he was saved from it by Underhill J’s guidance in Nesbitt. His decision was correct on the basis of the facts that he had found and Bean J in the appeal tribunal was right to dismiss the Secretary of State’s appeal.

We comment that it was part of Mr Tolley’s submissions in his skeleton argument that we could safely endorse the Fleming and Bottrill guidance, including that tribunals could and should have regard to the putative employee’s controlling shareholding, because tribunals “can be trusted to use their experience and expertise to determine these issues in a practical and common sense manner.” The evidence of the two cases under appeal gives us no confidence in that regard. The Bottrill guidance led to what was, with respect, an irrational decision in Mr Neufeld’s case; and it nearly did the same in Mr Howe’s case. There was plainly a need for clarification of the guiding principles. We hope that our judgment may go some way towards providing it.

We dismiss both appeals.