England and Wales Court of Appeal (Civil Division) Decisions

You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> Procter & Anor v Procter & Ors [2021] EWCA Civ 167 (12 February 2021)
URL: http://www.bailii.org/ew/cases/EWCA/Civ/2021/167.html
Cite as: [2022] STC 157, [2021] EWCA Civ 167, [2022] STI 39, [2022] 1 WLR 2277, [2021] BTC 34, [2021] Ch 395, [2021] EWCA Civ 1804, [2021] 2 WLR 1249, [2022] 1 All ER 358, [2021] WLR(D) 97

[New search] [Printable PDF version] [Buy ICLR report: [2021] 2 WLR 1249] [Buy ICLR report: [2022] 1 WLR 2277] [Buy ICLR report: [2021] Ch 395] [View ICLR summary: [2021] WLR(D) 97] [Help]

		Neutral Citation Number: [2021] EWCA Civ 167
		Case No: A3/2019/2493

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
BUSINESS AND PROPERTY COURTS IN LEEDS
PROPERTY TRUST AND PROBATE LIST
His Honour Judge Davis-White QC (sitting as a judge of the High Court)
C3OLS690

Royal Courts of Justice
Strand, London, WC2A 2LL

12th February 2021

B e f o r e :

LORD JUSTICE LEWISON
LORD JUSTICE ARNOLD
and
LORD JUSTICE NUGEE
____________________

Between:

	(1) PHILIP JOHN PROCTER (2) JAMES GEOFFREY PROCTER	Appellants
	- and -
	(1) SUZANNE ELAINE PROCTER	1st Respondent
	(2) GEORGE KNOWLES	2nd Respondent
	(3) WOMBLE BOND DICKINSON (TRUST CORPORATION) LIMITED	3rd Respondent

____________________

MR EDWARD PETERS (instructed by Ebery Williams) for the Appellants
MR BRUCE WALKER (instructed by Grays Solicitors LLP) for the 1st Respondent
The 2nd and 3rd Respondents did not appear and were not represented
Hearing date : 2nd February 2021
____________________

HTML VERSION OF APPROVED JUDGMENT
____________________

Covid-19 Protocol: This judgment was handed down remotely by circulation to the parties' representatives by email, release to BAILII and publication on the Courts and Tribunals Judiciary website. The date and time for hand-down is deemed to be 10:30am on Friday 12th February 2021.

Lord Justice Lewison:

Introduction

The issue on this appeal is whether Philip and James Procter and their sister Suzanne are entitled to a tenancy protected by the Agricultural Holdings Act 1986.

In the course of deciding a multiplicity of issues in what he called a "war (possibly of attrition)" between these siblings, the trial judge (HHJ Davis-White QC) decided that no such tenancy had been created. His judgment is at [2019] EWHC 1199 (Ch).

For the purposes of this appeal the relevant facts can be considerably simplified. The land in question is a family farm in Yorkshire. It consists of about 600 acres of arable land, four farmhouses, various agricultural buildings and a golf course. The golf course extends over some 128 acres of the land. Some former agricultural cottages are let on assured shorthold tenancies to outsiders. The freehold of most of the land was held by the trustees of a will trust. By 1994 the trustees were the parents of the siblings, Geoffrey Procter and Jean Procter, together with Philip Procter. Geoffrey and Jean Procter have since died. Since 2014 Philip Procter has been the sole trustee and freeholder. Although the legal estate has been in the trustees throughout the relevant events, the beneficial interests have been variously held by different trusts.

The farming itself was carried on by a partnership. The partnership began in April 1979. The original partners were Geoffrey Procter, Jean Procter, Philip Procter and Suzanne Procter. In 1994 the partners were Geoffrey Procter, Jean Procter, Philip Procter, James Procter and Suzanne Procter. There was thus a partial, though not a complete, overlap between the identity of the freeholders and the identity of the partners. The partnership itself was regulated by a deed of partnership dated 1 October 1980. At the date of the trial, the only remaining partners were Philip and James Procter, Suzanne Procter having resigned from the partnership in 2010.

There had been written tenancy agreements relating to some of the land farmed by the partners; but they all came to an end in or before 1994. It was alleged that the tenancy in issue on this appeal came into existence at that time. Having considered the evidence in detail, the judge concluded that, if the ordinary principles applied, the partnership would have been entitled to a tenancy created by conduct because (a) there was an intention to create legal relations; (b) rent was treated as having been paid by various accounting adjustments and (c) the partnership was in exclusive possession of the land. His detailed reasoning is in paragraphs [222] to [250] of his judgment. On the basis of the judge's findings the original tenants were Geoffrey Procter, Jean Procter, Philip Procter, James Procter and Suzanne Procter. Geoffrey and Jean Procter have since died with the consequence that, if there was a tenancy, the effect of the right of survivorship will have resulted in the tenants being Philip, James and Suzanne Procter.

Despite the judge's findings, the claim to a tenancy protected by the Agricultural Holdings Act failed because:

i) At common law, it was not possible to create a tenancy where there was an overlap between putative landlords and putative tenants.

ii) There was no lease in writing, with the consequence that the grant was not validated by section 72 of the Law of Property Act 1925.

iii) If, contrary to the judge's view, there was a tenancy, it was not a tenancy granted for the best rent reasonably obtainable; and therefore took effect as a tenancy at will under sections 53 and 54 of the Law of Property Act 1925. The judge did not find it necessary to decide whether a tenancy at will attracted the protection of the Agricultural Holdings Act.

iv) If the arrangements amounted only to a licence, then because of the overlap in identity between licensors and licensees, the latter could not have had exclusive occupation as against the former. Since that is an essential ingredient in order for a licence to attract the protection of the Agricultural Holdings Act, the claim to protection failed.

Can A, B and C grant a lease to A, B, C, D and E?

The first legal question posed by this appeal can be summarised thus. The freeholders are A, B and C. Does the law permit them to grant a tenancy to A, B, C, D and E? As Lord Tomlin memorably put it in Hillas & Co Ltd v Arcos Ltd (1932) 147 LT 503:

"… the problem for a court of construction must always be so to balance matters that, without violation of essential principle, the dealings of men may as far as possible be treated as effective, and that the law may not incur the reproach of being the destroyer of bargains."

I do not think that that general approach is confined to questions of construction. As a broad and over-simplified generalisation, the courts try to oil the wheels of commerce, rather than to throw grit into the engine.

There are two aspects of a tenancy that must be considered: both the estate in land created by the grant of a tenancy, and the contractual relation between landlord and tenant.

In the case of a written tenancy agreement, there is probably no problem. That is the result of sections 72 and 82 of the Law of Property Act 1925. Section 72 provides (so far as material):

"(1) In conveyances made after [12 August 1859], personal property, including chattels real, may be conveyed by a person to himself jointly with another person by the like means by which it might be conveyed by him to another person.

(2) In conveyances made after [31 December 1881], freehold land, or a thing in action, may be conveyed by a person to himself jointly with another person, by the like means by which it might be conveyed by him to another person; and may, in like manner, be conveyed by a husband to his wife, and by a wife to her husband, alone or jointly with another person.

(3) After the commencement of this Act a person may convey land to or vest land in himself.

