Edwards v Kumarasamy

This appeal concerns a repairing covenant implied into a subtenancy of a residential flat by section 11 of the Landlord and Tenant Act 1985. It raises two issues of interpretation relating to that section, and an issue of more general application as to the need for notice before a landlord can be liable under a repairing covenant.

By a lease (“the Headlease”) dated 28 April 2006, the freeholder of a small block of flats known as Oakleigh Court, Boston Avenue, Runcorn (“the Building”) let Flat 10 in the Building (“the Flat”) for a term of 199 years from 1 January 2006 at a rent of £195 per annum, for a premium of £130,000. The extent of the Flat demised by the Headlease was defined by “the plastered coverings and plaster work” of the external and internal “walls and partitions” and ceilings, and “the floorboards and surfaces of the floors”. Congruently, the demise expressly excluded any of main timbers and joists, and the “framework”, of the Building, and it also excluded “the walls or partitions therein”, except “the plastered surfaces thereof”. The demise of the Flat also included certain rights “for all purposes incidental to the occupation and enjoyment of the Flat”, and those rights included the right to use “the entrance hall lift staircases and landings … giving access to the Flat”, the right to use an “access road” and a specific space in a parking area in the curtilage of the Building, and the right to use the communal dust bins.

As is normal under a long lease of a flat, the Headlease contained provisions whereby the freeholder covenanted to provide certain services, and provisions whereby the headlessee covenanted to pay a service charge for those services. Those services included “keeping in good and substantial repair” (i) “all entrances passages landings stairs fire escapes Bin Store (if any) and other parts of the Building intended to be enjoyed or used by the owners or occupiers of the Building in common with others”, and (ii) other areas in the Building not “capable of being let as flats”. However, “[i]n the case of any item of disrepair”, it was stipulated that the freeholder “will not be liable for breach of this covenant until the [headlessee] has given written notice thereof to the [freeholder] and the [freeholder] has had a reasonable opportunity to remedy the same”.

The Headlease is and has at all material times been vested in the appellant, Mr Kumarasamy. By a subtenancy dated 6 April 2009 (“the Subtenancy”), Mr Kumarasamy granted to the respondent, Mr Edwards, a tenancy of the Flat for a term expiring on 5 October 2009 (although the tenancy was liable to be continued as a periodic tenancy, as it was an assured shorthold tenancy, but nothing hangs on that for present purposes). The Subtenancy included a grant of “the right to use, in common with others, any shared rights of access, stairways, communal parts, paths and drives” of the Building.

The Subtenancy contained a covenant by the subtenant, Mr Edwards, (i) to keep the Flat in good and tenantable condition, repair and decorative order, “items which the [Headlessee as] landlord is responsible to maintain … excepted”, and (ii) to permit Mr Kumarasamy and his agents to enter the Flat after giving 24 hours’ notice in order (a) to view its state of repair “and to execute repairs and other works upon the [Flat] or other properties” and (b) to show it to prospective new tenants or purchasers.

It is rightly common ground that section 11(1) of the 1985 Act, which cannot be contracted out of (see section 12(1)), applies to the Subtenancy. It is in these terms:

Subsection (6) of section 11 implies into any tenancy to which subsection (1) applies “a covenant by the lessee that the lessor, or any person authorised by him in writing, may at reasonable times of the day and on giving 24 hours’ notice in writing to the occupier, enter the premises comprised in the lease for the purpose of viewing their condition and state of repair”.

In Campden Hill Towers Ltd v Gardner [1977] QB 823, the Court of Appeal had to consider the application of the predecessor of section 11(1)(a), namely section 32(1)(a) of the Housing Act 1961, which was in effectively identical terms to section 11(1)(a), to a tenancy of a third floor flat in a large block of flats. Megaw LJ, giving the judgment of the court, said at p 834 that “[a]nything which, in the ordinary use of words, would be regarded as a part of the structure, or of the exterior, of the particular ‘dwelling house’ [sc the third floor flat], regarded as a separate part of the building, would be within the scope of paragraph (a).” However,as he went on to explain at pp 834-835, “other parts of the outside walls and other parts of the structure of the block” are “not ‘of the dwelling house’, and the paragraph expressly and deliberately uses the limiting words, as defined in the section itself, relating the paragraph to ‘the dwelling house’”. It has not been suggested on this appeal that this analysis is wrong; and in my view it is clearly right.

When Parliament repealed section 32 of the 1961 Act and replaced it with section 11 of the 1985 Act, it did not make any amendments of practical significance. However, section 116(1) and 116(2) of the Housing Act 1988 added some new subsections to section 11 of the 1985 Act, part of whose purpose would appear to have been to modify the effect of some of the reasoning in Campden Hill. In particular, new subsections (1A) and (1B) were added to the following effect:

The Building has a ground and two upper floors, and it appears that there are four flats on each floor, the Flat being on the second floor. The Building has a main entrance door which leads into a front hallway from which access can be got to the ground floor flats and to the lift and staircase which serve the two upper floors. The flats on the upper floors are accessed from hallways, leading from the lift and staircase. There is a car park in front of the Building, and, between the car park and the front door to the Building, there is a paved area, which is part of what is referred to in the Headlease as “the access road”, which is the only or principal means of access to the Building. The paved area, which is between three and four metres in length, is covered by paving stones. The paved area is also used by occupiers as a means of access to the communal dustbins which are sited in the car park outside the Building.

