Berrisford (FC) v Mexfield Housing Co-operative Limited

Mexfield Housing Co-Operative Ltd (“Mexfield”) is a fully mutual housing co-operative association, which was founded by a bank as part of a mortgage rescue scheme, ie with a view to buying mortgaged properties from individual borrowers who are in difficulty, and then letting the properties back to them. In that capacity, it acquired a number of residential properties, which it then let out to the former owner-mortgagors, who, in the normal way for tenants of a fully mutual housing co-operative, were required by its rules to be members of Mexfield.

One of those properties is 17 Elton Avenue, Barnet (“the premises”), which, during 1993, Mexfield purchased from Ms Ruza Berrisford, and, by a written agreement made on 13 December 1993, agreed to let it back to her. The agreement was described as an “Occupancy Agreement” (“the Agreement”), and it started by reciting that, in anticipation of her occupation of the premises, Ms Berrisford had become a member of Mexfield. Clause 1 then provided as follows:

“[Mexfield] shall let and [Ms Berrisford] shall take the [premises] from 13 December 1993 and thereafter from month to month until determined as provided in this Agreement.”

Clause 2 stipulated that the rent was to be payable weekly in advance, and the fourth schedule stated that it was to be £89 per week, subject to annual increase in accordance with the Retail Price Index plus 2%. Clause 2(4) stated that any increase in rent was only to take effect after “[a]t least one month’s notice in writing”.

Clauses 3 and 4 contained a number of covenants of a familiar nature to those who are conversant with professionally drafted residential tenancy agreements. Clause 3 was concerned with Ms Berrisford’s obligations, and it included, in subclause (1), an obligation to take possession of the premises and to pay the rent, in subclause (9), an obligation to use the premises as her “only or main residence”, and, in subclause (11), an absolute obligation not to assign, sublet or part with possession or occupation of the whole or any part of the premises.

The only provisions of the Agreement which expressly provided for its determination were clauses 5 and 6, which were in these terms:

“5.

This Agreement shall be determinable by [Ms Berrisford] giving [Mexfield] one month’s notice in writing.

6.

This Agreement may be brought to an end by [Mexfield] by the exercise of the right of re-entry specified in this clause but ONLY in the following circumstances:

a)

If the rent reserved hereby or any part thereof shall at any time be in arrear and unpaid for 21 days …

b)

If [Ms Berrisford] shall at any time fail or neglect to perform or observe any of the [terms of] this Agreement which are to be performed and observed by [her]

c)

If [Ms Berrisford] shall cease to be a member of [Mexfield]

d)

If a resolution is passed under … [Mexfield’s] Rules regarding a proposal to dissolve [Mexfield]

THEN in each case it shall be lawful for [Mexfield] to re-enter upon the premises and peaceably to hold and enjoy the premises thenceforth and so that the rights to occupy the premises shall absolutely end and determine as if this Agreement had not been made … .”

Because Mexfield was a mutual housing association, any residential tenancy granted to one of its members attracted no statutory protection, save the very limited protection accorded by the Protection from Eviction Act 1977 (“the 1977 Act”). For present purposes, the 1977 Act relevantly contains provisions which (i) preclude a residential property owner from physically excluding or evicting an existing or former licensee or tenant from the property without an order of the court, and (ii) entitle a residential occupier under a periodic tenancy to at least four weeks’ notice to quit.

Ms Berrisford remained in occupation of the premises, complying with her obligations under the Agreement, until (apparently through no fault of hers) she fell behind with her rent. Mexfield could have invoked clause 6(a), but it did not do so, presumably because it is a forfeiture provision, and Ms Berrisford soon paid off the rent arrears, so it would have been a foregone conclusion that she would have obtained relief from forfeiture. Rather than relying on clause 6(a), what Mexfield did was to serve a notice to quit on Mrs Berrisford on 11 February 2008, expiring on 17 March 2008. Mexfield then brought proceedings for possession in the County Court, arguing that, despite the apparent limited circumstances in which, and the limited method by which, it could terminate the Agreement (sc. under clause 6), it nonetheless was entitled to put an end to Ms Berrisford’s tenancy by serving a notice to quit.

The evidence advanced on behalf of Ms Berrisford suggests that, in the past five years or so, Mexfield, or its mortgagee, came under some financial pressure, and that, as a result of purchasing its mortgage debt, Mexfield is now effectively owned and controlled by “a businessman”, who is seeking to pursue the claim for possession for “commercial reasons”. Mexfield’s evidence is that it is run by a committee of management, and nobody else. For present purposes, it is unnecessary to address this dispute, but it is only fair to Mexfield to record that (i) even if Ms Berrisford’s version of events is correct, Mexfield has done nothing wrong, (ii) this is a test case, as the Agreement is a standard form used by Mexfield, and (iii) Mexfield has agreed that it will enter into a fresh agreement with Ms Berrisford if her current appeal fails.

The essence of Mexfield’s case was that the effect of a number of cases, culminating in the decision of the House of Lords in Prudential Assurance Co Ltd v London Residuary Body [1992] 2 AC 386, is that an arrangement such as that embodied in the Agreement could not be a valid tenancy as a matter of law. Accordingly, Mexfield contended, Ms Berrisford had become a periodic (either weekly, because she paid a weekly rent, or monthly, because that was the parties’ apparent intention) tenant of the premises by virtue of the payment and acceptance of rent since 1993, and, pursuant to well-established and unchallenged principles, Mexfield was entitled to determine such a tenancy with at least one week’s (or one month’s) notice in common law, but required to be at least four weeks’ notice by the 1977 Act.

At first instance, His Honour Judge Mitchell refused Mexfield’s application for summary judgment. However, on appeal, Peter Smith J, and, on appeal from him, the Court of Appeal, accepted, albeit reluctantly, Mexfield’s argument, and made an order for possession – see [2009] EWHC 2392 Ch and [2010] EWCA Civ 811; [2011] 1 Ch 244. In the Court of Appeal, in three clear and illuminating judgments, Mummery and Aikens LJJ considered that they were bound by authority to make such an order, but Wilson LJ held that the contractual limitation on Mexfield’s right to determine the Agreement was enforceable by Ms Berrisford.

The argument for Ms Berrisford, as developed by Mr Wonnacott in his conspicuously clear and learned submissions, is to this effect:

I shall consider those five points in turn.

The first point turns on the interpretation of the Agreement. At any rate at first sight, it seems hard to quarrel with Mr Wonnacott’s contention that it continues until either the tenant serves notice under clause 5 or the landlord can and does exercise its rights under clause 6. After all, clause 1 provides that the Agreement is to subsist “until determined as provided in this agreement”, and the only provisions which deal with determination are clauses 5 and 6.

However, on behalf of Mexfield, Mr Gaunt QC relies on the fact that clause 1 describes the tenancy as being “from month to month”, and says that this carries with it a right in the landlord to determine the Agreement on one month’s notice. This was not a point raised by Mexfield in the courts below, but it is a pure point of law, and no prejudice could be caused to Ms Berrisford by the fact that it was taken for the first time in this court. We therefore permitted Mexfield to argue that it has the right to determine the Agreement on one month’s notice.

Having now heard the argument, which was advanced in attractive and well-modulated terms, I would reject it. Before considering the argument by reference to the terms of the Agreement, it is, of course, necessary to consider whether there is anything in the surrounding circumstances relevant to the interpretation of the Agreement. Beyond the fact that Mexfield was a co-operative housing association, of which Ms Berrisford was a member, and that the purpose of the Agreement was to provide Ms Berrisford with a home (all of which is anyway plain from the provisions of the Agreement), it does not appear to me that there are any relevant facts, save that the mortgage rescue background tends to support the notion that Ms Berrisford’s right of occupation was not intended to be precarious. (There was some argument as to whether the Agreement was based on guidance given by the Housing Corporation, and there was an application to adduce further evidence in that connection. I consider that whether or not Mexfield was following such guidance is irrelevant to any issue on this appeal. I would therefore refuse the application, and will say no more about that aspect).

Turning to the language of the Agreement, Mr Gaunt’s central point is that the inclusion of the words “from month to month” mean that the Agreement was intended to be a monthly tenancy, and that, effectively by definition, a monthly tenancy can be determined by either side with one month’s notice. I accept that, in the absence of any indication to the contrary, a tenancy granted “from month to month” is a monthly tenancy, and that, again in the absence of any indication to the contrary, a monthly tenancy may be determined by either party giving one month’s notice to the other. For instance, as explained later in this judgment, it is clear that parties to a monthly tenancy could agree on a bar for a specific period on the right of a landlord to serve notice to quit, and they can also agree a longer (or indeed a shorter) period of notice than the one month which would normally be implied.

As observed by Lord Clarke a tenancy agreement has to be interpreted in the same way as any other written contract, so the precise rights and obligations of the parties under it must depend on the terms which the parties have agreed and the circumstances in which they were agreed. However, in some circumstances, there may be principles of law which result in the parties’ intention being frustrated or modified, and, as is clear from the reasoning in Street v Mountford [1985] AC 809, the legal consequences of what the parties have agreed is a matter of law rather than dependent on what the parties intended.