(4) Two or more persons (whether or not being trustees or personal representatives) may convey, and shall be deemed always to have been capable of conveying, any property vested in them to any one or more of themselves in like manner as they could have conveyed such property to a third party; provided that if the persons in whose favour the conveyance is made are, by reason of any fiduciary relationship or otherwise, precluded from validly carrying out the transaction, the conveyance shall be liable to be set aside."

It will be noted, however, that section 72 (1) only applies to "conveyances". In Rye v Rye [1962] AC 496 the House of Lords held that a "conveyance" had to be in writing.

Section 82 provides:

"(1) Any covenant, whether express or implied, or agreement entered into by a person with himself and one or more other persons shall be construed and be capable of being enforced in like manner as if the covenant or agreement had been entered into with the other person or persons alone."

Before delving further into the authorities, it is necessary to sketch out some legal history. In the Middle Ages freehold land was transferred by feoffment and livery of seisin. A person who had seisin of land was liable for feudal dues and services. In order to evade liability for feudal dues, conveyancers invented the device of conveying to a person (by feoffment and livery of seisin) on terms that the land was to be held "to the use of" another. In response to this, the Chancellor, and subsequently the courts of equity, began to develop the concept of the separation of legal and beneficial interests in land. The legal interest was in the person who had seisin, while the beneficial interest was in the "cestui que use". The Crown took the view that this device amounted, in effect, to tax evasion. Parliament therefore passed the Statute of Uses 1535 to nullify the effect of the conveyance device. It operated by transferring seisin to the cestui que use, thus converting an equitable estate into a legal estate. One side effect of this change was that it became possible for A to transfer land to A and B. The machinery by which this was done was for A to transfer the land to X (with livery of seisin) to the use of A and B. The statute then converted the beneficial interests of A and B into legal estates. By this indirect means A could, for all practical purposes, convey land to A and B.

The Statute of Uses only applied where a person was seised of land. Seisin was a term limited to freeholds (including leases for lives). It did not apply to leaseholds which were granted for fixed terms or to periodic tenancies. They were, in any event, "chattels real" in the eyes of the common law. To remedy this perceived defect, section 21 of the Law of Property Amendment Act 1859 permitted the assignment of personal property (including chattels real) by A to A and B. It did not in terms apply to the creation (as opposed to the assignment) of leases. Section 21 is now found in section 72 (1) of the Law of Property Act 1925.

The next reform came with section 50 of the Conveyancing Act 1881, the purpose of which was to simplify and improve the practice of conveyancing. Section 50 permitted the conveyance of freehold land by A to A and B (as well as conveyances between spouses). Section 50 is now to be found in section 72 (2) of the 1925 Act. Accordingly, what had formerly been possible indirectly by virtue of the Statute of Uses now became possible to achieve by direct conveyance. Nevertheless, the Statute of Uses remained in force until it was ultimately repealed by section 207 of and Schedule 7 to the Law of Property Act 1925.

There are two other reforms that I need to mention which are necessary in order to understand some of the early authorities. At common law if a person held two estates in the same land, the inferior estate would be merged in the higher estate. As Sir William Blackstone put it (2 Blackstone Commentaries p. 177):

"Whenever a greater estate and a less coincide and meet in one and the same person, without any intermediate estate, the less is immediately annihilated; or in the law phrase, is said to be merged, that is, sunk or drowned, in the greater."

The strict application of this principle sometimes produced surprising results. Thus where a tenant held under a lease for 20 years, and acquired a lease of the reversion granted for a term of one year, the 20 year term merged in the one year term. The common law was to some extent mitigated where a person held two interests held in separate rights: Jones v Davis (1860) 5 H & N 766. In that event, lease and reversion could exist together. As Lord Kenyon CJ put it in Webb v Russell (1789) 3 Durn & E 393:

"… nothing is clearer than that a term which is taken in alieno jure is not merged in a reversion acquired suo jure."

Sir William Blackstone makes the same point at 177:

"But they must come to one and the same person in one and the same right; else if the freehold be in his own right, and he has a term in right of another (en auter droit) there is no merger."

Courts of equity, on the other hand, took a different view. As Cozens-Hardy LJ explained in Capital and County Bank Ltd v Rhodes [1903] 1 Ch 631, 652:

"… there was prior to the Judicature Act, 1873, a great difference between the Courts of Law and the Courts of Equity on the subject of merger. The rule of the former was rigid, that whenever a term of years and a freehold estate, whether for life or in fee, immediately expectant upon the term, vested in the same person in his own right, the term was merged in the freehold, whatever may have been the intention of the parties to the transaction which resulted in the union. The Courts of Equity, on the other hand, in many cases treated the interest which merged at law as being still subsisting in equity. They had regard to the intention of the parties, and, in the absence of any direct evidence of intention, they presumed that merger was not intended, if it was to the interest of the party, or only consistent with the duty of the party, that merger should not take place."

We were shown illustrations of this principle in Chambers v Kingham (1878) 10 Ch D 743. As a result of section 25 (11) of the Judicature Act 1873 and section 185 of the Law of Property Act 1925, the equitable rules now prevail.

The second reform relates to co-ownership. Before 1926 land could be held at common law by co-owners either as joint tenants or as tenants in common. (In this context the phrases joint tenants and tenants in common are not limited to leasehold land, but simply refer to tenure). The essence of a joint tenancy is that each joint tenant holds the whole of the legal interest jointly and holds nothing separately. By contrast in the case of a tenancy in common, the co-owners hold the title in undivided shares. The shares are said to be undivided because the land itself has not been divided. But each tenant in common has his own separate and distinct interest which he can assign either to a third party or to his co-owner. Before 1926 a tenancy in common could exist either at law or in equity (or both). In addition, by severing a joint tenancy at law one of the joint tenants could turn it into a tenancy in common at law. But section 34 (1) of the Law of Property Act 1925 prohibited the creation of a tenancy in common at law; and 36 (2) prohibited the severance of a joint tenancy at law. From then on, the only permitted species of co-ownership of the legal title was as joint tenants. Any tenancy in common could arise only in equity.

Rye v Rye

Rye v Rye was the principal case cited to the judge on the first legal question; and it is worth considering it a little further. Two brothers were the freehold owners of property in Golden Square where they also carried on business as partners. In 1942 they agreed that a tenancy should be granted to the firm at a rent to be paid out of the partnership assets. The agreement was an oral one. The case was argued entirely on the basis of the Law of Property Act 1925. It was common ground that before the 1925 Act came into force it was not possible for the two brothers, as freeholders, to grant a tenancy to themselves as partners: Viscount Simonds at 504; Lord MacDermott at 507. Despite the fact that there was no recorded argument on the point and no citation of early authorities, their Lordships did discuss the position at common law. It is, perhaps, also worth pointing out that counsel for the successful respondent appears to have accepted that if a person was acting in two capacities (e.g. as beneficial owner of land and as charity trustee) he could grant a lease to himself: p 501. So that point was not in issue.

Viscount Simonds said at 505 that it was not suggested that A and B could do what A himself could not do. He therefore proceeded to consider whether A could grant a lease to himself.