On 1 July 2010, Mr Edwards was taking rubbish from the Flat to the communal dustbins, when he tripped over an uneven paving stone on the paved area. As a result, he suffered an injury to his right hand (which resulted in the exacerbation of pre-existing neuropathic pain for some 18 months) and to his right knee (which involved soft tissue injury lasting some four months). He issued proceedings against Mr Kumarasamy contending that his injury was caused by Mr Kumarasamy’s failure to keep the paved area in repair, in breach of the covenants implied into the Subtenancy by section 11(1)(a) and 11(1A)(a) of the 1985 Act.

The claim was heard by Deputy District Judge Gilman, who accepted Mr Edwards’s case both on the facts and on the law, and awarded him £3,750 damages. While he did not challenge the Deputy District Judge’s conclusions on fact or quantum, Mr Kumarasamy appealed against the conclusion that he was liable to Mr Edwards under the statutory repairing covenant as a matter of law. Her Honour Judge May QC allowed his appeal on two grounds, namely (i) the paved area was not within the ambit of the section 11 covenant, and (ii) even it had been, Mr Kumarasamy could not have been liable as he had had no notice of the disrepair. Mr Edwards was permitted to bring a second appeal, and the Court of Appeal allowed his appeal, disagreeing with Judge May on both grounds, for reasons given by Lewison LJ, with whom Sir Terence Etherton C and Christopher Clarke LJ agreed - [2015] Ch 484.

Mr Kumarasamy now appeals to this court.

In a case such as this, where the “dwelling-house” in question forms “part only of a building”, section 11(1A)(a) requires section 11(1)(a) to be read as if it required a landlord “to keep in repair the structure and exterior of any part of the building in which [he] has an estate or interest”. As Lewison LJ said in para 6 of his judgment, when discussing the argument then advanced by counsel then appearing for Mr Kumarasamy:

In the light of that analysis this appeal raises three questions. The first is whether, to quote again from Lewison LJ, “the paved area which leads from the front door to the car park [can] be described as part of the exterior of the front hall” within section 11(1A)(a). The second question is whether Mr Kumarasamy had an “estate or interest” in the front hall within section 11(1A)(a). The third question is whether Mr Kumarasamy could be liable to Mr Edwards for the disrepair in question notwithstanding that he had had no notice of the disrepair in the paved area before Mr Edwards’s accident.

The respondent, Mr Edwards, can only succeed if all three questions are answered in the affirmative, as the Court of Appeal held that they were. The first and second questions are of some significance in relation to the application of section 11, as they concern, in the first case, the extent of the physical property falling within section 11(1)(a), and, in the second case, the nature of the estate or interest which falls within section 11(1A)(a). The second question is particularly relevant to the liability to a subtenant of a flat of a landlord who has a headlease of that flat. The third question also is of importance to the application of subsections (1)(a) and (1A)(a) of section 11, but it is of much wider significance, as it relates to the extent of the need for notice of a want of repair before a landlord can be liable for disrepair under a repairing covenant, whether under section 11 or otherwise.

In my view, it is not possible, as a matter of ordinary language, to describe a path leading from a car park (which serves the building and can be said to be within its curtilage) to the entrance door which opens directly onto the front hall of a building, as “part of the exterior of the front hall”. It is hard to see how a feature which is not in any normal sense part of a building and lies wholly outside that building, and in particular outside the floors, ceilings, walls and doors which encase the front hall of the building, can fairly be described as part of the exterior of that front hall. The paved area may be said to abut the immediate exterior of the front hall, but it is not part of the exterior of the front hall, as a matter of normal English. Unless the natural meaning of the words of a statutory provision produces a nonsensical result, or a result which is inconsistent with the intention of the legislation concerned, as gathered from admissible material, the words must be given their ordinary meaning. (I should perhaps add that in many cases, particularly when the words are read in their context, they can have more than one ordinary meaning, and it is then for the court to decide which of those meanings is correct.)

There is some force in the argument that a purposive approach to the words of section 11(1A)(a) suggests that they should be given a wide, rather than a narrow, effect, as one might have expected that Parliament intended those parts of a building or its curtilage which are not included in an individual residential demise, and which are in any way enjoyed by the tenant in question, would be within the ambit of the landlord’s statutory repairing covenant. However, given that the section imposes obligations on a contracting party over and above those which have been contractually agreed, one should not be too ready to give an unnaturally wide meaning to any of its expressions. Quite apart from that, the fact that one might have expected words in a statute to cover a particular situation is not enough to justify giving those words an unnatural meaning in order to ensure that they do so. In this case, such a wide reading would be very difficult to reconcile with the wording of section 11(1A)(a), especially in the light of the limitation to “the building”. Further, the fact that section 11(1)(a) is specifically extended to cover “drains, gutters and external pipes” tends to support the notion that when it refers to the “exterior”, the word is to be given a natural, rather than an artificially wide, meaning.

