Freely Available British and Irish Public Legal Information
[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
United Kingdom Journals
You are here: BAILII >> Databases >> United Kingdom Journals >> Occupation for Life: Satisfying the Equity
URL: http://www.bailii.org/uk/other/journals/WebJCLI/1995/issue5/todd5.html
Cite as: Occupation for Life: Satisfying the Equity
[New search] [Printable RTF version] [Help]
Occupation for Life: Satisfying the Equity
Nicola Glover LLB
University of Manchester,< mllxeneg@fs1.ec.man.ac.uk>
and
Paul Todd MA BCL
Cardiff Law School,< slapnt@cf.ac.uk>
Copyright © 1995 Nicola Glover and Paul Todd.
First Published in Web Journal of Current Legal Issues in association with Blackstone Press Ltd.
Summary
Suppose that A, who has legal title to a house, represents to B that B will be allowed to occupy the house for his or her life, and B relies on that representation to his or her detriment. There is no problem, on the principles enunciated by Lord Bridge in Lloyds Bank plc v Rosset [1991] 1 AC 107, eg at p 132, in granting B an interest, by way of constructive trust or proprietary estoppel. When it comes to quantifying the interest, however, the courts are prevented from reaching an equitable solution by legislation which had its origins over a century ago, and was intended to deal with an entirely different problem. To judge from the number of recent cases, the situation must be quite a common one.
Web JCLI | [1995] 5 Web JCLI | Download this file.
Contents
- Introduction
- The Settled Land Act 1925
- Life interest under a trust.
- A personal right.
- Comparing trusts and estoppels.
- The constructive trust.
- Estoppel.
- Conclusion on trust and estoppel.
- The Settled Land Act 1925
Introduction
The obvious solution to the problem, at any rate where B's detrimental reliance is substantial (1), would be to prevent A from reneging on his or her representation, by granting B the right to occupy the property for his or her life. A should not be able to avoid the obligation by selling the property, and B's interest should be capable of binding a third party, with notice in the case of unregistered land, or as an overriding interest under s.70(1)(g) of the Land Registration Act 1925, if B is in actual occupation. Yet recent decisions, culminating with that of the Court of Appeal in Costello v Costello (1994) 27 HLR 12, suggest that this route is not available. Either B is restricted to a personal right, not binding on third parties, or B will have all the powers granted to a tenant for life under the Settled Land Act 1925, powers which are far greater than were ever envisaged by either party. B either gets too little, or too much. The Settled Land Act 1925 prevents the courts from giving effect exactly to A's representation.The Settled Land Act 1925
It seems most unlikely that the draftsmen of the 1925 legislation intended to create such difficulties. They are rather an accidental consequence of legislation, which originated with the Settled Land Act 1882, whose purpose had nothing to do with the problem under discussion; it was enacted to curtail the strict settlement, which had developed as a means of keeping landed estates within the same family, but was giving rise to difficulties in the latter part of the nineteenth century.
The detailed mechanism of the strict settlement is beyond the scope of this article (but see generally Harvey 1973). Essentially, its purpose was, by re-settling the land every generation, to ensure that the person in possession of the land was always a life tenant (Harvey 1973, pp 20-24). A tenant in tail in possession could bar the entail, creating a saleable fee simple interest. Obviously, this would have defeated the purpose of the settlement, so it was essential to ensure that the tenant in tail never got possession.
The problem was that, in the absence of express provision in the settlement, the powers of the life tenant in possession were extremely limited. Not only was he (as intended) unable to sell any part of the land, but (because the length of his interest was uncertain) he was also unable to lease it, or (because of the doctrine of waste) to develop the land, for example by quarrying and tree-felling (Burn 1994, pp 269- 273). By the latter part of the nineteenth century this was creating difficulties, stultifying land use at a time when agricultural land values were declining, and when it would often have been more profitable to turn some or all of the land over to industrial use.
The cure for the problem was legislation. The Settled Land Act 1882, amended in small particulars in 1884, 1887, 1889 and 1890, gave extensive powers, including sale of the fee simple, to the tenant for life in possession, whatever the provisions of the settlement (for a summary, and comparison between the 1882-90 and the 1925 legislation, see Harvey 1973, pp 62-67). If the land was sold, all interests in it were overreached by the purchasers, as long as the purchase money was paid to trustees of the settlement.
The 1925 legislation continued the policy of the earlier Acts, but also required the legal estate in fee simple to be vested in the life tenant in possession. This was necessary because of the restriction placed on possible legal estates by the Law of Property Act 1925, s 1. A settlement typically required a life interest followed by an entail, with ultimate remainders over, so that in the absence of some legislative provision nobody would have had a fee simple absolute in possession, and hence a legal estate in the land.
The effect of the Settled Land Act 1925, therefore, is that a life tenant is given extensive powers, including the power to sell the land, and is entitled to call for the fee simple to be vested in him or her. Given the mischief with which the legislation was intended to deal, to give such powers to the tenant for life is a rational approach. Once it is accepted also that the fee simple estate must be vested somewhere, the tenant for life (and certainly not the remainderman) is the appropriate person. However, to return to the problem with which this article is concerned, for B to be treated as a life tenant within the Settled Land Act is to give him or her far more than the parties ever intended. There is also no justification for vesting the fee simple in B, rather than A. Yet for B to be granted less than a right to a home for life is to give less than was intended. There are two issues: first, whether there is any way of avoiding this, of granting B no more nor less than was intended, and allowing A to retain the fee simple; secondly if so, whether it is sensible for the law to require the avoidance device to be used.