In this case, it seems clear that the parties intended that the arrangement created by the Agreement should only be determinable pursuant to clause 5 or clause 6. As already mentioned, that seems to follow from the closing words of clause 1. However, Mr Gaunt argues that, by virtue of the fact that a tenancy from month to month is determinable by either party on one month’s notice by either party, the Agreement impliedly provided for determination by Mexfield on a month’s notice.

That ingenious argument appears to me to suffer from several flaws. First, such an implied right of determination does not naturally fall within the scope of the natural meaning of the expression “determined as provided in this Agreement”. Secondly, it would mean that clause 5 is redundant. Thirdly, it seems scarcely consistent with the words “but ONLY” in clause 6. Fourthly, if Mexfield does have the right to determine on one month’s notice, there seems to have been little point in giving it the right to forfeit in clause 6. Finally, given the circumstances in which the Agreement was entered into, it seems unlikely that Ms Berrisford’s security was intended to be so tenuous as to be determinable by Mexfield on one month’s notice at any time from the day the Agreement was made.

Accordingly, if a monthly tenancy is automatically determinable by either party giving the other one month’s notice, as Mr Gaunt contends, it seems to me clear that the Agreement did not give rise to a monthly tenancy, unless it is impossible to conclude, as a matter of law, that it did not. In my view, it is perfectly possible to avoid such a conclusion. The words “from month to month” could quite properly have been included to indicate that, if Ms Berrisford wished to serve notice under clause 5, or (more questionably) if Mexfield wished to serve notice under clause 2(4), such a notice must take effect on the thirteenth day of a month.

However, it is right to add that, for reasons given later in this judgment, I do not accept Mr Gaunt’s contention that, if the Agreement gave rise to a monthly tenancy, it would have to be automatically determinable on one month’s notice by either party.

Accordingly, whether or not the Agreement gave rise to a monthly tenancy, I consider that, as a matter of contractual interpretation, the effect of clauses 1, 5 and 6 is that that tenancy can only be determined by Ms Berrisford pursuant to clause 5 or by Mexfield pursuant to clause 6, and in no other way (save consensually, viz by surrender).

I turn to the second issue, namely whether an arrangement, which can only come to an end by service of one month’s notice by the tenant, or by the landlord invoking a right of determination on one or more of the grounds set out in clause 6, is capable, as a matter of law, of being a tenancy in accordance with its terms. Mr Wonnacott accepts that it is not so capable. His concession is supported both by very old authority and by high modern authority.

It seems to have been established for a long time that an agreement for an uncertain term cannot be a tenancy in the sense of being a term of years. In Say v Smith (1563) Plowd 269, 272, Anthony Brown J said that “every contract sufficient to make a lease for years ought to have certainty in three limitations, viz in the commencement of the term, in the continuance of it, and in the end of it … and words in a lease, which don’t make this appear, are but babble.”

That is consistent with what was stated in Bracton’s De Legibus et Consuetudinibus Angliae, written in the mid-thirteenth century. It is there stated that a grant of land “until you have taken 40 pounds” would be a “free tenement” (which could not be created without certain strict formalities), rather than a term certain (which did not require such formalities), “because it cannot be known how long it may take for so many pounds to be raised from [the] land, because the term is uncertain and undetermined” - Bracton on the Laws and Customs of England (trans Professor E Thorne) (1977), vol 3, p 50 (f176b). This statement was referred to with approval by Sir Edward Coke in Co Litt 42a (1628), and much the same is stated in Brook’s New Cases (1554/5) pl 462. So too in The Bishop of Bath’s Case (1605) 6 Co Rep 34b, 35b, Coke stated that a letting expressed to last until a certain amount of money had been “levied of the issues and profits … is but a lease at will without livery” – i.e. without the formalities required for the establishment of a freehold interest.

Much more recently, in Lace v Chantler [1944] KB 368, the Court of Appeal held that a purported letting “for the duration [of the Second World War]” could not take effect as “a good tenancy for the duration of the war” as it was for an uncertain term, and that it was consequently ineffective. This decision was distinguished by a subsequent Court of Appeal in Ashburn Anstalt v Arnold [1989] Ch 1, where it was held that a right to occupy premises until the owner gave one quarter’s notice certifying he needed the premises for redevelopment created a tenancy binding on third parties.

Less than 20 years ago, the House of Lords approved and applied Lace [1944] KB 368, and disapproved and overruled Ashburn [1989] Ch 1. In Prudential [1992] 2 AC 386, land was sold in 1930 by the owner, Mr Nathan, to the London County Council, who immediately leased it back to him at a weekly rent until the land “is required by the council for the purposes of the widening of” the road. The House of Lords held that this arrangement was incapable of creating a tenancy, as it was for an uncertain, potentially perpetual, duration. Lord Templeman (with whom the other members of the House of Lords agreed, albeit with reluctance in most cases) said at [1992] 2 AC 386, 394F, that there had been “500 years of judicial acceptance of the requirement that a term must be certain applies to all leases and tenancy agreements”.

The position with regard to periodic tenancies containing a fetter on the right of either or both parties to serve a notice to quit seems to be much the same. The concept of a periodic tenancy appears to have originated in a suggestion in Burgh v Potkyn (1522) YB 14 Hen 8 f10 pl 6, but it does not seem to have been accepted by the courts until the end of the 17th century, and then only on special facts - see Taylor v Seed (1696) Comb 383. Accordingly, there is not the long established learning which there is in relation to terms of uncertain duration.

In Doe d Warner v Browne (1807) 8 East 165, an agreement provided that a tenancy at £40 per annum would not be determined so long as the tenant paid the rent and did not harm the landlord. At p 166, Lord Ellenborough CJ said in argument that it would be “inconsistent with, and repugnant to” a “tenancy from year to year” that “it should not be determinable at the pleasure of either party giving the regular notice”. In his judgment on the following page, Lawrence J said much the same thing, tantalisingly (as it appears to have been in an unreported case) suggesting that Lord Mansfield had held otherwise, in a view which had “been long exploded.”

Mr Browne subsequently obtained equitable relief from Lord Eldon LC – see Browne v Warner (1807) 14 Ves Jun 156 and (1808) 14 Ves Jun 409. However, these reports only cover the grant of interlocutory relief, and the precise basis for its grant is not entirely clear. In those circumstances, although Mr Wonnacott seeks to derive assistance from Lord Eldon’s observations (as did Malins V-C in In re King’s Leasehold Estates (1873) LR 16 Eq 521, 526-527), I do not think that any conclusions can be safely drawn from his decision, save, perhaps, that any conclusion which may be derived from Doe v Browne 8 East 165 may not be as simple as more modern authorities suggest.

Despite what was said in Doe v Browne 8 East 165, the Court of Appeal in Breams Property Investment Co Ltd v Stroulger [1948] 2 KB 1, 6 held that an agreement by a landlord in a periodic tenancy not to serve notice to quit for three years unless it required the premises for its own use was valid. In In re Midland Railway Co’s Agreement, Charles Clay & Sons Ltd v British Railways Board [1971] Ch 725, the Court of Appeal held that an agreement by a landlord not to determine a half-yearly tenancy until the premises were needed for the purposes of its undertaking was valid. The Court distinguished Lace [1944] KB 368 on the ground that it did not concern a periodic tenancy, and derived assistance from Breams [1948] 2 KB 1.

In Prudential [1992] 2 AC 386, the House of Lords overruled Midland Railway [1971] Ch 725, effectively on the basis that a fetter of uncertain duration on the service of a notice to quit in relation to a periodic tenancy was as objectionable to the concept of a tenancy as was the existence of an uncertain term. It was not, however, suggested by Lord Templeman that Breams [1948] 2 KB 1 was wrongly decided.

Following the decision of the House of Lords in Prudential [1992] 2 AC 386, the law appeared clear in its effect, intellectually coherent in its analysis, and, in part, unsatisfactory in its practical consequences. The position appears to have been as follows. (i) An agreement for a term, whose maximum duration can be identified from the inception can give rise to a valid tenancy; (ii) an agreement which gives rise to a periodic arrangement determinable by either party can also give rise to a valid tenancy; (iii) an agreement could not give rise to a tenancy as a matter of law if it was for a term whose maximum duration was uncertain at the inception; (iv) (a) a fetter on a right to serve notice to determine a periodic tenancy was ineffective if the fetter is to endure for an uncertain period, but (b) a fetter for a specified period could be valid.

If we accept that that is indeed the law, then, subject to the point to which I next turn, the Agreement cannot take effect as a tenancy according to its terms. As the judgment of Lady Hale demonstrates (and as indeed the disquiet expressed by Lord Browne-Wilkinson and others in Prudential [1992] 2 AC 386 itself shows), the law is not in a satisfactory state. There is no apparent practical justification for holding that an agreement for a term of uncertain duration cannot give rise to a tenancy, or that a fetter of uncertain duration on the right to serve a notice to quit is invalid. There is therefore much to be said for changing the law, and overruling what may be called the certainty requirement, which was affirmed in Prudential [1992] 2 AC 386, on the ground that, in so far as it had any practical justification, that justification has long since gone, and, in so far as it is based on principle, the principle is not fundamental enough for the Supreme Court to be bound by it. It may be added that Lady Hale’s Carrollian characterisation of the law on this topic is reinforced by the fact that the common law accepted perpetually renewable leases as valid: they have been converted into 2000-year terms by section 145 of the Law of Property Act 1922.