This conclusion seems to me to be consistent with the approach of the Court of Appeal in Campden Hill, where, as explained above, the natural meaning was adopted, and an unnatural wide meaning was rejected, when interpreting the words “structure and exterior of the dwelling-house” in what is now section 11(1)(a). As Mr Rainey QC said in his submissions on behalf of Mr Kumarasamy, the decision of the Court of Appeal in this case, although on a different subsection 11, is hard to reconcile with the reasoning in Campden Hill.

Instead, the Court of Appeal in this case relied on Brown v Liverpool Corpn [1969] 3 All ER 1345, where the premises consisted of a terraced house to which access was obtained from the street through a gate, down some steps and along a two metre path which led to the front door of the house. The court held that the steps were part of the exterior of the dwelling-house for the purpose of section 32(1)(a) of the 1961 Act. Danckwerts LJ said at p 1346 that, as the steps were “the means of access” to the dwelling-house in question, they were “plainly part of the building”. Salmon LJ at p 1346 agreed, but thought the case was not “by any means free from difficulty, or, indeed, from doubt” and emphasised that his decision was based “on the particular facts of this case” and not on “any general principle of law”. Sachs LJ at p 1347 said that the case had “caused [him] no little difficulty”, that he had “considerable hesitation” and that the argument was “a very close run thing”; while he accepted that the covenant did not apply to “those parts of the demise that are not part of the building itself”, he considered that the issue was “one of degree and fact”, and that the judge had been “entitled” to conclude that the steps were within the covenant.

In my view, that decision was wrong. The fact that a piece of property is a necessary means of access to a building cannot be sufficient for it to constitute part of the exterior of that building. Steps separated from the outside of a building by a two metre path cannot, as a matter of ordinary English, be said to be part of the exterior of that building. And the passages I have quoted from the brief judgments of Salmon and Sachs LJJ get close to impliedly acknowledging that simple proposition. I note a degree of understandable scepticism in the subsequent Court of Appeal decision of Hopwood v Cannock Chase District Council [1975] 1 WLR 373, which I consider was rightly decided, about the reasoning and conclusion in Brown. Indeed, it is very difficult to reconcile the approach of the Court of Appeal in Brown with that in Campden Hill (where I note that Brown and Hopwood were both cited in argument).

In the light of this conclusion, it is strictly unnecessary to consider the other two issues raised by the appeal. However, as they have been fully argued, and one of them is certainly of some significance (and was in my opinion wrongly resolved by the Court of Appeal) and the other is not without significance, it is right to address them. I shall do so on the assumption (contrary to what I have just concluded) that the paved area is part of the exterior of the front hall of the Building.

Under the Headlease, Mr Kumarasamy was granted a right of way over the front hall, and, as a matter of property law, a right of way over land constitutes an interest in that land, although it does not constitute an estate in that land - see subsections (1), (2)(a) and (3) of section 1 of the Law of Property Act 1925. It is true that the subsequent grant of the Subtenancy effectively deprived Mr Kumarasamy of any practical benefit from the easement so long as it continued. However, that does not alter the fact that, just as he retained his leasehold interest in the Flat, he retained his leasehold easement over the front hall, even though he had sublet the Flat and the easement to Mr Edwards (and any doubt about this is put to rest by section 1(5) of the 1925 Act). Therefore, there is obvious force in the argument, which Lewison LJ had little hesitation in accepting, that Mr Kumarasamy had an “interest” in the front hall (and indeed in the paved area), within the meaning of section 11(1A)(a).

On behalf of Mr Kumarasamy, it is argued that, at least for the purposes of section 11(1A)(a), he nonetheless did not have an “interest” in the front hall once he had effectively disposed of that right of way to Mr Edwards under the Subtenancy. There is obvious practical attraction, at least at first sight, in the contention that is unlikely that Parliament can have intended that the headlessee of a single flat, whose interest in the common parts is simply as a means of access to and egress from the flat, should have an implied liability to his subtenant of the flat to repair the common parts. After all, during the currency of the subtenancy, the headlessee will have little reason to go onto the common parts and will enjoy very limited, if any, rights of any practical value over them in his own right, because, when he visits the flat, it will normally be as an invitee of the subtenant.

However, on closer analysis, I do not consider that contention can be right. First, there would have to be a powerful reason not to give the word “interest”, when it appears in a property statute, its normal meaning in law. Secondly, if the word is to be given a limited meaning, it is hard to identify a satisfactory way to cut it down, which is consistent with the general policy of section 11. The only possible way of excluding the common parts of the Building in the present case from the ambit of Mr Kumarasamy’s statutory liability to Mr Edwards, would be to limit the word “interest” to an interest in possession. However, quite apart from the fact that this would involve reading words into a statute when it does not appear to be necessary, such an interpretation would scarcely be consistent with the liability of a landlord under subsections 11(1)(a) and 11(1A)(a), which impose repairing obligations for items demised to the tenant, which, ex hypothesi, are not in the possession of the landlord.