He came to the conclusion that A could not grant a lease to himself for several reasons. He first said that he found the idea of a lease granted by A (as freeholder) to himself as a "strange conception", citing authority that described it as fanciful and whimsical. I take that to mean that it would be pointless. One further objection was that a lease is in one aspect contractual, and a person cannot contract with himself. So no covenants could be attached to such a lease. Another reason was that there had to be an acceptance of the thing demised and a man could not accept from himself something that he already owned. A fourth reason was that the lease once granted would immediately merge in the higher estate out of which it had been created. A fifth reason was that A could not distrain on his own goods. It is to be noted that all Viscount Simonds' reasons were objections only to the grant of a lease by A to A. They applied to a lease from A and B to A and B only because it was not suggested that there was any difference between the two; and he certainly did not consider what would happen in the case of a lease from A and B to A, B and C.

Lord MacDermott said at 507:

"At common law a person could not convey real or personal estate either to himself or to himself and another."

He did not explain why he took that view, and cited no authority in support of it.

Lord Radcliffe agreed that the appeal should fail, but his reasons differed from those of the other Law Lords. He began at 510 by saying that in his view section 72 (3) of the 1925 Act did allow a person to grant a lease to himself in the sense that a term of years is not incapable of being created by such a transaction. He went on to say that, although the 1925 Act was intended to change the law, the intended change was "merely a technical one bearing on the necessary form" of written instruments and had "nothing to say one way or the other about the substantial validity of a transaction which is absurd in itself". At 511 he pointed out that before the 1925 Act a man was able to convey to himself through the indirect means of the Statute of Uses (as I have already explained). Under the Statute of Uses Lord Radcliffe was of the view that a person acting in one capacity could convey to himself in another capacity.

Having decided that section 72 (3) did not apply, Lord Radcliffe said at 512:

"Down to the Act of 1925 I take it that it would have been said without qualification that a man cannot make himself his own tenant. The contractual relationship which was the almost inescapable concomitant of a tenancy would have been regarded as precluding such a transaction. Terms of years for securing such things as jointures or portions could, I suppose, be created without covenants, but they were exceptional devices. There is nothing in the Act itself that removes this difficulty by making it possible for a man to enforce contractual obligations against himself. I do not feel sure that the same result would necessarily be reached in the case of two persons seeking to demise to themselves by deed, for section 72 (3) would, I think, be able to pass a legal interest by demise and it might be possible to express the required contractual obligations in the form of joint and several covenants, so that each single person covenanted separately with himself and the other. It seems that section 82 (1) of the Act would then convert such a covenant into an effective obligation. I should not like to put this possibility out of court in the disposal of the present case, for there is a practical advantage in allowing persons who own land as tenants in common to make a valid demise of it to themselves in another capacity."

The stumbling block for Lord Radcliffe seems therefore to have been the problem that section 82 could not validate the obligations contained in the putative tenancy. But he left open the possibility of a lease granted by co-owners in one capacity to themselves in another capacity; and, a fortiori, a grant of a lease by A and B to A, B and C.

Finally, Lord Denning said at 513:

"Is it possible for a person to grant a tenancy to himself? or for two persons to grant a tenancy to themselves? At common law it was clearly impossible. Nemo potest esse tenens et dominus. A person cannot be, at the same time, both landlord and tenant of the same premises: for as soon as the tenancy and the reversion are in the same hands the tenancy is merged, that is, sunk or drowned, in the reversion; see Blackstone's Commentaries (1766 edition), vol. II, p 177. Neither could a person at common law covenant with himself, nor could two persons with themselves. Neither could one person covenant with himself and others jointly. Such a covenant, said Pollock C.B., is "senseless," see Faulkner v Lowe."

Lord Denning continued at 514:

"But what is the position when a person grants a tenancy by writing to himself (A lets to A), or two persons grant it to themselves (A and B let to A and B): and there are the usual express or implied covenants (A covenants with A: or, A and B jointly covenant with A and B jointly)? Such a tenancy does not come within section 72 (4): nor do the covenants come within section 82 (1): because both those subsections, as I read them, only apply where one of the persons, at any rate, is not on both sides. But in the tenancy we are now considering, the persons are the same on both sides. Such a tenancy, if it is to be valid at all, must be validated by section 72 (3): but the trouble is that there is nothing there to validate the covenants: because by no possibility can section 82 (1) be made to cover them." (Emphasis added)

He considered that it was not possible to split up a single transaction into the grant of a tenancy on the one hand and the entry into covenants on the other. As he put it at 514:

"The tenancy must stand or fall with the agreement on which it is founded and with the covenants contained in it: and as they fall, so does the tenancy."

Once again, therefore, it was the lack of enforceable contractual obligations that was the real problem. I should note at this point that Rye v Rye does not suggest that section 82 is confined to written agreements.

Lord Denning raised a further objection relating to notice to quit. He said at 514:

"And what about notice to quit? If A grants a tenancy to himself A, can he mutter a notice to quit to himself and expect the law to take any notice of it? Or, if A and B grant a yearly tenancy to themselves A and B, can there be a notice to quit unless both agree? Of course not. So that, instead of a yearly tenancy, it becomes a life-long tenancy determinable only by the agreement of both. Which is absurd."

Rye v Rye was a case in which there was complete identity between landlords and tenants. Only Lord MacDermott said that at common law A could not grant a lease to A and B. Clearly, that was an obiter comment and its correctness has been challenged on this appeal. Lords Radcliffe and Denning seem to have contemplated that the position might be different if there is not complete identity between landlords and tenants; or if the lease is from co-owners in one capacity to themselves in a different capacity.

The earlier common law

Having looked at Rye v Rye, where the common law was not argued, it is now necessary to examine more closely what the common law was. From the perspective of the common lawyers, there were two potential problems that had to be confronted. The first was the proprietary problem that, as I have explained, the common law took the view that where two estates coincided in the same hands and in the same right, the lower estate merged in the higher.

The principles of merger primarily apply to interests in land which have already been created, and subsequently come into the same hands in the same right. The fact that the equitable rules about merger now prevail is not necessarily an answer to the question whether the inferior interest can be created by carving it out of the superior one. Again, I look at this in the first place from the perspective of proprietary rights.

This question was addressed by Millett LJ in Ingram v IRC [1974] 4 All ER 395. Although it was a dissenting judgment, his view was subsequently upheld by the House of Lords. The case concerned a tax planning scheme. In simple terms, the taxpayer transferred the freehold in property to her solicitor as her nominee. At the taxpayer's direction, the nominee then granted a lease of the property back to her. Once that lease had been granted, the nominee (again at the taxpayer's direction) transferred the freehold to the ultimate intended beneficiaries. One of the questions that arose was whether it was possible for A to grant a lease to B where B was A's nominee (or vice versa). Millett LJ addressed this question in detail at 419 to 427. The whole of his reasoning repays careful study. But for the purposes of this judgment I pick out a few parts. He quoted that part of Lord Denning's speech in Rye, which I have already quoted, and continued at 420:

"These two reasons correspond to the dual character of a lease in English law as both contract and conveyance. A man cannot convey to himself; and he cannot contract with himself. But he can convey to a nominee for himself, and if he can contract with a nominee for himself there is no reason why he should not be able to grant a lease to a nominee for himself."