Life interest under a trust.
The definition of a settlement for the purposes of the Settled Land Act 1925 is contained in section 1. For present purposes we are concerned only with s 1(1)(i), which applies whenever the land stands "limited in trust for any persons by way of succession". Where a trust is created expressly, this section seems always to apply in any case where the trustees are directed to allow somebody (whom we shall call X) to reside in the property rent-free for his or her lifetime, or for as long as X wishes, or until some other event, such as X's marriage. The most recent decision is that of the Court of Appeal in Costello v Costello (1994) 27 HLR 12, where land was held on trust for the plaintiff, subject to the following condition:
"For so long as [the defendant and her husband] shall desire [the plaintiff] will permit them to occupy the said property for the remainder of their lifetime...."
The Court of Appeal, taking the view that it was bound by previous authorities, and in particular Bannister v Bannister [1948] 2 All ER 133, held that the defendant was tenant for life under the Settled Land Act, with the consequence that she would be entitled to sell the property and apply the proceeds to the purchase of another house. Hence, the Settled Land Act reasoning was part of the ratio of the case.(2)
It can be argued that the phrase "limited in trust" implies an active limitation by the settlor (Harvey 1973, p 54), in which case a life tenancy arising under a constructive trust ought to be excluded, although there will inevitably be a succession whenever there is a life tenancy. If that interpretation were correct then many of the problems considered here would disappear, but it appears not to be supported by the authorities. In Bannister v Bannister [1948] 2 All ER 133, the plaintiff, who was the defendant's brother in law, was able to obtain two cottages from the defendant, at well under their market value, on the understanding that after the sale the defendant would be able to continue to live in one of the cottages rent free for as long as she wished. When the plaintiff later claimed possession of the cottage, the Court of Appeal held that he held the cottage as constructive trustee of the defendant for her life. Scott LJ observed (at p 137) that a "trust in this form has the effect of making the beneficiary a tenant for life within the meaning of the Settled Land Act", a conclusion which was technically obiter as she was concerned only to resist an action for possession by the plaintiff.
A similar conclusion was reached by two members of the Court of Appeal in Binions v Evans [1972] 1 Ch 359, where the defendant (Mrs Evans) was again concerned to resist a possession action by the plaintiffs (Mr and Mrs Binions). The vendors had, by an agreement in writing, granted the defendant a right to occupy a cottage rent-free for the rest of her life, and the plaintiffs had purchased the cottage expressly subject to her rights. The decision was that the plaintiffs held the house on trust for the defendant for her life, and although, as in Bannister v Bannister, the discussion was technically obiter, the effect of the Settled Land Act was again considered. Only Lord Denning MR (at p 366) took "limited in trust" to mean "expressly limited", which would have excluded the constructive trust from the operation of the Settled Land Act. Megaw and Stephenson LJJ, respectively at p 370 and p 372, simply adopted Scott LJ's views from Bannister v Bannister, but it is not clear that they rejected Lord Denning MR's views. Whereas Lord Denning clearly reasoned by way of a constructive trust, which arose only on the sale of the land to the plaintiffs, Megaw and Stephenson LJJ thought that the settlement arose from the earlier (written) agreement between the vendors and the defendant (this point is also made by Goff LJ, in Griffiths v Williams (1978) 248 EG 947, at p 949 (col 2)). It is arguable that the agreement indeed operated as an express limitation in trust, in which case there is no necessary inconsistency between Lord Denning MR's views on the operation of section 1 and those of his brethren.(3) In Ungurian v Lesnoff [1990] 1 Ch 206, however, Vinelott J (at pp 224-226) thought that there was a conflict, and preferred the views of Megaw and Stephenson LJJ to those of Lord Denning MR. Although the issue has not been resolved beyond doubt, therefore, the weight of authority is in favour of a life tenancy under a constructive trust being caught by the Act.
A personal right.
From the definition of settled land already considered, it follows that if there is no trust then there is no settlement within the Act. A purely personal right to occupy ought not to be caught by the Act, therefore: Re Anderson [1920] 1 Ch 175 (criticised by Harvey 1973, p 84).(4) It further follows that a contractual or estoppel licence allowing B the right to reside for his or her life does not invoke the Settled Land Act provisions. In Matharu v Matharu (1994) 26 HLR 648 (Welstead 1995), the Court of Appeal held that the defendant, who had successfully argued on the basis of a proprietary estoppel, had no beneficial interest, but merely a licence for life. It was not even argued that the Settled Land Act ought to apply. A similar order was made in Inwards v Baker [1965] 2 QB 29, but there are difficulties with this case: see further below.
It might therefore be thought that the solution for the courts to adopt is to use, as far as possible, estoppel rather than constructive trust reasoning (a view that would be endorsed, but for different reasons, by Hayton 1990). This would get around the problem of being forced to award B too much. This assumes, however, that estoppel and trusts reasoning is interchangeable, whereas we would suggest that this is not in fact the case. We would also suggest that estoppel reasoning gives B too little protection, so that even if it were possible for the courts to choose to adopt estoppel reasoning, this would not be a satisfactory solution.
Comparing trusts and estoppels.
The constructive trust.