Thirdly, if the headlessee has no liability to a subtenant for disrepair in the common parts, the subtenant would be without any contractual remedy for damage suffered as a result of such disrepair. It is true that he may have a remedy against the headlessor or freeholder of the building under section 4 of the Defective Premises Act 1972, but that would be of very limited value. (I note that a similar argument based on the Occupiers Liability Act 1957 does not seem to have impressed the House of Lords in Liverpool City Council v Irwin [1977] AC 239 - see at pp 254 and 257, per Lord Wilberforce and Lord Cross of Chelsea respectively). On the other hand, if the subtenant has a claim for disrepair against the headlessee, the headlessee can normally expect to pass on the claim to the freeholder.

Fourthly, quite apart from his rights against the headlessor, it is not as if the headlessee would be without protection in such a case. When subsection (1A) was introduced by the 1988 Act, subsection (3A) was also introduced, and it was to the following effect:

Mr Rainey contends that the reasoning of Jacob LJ, giving the judgment of the Court of Appeal in Niazi Services Ltd v van der Loo [2004] 1 WLR 1254, assists the argument that Mr Kumarasamy retained no interest in the common parts of the Building after he had sublet the Flat. Niazi was another case where a subtenant of a flat sought to invoke section 11 against the headlessee whose headlease included no other property in the building. However, that case was concerned with whether the headlessee was liable under section 11(1A)(b) for a defect in the water supply to the top floor flat in a building, owing to “inadequate supply upstairs when water was being drawn downstairs”. The actual decision and reasoning are of no assistance in this case, which is of course concerned with section 11(1A)(a). It is true that, in para 21 of his judgment, Jacob LJ referred to section 11(1A)(a) and said that “the lessor's extended liability is limited to the obligation to keep in repair the structure and exterior of any part of the building in which he has an estate or interest” and that in that case, the headlessee “has no estate or interest in any part of the building except the top floor flat”. However, he had no reason to consider, and presumably was not considering, whether the headlessee had a right of way over the staircase leading to the top floor flat, or (if there were any) other common parts of the building. If he was directing his mind to that point, he was wrong in what he said.

Where a landlord or a tenant (or anyone else) covenants to keep premises in repair, the general principle is that the covenant effectively operates as a warranty that the premises will be in repair. That principle has been laid down in a number of cases, which were discussed and applied by the Court of Appeal in British Telecommunications Plc v Sun Life Assurance Society Plc [1996] Ch 69. Accordingly, as soon as any premises subject to such a covenant are out of repair, the covenantor is in breach, irrespective of whether he has had notice of the disrepair, or whether he has had time to remedy the disrepair. However, this general principle is subject to exceptions, which are based on normal principles applicable to the interpretation of contracts. The most obvious exception is where the covenant is qualified by an express term, like the freeholder’s covenant in the Headlease in this case - see the end of para 3 above.

A further exception to the general principle, which is relevant in the present case is the rule (which I shall refer to as “the rule”) that a landlord is not liable under a covenant with his tenant to repair premises which are in the possession of the tenant and not of the landlord, unless and until the landlord has notice of the disrepair. The rule has been slightly differently expressed in different cases, but it is based on the normal principle upon which a term is implied into a contract, namely obviousness or necessity. (Accordingly, in accordance with normal principles governing the implication of terms, it could not be invoked where the parties had expressly agreed that the landlord is to be liable for such disrepair irrespective of whether or not he had had notice of it.)

This rule was first formally expressed in Makin v Watkinson (1870) LR 6 Ex 25 (although it was voiced in an interlocutory observation in Moore v Clark (1813) 5 Taunt 90, 96 by Sir James Mansfield CJ and Gibbs J). In Makin, a building had been demised under a lease which contained a covenant by a landlord to keep the main walls and roofs in repair. Channell and Bramwell BB considered that commercial necessity justified implying a term that the obligation to repair only arose once the landlord had had notice of the disrepair. Bramwell B at p 28 said that he was “irresistibly driven” to hold that the parties cannot have intended that a landlord “should keep in repair that of which he has no means of ascertaining the condition”. He explained this at p 30 by reference to the general proposition that “when a thing is in the knowledge of the plaintiff, but cannot be in the knowledge of the defendant, but the defendant can only guess or speculate about the matter, then notice is necessary”. Channell B said much the same at pp 27-28.

Sir Richard Collins MR took the same view in Tredway v Machin(1904) 91 LT 310, 311, where he said that the rule is based on the fact that “the landlord is not the occupier of the premises, and has no means of knowing what is the condition of the premises unless he is told, …, whereas the occupier has the best means of knowing of any want of repair”. Brett J pithily explained the rule thus in The London and South Western Railway Co v Flower (1875) LR 1 CPD 77, 85: “where there is knowledge in the one party and not in the other, there notice is necessary”.