However, I would not support jettisoning the certainty requirement, at any rate in this case. First, as the discussion earlier in this judgment shows, it does appear that for many centuries it has been regarded as fundamental to the concept of a term of years that it had a certain duration when it was created. It seems logical that the subsequent development of a term from year to year (ie a periodic tenancy) should carry with it a similar requirement, and the case law also seems to support this.

Secondly, the 1925 Act appears to support this conclusion. Having stated in section 1(1) that only two estates can exist in land, a fee simple and a term of years, it then defines a term of years in section 205(1)(xxvii) as meaning “a term of years … either certain or liable to determination by notice [or] re-entry”; as Lord Templeman said in Prudential [1992] 2 AC 386, 391B, this seems to underwrite the established common law position. The notion that the 1925 Act assumed that the certainty requirement existed appears to be supported by the terms of section 149(6). As explained more fully below, this provision effectively converts a life tenancy into a determinable term of 90 years. A tenancy for life is a term of uncertain duration, and it was a species of freehold estate prior to 1926, but, in the light of section 1 of the 1925 Act, if it was to retain its status as a legal estate, it could only be a term of years after that date. Presumably it was converted into a 90-year term because those responsible for drafting the 1925 Act thought it could not be a term of years otherwise.

Thirdly, the certainty requirement was confirmed only some 20 years ago by the House of Lords. Fourthly, while not a very attractive point, there is the concern expressed by Lord Browne-Wilkinson, namely that to change the law in this field “might upset long established titles” – [1992] 2 AC 386, 397A. Fifthly, at least where the purported grant is to an individual, as opposed to a company or corporation, the arrangement does in fact give rise to a valid tenancy, as explained below. Finally, it has been no part of either party’s case that the Agreement gave rise to a valid tenancy according to its terms (if, as I have concluded, it has the meaning for which Mr Wonnacott contends).

While Mr Wonnacott accepts that the arrangement contained in the Agreement would not be capable of constituting a tenancy in accordance with its terms, he contends that, at any rate before 1926, the arrangement would have been treated by the court as a tenancy for the life of Ms Berrisford, determinable before her death by her under clause 5, or by Mexfield under clause 6.

There is much authority to support the proposition that, before the 1925 Act came into force, an agreement for an uncertain term was treated as a tenancy for the life of the tenant, determinable before the tenant’s death according to its terms. In Bracton (op cit) vol 3, p 50 (f176b), it will be recalled that the grant of an uncertain term was held to give rise to a “free tenement”, provided that the formalities had been complied with. The nature of this free tenement would appear to be a tenancy for the life of the grantee. That is clear from what was said in Littleton on Tenures (1481/2) vol 2, section 382 namely:

“[I]f an abbot make a lease to a man, to have and to hold to him during the time that he is abbot … the lessee hath an estate for the term of his owne life: but this is on condition … that if the abbot resign, or be deposed, that then it shall be lawful for his successor to enter.”

In Co Litt vol 1, p 42a, it is similarly stated that if an estate is granted to a person until, inter alia, she marries, or so long as she pays £40 “or for any like incertaine term”, “the lessee hath in judgment of law an estate for life determinable if [the formalities of creation are satisfied]”. This passage was quoted and applied by North J in In re Carne’s Settled Estates [1899] 1 Ch 324, 329. The same point was made in Sheppard’s Touchstones on Common Assurances, where it is said that “uncertain leases made with … limitations … may be good leases for life determinable on these contingents, albeit they be no good leases for years” – 7th ed (1821), vol 2, p 275.

In Doe v Browne 8 East 165, 166-167, Lord Ellenborough CJ and Lawrence J, both of whom rejected the contention that an agreement which was to continue so long as the tenant paid the rent and did not harm the landlord’s business could be a valid term of years, said that it could be “an estate for life”, but that it failed to achieve this status because the necessary formalities had not been complied with. Such formalities have now largely been done away with, and they normally only require a written, signed document. As Lord Dyson’s reference to Joshua Williams’s 1920 textbook shows, the perceived legal position right up to the time of the 1925 property legislation was that terms of uncertain duration were converted into determinable terms for life.

On this basis, then it seems clear that, at least if the Agreement had been entered into before 1 January 1926 (when the 1925 Act came into force), it would have been treated by the court as being the grant of a tenancy to Ms Berrisford for her life, subject to her right to determine pursuant to clause 5 and Mexfield’s right to determine pursuant to clause 6.

Mr Gaunt relies on more recent authorities to support a contention that an agreement for an uncertain term was only regarded as creating a tenancy for life if, on a fair reading of the agreement, that was what the parties to the agreement intended. That does indeed seem to have been the approach of Sir George Jessel MR in Kusel v Watson (1879) 11 Ch D 129, but that is of limited value as the agreement in that case could not have created a tenancy for life, as it was created by a lessee. However, Mr Gaunt’s contention gets rather more support from the reasoning of the Court of Appeal in Zimbler v Abrahams [1903] 1 KB 577, 582 (per Vaughan Williams LJ) and 583 (per Stirling and Mathew LJJ).

In my judgment, however, there are three answers to that contention. The first is that the reasoning in Zimbler [1903] 1 KB 577 is not strictly inconsistent with Mr Wonnacott’s analysis: if, as a matter of interpretation, the agreement in that case did involve the grant of a tenancy for life, then there was no need to invoke Mr Wonnacott’s analysis, but that does not mean that the analysis is wrong. Secondly, if Zimbler [1903] 1 KB 577 did proceed on the assumption that an agreement which purported to create a tenancy for an uncertain term could not give rise to a tenancy for life unless it was the parties’ intention to do so, it was wrong, as it would have been inconsistent with the authoritative dicta relied on by Mr Wonnacott, in particular the clear statement in Littleton, vol 2, section 382. (I also note that neither counsel in Zimbler [1903] 1 KB 577 relied on the point made by Mr Gaunt: see pp 578-580). Thirdly, even if an agreement which creates an uncertain term could only have resulted in a tenancy for the life of the tenant if that was the intention of the parties, I consider that, on a true construction of the Agreement, it was intended that Ms Berrisford enjoy the premises for life – subject, of course, to determination pursuant to clauses 5 and 6. I have in mind in particular clause 6(c), which will apply on Ms Berrisford’s death, the fact that her interest is unassignable, and the fact that it was intended to ensure that she could stay in her home.

The next step in Mr Wonnacott’s argument is that, given that the Agreement would have given rise to a tenancy for life prior to 1926, the effect of section 149(6) of the 1925 Act (“section 149(6)”) is that the Agreement is now to be treated as a term of 90 years determinable on the death of Ms Berrisford, subject to the rights of determination in clauses 5 and 6.

Section 149(6) provides:

“Any lease ... at a rent ... for life ... or any contract therefor, made before or after the commencement of this Act, ... shall take effect as a lease ... or contract therefor, for a term of 90 years determinable after the death ... of the original lessee ... by at least one month’s notice in writing given to determine the same on one of the quarter days applicable to the tenancy…”

As already mentioned, the 1925 Act began by limiting the number of permissible legal estates in land to two, a fee simple and a term of years. Accordingly, it was necessary for the statute to deal with interests, such as estates for lives, which had previously been, but no longer were, valid legal estates. Hence one of the reasons for section 149(6). However, it is clear from its terms that the section was not merely concerned with preserving life interests which existed prior to 1 January 1926: it also expressly applies to life interests granted thereafter. Therefore, says Mr Wonnacott, the section converts an arrangement such as the Agreement, which, according to the common law, is a life tenancy into a 90-year term.

The first argument which might be raised against this contention is that the Agreement was not a “lease … for life”, merely a contract which would have been treated by established case law as such a lease. I do not consider that can be right. Apart from not being consistent with the wording of section 149(6), it would mean that an agreement such as that described in Littleton section 382, which existed as a continuing valid determinable life estate on the 1 January 1926, would have lost its status as a legal estate, as it would not have been saved by section 149(6): that cannot have been the legislature’s intention.

Mr Gaunt contends that section 149(6) is concerned with tenancies which automatically end with the tenant’s death, not with tenancies which can be determined on the tenant’s death, and, in this case, the effect of clause 6(c) is that the tenancy can be determined, not that it automatically determines, on the tenant’s death. I accept that section 149(6) only applies to tenancies which automatically determine on death, and I am prepared to assume that clause 6(c) can only be invoked by service of a notice. However, the argument misses the point, because the Agreement is (or would be in the absence of sections 1 and 149 of the 1925 Act) a tenancy for life, not because of the specific terms of, or circumstances described in, clause 6(c), but because it is treated as such by a well-established common law rule.