Having discussed tenancies by attornment, he pointed out that as a property right, a lease could arise independently of contract. He then said at 422:

"In so far as a lease is a conveyance, that is to say in so far as it lies in grant, there is no difficulty in the proposition that a man can vest a term of years in a nominee for himself. There is no question of the same person being at the same time both landlord and tenant at law, for the two legal estates are vested in different persons; while the rule that a man cannot be both landlord and tenant does not apply in equity, which allows the question of merger to be governed by intention." (emphasis added)

Millett LJ then examined the lease as contract. He held that there had to be at least two parties to a contract ("the two-party rule"), with the consequence that a person could not contract with himself. But he held that a trustee could contract with a beneficiary, even if he held the benefit of the contract on trust for that beneficiary. The contract would be valid in law, even though its enforcement might be subject to equitable defences or procedural objections (such as circularity of action). The upshot was that, in his view, the lease was a valid lease. In the House of Lords the case was decided on a different point, but Lord Hoffmann expressly approved the dissenting judgment of Millett LJ on this point: [2000] 1 AC 293, 305.

Thus far there would seem to be no conceptual problems from a proprietary perspective in a person being simultaneously entitled to different interests in the same land if the common law or equitable merger rules permitted this. The consequence of applying the equitable rules would carry the consequence that a person could be entitled to exclusive possession as against himself. Nor would there seem to be any conceptual problems in a contract provided that the alleged contract did not infringe the two-party rule.

I must now consider some of the older cases in which there was not a complete overlap between those alleged to be landlords and those alleged to be tenants. The earliest of the cases that we were shown is Doe d Colnaghi v Bluck (1838) 8 C & P 464. Mr Colnaghi was the lessee of a house in Cockspur Street. He was adjudicated bankrupt; but he continued to trade at the direction of the trustee in bankruptcy. Following his discharge from bankruptcy he went into partnership with Mr Bluck. The books of the partnership showed that rent had been paid by the firm to Mr Colnaghi. Following the dissolution of the partnership the issue was whether Mr Colnaghi could recover possession from Mr Bluck. If it had been legally impossible for Mr Colnaghi to have granted a tenancy to himself and Mr Bluck, that would have been the short answer. Moreover, the report does not mention the grant of a tenancy by deed. But Tindal CJ did not decide the case on points of law. He left it to the jury. If there had been a deed of grant it is difficult to see what issue would have been left to the jury. At 468 he directed the jury that "if Mr Bluck, as one of the partners in the firm of Colnaghi & Bluck, became tenant to Mr Colnaghi", he was bound to give up possession. He left it to the jury to decide whether there had been a letting; and they decided that there had been. Tindal CJ clearly considered it to be legally possible that Mr Colnaghi had granted a tenancy to the firm of which he was one partner by informal means.

Rowley v Adams (1844) 7 Beav 548 contains only a brief mention of the question. A and B were business partners. They bought a house out of partnership assets. They entered into a new partnership consisting of A, B and C. The house was not an asset of the new partnership. A and B stipulated for a rent to be paid by the new partnership. It is not clear from the report whether the stipulation was contained in a deed. Lord Langdale MR described this arrangement at 549 as follows:

"There was, therefore, in two partners, an ownership of the property and a right to receive rent, and in the three partners a possession for which rent was paid; and although the three used the possession for the purposes of the trade of the three, including the true owners, yet I think that the right of the two owners, to whom as landlords the rent became due, must be considered as real estate."

The significance of these observations is that (a) there was no conceptual difficulty in separating A and B's rights in right of their freehold ownership and the rights of A, B and C in right of the grant of possession and (b) that there was no apparent hesitation in calling A and B "the landlords" or the payment that they received "rent". This is, to some extent, mirrored by section 205 (1) (xix) of the Law of Property Act 1925 where "possession" is defined as including receipt of rent and profits. Although the case went on appeal to the House of Lords ((1849) 11 HL Cas 725) nothing of substance was said about the point.

Rogers v Harvey (1858) 5 CBNS 5 concerned entitlement to vote. In order to be entitled to vote it was necessary (at least) to be a tenant. In that case A was the lessee of a mill. He formed a partnership with his three sons, B, C and D in which all four were partners. There was no written partnership agreement; but the partnership paid rent to A. B, C and D claimed to be tenants and hence entitled to vote. The Court of Common Pleas upheld their claim. Cockburn CJ said at 12:

"[A], being assignee of the lease of the premises in question, and entering into partnership with his three sons, demises the premises to the partnership at will in undivided shares: and, though true it is that a man cannot become tenant to himself, so that his share would merge in the rest, I see no reason why the demise should not enure for the benefit of the others; and then the three sons (the claimants) would be tenants in common, and all (the value being sufficient) entitled to be registered as tenants within the 27th section of the Reform Act."

Williams J said at 13:

"I think it is impossible to avoid coming to the conclusion that the three sons of [A] had an interest in the mill and premises in question at the lowest as tenants at will to their father."

Crowder J said at 14:

"It was clearly the intention of the parties that the three sons should be jointly and equally liable to the rent with the father: and it is found as a fact that the rent was regularly paid out of the partnership funds. There was, therefore, a tenancy of some sort; and, the value being sufficient to confer the franchise upon each of the partners under the 27th section of the Reform Act…"

Byles J said at 15:

"The result of the transaction between [A] and the claimants here is, that there was a sub-demise of the premises to the three sons as partners and tenants in common of undivided shares with their father."

Thus as I read this decision, the transaction between A, B, C and D amounted to a sub-demise of a tenancy in common at law; but A's undivided share merged in his superior leasehold interest.

In Napier v Williams [1911] 1 Ch 361 A and B and C were will trustees. They executed a lease of property to A by deed. In the lease A covenanted with A, B and C to keep the property in repair and not to assign without consent. A subsequently assigned the term to a company which issued debentures over the lease. The question was whether the trustees of the debenture trust deed were bound by the covenants. It is important to note that Warrington J said clearly that the lease "is certainly not void in law" although there was doubt about the nature of the lessee's interest. He considered that question in an obiter passage at 367-8 as follows:

"Sect. 50 of the Conveyancing Act, 1881, by virtue of the definition of conveyance applies to a lease of freehold land, but only enables such a lease to be made by one to himself jointly with another, and has no reference to a lease by two or more to one of themselves alone. What is the effect then at common law of such a demise as the present? The lessee is already seised in fee per mie et per tout, and the demise by himself can have no effect, for any term granted by himself would merge in the fee. The other two joint tenants could make an effectual demise of their two-thirds, but they would thereby sever the joint tenancy. I think that the effect must be that the joint tenancy is severed during the term, and that the lessee is entitled to the two-thirds of the land by virtue of the lease, remaining seised of his one-third for his original estate in fee. If this is the right view I can see many serious difficulties in the plaintiffs' way, but I prefer to dispose of the case on other grounds, and will, therefore, treat the matter as if there were no question arising on the form of the demise itself."

Clearly, then, he must have considered it legally possible for B and C to create a leasehold estate at least in A, even if A's purported grant to himself merged in his existing interest. Again, however, the answer seems to turn on the proposition that (at that time) it was possible to create a tenancy in common at law. He went on, however, to hold that despite the existence of the lease the covenants contained in it were unenforceable. I shall return to that in the context of the "two-party" rule. Importantly, he also held that if the covenants were not performed, then the lease could be forfeited for breach of covenant. That reinforces his conclusion, earlier in his judgment, that a valid lease (of some sort) had been created.