In an Irish appeal, Murphy v Hurly [1922] 1 AC 369, the House of Lords had to consider the basis for the rule, which, on the facts, they held did not apply in that case. At p 375, Lord Buckmaster said that the rule had to be considered by reference to “the actual facts existing in each case”, and it was based “upon the consideration whether the circumstances are such that knowledge of what may be required to be done to comply with the covenant cannot reasonably be supposed to be possessed by the one party while it is by the other”. At p 385, Lord Atkinson described “the presumption upon which the right to notice is stated to depend” as being “that the tenant being in occupation has a full opportunity of seeing and knowing the condition of the premises he occupies and their need of repair, while the landlord has no such opportunity”. Lord Sumner said at p 387-388 that the reason for the rule was “(1) that the tenant is in occupation and the landlord is not; (2) that the tenant, therefore, has the means of knowledge peculiarly in his own possession …; and (3) … the repairs of dwelling-houses … are … not … such as to demand of the landlord incessant vigilance …”

Morgan v Liverpool Corpn [1927] 2 KB 131 was a case like the present, in that it involved a statutorily implied liability on a landlord of a dwelling (in that case a house) to keep the dwelling fit for human habitation and in good repair. Lord Hanworth MR at pp 141-142 said that the fact that the liability originated in statute did “not put it on higher authority” than a contractually agreed covenant. Atkin LJ at p149 took the same view saying that the statutory obligation was “imposed as a contractual term and as such it appears to be only available to the tenant because it is a term of the tenancy”. Lawrence LJ agreed.

At p 143, Lord Hanworth expressed the rule in these terms: “it is the duty of the tenant to inform the landlord, if there is to be a responsibility in respect of a breach of his covenant enforced against the landlord”. At p 150, having described the reason for the rule as “obvious”, Atkin LJ explained that, as “[t]he landlord has given the tenant exclusive occupation of the house” and “therefore, is not in a position to know whether the house is in repair or out of repair, and … it would be quite contrary to justice to impose an obligation to repair of this kind upon a landlord in respect of matters of which he has in fact no knowledge”. Lawrence LJ said at p 153 that the “foundation” of the rule is that “the tenant in occupation is generally in a far better position to know of any want of repair”.

At pp 150-151, Atkin LJ referred to the fact that the statute involved gave the landlord “a right of access”, but said that this was “quite insufficient to redress the injustice that would arise from imposing this obligation [sc an obligation to remedy disrepair of which he had no notice] upon the landlord”, and Lord Hanworth and Lawrence LJ took the same view.

McCarrick v Liverpool Corpn[1947] AC 219 was another case which involved a statutorily implied covenant by a landlord to keep a demised house fit for habitation. The appeal was treated as an appeal against the decision in Morgan, and all five members of the House of Lords agreed with the reasoning of Atkin LJ, both on the applicability of the rule to a statutorily implied covenant and on its applicability even in a case where the landlord had the right to enter and inspect the premises (see at pp 223, 226, 229, 230, and 231-232, per Lord Thankerton, Lord Porter, Lord Simonds, Lord Macmillan and Lord Uthwatt respectively). At p 226, Lord Porter cited with approval Lord Sumner’s explanation in Murphy for the rule. Lord Uthwatt explained at p 232 that it was unreasonable from the point of view of the tenant, as well as that of the landlord, if performance of the landlord’s covenant to repair premises in the possession of the tenant was not subject to the landlord having notice of the disrepair, adding that “[t]he only part the tenant is on this basis required to play in performance is, that knowing what he wants, he should say so”.

In O’Brien v Robinson [1973] AC 912, the House of Lords confirmed that the rule applied to a covenant to repair implied into a tenancy by section 32(1)(a) of the 1961 Act. The arguments largely reflected those considered in McCarrick, and the outcome was the same, in that it was unanimously decided that the reasoning in Morgan was correct and applied in that case. It was also made clear that the rule applied to defects which the tenant did not know about, and even to those which he could not reasonably be expected to discover - see at pp 925, 930, per Lord Morris of Borth-y-Gest and Lord Diplock respectively. At p 926, Lord Morris (with whom Lord Cross agreed) also considered that, where the rule applies, a landlord will be liable once he has notice of the defect, even if that notice does not emanate from the tenant. However, Lord Diplock, with whom Lord Simon of Glaisdale and Lord Reid (as well as Lord Cross) agreed, preferred to keep that point open.

Two preliminary questions arise in relation to the applicability of the rule to lettings of flats. The first question is whether, where the landlord of a flat agrees to repair the structure and exterior, the applicability of the rule to the structure and exterior of the flat itself may in some cases depend on whether or not the demise is limited to the internal surfaces of the walls, ceilings and floors (as it is under the Headlease in the present case). In my view, the rule would apply but only to the extent that the structure is included in the demise. If a part of the structure included within a tenant’s letting is out of repair, then the tenant is in possession of that part of the structure and the landlord is not. Accordingly, the rule would apply to the landlord’s obligation to repair that part of the structure. However, if that part of the structure is excluded from the demise, it would not be in the possession of the tenant (indeed it would presumably be in the possession of the landlord) and so the rule would not apply.