In Bannister v Bannister [1948] 2 All ER 133, the plaintiff A obtained title to property by making a promise, and equity held him to that promise in order to prevent a fraud. Lord Denning MR's view in Binions v Evans was similar [1972] 1 Ch 359. A constructive trust can also arise, however, under Lord Bridge's first category in Lloyds Bank v Rosset [1991] 1 AC 107, at pp 132. Lord Diplock analysed this situation in Gissing v Gissing [1971] AC 886, at pp 904-905, in terms of a declaration of trust by A for B. The declaration would (if oral) be void for want of writing, which is required by the Law of Property Act 1925 s 53(1)(b), unless it were acted upon by B, so as to render it inequitable to allow A to deny the trust. It would then take effect as a constructive trust, to which, by virtue of s 53(2), s 53(1)(b) has no application. The explanation in Lloyds Bank v Rosset, at p 129, is similar: see also Re Densham [1975] 1 WLR 1519.
This analysis assumes that all the incidents of an express trust are present, even if the trust which may be enforceable is technically termed a constructive trust.(5) Certain consequences follow from this. First, a statement that might suffice to support an estoppel would not necessarily support an express declaration of trusteeship. For the former, statements as to future intention will suffice, and indeed, the property need not even be identified (see further below). Before a trust can be inferred, however, A must evince an intention to deal with the property in such as way as to deprive himself of his or her beneficial ownership, and to declare that he or she will hold it for that time forward on trust for B (see, eg, Richards v Delbridge (1874) LR Eq 11, at p 14). The property must be identified as existing property (Re Ellenborough [1903] 1 Ch 697), and an immediate irrevocable intention must be displayed.
On this basis, it is arguable that Ungurian v Lesnoff [1990] 1 Ch 206 was wrongly categorised as a constructive trust. Mrs Lesnoff, who was a Polish academic, gave up a flat in Poland of which she could have remained in occupation for life, her Polish nationality and her career, in order to live with Mr Ungurian. Ungurian bought a house in London, registered in his sole name, in which he and Mrs Lesnoff lived as if married for four years. During that time Mrs Lesnoff installed or supervised the installation of central heating, and the re-wiring and re-plumbing of the house, in addition to other works of improvement and redecoration. As in the cases already considered, Mrs Lesnoff was concerned only to resist an action for possession by Mr Ungurian, so that Vinelott J's comments on the application of the Settled Land Act (above) were obiter dicta.
Vinelott J held, on the basis of conversations between the parties upon which Mrs Lesnoff had relied, that Mrs Lesnoff had an interest under a constructive trust. However, presumably because he felt that to give Mrs Lesnoff the fee simple would be more than was intended, Vinelott J did not think that Mr Ungurian's words when handing over a cheque for the balance of the purchase price: "This is your house; you had better sign it", constituted a declaration of trusteeship (at pp 214 and 220). He relied instead on the Beirut conversations: "We will have to look for and buy a house for us in London so that you will feel secure and happy, having lost your house in Poland", and "You'll have to decide and find the house which you like. I want you to feel that you have something to rely on if anything happens to me", to conclude that Mrs Lesnoff had a life tenancy by way of trust (at pp 221-222). Clearly, however, these statements could not possibly amount to a declaration of trust. They did not indicate an irrevocable commitment on the part of Ungurian, but were merely statements of future intent. Indeed, the trust property had not at that stage been identified. As a trust, this should therefore have failed as an attempt to create a trust of future property, ie, as in Re Ellenborough [1903] 1 Ch 697. Not only could the case have been decided on estoppel principles, therefore, but we would contend that it should have been.(6)
Not only must the trust property be existing and identifiable, but A must display an immediate and irrevocable intention to declare himself or herself trustee. A statement as to future intention ought not to suffice. Both types of statement were present in Hammond v Mitchell [1991] 1 WLR 1127. On its own, Hammond's promise (at p 1131):
"Don't worry about the future because when we are married it will be half yours anyway and I'll always look after you and [the boy]"
is a statement of future intent which ought not to lead to the inference of a trust. However, he also said (at p 1131):
"I'll have to put the house in my name because I have tax problems due to the fact that my wife burnt all my account books and my caravan was burnt down with all the records of my car sales in it. The tax man would be interested, and if I could prove my money had gone back into a property I'd be safeguarded."
This is similar to the statements in Eves v Eves [1975] 1 WLR 1338 and Grant v Edwards [1986] Ch 638, where A was held to have declared himself trustee for B. These cases were criticised by Gardner (Gardner 1993, pp 264-265) on the grounds that A did not intend this. We have argued elsewhere (Glover and Todd 1995) that Gardner was wrong to concentrate on A's subjective intention, and that Vicky Mitchell might reasonably have translated this statement as: "although legal title must be vested in myself alone, this is for purely formal reasons, and the reality is that the property is to belong to both of us." It is difficult to imagine a clearer declaration of trusteeship, if the test is what Vicky might reasonably think. In any case, Hammond v Mitchell differs from Grant v Edwards and Eves v Eves, in that it is probable that Hammond really did intend Mitchell to have a share, and we would therefore suggest that Waite J correctly regarded Hammond as being a trustee for Mitchell.
Where B's interest arises under a trust, its quantification depends on A's representation, rather than on B's reliance on it. For the Bannister v Bannister type of constructive trust (or Binions v Evans on Lord Denning MR's reasoning), the purchaser must be bound to the full extent of his representation for the fraud to be prevented. In the case of Rosset first category constructive trusts, but for want of writing, there would be an enforceable trust, so that the only relevance of reliance is to render it inequitable for A to deny the existence of the trust. Reliance, in other words, acts only to overcome the fraud hurdle, in order to trigger s 53(2), and it should therefore follow that once the requisite reliance has occurred, the trust ought to be enforceable in full. Thus, the quantification of B's interest should depend on the declaration and not on the reliance. In Gissing v Gissing, Lord Diplock (at p 905), adopting the reasoning based on s 53(2), also suggested that effect is given to the original declaration.(7) This argument also removes (assuming the declaration to be clear) any discretionary element in the quantification.