It is sufficient to note at this stage that the idea of a leasehold term without enforceable covenants was not a novelty in the common law. Lord Millett explained in Barrett v Morgan [2000] 2 AC 264, 271:

"Formerly the extinguishment of the tenancy by surrender also extinguished the reversion to any subtenancy, so that the remedy for the rent and the covenants attached to the reversion ceased with the reversion to which they were annexed. The subtenant held the property as tenant of the head landlord for the residue of the term of the extinguished tenancy but without privity of estate and accordingly without any obligation to pay the rent or perform the tenant's covenants: see Webb v Russell (1789) 3 Durn & E 393. This unsatisfactory state of affairs was remedied by statute in two stages."

A similar result can come about in modern law. If the liquidator of an insolvent company disclaims a lease, leaving a sub-tenant in possession, the sub-lease does not come to an end. The sub-tenant holds as if his sub-lease had continued; but the landlord cannot directly sue on the covenants contained in the sub-lease. Instead, the landlord is entitled to distrain or forfeit if the sub-tenant fails to pay the rent or is in breach of covenant: see Hindcastle Ltd v Barbara Attenborough Associates Ltd [1997] AC 70, 89.

The upshot of these cases seems to me to be that, looking at the question from a proprietary angle, there was no insuperable bar, at common law, to the creation of a tenancy where there was a partial overlap between landlords and tenants (though not complete coincidence). But that depended, at least in part, upon the ability to create tenancies in common at law. Since the Law of Property Act 1925 that has no longer been possible. The only form of co-ownership that can exist at law is the joint tenancy.

Since the only species of co-ownership that can now exist in England and Wales is a joint tenancy, the problem of merger no longer applies. A joint tenancy of a leasehold interest held jointly by A, B and C will not merge in a freehold interest held jointly by A and B. The reason for that is that since the joint tenancy in each interest is incapable of severance at law, there is not the necessary coincidence of interest in the two interests in land. Accordingly, the common law does not mandate a merger; and in equity merger depends on intention. Where, as in our case, a tenancy is granted by one partner to a partnership for the benefit of the partnership business, the mutual duty of good faith would, in my judgment, in any event preclude merger.

In Lie v Mohile [2014] EWCA Civ 728, [2014] 2 P & CR 13 at [16] Patten LJ said (obiter) that:

"the common law rule [was] that a deed by A in favour of A plus B had the effect of vesting the legal estate in B alone…"

Mr Walker argued that in so describing the common law rule, Patten LJ confined it to deeds. But that would be inconsistent with (at least) the decision of the Court of Common Pleas in Rogers v Harvey and Tindal CJ's direction to the jury in Doe d Colnaghi v Bluck. No doubt Patten LJ expressed himself in that way because he was pointing the contrast with section 72 of the Law of Property Act 1925 which, as the result of Rye v Rye, does only apply to written instruments.

Accordingly, although the common law set its face against the grant of a tenancy by A and B to A and B, I do not consider that there was an absolute rule that prohibited the grant of a tenancy by A to A and B; or by A and B to A, B and C.

The two-party rule

Subject to section 82 of the Law of Property Act 1925, the "two-party rule" is well established. Grey v Ellison (1856) 1 Giff 438 was a case in which the question was whether an insurance company was entitled to deduct the cost of an insurance premium from an annuity. The insurance company had not in fact taken out a policy; but had carried the risk itself. The annuity and the underwriting of risks were carried out by two different departments within the company. In those circumstances Stuart V-C held that it was not entitled to deduct anything. He went on to say at 444:

"If a man were so fanciful as to grant a lease to himself of his own house, with a covenant that he should quietly enjoy, and a covenant that he should pay to himself a rent for his own house, and chooses to conduct it in the way of having two departments, that is, that he will draw cheques upon himself upon his own account for rent, and pay them into another account of his own at his bankers—it would be a mere whimsical transaction; but it would be futile and an abuse of language to say that it came within the law of contract, or within any fiscal regulation respecting stamps."

This is a clear illustration of the "two-party rule". I do not consider that it goes further than that.

Nevertheless, so far as the contractual aspect of a tenancy is concerned, the common law did go further than the "two-party" rule. The common law regarded it as impossible for A to contract with himself and others. That is illustrated by Faulkner v Lowe (1848) 2 Ex 595. That case was a claim in covenant. It was thus a purely contractual claim. A, B and C maintained a joint account. A asked B and C to lend him money; and covenanted with B and C that he would repay the loan with interest to A, B and C or their survivors. The Court of Exchequer held that no action lay on the covenant. The judge pointed out that if B and C died, the effect of the covenant would be that A was indebted to himself. Pollock CB is recorded as having said that the covenant was "senseless" and that he did not know "what is meant, in point of law, by a man paying himself".

In Ellis v Kerr [1910] 1 Ch 529 A assigned an insurance policy to three trustees: B, C and D. A, B and C covenanted with the trustees to pay the premiums due on the policy if A did not do so. A new trustee, E, brought an action against A, B and C for a declaration that they were jointly liable to pay the premiums. Warrington J said at 534:

"The substance of the objection is that a man cannot make a contract with himself. One would have thought it only required to be stated to be self-evident that it makes no difference that he joins in that contract with himself some other person either as covenantor or covenantee, if the obligation on the one side, or, as the case may be, the right to enforce that obligation on the other side, is joint."

He considered previous authority and concluded at 536:

"Now the result of that case, I think, is plainly to determine that as a matter of substance an obligation by a man to pay himself, or to pay himself and another, is one which is in fact not an obligation in the eye of the law."

He went a little further in holding that if the covenant was void at law, no action could be maintained on the covenant in equity (although he did not rule out some other form of equitable remedy).

In the following year, Warrington J decided Napier v Williams, to which I have already referred. As noted, he held that the lease was not void; but then went on to consider the status of the covenants contained in the lease. In that respect, he followed and applied his previous decision in Ellis v Kerr. Since the covenants were void at law, it followed (a) that they could not be burdens on the estate created by the lease; and (b) the covenants were not enforceable in equity. But, as I have said, he decided that if the covenants were not performed, then the landlord could forfeit for breach of condition.

In this respect, the common law has been altered by section 82 of the Law of Property Act 1925. As we have seen, that provides that any agreement entered into by a person with himself and one or more other persons is construed and is capable of being enforced as if the agreement had been entered into with the other person or persons alone. It is not suggested that the section applies only to written agreements. The point of section 82 was explicitly to reverse cases like Napier v Williams: see Lie v Mohile at [16]. Accordingly, in terms of contract, an agreement by A and B to let property to A, B and C would appear to be enforceable (at least):

i) By A and B against C

ii) By C against A and B.

Common to these permutations is that C has fully enforceable contractual rights and obligations.

Exclusive possession

As Mr Walker correctly submitted, it is one of the essentials of a tenancy that the tenants have exclusive possession. This was not a necessary condition of a tenancy introduced for the first time by Street v Mountford [1985] 1 AC 809. Rather, Street v Mountford restored what had been regarded as orthodoxy after the development of the law by Lord Denning in a series of decisions in the 1960s and 1970s: see Ramnarace v Lutchman [2001] UKPC 25, [2001] 1 WLR 1651 at [15]. Lord Templeman was clear on the point in Street v Mountford, saying at 816B that Mrs Mountford was seeking "to re-establish and re-affirm the traditional view that an occupier of land for a term at a rent is a tenant providing the occupier is granted exclusive possession". It cannot, in my judgment, be supposed that the distinguished judges who decided the 19^th century cases to which I have referred were unaware of the necessity of exclusive possession before a tenancy could be created.