It is also suggested that section 149(6) does not apply to arrangements such as the Agreement which are determinable in circumstances other than the tenant’s death – e.g. on the grounds set out in clause 6. I can see no reasons of principle for accepting that contention, and it appears to me that there are strong practical reasons for rejecting it. The common law rule that uncertain terms were treated as life tenancies applied, almost by definition, to arrangements which determined in other events, and it is hard, indeed impossible, to see why the rule should be limited to cases where an event automatically determines the term, as opposed to cases where an event entitles the landlord to serve notice to determine the term. In each case, the term is uncertain. At least one of the reasons the common law treated uncertain terms as tenancies for lives was, as I see it, to save arrangements which would otherwise be invalidated for technical reasons, and I find it hard to accept that the modern law requires the court to adopt a less benevolent approach to saving contractual arrangements.

I have read what Lord Walker has written about section 149(6) and Bass Holdings Ltd v Lewis [1986] 2 EGLR 40, and respectfully agree with him.

It is strongly pressed by Mr Gaunt that the conclusion that the Agreement gives rise to a valid tenancy for 90 years determinable on the tenant’s death would be inconsistent with high modern authority. In particular, he said that such a conclusion would be contrary to the outcome in Lace [1944] KB 368, and inconsistent with clear dicta of Lord Greene MR in that case and of Lord Templeman in Prudential [1992] 2 AC 386. I accept the factual basis for that argument, but do not agree with its suggested conclusion.

The fact is that it was not argued in either of those two cases that the arrangement involved would have created a life tenancy as a matter of common law, and that, following section 149(6), such an arrangement would now give rise to a 90-year term, determinable on the tenant’s death (and Mr Wonnacott was kind enough to point out that such an argument would not have assisted, and may even have harmed, the unsuccessful respondent’s case in Prudential [1992] 2 AC 386). Some of the statements about the law by Lord Greene and Lord Templeman can now be seen to be extravagant or inaccurately wide, but it is only fair to them to repeat that this was, at least in part, because the tenancy for life argument was not raised before them.

For the reasons given, I accept Mr Wonnacott’s case that (i) the arrangement contained in the Agreement could only be determined in accordance with clauses 5 and 6, and not otherwise, (ii) such an arrangement cannot take effect as a tenancy in accordance with its terms, but (iii) by virtue of well-established common law rules and section 149(6), the arrangement is a tenancy for a term of 90 years determinable on the tenant’s death by one month’s notice from the landlord, and determinable in accordance with its terms, i.e. pursuant to clauses 5 and 6.

I indicated earlier in this judgment that this conclusion would apply irrespective of whether the purported tenancy created by the Agreement was simply for an indeterminate term or was a periodic tenancy with a fetter on the landlord’s right to determine. There is no difficulty if the former is the right analysis. If the latter is correct, then there is a monthly tenancy which the landlord is unable to determine unless he can rely on one or more of the grounds in clause 6. In Breams [1948] 2 KB 1, the Court concluded that a periodic tenancy with a fetter on the landlord’s right to determine for three years was valid. It seems to me that the term thereby created was equivalent to a fixed term of three years (subject to a restricted right of determination in the landlord and an unrestricted right of determination by the tenant) followed by a periodic tenancy.

Accordingly a periodic tenancy with an invalid fetter on the landlord’s right to determine should be treated in the same way as a tenancy for a fixed, if indeterminate, term. That seems to me to be justified in principle, logical in theory, and it ensures the law in this area is the same for all types of tenancy, whether or not periodic in nature (which was, I think, part of the reasoning in Prudential [1992] 2 AC 386). On that basis, even if the tenancy created by the Agreement was a monthly tenancy with an objectionable fetter, it seems to me that it would have been treated as a life estate under the old law (subject to the right to determine in accordance with the terms of the fetter), and so would now be a tenancy for 90 years.

Ms Berrisford is still alive, and it is common ground that she has not served notice under clause 5 and that Mexfield is not relying on clause 6. In those circumstances, it follows that Ms Berrisford retains her tenancy of the premises and that Mexfield is not entitled to possession.

This conclusion renders it unnecessary to consider two alternative arguments, which were raised by Mr Wonnacott, namely that (i) if the Agreement did not create a tenancy, it nonetheless gave rise to a binding personal contract between Mexfield and Ms Berrisford, which Ms Berrisford is entitled to enforce against Mexfield so long as it owns the premises, or (ii) if the Agreement created a periodic tenancy with an impermissible fetter on the right of the landlord to serve notice to quit, the fetter is nonetheless enforceable as against Mexfield so long as it is the owner of the premises.

However, having heard full submissions on those two arguments, I incline fairly strongly to the view that, if Ms Berisford had failed in establishing that she had a subsisting tenancy of the premises, she would nonetheless have defeated Mexfield’s claim for possession on the ground that she is entitled to enforce her contractual rights.

If the Agreement does not create a tenancy for technical reasons, namely because it purports to create an uncertain term, it is hard to see why, as a matter of principle, it should not be capable of taking effect as a contract, enforceable as between the parties personally, albeit not capable of binding their respective successors, as no interest in land or other proprietary interest would subsist.

The argument to the contrary rests in part on authority and in part on principle. So far as authority is concerned, the point at issue was specifically addressed and rejected by Lord Greene in Lace [1944] KB 368, 371-372 in these terms:

“[Counsel] argued that the agreement could be construed as an agreement to grant a licence. In my opinion, it is impossible to construe it in that sense. The intention was to create a tenancy and nothing else. The law says that it is bad as a tenancy. The court is not then justified in treating the contract as something different from what the parties intended … . That would be setting up a new bargain which neither of the parties ever intended to enter into.”

So, too, in Prudential [1992] 2 AC 386, it appears that Lord Templeman treated as void a fetter for an indefinite period on the right of the landlord under a periodic tenancy to serve a notice to quit.

It does not seem to me that the observations of Lord Greene, although they are strongly expressed views of a highly reputable judge, can withstand principled analysis. As Lord Templeman made clear in Street [1985] AC 809, while the parties’ rights and obligations are primarily determined by what they have agreed, the legal characterisation of those rights is ultimately a matter of law. If the Agreement is incapable of giving rise to a tenancy for some old and technical rule of property law, I do not see why, as a matter of principle, that should render the Agreement invalid as a matter of contract.

The fact that the parties may have thought they were creating a tenancy is no reason for not holding that they have agreed a contractual licence any more than in Street [1985] AC 809, the fact that the parties clearly intended to create a licence precluded the court from holding that they had, as a matter of law, created a tenancy. So, too, as Mr Wonnacott points out, in Milmo v Carreras [1946] KB 306, the Court of Appeal (led by Lord Greene) held that what was plainly stated and understood by the parties to be an underlease operated as an assignment of the lease as a matter of law, because the duration of the purported underlease equalled or exceeded that of the lease.

Mr Gaunt relies on Street [1985] AC 809 to support another argument, namely that the Agreement could not amount to a licence because it granted the occupier exclusive possession, which is the hallmark of a tenancy. In my view, there is nothing in that argument. The hallmarks of a tenancy include the grant of exclusive possession, but they also include a fixed or periodic term. That was emphasised by Lord Templeman in Street [1985] AC 809 at several points in his judgment, where he referred to a tenancy having to be for “a term of years absolute”, “a fixed or periodic term certain”, or (in a formulation which he approved and adopted) “for a term or from year to year or for a life or lives” – [1985] AC 809, 814E-F, 818E, and 827C and E. Further, as Lord Templeman made clear more than once, the rule that an occupier who enjoys possession is a tenant admits of exceptions, even where the occupier has been granted a fixed or periodic term – see at [1985] AC 809, 818E-F and 823D-E.

It has been suggested (although not in argument before us) that the notion that the Agreement could give rise to a contractual licence if it cannot be a tenancy is somehow inconsistent with the reasoning of the House of Lords in Bruton v London & Quadrant Housing Trust [2000] 1 AC 406. In that case, Lord Hoffmann said that an agreement can give rise to a tenancy even if it does not create “an estate or other proprietary interest which may be binding upon third parties”: p 415. That statement was made in circumstances where a housing trust, which had been granted a licence by a local authority to use a block of flats, agreed that a Mr Bruton could occupy one of the flats. The point being made by Lord Hoffmann was that the fact that the trust was only a licensee, and therefore could not grant a tenancy binding on its licensor, did not prevent the agreement with Mr Bruton amounting to a tenancy as between him and the trust. The tenancy would thus have been binding as such not only on Mr Bruton and the trust, but also on any assignee of Mr Bruton or the trust. Bruton [2000] 1 AC 406 was about relativity of title which is the traditional bedrock of English land law. Lord Hoffmann’s observations in that connection have no bearing on a case where the nature of the agreement is such that it cannot, as a matter of law, be a tenancy even as between the parties.

If the Agreement cannot give rise to a tenancy, then, if it is not a contractual licence, the only right that Ms Berrisford could claim would be that of a periodic tenant on the terms of the written Agreement in so far as they are consistent with a periodic tenancy, because she has been in possession purportedly under the Agreement, paying a weekly rent to Mexfield. It is worth briefly considering why she would be a periodic tenant on this basis, not least because it is Mexfield’s contention that this is the right analysis.