Where B's interest arises under a constructive trust, the conventional view is that constructive trusts, like their express counterparts, create full proprietary interests binding on third parties (eg Ferguson 1993, pp 120-121).(8) The majority view in Binions v Evans was that the plaintiff purchasers were bound by an existing Bannister v Bannister trust. In principle also, Rosset first category constructive trusts have all the incidents of an express trust apart from writing, so it would be difficult to justify adopting a different position for them regarding third parties. However, until s 53(2) is satisfied there cannot be an enforceable trust, and hence no equitable interest. It follows that a third party will be bound only if he or she takes after s 53(2) is satisfied (Hayton 1990, at p 380).
Estoppel
Estoppels differ from constructive trusts in a number of significant respects. Whereas a trust is based on an irrevocable commitment by A, estoppel is based on the notion of A misleading B, and A is prevented from going back on his assurances only to the extent necessary to do justice to B. The commonly- accepted formula for proprietary estoppel, adopted by Oliver J in Taylors Fashions Ltd v Liverpool Victoria Trustees Co Ltd [1982] QB 133, at pp 151-152, (9) requires only that "it would be unconscionable for a party to be permitted to deny that which, knowingly or unknowingly, he has allowed or encouraged another to assume to his detriment", and in Moorgate Mercantile Co Ltd v Twitchings [1975] QB 225, Lord Denning MR took the view (at p 241) that encouragement can be by words or conduct. This is less stringent than the requirements for a declaration of trusteeship. A statement as to future intent will suffice, even where the property is not as yet identified (Re Basham [1986] 1 WLR 1498, at pp 1508-1509).
It is however also necessary for B to have relied on A's representations, or other form of encouragement (requirements for reliance being set out by Mustill LJ in Grant v Edwards [1986] Ch 638, at p 652). Ferguson tentatively argues (Ferguson 1993, pp 116-120) that the requirements for detrimental reliance are less stringent for proprietary estoppel than they are for a constructive trust within the Rosset first category, but we doubt whether she has conclusively established this distinction. She suggests (at p 119) that the reversal of onus of proof in Greasley v Cooke [1980] 1 WLR 1306, per Lord Denning MR at p 1311, where the Court of Appeal held that once it was shown that B had relied on the assurances given to him or her, the burden of proving that she had acted to her detriment was shifted back on to A, applies only to estoppels and not to constructive trusts. However, her main authority, a neutral remark by Nourse LJ in Grant v Edwards [1986] Ch 638, at p 650, is not strong. Another possible distinction might be that whereas reliance for estoppel purposes can be any detrimental reliance, for a constructive trust it must be linked in some way to the property. This might, after all, lead more readily to the inference that an interest in the property was to be created. However, Hammond v Mitchell [1991] 1 WLR 1127 is authority against such a distinction, since Vicky Mitchell's reliance was unconnected with the property in which she obtained a half share. In any case, since for this type of constructive trust, all reliance does is to get over the fraud hurdle, we would suggest that there is no need for such a link.
When it comes to quantification, however, there is a clear difference between interests arising under an estoppel and those arising under a constructive trust. For estoppels, the quantification of B's interest does not depend on A's representation alone, but also on the extent of B's reliance. In Crabb v Arun District Council [1976] Ch 179, Scarman LJ (at p 198) took the view that quantification should be on the basis of "the minimum equity to do justice to the plaintiff". A's representation is undoubtedly relevant in that B cannot get more than was contemplated by the parties. In Baker v Baker (1993) 25 HLR 408, Edward Baker provided over £30,000 towards the purchase of a house which was conveyed into the name of Peter Baker, and moved out of a council house in reliance upon Peter's representation that Edward would have a "granny room" rent free in the house for the rest of his life (see p 410). Edward Baker left the property, after a row, only eight months later. He unsuccessfully claimed a beneficial interest by way of resulting trust, but an alternative estoppel claim succeeded, the starting point for quantification being the value of what was represented, ie, the right to occupy the granny room rent free for life. B's claim to the return of the money he had paid, a far greater amount than the value of this occupation right, was rejected.
However, while B's entitlement cannot be more than would be accorded on the basis of A's representation, it can certainly be less, as in Dodsworth v Dodsworth (1973) 228 EG 1115, discussed in greater detail below, and in satisfying the equity it may also be appropriate to consider the extent of B's reliance.
The main difference between estoppel and trust interests, however, is that as Ferguson states (Ferguson 1993, at p 122), "the consensus of opinion favours" the view that estoppels do not bind third parties, at any rate until the extent of the interest has been crystallised by a court (see also Hayton 1990, pp 380ff; Battersby 1991). An estoppel solution may therefore give B inadequate protection if A sells the land.
Battersby argues that estoppels take their status from whatever interest they create, and before a court order they have created nothing, not being proprietary interests in themselves. Hayton argues that they should not bind a third party at that stage because they are inchoate interests only. It can further be argued that, since until a court order has been obtained it is by no means clear that an interest in land will be granted at all (it was not in Baker v Baker (1993) 25 HLR 408 or Dodsworth v Dodsworth (1973) 228 EG 1115), the estoppel at this stage is a "mere equity", to adopt the terminology of the House of Lords in National Provincial Bank Ltd v Ainsworth [1965] AC 1175, rather than an equity which is ancillary to or dependant upon an equitable estate or interest in land, such as was held in Blacklocks v J B Developments (Godalming) Ltd [1982] Ch 183, at p 196, to constitute an overriding interest under the Land Registration Act 1925 s 70(1)(g). This argument is considered by Simon Baughen (Baughen 1994, at p 154), who is concerned however to argue the opposite point of view, that estoppel rights are capable of binding third parties (cf. Todd 1981; Battersby 1991, pp 39-44).