I have already concluded that there is no conceptual problem in the express grant of a tenancy by A and B to A, B and C, even if the grant is not a written one. In such a case the grant will necessarily entail that A, B and C have exclusive possession as against A and B. That cannot be a legal impossibility, not least because it is precisely what section 72 contemplates.

Why it is not a legal impossibility is, in my judgment, explicable as follows. First, possession is a single possession, although it may be exercised by several persons jointly: J A Pye (Oxford) Ltd v Graham [2002] UKHL 30, [2003] 1 AC 419 at [41]. In that case, Lord Hope explained at [70]:

"The general rule, which English law has derived from the Roman law, is that only one person can be in possession at any one time. Exclusivity is of the essence of possession. The same rule applies in cases where two or more persons are entitled to the enjoyment of property simultaneously. As between themselves they have separate rights, but as against everyone else they are in the position of a single owner."

Thus as against A and B, A, B and C are in the position of a single owner. Second, following the grant of the tenancy, A and B are entitled to possession in two capacities; and to two different types of possession. As joint landlords they are entitled to possession in the sense of receipt of rents and profits (as contemplated by section 205 of the Law of Property Act 1925); and as joint tenants they are entitled, together with C, to physical possession of the land. Put another way, they have exchanged their right to physical possession of the land for symbolic possession in the shape of receipt of rents and profits.

Even if that analysis is correct, however, Mr Walker argued that it cannot apply to a case in which there has been no express grant (whether written or oral) but the existence of a tenancy has to be inferred from the facts on the ground. If a tenancy is expressly created then exclusive possession is a consequence of the grant. But if a tenancy is to be inferred from conduct, exclusive possession is a necessary pre-condition of the inference, rather than a consequence. If A and B continue to exercise rights over the land itself, how can it be said that they have in some way deprived themselves of exclusive possession?

The first answer to this conundrum seems to me to lie in the proposition that A and B have exchanged their right to possession in the physical sense for the right to receive rents and profits which are payable by A, B and C for the whole of the land. The second is that A and B own the freehold in one capacity, while A, B and C occupy the land in a different capacity.

In Churcher v Martin (1888) 44 Ch D 312 A conveyed land to B, C and D as trustees on charitable trusts. For technical reasons the deed of conveyance was void. But A continued to receive the rent. A died, leaving his property to B and appointing B, C and D as his executors. B, C and D entered into possession and applied the income in accordance with the trusts of the void deed. Thus although B was the owner of the land under A's will, he took possession as one of three trustees. Following B's death, his executors claimed possession. The issue was whether the trustees had acquired title by adverse possession despite the fact that B was both the paper owner and one of the trustees. Kekewich J held that the claim to adverse possession succeeded. He said at 319:

"If [B] had been the sole trustee the aspect of the case would have been very different, but here he was one of several; receiving rents and performing acts of ownership without the slightest reservation or indication of beneficial claim, and unless I am bound to regard the physical identity of the beneficial owner with one of the trustees as prevailing to the disregard of the facts and substance of the story, I must treat [B] the beneficial owner as excluded from possession for twelve years and upwards, notwithstanding that [B] the trustee took an active part in the management of the estate. Am I so bound? Trustees are not a corporate body; there is no aggregate existence independent of the individual members; but, on the other hand, they are joint owners of the trust property, and they cannot act otherwise than jointly, even though frequently and for many purposes one member of the body represents the others. It seems to me consistent with principle to hold that the joint possession excludes that of any one of the joint possessors on his own behalf, and no authority was cited or has occurred to me inconsistent with that view. My conclusion therefore, is that the accident of [B's] beneficial interest did not operate to defeat the title of the trustees which he intended to preserve."

Churcher v Martin was recently referred to, without any hint of disapproval, in the Privy Council case of Bannerman Town, Millars and John Millars Eleuthera Association v Eleuthera Properties Ltd [2018] UKPC 27. The significance of dual capacities in which land is held is also inherent in the view that the common law took of the operation of merger. If the same person held both term and reversion in different capacities (particularly where one of them was a representative capacity), no merger took place. In such a situation, the correct analysis must be that in his capacity as holder of the term the lessee was entitled to exclusive possession as against the holder of the reversion (even though he himself was that holder).

In addition, Mr Walker's submission is not consistent with either Rogers v Harvey or Tindal CJ's direction to the jury in Doe d Colnaghi v Bluck. In each of those cases exclusive possession was inferred from the facts, despite the fact that the putative landlord exercised rights over the land.

As a practical illustration of the significance of differing capacities, consider a case in which A (being both trustee and partner) incurs expenditure on the land. If that expenditure is incurred by him as trustee, he must look to the trust assets for his indemnity. If, on the other hand, he incurs the expenditure as partner, it is to the partnership assets that he must look. Moreover, in a case such as this where the freeholders are will trustees, and the tenants are farming partners, the trustees qua trustees may have powers of leasing but no power to carry on a business.

Return to Rye v Rye

It is convenient at this point to return to Viscount Simonds' objections to the grant of a lease by A and B to A and B in order to consider whether those same objections apply to a lease granted by A and B to A, B and C.

The first objection was, in effect, that such a grant would be pointless. But in the case of a lease by A and B to A, B and C, that objection does not apply. C had no right to enter the land before the grant. Following the grant he has joint possession of it, together with A and B. I have already suggested some ways in which a person may act in dual capacities. So there may be perfectly sound reasons for wishing to grant such a lease. Lord Radcliffe took the same view, at least where landlord and tenant held in different capacities.

The second objection was that A and B could not contract with A and B. But, as a result of section 82 of the Law of Property Act 1925, A and B can contract with A, B and C. So that objection does not apply.

The third objection was that the lease, once granted, would merge in the freehold. But a lease vested in A, B and C would not merge (either at law or in equity) with a freehold held by A and B alone. So that objection does not apply.

The fourth objection was that A could not distrain on his own goods. The traditional common law view about rent was that it was a thing that issued out of the land. It is for that reason that the landlord is, in principle, entitled at common law to distrain on any goods that he finds on the land, whoever they belong to. There are, of course, exceptions to this general rule some of which are creations of the common law and some of which owe their origin to statutory intervention. At common law, therefore, the landlord may distrain on goods of a sub-tenant or even of a stranger (for a well-known example of the latter see Exall v Partridge (1799) 8 Term Rep 308). So ownership of the goods is not a reason to deny the landlord the remedy of distress. Moreover, even if the landlord cannot (or would not) distrain on goods which belong to him alone, there is no reason why he should not distrain on goods that belong jointly to him and someone else.

I do not, therefore, consider that Viscount Simonds' objections apply to a lease from A and B to A, B and C.

As far as Lord McDermott's observation is concerned, it is correct that at common law A could not contract with A and B. But that is not the same as saying that A could not have granted a leasehold interest to A and B which, at any rate, gave B an interest in the land.