It would be because the law will infer a contract on these terms from the actions of the parties, namely the terms they purported to agree in the Agreement, and Ms Berrisford’s enjoyment of possession and payment of rent. But the ultimate basis for inferring a tenancy (and its terms) is the same as the basis for inferring any contract (and its terms) between two parties, namely what a reasonable observer, knowing what they have communicated to each other, considers that they are likely to have intended. Given that no question of statutory protection could arise, it seems to me far less likely that the parties would have intended a weekly tenancy determinable at any time on one month’s notice than a licence which could only be determined pursuant to clauses 5 and 6.

Since writing this, I have read what Lord Mance and Lord Clarke have written in connection with this point, and I respectfully agree with them. It is also interesting to read Lord Hope’s judgment, which demonstrates that the Scottish courts have also encountered difficulty when grappling with interests of uncertain duration, and seem to have come up with a similar answer.

That leaves Mr Wonnacott’s further alternative argument, namely that, if Mexfield is right and there is a periodic tenancy, then, even if the fetter on the landlord’s right to serve a notice to quit is objectionable in landlord and tenant law, it can be enforced as between the original parties as a matter of contract. That was the basis on which Wilson LJ felt able to find for Ms Berrisford in the Court of Appeal. I prefer to say nothing about that point: I have already dealt with one alternative reason for allowing this appeal, so considering this argument would involve making two successive counter-factual assumptions, rarely a satisfactory basis for deciding a point of law.

In these circumstances, I would allow Ms Berrisford’s appeal, and discharge the order made against her. It is only right to repeat that the Court of Appeal and Peter Smith J were bound by authority which made it impossible for them to reach the same conclusion as I have done on the points on which I would allow the appeal.

For the reasons given by Lord Neuberger, with which I entirely and respectfully agree, I too would allow this appeal. I wish to add just a few words about the position in Scotland, as there are significant differences between the way English and Scots law treat agreements of the kind that are in issue in this case.

The first difference relates to the status of Mexfield as compared with the status that a similar body has in Scotland. It is a fully mutual housing association within the meaning of section 1(2) of the Housing Associations Act 1985 and section 5(2) of the Housing Act 1985. It cannot create an assured tenancy in England: section 1(2) of and paragraph 12(1)(d) of Schedule 1 to the Housing Act 1988. Nor can it create a secure tenancy there, because it is registered under the Industrial and Provident Societies Act 1965. A housing association is not a landlord for the purpose of creating a protected or statutory tenancy: Rent Act 1977, sections 15(1) and 15(3). So its members have no statutory protection except that which is given to them by the Protection from Eviction Act 1977.

In Scotland a fully mutual co-operative housing association which meets the conditions for registration set out in sections 58 and 59 of the Housing (Scotland) Act 2001 is eligible for registration as a social landlord under Part 2 of that Act. Mexfield would meet these criteria if it was providing housing services in Scotland, as they extend to bodies established for the purpose of providing houses for occupation by members of that body where the rules restrict membership to persons entitled to occupy a house provided by that body: section 58(2)(b). It is the normal practice for eligible bodies to apply for registration. There were 211 registered social landlords in Scotland in 2011: see the annual report of the Scottish Housing Regulator. Where the landlord is a registered social landlord which is a co-operative housing association and the tenant is a member of the association the tenancy is a Scottish secure tenancy: section 11(1)(d). The Protection from Eviction Act 1977 does not extend to Scotland: section 13(3). But Ms Berrisford would have had statutory security of tenure if the house which she is occupying under the Occupancy Agreement was in Scotland and her landlord had been registered under the 2001 Act.

The other difference relates to the way the common law treats such agreements if statutory security of tenure is not available. The starting point in Scots law is that a lease is a contract which gives the tenant a personal right to the subjects: Gordon, Scottish Land Law, 3rd ed (2009), vol 1, para 18-136. A right of that kind is capable of being created simply by agreement between the parties. The original parties to the contract are free to regulate the arrangements that are to apply between themselves as they wish. They will be held to the terms of their contract so long as the original proprietor of the premises retains ownership of the property and the original tenant remains in occupation of it.

The situation becomes more complicated where the ownership of the premises passes to someone else. In that event the tenant will need to have acquired a real, or proprietary, right if he wishes to enforce his exclusive right to remain there against the new owner. The circumstances in which the lease will confer on the tenant a real right were set out in the Leases Act 1449, which remains in force: Gordon, para 18-137. To obtain the benefit of the statute, which refers to the takers of the lands having “terms and years thereof”, there must be a term when it is to come to an end. The lease must be for a definite duration, such as a number of years or the lifetime of the grantor or the grantee: Hunter on Landlord and Tenant, 4th ed (1876), vol 1, p 461. A real right may also be obtained by registering the lease in the Land Register of Scotland if it is to endure for 20 years or more under the Registration of Leases (Scotland) Act 1857, as amended by the Land Tenure Reform (Scotland) Act 1974, section 18 and Schedule 6. But the grant of a lease which could extend to more than 20 years is prohibited in the case of property which is to be used as or as part of a private dwelling-house: 1974 Act, section 8(1).

In Carruthers v Irvine 1717 Mor 15195 the period for which the lease was granted was expressed by the words “perpetually and continually as long as the grass groweth up and the water runneth down”. The grantor died and his heir sought to remove the tenant on the ground that the lease did not say when it was to come to an end. His claim failed because the court found that “by the meaning of [the] parties the contract was intended to be a perpetual right to the tenant and his successors”. This did not meet the requirements of the 1449 Act, and it was admitted that the tenant would not have been able to enjoy that right in a question with a singular successor of the grantor: see Hunter, p 462. But the personal right against the heir under the contract was not affected. In Crighton v Lord Air 1631 Mor 11182 the grant was to the tenant and his heirs and successors for five years and after that a further five years and then five years for ever. The argument that the lease was a nullity because it did not say when it was to come to an end was repelled. It was noted that the grantor might have objected on this ground in question with a singular successor of the grantee. But it was held that he could not do so in a question with the grantee’s heirs, as he had bound himself by the words of the grant never to remove the grantee’s heirs.

These cases show that as between the original parties a lease may be granted for an indefinite period: Rankine, The Law of Leases in Scotland, 3rd ed (1916), p 115; Gordon, para 18-11. The grantee will obtain security of tenure in a question with singular successors of the landlord if the lease sets out the term when it is to come to an end. If no term of endurance at all is specified in the agreement, or it is for an uncertain term or it is potentially perpetual, it will not meet the requirements of the Act. But this does not mean that effect cannot be given to the personal obligation. If the issue arises as between the original parties to the agreement and the agreement does not provide for this, a term after which either party can bring it to an end will be implied by law. In Redpath v White 1737 Mor 15196 there was no such term, but the court accepted the argument that it was open to it to fix the time of endurance. The starting point is that a lease of that kind is regarded as good for one year only. But if there are words in the lease which show that it was the intention of the parties that it should continue for more than one year, the court will select the minimum period that the words admit of: Erskine, Principles of the Law of Scotland, 21st ed (1911), II, vi, 10; Rankine, p 115. As there were clauses in the lease which showed that it was intended to last for more than a year, the court in Redpath substituted a period of two years.

Where the term of the lease expires and the grantee remains in occupation without the parties having entered into a new agreement its terms are prolonged from year to year, or a shorter period if that is what the lease indicates, under tacit relocation: Rankine, p 602; Gordon, para 18.25. Once it is under tacit relocation it is open to either party to bring the tenancy to an end by notice to quit. The right to do so is implied by law, so any term of the contract which is inconsistent with tacit relocation is prima facie unenforceable: Gloag on Contract, 2nd ed (1929), p 733. But tacit relocation is excluded as between the original parties to the lease if the parties make a bargain as to the terms on which the tenant is to stay on in occupation of the premises: Buchanan v Harris & Sheldon (1900) 2 F 935, 939, per Lord Adam. It has also been held to be excluded if the parties to the lease provide expressly by their contract that tacit relocation is not to apply to it: MacDougall v Guidi 1992 SCLR 167; see also Stair Memorial Encyclopaedia, Landlord and Tenant, para 364.

Applying these various points to this case, clause 1 of the Occupancy Agreement states that it will continue from month to month until determined as provided in the agreement. This is an indefinite period, so the agreement is not capable of conferring on the tenant a real right under the Leases Act 1449. But there would be no need in Scots law for the court to imply any period in place of what is provided for in clause 1. This is because the dispute in this case is between the original parties to the agreement and because the circumstances in which it may be brought to an end are sufficiently set out in the contract. I agree with Lord Neuberger’s analysis of what, as a matter of contractual interpretation, is its effect: see para 22, above. Clause 5 provides that the agreement is determinable by the member on giving the Association one month’s notice in writing. Clause 6 provides that it may be brought to an end by the Association but only in the circumstances that that clause sets out. The possibility of the Association bringing the agreement to an end by serving one month’s notice in writing is excluded by the terms of the agreement. The question whether Ms Berrisford’s agreement with Mexfield, which has now endured for more than 20 years, has the effect of excluding the implied right of the landlord to terminate under the rules relating to tacit relocation is not easy to answer. But, as she would have had security of tenure under the statute, it is not one that would need to be addressed if her landlord had been registered under the 2001 Act.