By contrast, it is clear that if the minimum equity to do justice to B requires the creation by the court of a proprietary interest, as in, eg, Pascoe v Turner [1979] 1 WLR 431, then from the time of the court order that interest will be capable of binding third parties. This, we suggest, is a probable explanation of Williams v Staite [1979] Ch 291, where a third party with notice was bound, but only after a judge had held that the defendants had an equitable right to live in the property for the rest of their lives.
If the consensus of opinion is as Ferguson states, it is not unanimously held, and indeed it can be argued (eg, Baughen 1994) that estoppels are capable of binding third parties even prior to a court order. There is authority for this position, and in particular it can be argued that in both E R Ives Investments Ltd v High [1967] 2 QB 379 and Inwards v Baker [1965] 2 QB 29, an interest created by an estoppel was held binding on a third party prior to crystallisation by the court. We would suggest, however, that both cases are explicable on other grounds, as are nearly all the other authorities cited by Baughen, and that the arguments of Hayton, Battersby and Ferguson are to be preferred.
In Inwards v Baker, the Court of Appeal held that the son had an estoppel licence to remain in a bungalow on land which was owned by his father, for the remainder of his life, having expended money on the property in reliance upon his father's representation that he would be able to live there. There was no discussion of the Settled Land Act, "[A] point which appears to have been overlooked", according to Russell LJ in Dodsworth v Dodsworth (1973) 228 EG 1115, at p 1115 (col 2), and the case appears superficially similar to Matharu v Matharu (1994) 26 HLR 648. The difficulty is that the estoppel was held binding not on the father, who had died several years before the case arose, but on the trustee under his will. So the case appears to be one of an estoppel binding a third party prior to its crystallisation by the court. (Note that the argument advanced by Battersby (Battersby 1991, p 43), that the court crystallised the defendant's interest, looks incorrect, as this would surely have occurred too late to bind the trustee.)
Even if the case is authority for this proposition, it does not assist the argument that courts should (on the assumption, which we doubt, that they can pick and choose) adopt estoppel reasoning to avoid the Settled Land Act, since there is House of Lords authority (National Provincial Bank Ltd v Ainsworth [1965] AC 1175) that the defendant's interest must have been an interest in land for it to bind the trustee. If it had been an interest in land, then it would have been a beneficial life tenancy, and should have fallen within the Settled Land Act anyway.
No doubt it is possible to argue that Inwards v Baker was wrongly decided, as being inconsistent with other authorities. We would suggest, however, that this is precisely the type of case that can best be explained in trust terms. The father's representation: "Why not put the bungalow on my land and make the bungalow a little bigger" can reasonably be interpreted as a declaration of trust by him, giving his son an equitable life tenancy. (10) It was a declaration of an irrevocable commitment, not merely a future intention, and the trust property was existing and identifiable. Although a declaration of trust in land normally requires writing, by virtue of the Law of Property Act 1925 s 53(1)(b), the son's reliance would have rendered it fraudulent for the father to renege. Thus, the father's declaration could take effect as a constructive trust, which requires no writing, by virtue of s 53(2) (as in Lloyds Bank plc v Rosset [1991] 1 AC 107 first category constructive trusts, discussed above).
If this is the correct explanation of Inwards v Baker, and we would suggest that it is the only one which works, then the son ought to have been a tenant for life under the Settled Land Act. Arguably, therefore, his interest is greater than the parties intended, although Battersby has argued otherwise, on the grounds that the son should have been able to move house during his lifetime, without losing his investment (Battersby 1991, p 43, n 28).
Ives v High [1967] 2 QB 379 also presents problems for the "consensus of opinion". Under a contract made in 1949, Westgate agreed to grant High a right of way over his land, the consideration from High being to allow Westgate to leave in place foundations which he (Westgate) had accidentally placed on High's land (the properties having been bombed during the second world war, obliterating the boundary). No legal right of way was granted, but High relied on the agreement to his detriment by building a house on the assumption that he would have the right of way. In 1950 Westgate sold the land to Wright, who had notice of the agreement, and stood by while High built a garage which could only be accessed over Wright's yard, and contributed to the re-surfacing of the yard. In 1962, the land was sold to the plaintiff company, expressly subject to High's rights. When the plaintiffs discovered that High had not registered any interest under the Land Charges Act, they sued High for trespass.
The Court of Appeal held that High's right of way was enforceable against the plaintiffs, despite non- registration, Winn LJ holding that High had an easement by estoppel which was binding on the plaintiffs, and which did not need to be registered. (11) If this reasoning is correct then estoppels can bind third parties even before the interest is crystallised by a court, but there are several other explanations, which we would suggest are preferable. First, if Lord Denning MR's interpretation of the Land Charges Act 1925 was correct (at p 395),
"an 'equitable easement' is a proprietary interest in land such as would before 1926 have been recognised as capable of being conveyed or created at law, but which since 1926 only takes effect as an equitable interest."