I should also note Lord Denning's point that the tenancy "must stand or fall with the agreement on which it is founded". If, as I consider, section 82 of the Law of Property Act 1925 validates the agreement, then it would follow that in Lord Denning's view the tenancy would stand. That said, as Millett LJ pointed out in Ingram, and as we have seen from the early cases, it is possible to have a tenancy without any enforceable covenants.

So far as Lord Denning's observations on the ability of joint landlords to serve notice to quit are concerned, they are, in my respectful opinion, simply wrong. In Hammersmith LBC v Monk [1992] AC 478 the House of Lords held that a contractual periodic tenancy held by two or more joint tenants continued only so long as they all agreed in its continuation; and that, accordingly, in the absence of any term in the tenancy agreement to the contrary, a periodic tenancy was determinable by a notice to quit given by one joint tenant without the concurrence of the other joint tenants. The principle is not confined to joint tenants. It applies also to joint landlords, as Lord Bridge explained at 484:

"… from the earliest times a yearly tenancy has been an estate which continued only so long as it was the will of both parties that it should continue, albeit that either party could only signify his unwillingness that the tenancy should continue beyond the end of any year by giving the appropriate advance notice to that effect. Applying this principle to the case of a yearly tenancy where either the lessor's or the lessee's interest is held jointly by two or more parties, logic seems to me to dictate the conclusion that the will of all the joint parties is necessary to the continuance of the interest." (Emphasis added)

That was the position in the nineteenth century too: Doe d Aslin v Summersett (1830) 1 B & Ad 135; Doe d Kindersley v Hughes (1840) 7 M & W 139 and Alford v Vickery (1842) Car & M 280. Had the House heard argument on the common law in Rye v Rye, this error would no doubt have been avoided.

With the benefit of far more extensive citation of authority than was given to the judge, I conclude that he was wrong to hold that the alleged tenancy was incapable of being created.

Tenancy at will?

Section 52 (1) of the Law of Property Act 1925 provides that all conveyances are void for the purposes of creating a legal estate unless made by deed. Section 52 (2) (d) provides that that rule does not apply to leases or tenancies not required by law to be made in writing. Section 53 (1) provides that, subject to the provisions relating to the creation of leases by parol:

"(a) no interest in land can be created … except by writing signed by the person creating … the same, or by his agent…"

Section 54 is the section dealing with the creation of leases. It provides:

"(1) All interests in land created by parol and not put in writing and signed by the persons so creating the same, or by their agents thereunto lawfully authorised in writing, have, notwithstanding any consideration having been given for the same, the force and effect of interests at will only.

(2) Nothing in the foregoing provisions of this Part of this Act shall affect the creation by parol of leases taking effect in possession for a term not exceeding three years (whether or not the lessee is given power to extend the term) at the best rent which can be reasonably obtained without taking a fine."

Section 55 provided that the law relating to part performance was unaffected. But part performance is no longer part of the law.

The judge held that the putative tenancy that he would have found in this case (had it been legally possible to create it) was not saved by section 54 (2) because he was narrowly persuaded that it was not "at the best rent which can be reasonably obtained". There is no appeal against that finding of fact. He therefore concluded that if a tenancy had been created it would have been a tenancy at will.

In many cases the purported grant of a leasehold interest which fails to comply with the formalities required by the Law of Property Act 1925 will be saved by the application of the principle formulated in Walsh v Lonsdale (1882) 21 Ch D 9. But that principle is founded on the existence of a specifically enforceable contract. Since the enactment of section 2 of the Law of Property (Miscellaneous Provisions) Act 1989 a contract for the creation of an interest in land must be made in writing by incorporating all the agreed terms in one document signed by all parties (or exchange of written contracts). Section 2 (5) (a) says that the rule does not apply to leases falling within section 54 (2) of the Law of Property Act 1925. Since the tenancy was, on the judge's findings, outside the scope of section 54 (2) it could not take effect in equity as a periodic tenancy.

At common law if a person entered land under a void lease, he became a tenant at will. But at common law, the court leaned away from a tenancy at will in such circumstances, and consequently when the tenant paid or agreed to pay any of the rent expressed to be reserved, he became a tenant from year-to-year on the terms of the void lease, so far as applicable to, and not inconsistent with, a yearly tenancy: Woodfall on Landlord and Tenant (Looseleaf ed) para 6.066. It was not argued before us that that approach applied on the facts of this case.

I proceed, therefore, on the basis that the tenancy was, therefore, a tenancy at will, both at common law and in equity.

Is a tenancy at will converted into a periodic tenancy?

Section 2 of the Agricultural Holdings Act 1986 provides (so far as material):

"(1) An agreement to which this section applies shall take effect, with the necessary modifications, as if it were an agreement for the letting of land for a tenancy from year to year unless the agreement was approved by the Minister before it was entered into.

(2) Subject to subsection (3) below, this section applies to an agreement under which—

(a) any land is let to a person for use as agricultural land for an interest less than a tenancy from year to year, or

(b) a person is granted a licence to occupy land for use as agricultural land,

if the circumstances are such that if his interest were a tenancy from year to year he would in respect of that land be the tenant of an agricultural holding."

So the first question under this head is: is a tenancy at will a letting of land for an interest less than a tenancy from year to year? Tenancies at will have been known for centuries. Coke on Littleton (Co Lit 55 a) states:

"Tenant at will is, where lands or tenements are let by one man to another to have and to hold to him at the will of the lessor, by force of which lease the lessee is in possession."

This description (which is taken directly from Littleton's Tenures) affirms not only that a tenancy at will is a letting, but also that it carries with it possession of the land. Lord Coke went on to explain that in the case of a letting of arable land held on a tenancy at will, if the lessee had sown a crop he was entitled to reap it despite the termination of the tenancy.

Apart from the fact that a tenancy at will is, as its name suggests, terminable at will it is in all other respects a tenancy. As Lord Millett put it in Ramnarace v Lutchman at [16]:

"A tenancy at will is of indefinite duration, but in all other respects it shares the characteristics of a tenancy."

One of those defining characteristics is that of exclusive possession. A further consequence of the creation of a tenancy at will was that the landlord was entitled to distrain for rent: Co Litt 57 b; Doe d Davies v Thomas (1851) 6 Exch 854; Morton v Woods (1869) LR 4 QB 293. That was a remedy which the common law reserved to situations in which there was a subsisting relation of landlord and tenant at law. In addition, as we have seen in Rogers v Harvey, a tenancy at will was sufficient to qualify a tenant at will to vote "as tenant". I do not consider that the obiter observations of Viscount Simonds in Wheeler v Mercer [1957] AC 416, 427 detract from this conclusion.

On the face of it, therefore, a tenancy at will is a letting of land for an interest less than a tenancy from year to year. There are cases in which tenancies at will have been held to fall outside the scope of section 2 or its predecessors. A tenancy at will created by an attornment clause in a mortgage is one such case (Steyning and Littlehampton BS v Wilson [1951] Ch 1018); as is a tenancy at will arising in favour of a purchaser under a contract of sale of the freehold let into possession before completion (Walters v Roberts (1980) 41 P & CR 210). But in both those cases the tenancy at will was entirely parasitic upon and ancillary to a wholly different legal relationship. As Danckwerts J said in Steyning at 1025 the 1986 Act (and its predecessor) was intended to apply to "true transactions between landlord and tenant in respect of agricultural holdings".