I have to confess that I have found it difficult to understand why English law finds it so difficult to hold that, if an agreement of this kind cannot for technical reasons take effect as a tenancy, it can be regarded as binding on the parties simply by force of contract. I appreciate the problems that would need to be faced if it was necessary for the agreement to have proprietary effect, which it would if the dispute had not been between the original contracting parties. As it is, however, the essence of the dispute between the parties in this case seems to me to be about the effect of the contract which they entered into. One might have expected it to be capable of being solved by applying the ordinary principles of the law of contract without having to resolve questions about the effect of the agreement on the parties’ proprietary interests or what the agreement is to be called. But I entirely understand that the contrary view is supported by a very substantial body of authority. It can by no means be lightly brushed aside, and I am persuaded that, for all the reasons that Lord Neuberger gives, it would not be appropriate for us to consider changing the law as to what constitutes what English law will hold to be a tenancy, at least in this case.

I also wonder whether the time has not now come for the legislative fetter which prevents mutual housing associations from granting protected or statutory tenancies in England and Wales to be removed, so that they are placed on the same footing as other providers of social housing as in Scotland. The reason that was given by the Minister of State in the Department of the Environment, the Earl of Caithness, for introducing an amendment to the Bill which became the Housing Act 1988 that provided that a fully mutual housing association cannot create an assured tenancy was that a statutory regime designed to regulate the relationship between landlord and tenant had little relevance in a situation “where, as is the nature of a co-operative, the interests of landlord and tenants as a whole are in effect indivisible”: Hansard (HL Debates), 3 November 1988, vol 501, col 395. That statement was repeated in the House of Commons by the Parliamentary Under Secretary of State, David Trippier, when the Lords amendment was approved: Hansard (HC Debates), 9 November 1988, vol 140, col 337. The facts of this case suggest that, as least so far as Mexfield is concerned, that happy state of affairs no longer exists. The assumption on which that measure was put through Parliament seems now to rest on doubtful foundations, as financial pressures may cause the parties’ interests to diverge to the detriment of the residential occupier. That is not something that this court can deal with. But I suggest that it might be considered in any future programme for the reform of housing law.

I respectfully concur in the cogent and comprehensive judgment of Lord Neuberger MR. I add a few words of my own as to the decision of the Court of Appeal in Bass Holdings Ltd v Lewis [1986] 2 EGLR 40. In that case the Court of Appeal held that the word “determinable”, as used in section 149(6) of the Law of Property Act 1925, means “liable to terminate automatically” on a person’s death, or some other uncertain event, rather than “capable of being terminated by notice” given after a person’s death, or some other uncertain event. With the exception of the case mentioned in the last paragraph of this judgment, Bass seems to be the only reported case that has made more than a passing reference to section 149(6).

The case was concerned with a standard-form tenancy agreement made between a brewery company and the licensee of a public house in Deptford. The tenancy was for three years and then continued as a tenancy from year to year, subject to provisions for termination set out in a schedule. Para 1 allowed the landlord to terminate the tenancy on six months’ notice. Para 4 provided that if the tenant died during the term the landlord could terminate the tenancy on 14 days’ notice, or three months if the tenant left a widow. This shorter notice was no doubt thought appropriate because of the exigencies of the on-licensed trade. The brief report does not state what notice the tenant had to give, but it seems inconceivable that that was not covered in another paragraph of the schedule. The issue was whether section 149(6) applied to the tenancy. Nourse LJ gave the main judgment, with which Glidewell LJ and Sir John Donaldson MR agreed. The court upheld the first-instance decision of Hoffmann J that section 149(6) did not apply.

“Determinable” can, in the vocabulary of the law, have either of the meanings mentioned at the beginning of this judgment. The three reasons given by Nourse LJ for preferring the narrower meaning are to my mind convincing. But I think there is a more powerful reason based not on the language of section 149(6) but on its purpose. It was intended to enable a commercial lease for life to exist as a legal estate under the new regime introduced by the Law of Property Act 1925. It was not intended to apply to leases or tenancies which did not need that sort of helping hand. Mr Lewis’s tenancy in Bass was well able to stand on its own feet as a legal estate without being converted into a term of 90 years determinable by notice after Mr Lewis’s death while still licensee of the public house. That point is, I think, reflected in the short concurring judgment of Sir John Donaldson MR.

Lord Neuberger did not find it necessary to refer to Bass for reasons mentioned in para 47 of his judgment. I agree with his reasoning, and also with what Lord Neuberger goes on to say in para 48. Even if section 149(6) were supposed to have applied to the tenancy in Bass, there would be no good reason to make a fundamental change in its commercial effect by disregarding the landlord’s wider power to terminate on six months’ notice.

I echo Lord Neuberger’s tribute to Mr Wonnacott’s clear and scholarly submissions. He did not refer to what seems to be the only other case in which section 149(6) has been considered by the Court of Appeal, that is Skipton Building Society v Clayton (1993) 66 P & CR 223. The facts are of some interest, illustrating the difficulties that can arise when a husband regularly forges his wife’s signature. The main point of law is whether the sale of a flat at a reduced price, with the retention of a rent-free licence for life, constituted a fine (that is, a premium) for the purposes of section 149(6). The Court of Appeal held that it did. There is nothing in the judgment of Slade LJ that conflicts with any of the reasoning in Lord Neuberger’s judgment.

Periodic tenancies obviously pose something of a puzzle if the law insists that the maximum term of any leasehold estate be certain. The rule was invented long before periodic tenancies were invented and it has always been a problem how the rule is to apply to them. In one sense the term is certain, as it comes to an end when the week, the month, the quarter or the year for which it has been granted comes to an end. But that is not the practical reality, as the law assumes a re-letting (or the extension of the term) at the end of each period, unless one or other of the parties gives notice to quit. So the actual maximum term is completely uncertain. But the theory is that, as long as each party is free to give that notice whenever they want, the legal maximum remains certain. Uncertainty is introduced if either party is forbidden to give that notice except in circumstances which may never arise. Then no-one knows how long the term may last and indeed it may last for ever.

These rules have an Alice in Wonderland quality which makes it unsurprising that distinguished judges have sometimes had difficulty with them. It is intriguing to read, in Doe d Warner v Browne (1807) 8 East 165, 167, that Lord Mansfield had once “thrown out” (obviously meaning “suggested”) the “notion of a tenancy from year to year, the lessor binding himself not to give notice to quit”. By that date the notion cannot have been “exploded” for very long. More recently, in Breams Property Investment Co Ltd v Stroulger [1948] 2 KB 1, the Court of Appeal held that it was not repugnant to the notion of a quarterly tenancy when the landlords promised that they would not terminate it within the first three years unless they required the premises for their own occupation, so the purchasers of the reversion could not give notice to quit until the three years were up. Breams Property can, however, be explained on the basis that although phrased as a quarterly tenancy with a restriction on the landlord’s right to serve notice to quit, in effect it simply turned the quarterly tenancy into a three-year term terminable by the tenant on notice before that, to be followed by a normal quarterly tenancy after that.

However, in In re Midland Railway Co’s Agreement, Charles Clay & Sons Ltd v British Railways Board [1971] Ch 725, a strong Court of Appeal, in a reserved judgment of the Court, went much further and upheld a term in a half-yearly tenancy which prohibited the landlords from serving notice to quit unless they required the property for their own undertaking. The Court accepted that the maximum term of a single term of years had to be certain before the lease took effect. Thus a letting “for the duration of the war”, as in Lace v Chantler [1944] KB 368, was invalid. This rule was so long established that it was now “too late to inquire why this aspect of the particular estate was considered essential to its existence or to question the doctrine”: [1971] Ch 725, 731-732. But it could be easily circumvented by granting a lease for (say) 90 years, terminable earlier than that should the uncertain event happen. So the rule “has an air of artificiality, of remoteness from practical considerations”: p 732. (Had Mr Wonnacott been around then forcibly to remind their Lordships that in fact such a lease to an individual was treated at common law as a lease for life, provided that the necessary formalities were complied with, and had therefore been converted by section 149(6) of the Law of Property Act 1925 into a lease for 90 years, their Lordships might have expressed themselves even more strongly.) In any event, they held that the rule only applied to an attempt to grant a lease or tenancy for a single and uncertain period. It was not repugnant to the nature of a periodic tenancy to place a curb on the landlord’s power to determine it, unless perhaps there was an attempt to prevent the landlord from ever doing so. They could not see any distinction between this case and the curb of limited duration in Breams Property.

In Ashburn Anstalt v Arnold [1989] Ch 1, the Court of Appeal went even further and upheld as a tenancy a grant of what was described as a licence to occupy premises rent free which the landlords could only determine if they certified that they were ready immediately at the end of the quarter’s notice period to demolish and redevelop the property. The Court considered that the vice of uncertainty was that neither party knew where they stood and the court did not know what to enforce. In this case there would be no doubt about whether the determining event had occurred. The parties should be held to their agreement even though this might not be a periodical tenancy as in the Midland Railway case.