This may seem a rather technical, or in some cases an almost capricious, distinction, but I believe that it follows from the various dicta which I have quoted from the cases concerning the rule. If the tenant is not in possession (and, a fortiori, if the landlord is in possession) of part of the structure which is out of repair, then there is no reason for excluding the general principle set out in para 29 above. The rule is in any event demonstrably based as much on principle as on practicality, given that, as was confirmed in O’Brien, it applies to disrepair to demised property even where the disrepair is not reasonably discoverable by the tenant. Further, the distinction between property let to the tenant and property not so let is one which leaves the law as to the applicability of the rule in a tolerably clear state, and clarity is self-evidently a desirable feature of any rule or principle.

The second question is rather more difficult in my view. It is whether a landlord, who has covenanted with one tenant to repair the structure but has let part of the structure to another tenant, can thereby automatically escape liability to the first tenant for disrepair of that part until he has had notice of that disrepair. Subject to one point, this question could be characterised as being whether the rule applies to property which is in the possession of neither the landlord nor the tenant - ie can the rule apply to property which has not been demised to the tenant? It can be said that the dicta in the cases do not speak with one voice on this question, as some appear to emphasise the unfairness of imposing an absolute liability on a landlord in circumstances where he is not in possession and therefore not in a position to know of any disrepair, whereas other dicta indicate that the rule also depends on the tenant being in possession and therefore in a position to know of the disrepair.

Given that one is concerned with an implied term, it may be dangerous to generalise (as the point discussed in paras 49-58 below demonstrates). However, I have concluded that the rule does not normally apply to premises which are not in the possession of the tenant. Most of the dicta describing the reason for the rule rely not only on the landlord’s lack of ability to know, but also on the tenant’s advantageous position; and some do so very strongly - see eg what was said in the earlier cases cited in para 32 above and the observations of Lord Atkinson and Lord Sumner in Murphy,and Lord Porter and Lord Uthwatt in McCarrick. Further, the dicta which do not refer to the tenant’s privileged position could well have been taking it for granted, as they were all in cases where the tenant was in possession. Further, as is suggested in some of the cases (in Makin (1870) LR 6 Ex 25, 27-28 per Channell B, in Flower (1875) LR 1 CPD 77, 82, in Murphy at pp 375, 392 per Lord Buckmaster and Lord Parmoor, and in McCarrick [1975] AC 219, 231 per Lord Uthwatt), it seems to me that the rule is an aspect of a wider principle described in these terms by Lord Abinger CB in Vyse v Wakefield (1840) 6 M & W 443, 452-453:

This would therefore suggest that it is not normally open to a landlord who has agreed to repair the structure, to invoke the rule against a tenant of a flat in relation to disrepair of part of the structure which the landlord has let to another tenant, or indeed were not in the possession of the landlord for some other reason. The only argument against this might be that the lettings of flats in a block on the same terms can be treated as a sort of scheme between (i) the landlord and (ii) the tenants as a group. I do not accept that is a valid analysis. Once it is determined that the rule only applies to property in the possession of the tenant, there is no warrant for implying it to any other property - unless of course it is justified by the terms of the particular tenancy and the surrounding circumstances. After all, it is normally open to a landlord to add a term expressly incorporating the rule (as was done in the Headlease in this case). In the light of section 12(1) of the 1985 Act, that cannot, I think, be done in relation to the covenant implied by section 11, but it may well be that a landlord could protect himself to some extent by imposing an obligation on the tenant to give notice to him of any disrepair which has come (or, possibly, even reasonably should have come) to the attention of the tenant.

I turn now to Mr Rainey’s submission that, in every case where a tenant relies on a covenant implied by section 11, a landlord is entitled to invoke the rule. Thus, even in relation to property which is undoubtedly in the possession of the landlord, he submits that section 11 cannot be relied on by a tenant in relation to any disrepair unless the landlord has had notice of the disrepair.

This submission is supported by Dowding and Reynolds in Dilapidations: The Modern Law and Practice 5th ed (2013-14), para 20-37 on two grounds, namely (i) the section 11 repairing covenant is most likely to apply to property which is either within the relevant demise or so close to the relevant demise as to be more easily observed by the tenant than the landlord, and (ii) the speeches in O’Brien are more consistent with that analysis.

I agree with Lewison LJ that this submission must be rejected. It is clear from Morgan, McCarrick and O’Brien that the repairing covenant implied by section 11 is to be interpreted and applied in precisely the same way as a landlord’s contractual repairing covenant.As I have sought to explain, the rule in relation to such covenants is that, until he has notice of disrepair, a landlord should not normally be liable for disrepair of property in so far as it is in the possession of the tenant. I can see no basis as a matter of principle for departing from the rule when it comes to covenants implied by section 11, which is what Dowding and Reynolds suggest. And, as a matter of practicality, once one departs from the clear rule, there is a real risk of uncertainty and confusion - for instance, it could be difficult to resolve whether, on particular facts, it is more likely that the tenant or the landlord should have noticed the disrepair.