This definition excludes a contract for an easement, in which case (if correct) the original 1949 agreement could have bound the plaintiffs even in the absence of registration. Battersby observes, however, that this interpretation "has no justification in the statutory provisions" (Battersby 1991, p 40), a sentiment with which we find it difficult to disagree. (12) It is also worth observing that in Huckvale v Aegean Hotels Ltd (1989) 58 P & CR 163, the Court of Appeal assumed without argument that a contract for an easement is registrable under the Land Charges Act.
Another possible explanation is Battersby's own (Battersby 1991, pp 41-42), that the plaintiffs were estopped by their conduct (leaving the foundations in place) from asserting the defendant's failure to register. (13) This depends on the plaintiffs continuing to keep the foundations in place, a problem which would be avoided by an explanation derived from views expressed by Megaw LJ in Binions v Evans [1972] Ch 359 (at p 371). Under the 1949 agreement Westgate had undertaken obligations towards High. He could not free himself from these obligations by selling the land, so the agreement remained binding on him. The plaintiffs knew of that agreement and indeed took subject to it. If they were now to deny High access they would be liable in tort for interference with contract, and the court would intervene to prevent them from interfering with High's contractual rights. Megaw LJ, however, expressed some doubts as to whether normal tortious principles apply in the law relating to land, presumably because the tort action could drive a coach and horses through the registration requirements of the Land Charges Act. The reasons for Megaw LJ's doubts are unclear, however, since he refers to Lord Upjohn's speech in National Provincial Bank Ltd v Ainsworth [1965] AC 1175, at p 1239, where there is no reference to a tort action. It is in any case clear from the authorities referred to by Smith (Smith 1977) that tort principles do apply in this area, although Smith (at pp 328-329) is unhappy with this development, preferring to reform the Land Charges and Land Registration Acts where appropriate. His argument is that tort principles are in their infancy, and that it is wrong to import them into an area where there are already established doctrines.
Our preferred explanation of Ives v High derives, like the tort solution, from Binions v Evans, but from the judgment of Lord Denning MR. The plaintiffs took expressly subject to High's rights, and there is no reason why they should not have become constructive trustees of the right of way on that basis alone. The possibility of a constructive trust arising in these circumstances has been kept alive by Fox LJ in Ashburn Anstalt v Arnold [1989] 1 Ch 1, at pp 23-26, and it enables Ives v High to be explained without resorting to the argument that the estoppel binds the third party, without adopting a tortuous interpretation of the Land Charges Act, and without driving a coach and horses through its registration requirements (because of the limitations placed on this analysis in Ashburn Anstalt). It is also worth noting that at least one of the criticisms made by M P Thompson (Thompson 1988, p 206) of this entire line of reasoning does not arise here. He objects in general to "the difficulty of analysing the ambit of the trust", but that would hardly have been a problem in Ives v High, where the nature and extent of High's interest could hardly have been clearer.
There are other difficulties in the way of the consensus view, but nothing, we feel, that is really substantial, especially when weighed against the convincing arguments of principle advanced by Hayton, Battersby and Ferguson. After Ashburn Anstalt, it is no longer possible to explain Errington v Errington [1952] 1 KB 290 as a contractual licence case, but Fox LJ (at p 17) did not doubt that the decision was correct. One solution he canvassed (at p 17) was an estoppel license binding on the widow, who was a third party. That would obviously create difficulties for the view we propound here, but he also suggested the alternative explanation of an estate contract (see also Todd 1981, p 352, where the formalities and non-registration difficulties of this solution are addressed).
Baughen also cites the pre-1925 cases of Plimmer v Mayor of Wellington (1884) 9 App Cas 699 and Dillwyn v Llewellyn (1862) De GF & J 517. Neither is a very convincing estoppel case, and both have been re-explained elsewhere (respectively Ferguson 1993, p 123; Todd 1981, p 357). That leaves Baughen with little convincing authority, whereas we would suggest, the argument for the "consensus of opinion" we have put forward here is quite strong.
Conclusion on trust and estoppel
It should be apparent from the above argument that trusts and estoppels can arise in different circumstances, and that it is not therefore open for the courts to pick and choose between them. Even if the courts could pick and choose, the estoppel solution is unsatisfactory because of the problems of binding third parties.
In any event, whatever the position before the court has determined the extent of the equity, it is impossible to avoid Settled Land Act issues, when the estoppel is crystallised. If it is crystallised as a licence only, then B's position remains weak against a third party. If on the other hand it is crystallised as an equitable life tenancy, then from that time onwards the Settled Land Act applies. A preference for estoppel reasoning would ultimately achieve nothing, therefore, the problems being merely postponed until crystallisation. It has already been observed that the basis of the remedy should be "the minimum equity to do justice to the plaintiff." To crystallise B's interest as a beneficial life tenancy may be to give B far more than was originally contemplated by the parties. In Dodsworth v Dodsworth (1973) 228 EG 1115, Russell LJ (at p 1115 (col 2)) refused to grant the defendants (Mary Dodsworth's brother and his wife) the life tenancy which they claimed on the basis of the Mary Dodsworth's representation, on the grounds that, as tenants for life under the Settled Land Act 1925, they would obtain more than had ever been contemplated. He observed that
"We do not think that it can be right to satisfy such an equity by conferring upon the defendants a greater interest in the property than was envisaged by the parties."
It appears, then, that the courts can only give B too little, or too much: either B gets a full beneficial interest, in which case the Settled Land Act is invoked, giving greater rights that the parties contemplated (except, on Battersby's view, above, of Inwards v Baker (Battersby 1991, p 43, n 28)), or B gets only a licence, in which case he or she is insufficiently protected against third parties. The question is whether there is any way of avoiding this problem.