Both the Midland Railway and Ashburn Anstalt cases were overruled by the House of Lords in Prudential Assurance Co Ltd v London Residuary Body [1992] 2 AC 386. There Mr Nathan had sold a strip of land between his shop and the road to the London County Council, which had let it back to him at a weekly rent, by an agreement which provided that the “tenancy shall continue until the . . . land is required by the council for the purposes of the widening of” the highway. His successors in title sought to enforce this restriction against the county council’s successors in title and the House of Lords held that they could not do so. It is understandable that the House of Lords should have taken the view that this was in effect a single term of uncertain, indeed potentially perpetual, duration and thus incompatible with the long-established rule of the common law against terms of uncertain duration. It is understandable that the House should have taken the same view of the Ashburn Anstalt case and overruled it. It was, as it seems to me, unnecessary for them to hold that a similar curb in an otherwise conventional periodical tenancy was similarly repugnant and thus to overrule Midland Railway but this is what they did.

Their Lordships were not invited to consider the argument which Mr Wonnacott has advanced in this case. This is not surprising, as it would have done the company which had bought the land from Mr Nathan no good. Mr Nathan was almost certainly already dead (and companies cannot have a lease for their own lives). Indeed, Mr Wonnacott’s argument does not appear previously to have been made in a case which concerns what would otherwise bear all the hallmarks of a periodic tenancy with a curb on the landlord’s power to determine it.

So we have now reached a position which is curiouser and curiouser. There is a rule against uncertainty which applies both to single terms of uncertain duration and to periodic tenancies with a curb on the power of either party to serve a notice to quit unless and until uncertain events occur. But this rule does not matter if the tenant is an individual, because the common law would have automatically turned the uncertain term into a tenancy for life, provided that the necessary formalities were complied with, before the Law of Property Act 1925. A tenancy for life was permissible at common law, although of course it was quite uncertain when the event would happen, but it was certain that it would. I suppose at the time of the hundred years’ war there was uncertainty both as to the “when” and the “whether” it would ever end. Be that as it may, a tenancy for life is converted into a 90 year lease by the 1925 Act.

As it happens, in the particular agreement with which we are concerned, it is not difficult to conclude that the parties did in fact intend a lease for life determinable earlier by the tenant on one month’s notice and by the landlords on the happening of certain specified events. So our conclusions are in fact reflecting the intentions of the parties. But it is not difficult to imagine circumstances in which the same analysis would apply but be very far from the intentions of the parties. And that analysis is not available where the tenant is a company or corporation. So there the court is unable to give effect to the undoubted intentions of the parties. Yet, as the Court pointed out in Midland Railway, it is always open to the parties to give effect to those intentions by granting a very long term of years, determinable earlier on the happening of the uncertain event. The law, it would seem, has no policy objection to such an arrangement, so it is difficult to see what policy objection it can have to upholding the arrangement to which the parties in fact came.

It is even more bizarre that, had the “tenancy for life” analysis not been available, the conclusion might have been, not that this was a contractual tenancy enforceable as such as between the original parties, but that it was a contractual licence, also enforceable as such between the original parties. This, as I understand it, is the difference between English and Scots law. I do not understand that it makes any difference to the result.

As will be apparent, I entirely agree with the reasoning and conclusions reached by Lord Neuberger on the first question: does Ms Berrisford have a subsisting tenancy? For that reason, I do not think it necessary to express an opinion on the alternative case in contract. But it seems to me obvious that the consequence of our having reached the conclusions which we have on the first issue is to make the reconsideration of the decision in Prudential, whether by this Court or by Parliament, a matter of some urgency. As former Law Commissioner Stuart Bridge has argued ([2010] Conv 492, 497):

“If the parties to a periodic tenancy know where they stand, in the sense that the contract between them is sufficiently certain, then that should be enough. If a landlord, in this case a fully mutual housing association, decides that its tenants should be entitled to remain in possession unless and until they fall into arrears with their rent or break other provisions contained in the tenancy agreement, it is difficult to see what policy objectives are being furthered in denying the tenant the rights that the agreement seeks to create.”

Quite so.

I too agree with Lord Neuberger’s comprehensive judgment, including his tribute to Mr Wonnacott’s exposition of the law. For the present, I (like Lord Neuberger: paras 33-35) proceed on the basis that an essential characteristic of a contractual tenancy is lacking if the contract provides for a series of periods indefinitely renewable unless and until some future event occurs which may never occur.

The decision in this appeal is therefore that the present Agreement between the parties, being for a term uncertain at inception, would have been treated until 1925 as involving the grant of a tenancy for Ms Berrisford’s life, and falls now, under the Law of Property Act 1925, section 149(6), to be treated as a tenancy for a term of 90 years determinable on Ms Berrisford’s death.

But for this conclusion, the Court would have had to consider the effect of an Agreement which was on its face for an uncertain term, but incapable in law of taking effect as such. Mr Gaunt QC for Mexfield submits that such an Agreement must take effect as a tenancy, but of a wholly different kind to any which the parties intended, in that it would be terminable at any time by Mexfield on a month’s notice. There is in my opinion neither attraction in nor need to accept that submission.

The Agreement was to run “from month to month until determined as provided in this Agreement” – an obvious reference to clauses 5 and 6, containing the parties’ agreement on the only permissible methods of contractual determination. To treat the Agreement as one under which Mexfield could terminate at the end of any month on a month’s notice would undermine an evidently fundamental element in the parties’ contract. I agree with Lord Clarke that ordinary principles of construction govern the Agreement’s true construction. I note that Lord Hobhouse made a similar point in relation to the agreement, held to be a tenancy, in Bruton v London & Quadrant Housing Trust [2000] 1 AC 406, 417; he cited in that connection Reardon-Smith Line Ltd v Yngvar Hansen-Tangen (The “Diana Prosperity”) [1976] 1 WLR 989 - a precursor, famous in commercial law, of the Investors Compensation and Mannai Investment cases.

The three characteristic hallmarks of a contractual tenancy, as distinct from a contractual licence, are (a) exclusive occupation, (b) rent and (c) a term which the law regards as certain: Street v Mountford [1985] AC 809, esp. 826E-F, per Lord Templeman. That case

“is authority for the proposition that a ‘lease’ or ‘tenancy’ is a contractually binding agreement, not referable to any other relationship between the parties, by which one person gives another the right to exclusive occupation of land for a fixed or renewable period or periods of time, usually in return for a periodic payment in money. An agreement having these characteristics creates a relationship of landlord and tenant to which the common law or statute may then attach various incidents”.

See Bruton v London & Quadrant Housing Trust [2000] 1 AC 406, 413E, per Lord Hoffmann. Only in “special circumstances” (not here relevant) will an agreement having these characteristics not involve a tenancy: see Street v Mountford, p 822B and Bruton, pp 410E, 411C-412B, 414-B-G and 417A-B.

On the hypothesis I am presently considering, those three characteristics were not all present. The basis for asserting that there was a contractual tenancy therefore falls away. But the contract was valid as such. There is no reason not to give it effect according to its terms. As a matter of legal categorisation, because it was not a tenancy, it can only involve a licence. Its terms precluded the giving by Mexfield of notice to terminate, except in circumstances falling within clause 6 of the Agreement.

To force the contract into the category of tenancy, by rewriting its essential terms to provide for a periodic monthly tenancy terminable on a month’s notice, would be to substitute for the Agreement that the parties have made a wholly different contract. It would be to treat the first two of the three characteristics of a tenancy mentioned above as sufficient by themselves and as displacing any need to satisfy the third. It would be to insist on terminology (such as the Agreement’s references to letting and taking possession “from month to month” and “this Tenancy”) over substance (the parties’ express limitation of the right to terminate and the consequent absence of an essential characteristic of a tenancy).

Like Lord Neuberger (para 69), I reserve my view on the position upon the hypothesis of a contract constituting a tenancy, but which was both subject to provisions restricting termination for an uncertain period and not capable of being treated as a tenancy for life at common law or a tenancy for 90 years under section 149(6) of the 1925 Act. In the light of what I have already said and on the law as it stands, this is an impossible hypothesis, since such a contract could not give rise to what the law would regard as a tenancy; it could however take effect between the parties according to its terms, although it would not have proprietary effect as against third parties: see paras 101-102 above. Parties can normally contract as they will, either inter se or indeed with third parties.

I agree that this appeal should be allowed. It seems remarkable to me that it is necessary to decide this appeal in 2011 by reference to jurisprudence developed over the centuries to the effect that an agreement for an uncertain term was treated as a tenancy for the life of the tenant, determinable before the tenant’s death according to its terms. It is a mystery to me why in 2011 the position of a tenant who is a human being and a tenant which is a company should in this respect be different. There is in my opinion much to be said for the view that the certainty rule should now be abandoned. However, I agree with Lord Neuberger that it is not necessary to abandon it for the purposes of deciding this appeal.