It is true that in many cases where section 11 applies, the tenant may be in a better position than the landlord to observe the disrepair, but that is not the basis on which the rule has been justified in the cases - eg it applies to disrepair which could not be reasonably discovered. And, quite apart from its uncertainty, given that the rule is justified by the normal principles governing an implied term, such a yardstick would not, I think, satisfy the requirement of necessity or of obviousness.

I accept that the speeches of Lord Morris and Lord Diplock in O’Brien contain nothing to suggest that there might be cases where a landlord could be liable under his statutorily implied covenant without having been given notice. However, I do not regard that as significant. They were concerned with a case where there could be no doubt but that the item which had fallen into disrepair (a ceiling in a room of the demised premises) was included in the demise to the tenant, and therefore on any view the landlord could claim the benefit of the rule.

The present case is different from the cases which have so far been decided in relation to the rule, because it is concerned with the application of a landlord’s repairing covenant to property which is not in the possession of either the landlord or the tenant, although it is property over which they each have a right of way as discussed in paras 23-28 above. However, in my judgment, the application of the reasoning upon which the rule is based justifies the conclusion that the landlord’s (assumed) obligation to repair the paved area is only triggered once he has notice of any disrepair for which the tenant would seek to make him liable.

As explained above, the landlord, Mr Kumarasamy, has a lease of a single flat which includes the right to use the front hall and paved area, and he has effectively sublet his right to use and occupy the flat and to use the hall and paved area to the tenant, Mr Edwards. In so far as the landlord had any right over the hall and paved area, he has effectively disposed of that right to the tenant for the term of the Subtenancy just as much as he has disposed of his right to use and occupy the Flat to the tenant for the term of the Subtenancy. During the term of the Subtenancy, it is the tenant who uses the common parts, not the landlord, just as it is the tenant who occupies the flat, not the landlord. It is true that the tenant does not enjoy exclusive possession of the common parts, but he is present on them every time he comes to or leaves the flat. The present issue is concerned with the relationship between a particular landlord and a particular tenant, and the landlord has effectively lost the right to use the common parts and the tenant has acquired the right to use them, for the duration of the Subtenancy.

It is true that the landlord has the right to use the common parts as against the freeholder, but that is irrelevant for present purposes, in the same way as the fact that he has the right to occupy the Flat as against the freeholder does not prevent him from invoking the rule against the tenant in relation to any part of the demised premises which he has covenanted to repair. It is also true that the landlord has the right to use the front hall to get access to the Flat in order to inspect and repair it, but that cannot deprive him of the right to invoke the rule, any more than his right to visit the Flat itself for those purposes would deprive him of the right to invoke the rule in relation to his repairing obligations in relation to the Flat.

To use the words of Collins MR in Tredway 91 LT 310, 311, as against the landlord, the tenant “has the best means of knowing of any want of repair” in the common parts, or, to adapt what Lord Atkinson said in Murphy[1922] AC 369, 385, the tenant “has a full opportunity of seeing and knowing the condition of the [common parts he uses] … and their need of repair, while the landlord has no such opportunity”. To adapt Atkin LJ’s formulation in Morgan[1927] 2 KB 131, 150, the landlord “is not in a position to know whether the [common parts are] in repair or out of repair”, whereas the tenant is, or, per Lawrence LJ in the same case at p 153,“the tenant [using the common parts] is generally in a far better position to know of any want of repair”.

Mr Rainey also argues that subsection (3A) of section 11 supports Mr Kumarasamy’s case that the rule shall be extended to a case such as this, as the landlord cannot be required to use “reasonable endeavours” to have repairs carried out until he knows of the relevant disrepair. I am unconvinced by this argument, as it seems to me to be circular. Nonetheless, there is something in the point that subsection (3A) shows that Parliament was concerned not to impose an unrealistically demanding duty on a landlord. And that provides a little further support for the conclusion that, in a case such as the present, the landlord is not in breach of his statutorily implied repairing obligation until he has notice of the disrepair.

The Court of Appeal reached a different conclusion. That was partly because they took the view that the rule only applied to disrepair within the demised premises themselves. But that is because all the cases so far have been concerned with such disrepair. There is no reason why the rule cannot be extended to cover a state of affairs not so far considered judicially, and, as just explained, it seems to me that the reasoning on which the rule is based means that it should be so extended in the present case. The potential harshness on a headlessee of a single flat of imposing a covenant to repair the common parts, which he has effectively transferred to the tenant his right to use, is mitigated by the need for notice of any disrepair before the covenant becomes activated (and see the end of para 27 above).