On the judge's findings this is a case of a true transaction between landlords and tenants, albeit one that is inferred from conduct.

Mr Walker objected that to reach this conclusion would have the radical effect that the most precarious form of tenure would be converted into one of the most secure. He pointed to the contrast with Part II of the Landlord and Tenant Act 1954 under which a tenant at will obtains no security of tenure. But Part II operates in a very different way to the Agricultural Holdings Act 1986. First, it expressly excludes from protection a tenancy granted for a term certain not exceeding six months (subject to some exceptions), whereas the 1986 Act would protect such tenancies. Second, the machinery for determination of tenancies is very different. It operates on the tenancy as granted, whereas the 1986 Act first converts the tenancy into a tenancy from year to year; and the provisions for termination then operate on that tenancy from year to year. Third, the policy underlying the protection of tenancies of agricultural holdings is different. As Lord Salmon explained in Johnson v Moreton [1980] AC 37, 52 it was seen as vital to the national economy that the level of production and efficiency of our farms should be maintained and improved. Thus the security of tenure with which tenant farmers were provided was:

"… not only for their own protection as an important section of the public …; it was for the protection of the nation itself."

Lord Hailsham described it as a "public interest introduced for the sake of the soil and husbandry of England". There is every policy reason, therefore, for holding that a tenancy at will created independently of another overarching legal relationship attracts the protection of section 2 of the 1986 Act.

The leading textbooks all take the view that a tenancy at will falls within the scope of section 2: Muir Watt and Moss on Agricultural Holdings (15^th ed) para 9.7; Rodgers on Agricultural Law (4^th ed) para 5.33; Scammell Densham and Williams on Agricultural Holdings (10^th ed) para 18.11; Woodfall on Landlord and Tenant (Looseleaf ed) para 21.021. In my judgment they are correct.

Mr Peters had a fall-back position on the basis that the partnership had a licence which was converted into a yearly tenancy. But on the view that I take, that issue does not arise.

Was the land an agricultural holding?

That brings me to the final issue which is: was the land an agricultural holding at all? The judge decided this issue in favour of Philip and James Procter; and that decision is put in issue by a Respondent's Notice.

This issue turns on Section 1 of the Agricultural Holdings Act 1986 which provides, so far as material:

"(1) In this Act "agricultural holding" means the aggregate of the land (whether agricultural land or not) comprised in a contract of tenancy which is a contract for an agricultural tenancy, not being a contract under which the land is let to the tenant during his continuance in any office, appointment or employment held under the landlord.

(2) For the purposes of this section, a contract of tenancy relating to any land is a contract for an agricultural tenancy if, having regard to—

(a) the terms of the tenancy,

(b) the actual or contemplated use of the land at the time of the conclusion of the contract and subsequently, and

(c) any other relevant circumstances,

the whole of the land comprised in the contract, subject to such exceptions only as do not substantially affect the character of the tenancy, is let for use as agricultural land."

The judge gave his decision succinctly on this point. What he said was:

"[270] First, Mr Walker relies upon the AST of Wide Open Farm and the occupation of Moor Park Farm by Jamie alone who is not involved actively in the Partnership. As regards these matters there is clear authority that it is not simply a question of the value or income derived from respective portions of land which is determinative. Assuming no deemed surrender, I would not have held that these matters prevented AHA protection applying.

[271] The more difficult question is that of the Golf Course which, on Mr Walker's calculation amounts to approximately 27% in acreage of the land said to be held on the AHA tenancy. It is not suggested that the nature of a Golf Course is such that it would amount to agricultural use. With some hesitation I accept the submissions of Ms Shea that the following matters considered against the whole mean that any tenancy would not have lost AHA protection by reason of part of the land being used for a Golf Course, even though it is a substantial portion of the land in question: (a) the Golf Course has been loss making compared with the farm; (b) it was developed as being ancillary to the farmland and as a diversification of the farm business; (c) it is operated as part of the Partnership's business; (d) the vast majority of the land is used for farming. I would also have taken into account, as an overall consideration, the Golf Course and the overall farming land within the Family Inheritance as a whole so that the proportion of the Golf Course to that land as a whole in acreage terms is closer to 20%."

The statutory test requires the trial judge to evaluate the "character" of the tenancy. That is in itself a very broad inquiry. He is to do so applying a very broad-textured test, which requires him to take into account all "relevant circumstances".

Mr Walker criticised the judge for having had regard to the fact that the golf course was loss-making. Although I agree that the fact that it was loss-making is not determinative, I cannot regard it as irrelevant to the inquiry to ask whether the partnership profit is substantially (or indeed solely) derived from agricultural activity. That ties in with the judge's third and fourth reasons.

One of the matters that he took into account was the fact that the golf course had been created as ancillary to the farm business and as part of farm diversification. He described the circumstances which led to that decision at [126] to [129] of his judgment. As the judge explained, the golf course was created as the result of a study conducted by the Agricultural Development and Advisory Service. Their report stated:

"93. The agricultural business is made up of about 500 acres of arable land plus an area of grassland, roads, buildings and ponds. Just over 100 acres of land is in set-aside and up to 35 acres per year is fallowed….

96. The area of land required for the proposed course is 120 acres and will utilise the set-aside land (100.5 acres) and take up most of the fallow land out of the rotation. The effect on the farm income will be very slight, being the reduced rate paid by set-aside land used for development compared with the permanent fallow. This is equivalent to £24 per acre. In addition there will be a one-off repayment of set-aside…[which] could amount to £1,612."

There are two important features of this. First, the report was compiled by an agricultural advisory board. Second, the golf course was created in large part on land that had already been taken out of active use for agriculture and was designated as "set-aside". If putting unproductive land to productive use were to change the character of a tenancy, that would be a serious disincentive to the efficient use of land.

The kind of judgment which the judge was called upon to make was an evaluative judgment based on a variety of factors. The approach of an appeal court to such a judgment was explained by Hoffmann LJ in Re Grayan Building Services Ltd [1995] Ch 241, 254:

"… the standards applied by the law in different contexts vary a great deal in precision and generally speaking, the vaguer the standard and the greater the number of factors which the court has to weigh up in deciding whether or not the standards have been met, the more reluctant an appellate court will be to interfere with the trial judge's decision."

Similarly, in Jeffery v British Council [2018] EWCA Civ 2253, [2019] ICR 929 Longmore LJ said at [136]:

"… it is well settled that an appellate tribunal will not interfere with a first instance evaluative judgment of this kind unless that tribunal took into account matters it should not have taken into account or failed to take into account matters it should have taken into account or made some error or was otherwise wrong."

Other judges might have reached a different conclusion on the facts; and the judge said that he reached his own conclusion "with some hesitation". But that is the kind of question on which reasonable people can differ. It is in precisely those circumstances that an appeal court should respect the balance struck by the trial judge; all the more so because he had been immersed in the evidence about the running of the farm over the course of a seven-day trial.

I would reject the point raised by the Respondent's Notice.

Result

I would allow the appeal.

Lord Justice Arnold:

I agree.

Lord Justice Nugee:

I also agree.