I can understand why in this appeal Mexfield sought to abandon its concession that, on the true construction of the contract, which is called “the Agreement”, it could not serve a notice to quit under the contract and that the only way the contract could come to an end was by Ms Berrisford serving a notice under clause 5 or by Mexfield exercising its rights under clause 6. One might have thought that, if the contract was not brought to an end in one of those ways it would continue, at least as a contract which governed the rights and obligations of the parties to it. Moreover, it would do so even if, by some quirk of the law, the parties had failed to create a tenancy. It was thus of some importance for Mexfield to abandon its concession.

As I see it, the ordinary principles governing the true construction of a contract apply to tenancy agreements and leases. The principles have been discussed in many cases, notably of course, as Lord Neuberger MR said in Pink Floyd Music Ltd v EMI Records Ltd [2010] EWCA Civ 1429, para 17, by Lord Hoffmann in Mannai Investment Co Ltd v Eagle Star Life Assurance Co Ltd [1997] AC 749, passim, in Investors Compensation Scheme Ltd v West Bromwich Building Society [1998] 1 WLR 896, 912F-913G and in Chartbrook Ltd v Persimmon Homes Ltd [2009] AC 1101, paras 21-26. I agree with Lord Neuberger (also at para 17) that those cases show that the ultimate aim of interpreting a provision in a contract is to determine what the parties meant by the language used, which involves ascertaining what a reasonable person would have understood the parties to have meant. As Lord Hoffmann made clear in the first of the principles he summarised in the Investors Compensation Scheme case at p 912H, the relevant reasonable person is one who has all the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time of the contract.

Here the language used is simple and straightforward. I agree with Lord Neuberger for the reasons he has given that the contract makes it clear that the two ways in which the contract could be brought to an end under the contract were those described in clauses 5 and 6. Since the contract has not been brought to an end in either of those ways, it must, at any rate for the purposes of the law of contract, remain on foot. Even if the contract somehow failed to create a tenancy so that Mrs Berrisford does not have an estate in the property, there is in my opinion no principle of the law of contract which prevents the terms of the contract having effect as between the parties to it.

In this regard I agree with the views of Lord Neuberger expressed at paras 57 to 64 above. Ms Berrisford has been living in the property for a considerable time and, except for a short period referred to by Lord Neuberger at para 7, has been paying rent at the rate provided for in the contract. It would to my mind be bizarre for the law to imply or infer a contract between the parties to the effect that there was a periodic tenancy between them at the contractual rate. That would mean that Mexfield can bring the contract to an end by giving one month’s notice to quit. I see no basis for such an inference or implication. It would be contrary to the express terms of the agreement, namely that the only way that Mexfield can determine it is under clause 6. There is no need for any process of implication or inference because the parties have expressly agreed the position. In particular, Ms Berrisford at no time agreed that Mexfield could give her a month’s notice to quit. If, as a matter of law, the parties have created a licence and not a tenancy, so be it.

I appreciate that the point just discussed is Ms Berrisford’s alternative case and is discussed as such by Lord Neuberger. I take it first because it seems to me to be of critical importance to ascertain the contractual position between the parties. It follows that, as I see it, even if the contract does not create a tenancy, it creates rights and obligations between the parties, so that in an appropriate case Ms Berrisford could in principle obtain an injunction against Mexfield for a threatened breach of contract. In the meantime the contract remains on foot.

On the question whether there is a valid tenancy between the parties, with some misgiving I shall assume that Mr Wonnacott’s concession that the contract is not capable, as a matter of law, of creating a tenancy in accordance with its terms because it is an agreement for an uncertain term is correct. However, I agree with Lord Neuberger for the reasons he gives at paras 36 to 42 that, at any rate before 1926, the arrangement between the parties would have been treated as a tenancy or lease for life, determinable before her death in accordance with its terms. I further agree with him (at paras 43 to 54) that the effect of section 149(6) of the Law of Property Act, which applies to arrangements made before and after the 1925 Act came into force, is that the lease for life is converted to a lease or contract for a term of 90 years determinable on Ms Berrisford’s death.

For these reasons, which are essentially the same as those given by Lord Neuberger, I agree with him that the appeal should be allowed. I am pleased to be able to arrive at this conclusion because any other conclusion would be contrary to the agreement freely entered into by the parties and, in particular, would be most unjust to Ms Berrisford.

I agree that this appeal should be allowed for the reasons stated by Lord Neuberger in his comprehensive judgment. My views can be summarised quite shortly. The starting point must be the true construction of the Agreement applying ordinary principles of contractual interpretation. In this respect, an agreement which purports to create a tenancy agreement is no different from any other agreement. For the reasons given by Lord Neuberger, I have no hesitation in concluding that, as a matter of construction, the Agreement provides that Mexfield can only determine it on one of the grounds specified in clause 6: it cannot determine it by simply serving one month’s notice to quit.

The next point is that there is nothing in the general law of contract which makes such an agreement void or unenforceable according to its terms. But this is not an ordinary contract. On its face, it purports to be a tenancy agreement ie an agreement which purports to grant an interest in land. It is notorious that the law of landlord and tenant is highly technical, not least because its development has been affected by rules of law of ancient origin. It is, therefore, necessary to turn to the law of landlord and tenant to see whether the fact that the Agreement purports to create a tenancy determinable by Mexfield, but only on the grounds specified in clause 6, requires a different approach to be adopted to the ascertainment of its meaning and effect.

As Lord Neuberger explains (paras 24 to 33), it seems to have been long established that an agreement for an uncertain term cannot be a tenancy in the sense of being a term of years; and a fetter on a right to serve a notice to determine a periodic tenancy is ineffective if the fetter is of uncertain duration. Such a fetter is “repugnant” to a periodic tenancy: see Doe d Warner v Browne (1807) 8 East 165, 166. As Lord Browne-Wilkinson said in Prudential Assurance Co Ltd v London Residuary Body [1992] 2 AC 386, 396, this “bizarre” outcome results from an application of “an ancient and technical rule of law” for the genesis of which there is no satisfactory rationale and for which there is no apparent useful purpose. Despite the concern expressed by Lord Browne-Wilkinson that to depart from a rule relating to land law which has been established for many centuries might upset long established titles, there is much to be said in favour of getting rid of the rule. But I think that, rather than the court introducing a change to such a fundamental tenet of the law of landlord and tenant, it would be better if this were done by Parliament after full consultation of interested parties of the kind that is routinely undertaken by the Law Commission.

At all events, as a result of Mr Wonnacott’s impressive and scholarly research (which was not placed before the Court of Appeal), it is clear that it is unnecessary to get rid of the uncertainty rule in this case. This is because before the enactment of the Law of Property Act 1925 (“the 1925 Act”), the tenancy purportedly created by the Agreement would have been treated as a tenancy for life, defeasible by determination on any of the grounds specified in clauses 5 and 6. Lord Neuberger has referred to some of the pre-1926 authorities at paras 37 to 39. The position is well summarised in the last edition of the standard work on land law before the 1925 legislation, Joshua Williams’ Law of Real Property, 23rd ed (1920), p 135 in these terms:

“Where land is given to a widow during her widowhood, or to a man until he shall become bankrupt, or for any other definite period of time of uncertain duration, a freehold estate is conferred, as in the case of a gift for life. Such estates are regarded in law as determinable life estates…” (emphasis added).

Accordingly, a periodic tenancy determinable on an uncertain event was treated as a defeasible tenancy for life. In disputing this proposition, Mr Gaunt’s principal submission was that, before the enactment of the 1925 Act, the question whether a periodic tenancy determinable on an uncertain event was a defeasible tenancy for life was one of construction of the particular agreement. But, as Lord Neuberger explains, it is clear from the authorities that this is incorrect. It was a rule of the common law that such a tenancy was automatically treated as a tenancy for life. It had nothing to do with the intention of the parties.

The effect of section 149(6) of the 1925 Act was to convert such a tenancy into a term for 90 years, subject to earlier termination in accordance with its terms. It follows that the Agreement is such a tenancy and all the terms of clause 6 apply with full force and effect. Mexfield cannot terminate the Agreement by serving a notice to quit as if this were a simple monthly tenancy without more.

This is a just result which plainly accords with the intention of the parties. But it may legitimately be said that it is not satisfactory in the 21st century to have to adopt this chain of reasoning in order to arrive at such a result. It is highly technical. There should be no need to have to resort to such reasoning in order to arrive at the result which the parties intended. That is why the radical solution of doing away with the uncertainty rule altogether is so attractive. There is the further point that the section 149(6) route to the right result can only be followed where the purported tenant is an individual and not a corporate entity. To treat an individual and a corporate entity differently in this respect can only be explained on historical grounds. The explanation may lie in the realms of history, but that hardly provides a compelling justification for maintaining the distinction today.

To conclude, in my view the answer to this appeal lies in the law of landlord and tenant and the appeal must be allowed. I do not find it necessary to address the alternative arguments advanced by Mr Wonnacott. I would, however, go so far as to say that, like Lord Neuberger (paras 57 to 62), I am strongly attracted by the submission that, if by reason of the uncertainty argument the Agreement did not create a tenancy, then it was enforceable as a contract according to its terms like any other contract.