Avoiding the problem
The requirement in s 1(1)(i) for the land to stand "limited in trust for any persons by way of succession" implies that B must have a full beneficial interest in the property. This suggests that the Settled Land Act might be avoided by a private purpose trust, such as that in Re Denley's Trust Deed [1969] 1 Ch 373, which, in Goff J's view, B has sufficient interest to enforce, but where B does not obtain a full beneficial interest. Where land is conveyed to trustees, to permit B to occupy the property, and subject thereto, to A, B has an interest capable of binding a third party. However, it is arguable that B is not a life tenant, having less than a full beneficial interest. It is not clear, for example, that B has any right to receive rent.
There are two problems with this approach. The first is that it is not clear that Denley is authority that such a trust can exist. In Re Grant's WT [1980] 1 WLR 360, Vinelott J (at p 370) analysed Denley as an ordinary trust for beneficiaries (also Millett 1985, pp 280-282), in which case B would have a full beneficial interest, and would be a tenant for life. The second problem is that the authorities (discussed above) simply do not suggest that a trust of this type avoids the Act. So a solution along these lines can also be ruled out.
Another possibility was canvassed in Griffiths v Williams (1978) 248 EG 947 where, as in Dodsworth v Dodsworth (1973) 228 EG 1115, the Court of Appeal was concerned to crystallise an estoppel interest, where the solution adopted was to grant Mrs Williams a long lease determinable at her death at a nominal rent (see also Thompson 1994, p 394). Under the Law of Property Act 1925 s 149(6) this would be converted into a 90-year lease, and this would avoid the Settled Land Act. (14) However, since we are assuming that A has represented to B that B can reside rent-free, the requirement for a nominal rent is unfair on B, and Griffiths v Williams depended on all the parties having consented to an order in those terms. Further, as was noted in Costello v Costello (1994) 27 HLR 12, at p 19, the device would no longer work because of the Leasehold Reform Act provisions which came into effect after the case had been decided.
In Dodsworth v Dodsworth (and see also Thompson 1994, p 394), it was suggested in argument that an immediate binding trust for sale would be a possible solution, subject to the consent of the defendants while in occupation of the bungalow as their home. (Before the appeal was heard, the plaintiffs had died intestate, so the administrators already held the property on trust for sale under the Administration of Estates Act 1925, s 33.) Since 1884, trusts for sale have been exempted from the provisions of the settled land legislation, the present provision being s 1(7) of the Settled Land Act 1925, a trust for sale being defined as "an immediate binding trust for sale" in the Law of Property Act 1925 s 205(1)(xxix). There is no doubt that the use of consents in this way could prevent the land being sold over the heads of the defendants (Russell LJ mentions Re Herklot's WT [1964] 1 WLR 583, and see also Re Inns [1947] Ch 576). However, Russell LJ observed that it would not protect the defendants' occupation. It would also have been necessary to prevent Mary Dodsworth's administrators (Mary herself having died) from obtaining possession in order to let the property at a rent, but in his view, any order to that effect would necessarily have had the effect of taking the trust for sale outside the statutory definition, in which event the Settled Land Act would have remained applicable. Presumably Russell LJ thought that it would be necessary to grant the defendants an interest, which would have prevented the trust for sale being "immediate" (a phrase on which there has been little litigation, but see Harvey 1973, p 73), but it is not entirely clear that this is correct. A licence enforceable against the administrators alone would have sufficed, which coupled with the consent requirement ought to have given the defendants sufficient protection.
Conclusion
It seems absurd that a right to occupy a home for life cannot be granted under a constructive trust or proprietary estoppel without invoking the full force of a statute whose purpose was to deal with an entirely different problem, which has no modern day relevance.
A solution to the problem was canvassed in the discussion of Dodsworth v Dodsworth, above, but the device is highly artificial, and the trust for sale is in reality a fiction. That fictitious trusts for sale are routinely used to deal with concurrent interests in land is not a good reason for extending their ambit, with all the attendant difficulties in ascertaining whether A's and B's interests are in realty or personalty.
We would therefore suggest that legislation is really necessary to resolve the problem. The most obvious solution might appear to be the simple repeal of the Settled Land Act, which no longer serves any useful purpose (the strict settlement is simply no longer a social problem, for the reasons outlined by Harvey 1973, p 4). However, this would not suffice, since the legal title must go somewhere, and it would still therefore be necessary to grant one of the parties the fee simple absolute in possession. In the situation addressed by this article, we would suggest that the appropriate person is A, the remainderman, rather than B, the tenant for life. However, the opposite would remain true for the traditional settlement, should anyone still wish to make one, and for many existing settlements. A legislative solution could be quite complex, therefore (see Thompson 1994, p 394, who argues for allowing land to remain unsaleable for one generation), but it is necessary if the absurd difficulties addressed in this article are to be avoided.
Bibliography
Battersby, G (1991) 'Contractual and Estoppel Licences as Proprietary Interests in Land' [1991] Conveyancer 36.
Baughen, S (1994) 'Estoppels over land and third parties. An open question?' 14 Legal Studies 147.
Burn, E H (1994) Cheshire & Burn's Modern Law of Real Property, 15th ed. (London: Butterworths).
Ferguson, P (1993) 'Constructive Trusts - A Note of Caution' 109 Law Quarterly Review 114.
Gardner, S (1993) 'Rethinking Family Property' 109 Law Quarterly Review 263.
Glover, N and Todd, P 'Inferring share of interest in home: Midland Bank v Cooke' [1995] 4 Web Journal of Current Legal Issues.