It is also suggested that it is inappropriate to extend the rule to a case where section 11(1A) applies, when Parliament had not included a need for notice when inserting that subsection into section 11, given that it had expressly limited the landlord’s liability under that subsection by inserting subsections (1B) and (3A) at the same time. I do not consider that to be a good point. There is nothing about the need for notice in section 11 as originally enacted and yet there is no doubt that the rule applied and applies to the covenant in section 11(1)(a); it seems to me that it would be positively surprising if it did not also apply to any subsequent extensions to the ambit of section 11(1)(a), unless of course it was expressly or by necessary implication excluded, which it is not. Further, as stated in para 53 above, the concern with practicality demonstrated by subsection (3A) appears to me to provide a little support for the requirement of notice in a case such as this.

Mr Benson QC, who appears for Mr Edwards, also submits that the implication of the rule in the present case would be inconsistent with the decision of the Court of Appeal in British Telecommunications. I do not agree. That case was concerned with disrepair to part of the exterior of a building on the fifth floor. It is true that the tenant in that case may have had rights in respect of that part, but it was not a right to be frequently present on, a right frequently to use physically (if not to occupy), the property out of repair, as in the present case. In any event, the issue was very different, namely whether, in a case where it was (rightly) common ground that the rule did not apply, a landlord would be in breach the moment disrepair occurs, or whether he would be in breach only after the expiry of a reasonable time to remedy the disrepair.

The Court of Appeal also relied on the fact that the law implied a right in Mr Kumarasamy, as a headlessee and tenant of the right to use the common parts, to go on to the common parts to repair them, invoking the decision in Newcomen v Coulson (1877) 5 Ch D 133. I do not consider that to be a good point for two reasons. First, a right of way does not necessarily carry with it a right to carry out repairs to the way: such an ancillary right only arises as a matter of implication, and is normally justified because the servient owner has no obligation to repair the way. As it is put in Gale on Easements 19th ed (2012), para 1-90, “[t]he ancillary right arises because it is necessary for the enjoyment of the right expressly granted”. In the present case, the Headlease, under which Mr Kumarasamy was granted the right to use the common parts, contains an obligation on the freeholder to keep the common parts in repair. Accordingly, I do not consider that it would be appropriate to imply such an ancillary right: it is not necessary for business efficacy, nor is it obvious. (It may well be that such a right could arise in extremis as Etherton J suggested in Metropolitan Properties Co Ltd v Wilson [2002] EWHC 1853; [2003] L & TR 226, paras 49-51, but that cannot possibly do for present purposes).

Secondly, even if a term such as that envisaged by the Court of Appeal could be implied, I do not see how it would help the argument that the rule should be displaced in this case. As mentioned above, it is well established that the fact that a landlord has the right to go into the demised premises to inspect and carry out repairs does not mean that the rule is displaced so far as disrepair to the premises is concerned. By the same token, even if the landlord had the right to repair the common parts, I fail to see why that should displace the rule if it would otherwise apply to disrepair of the common parts.

Finally, I should say that, where a flat is let under a tenancy to which section 11 applies, by a landlord who owns the building in which the flat is situated, it seems to me likely that, in so far as the statutory covenant extends to repairing the common parts, it would not normally be subject to the rule. That is because such landlord would ordinarily be in possession of the common parts. Indeed, it may be that the rule would not apply in any case where the landlord is headlessee of more of the building than the single flat he has sublet, as he would have exercisable rights over the common parts in his capacity of headlessee of property other than the flat in question. However, those issues have, understandably, not been even touched on in argument, and it would be wrong to express a concluded view on them.

I would therefore allow this appeal, on the ground that, although he had a sufficient “interest” in the front hallway and paved area for the purposes of section 11(1A)(a), Mr Kumarasamy was not liable for the disrepair which caused Mr Edwards’s injury, as (i) he could only be liable if the paved area was “part of the exterior of the front hall” and it was not, and (ii) he could only be liable if he had had notice of the disrepair before the accident and he did not.

Her Honour Judge May QC reached the correct conclusion on these two points (although, reflecting the way that the case was argued before her by counsel other than those appearing before this court, she slightly mischaracterised the first point). Accordingly, she dismissed Mr Edwards’s claim, and I would do so too.

I agree that the appeal should be allowed for the reasons given by Lord Neuberger. My only reservation concerns a part of his judgment which does not relate directly to the issues in the appeal, and on which we have heard no argument.

In paras 40-44, he considers the application of the “rule” to cases where (unlike the present) parts of the external structure have been included in the relevant demise, or in a demise by the same lessor to another tenant. While I understand the logic of his observations (even if somewhat “technical”, as he says), I am not convinced that it is safe to lay down a general rule for all such cases. As he rightly says (para 30) the question ultimately depends on ordinary principles for the implication of terms, such as obviousness or necessity. I would prefer not to consider such issues in the abstract without regard to all the circumstances, including the commercial or practical reasons which might have led to the grant in a particular case. I doubt in any event that it is a problem likely to arise often in practice. For the moment I would prefer to reserve my position.