Harvey, B (1973) Settlements of Land (London: Sweet & Maxwell Modern Legal Studies Series).
Hayton, D (1990) 'Equitable Rights of Cohabitees' [1990] Conveyancer 370.
Millett, P J (1985) 'The Quistclose Trust: Who can enforce it?' 101 Law Quarterly Review 269.
Smith, R J (1977) 'The Economic Torts: Their Impact on Real Property' 41 Conveyancer (NS) 318.
Thompson, MP (1988) 'Leases, Licences and the Demise of Errington' [1988] Conveyancer 201.
Thompson, MP (1994) 'Accidental Settlements' [1994] Conveyancer 391.
Todd, P N (1981) 'Estoppel Licences and Third Party Rights' [1981] Conveyancer 347.
Welstead, M (1995) 'Proprietary Estoppel - A Flexible Familial Equity' [1995] Conveyancer 61.
Footnotes
(1) For estoppels the principle is that quantification should be on the basis of "the minimum equity to do justice to the plaintiff". In satisfying the equity it may therefore be appropriate to consider the extent of B's reliance. See further below. Back to text.
(2) For a review of the authorities generally, see Harvey 1973, pp 82-87. At p 86, Harvey describes Re Varley (1883) 68 LT 665, the only case where the Settled Land Act (in this case 1882) did not apply, as a "rogue" decision which is inconsistent with the mainstream of authorities. Back to text.
(3) A problem with the majority view is that it is difficult to see why the sale to the plaintiffs was not caught by the "paralysing section": Settled Land Act 1925, s 13. Back to text.
(4) See also Chandler v Kerley [1978] 1 WLR 693, where the effect of the Settled Land Act upon a contractual licence was discussed. Back to text.
(5) In Midland Bank plc v Cooke, CA, 7 July 1995, the Court of Appeal also used a constructive trust to quantify B's interest, it being otherwise clear that there was a trust. We suggest that there are problems with this approach (Glover and Todd 1995). Back to text.
(6) It may be objected that the representations in Bannister v Bannister [1948] 2 All ER 133 and Binions v Evans [1972] 1 Ch 359 were also made before the trustee obtained the trust property, but the juristic basis of those decisions is different. There, the trustee obtained the property by fraud. For Ungurian to have reneged on the representations he made to Lesnoff may well have constituted a fraud on Lesnoff, but it cannot be said that he obtained the property by fraud. Back to text.
(7) This can most clearly be seen in Hammond v Mitchell [1991] 1 WLR 1127, where Vicky Mitchell obtained a half share, based on Hammond's representations, although she had contributed nothing at all to the house itself, and it would not have been possible to justify a half share on the basis of her detrimental reliance.
All this assumes that B's share can be inferred from A's declaration. There are cases, such as Stokes v Anderson [1991] 1 FLR 391, and Passee v Passee [1988] 1 FLR 263, where although a trust is declared, B's share is not agreed at the outset, but is to be settled at some time in the future, depending on B's contribution. In this case the contribution is obviously relevant to the quantification, but for there to be certainty of subject matter, we would contend that there has to be some formula, irrevocably agreed from the outset, for calculating B's share. In Cowcher v Cowcher [1972] 1 WLR 425, Bagnall J observed, at p 430, that "[t]here can be no trust unless the nature and quantum of the several beneficial interests are certain at the inception, and during the whole life, of the trust".
Where no share can be inferred from A's declaration, there is authority that the courts presume a half share: Eves v Eves [1975] 1 WLR 1338, per Brightman J at p 1345; Grant v Edwards [1986] Ch 638, at p 657, but the presumption can be rebutted in appropriate circumstances (it was rebutted by Brightman J in Eves v Eves). Back to text.
(8) See also Re Sharpe [1980] 1 WLR 219, but note Browne- Wilkinson J's reservations about the position of a purchaser at p 226. Back to text.
(9) This has been accepted as broadening the approach adopted by Fry J in Willmott v Barber (1880) 15 Ch D 96, at pp 105-106. Back to text.
(10) It is not necessary to say: "I declare myself trustee", and informal words will suffice. In Paul v Constance [1977] 1 WLR 527, at p 531, the statement, "The money is as much yours as mine" was held sufficient by the Court of Appeal.
It may be objected that the declaration is too vague, since it is as consistent with a fee simple as with a life interest, but Lord Denning MR's interpretation (at p 37) that it "was to be his home for his life or, at all events, his home as long as he wished it to remain his home" is equally applicable whether a trust or an estoppel is created. Back to text.
(11) The reasoning, which is based on Wright's acquiescence and not on the original agreement, is set out at p 405. Inwards v Baker is cited as authority. Back to text.
(12) Lord Denning MR's definition relates to class D(iii), an alternative argument based on class C(iv) being rejected on the grounds that there was no contract to convey a legal estate (at p 395). We confess some difficulty in comprehending Battersby's contrary view on the C(iv) argument (Battersby 1991, p 41, n 16), given the definition of a legal estate in the Law of Property Act 1925 s 1(1). Back to text.
(13) Presumably Battersby means the 1949 contract. He observes that his argument is not unlike that adopted in Taylors Fashions Ltd v Liverpool Victoria Trustees Co Ltd [1982] QB 133. Back to text.
(14) See also Re Catling [1931] 2 Ch 359, where Bennett J held, on the basis of the Settled Land Act 1925, s 20(1)(iv), that a yearly tenant at stlg1 per annum was not a tenant for life. The case is criticised by Harvey 1973, p 83. Back to text.