City of London v Various Leaseholders of Great Arthur House

The issue on this appeal is the extent to which the lessees of Great Arthur House are liable, under the terms of their leases, to contribute to the cost of carrying out substantial works to the structure and exterior of the building.

Great Arthur House is a Grade 2 listed building, consisting of 120 flats, on the Golden Lane Estate in London EC1. It was constructed in 1957. The flats are held under leases originally granted pursuant to the right to buy. Each lease was granted for a term of 125 years from 10 May 1982. The reversion is held by the City of London Corporation. The leases with which we are concerned were granted between 1983 and 2015. The leases all contain a covenant requiring the lessee to pay to the Corporation a reasonable part of the costs of carrying out specified repairs and of insuring against risks involving specified repairs. The expression “specified repairs” is defined by the leases as meaning:

“… repairs carried out in order:

(i)

to keep in repair the structure and exterior of the premises and of the Building in which they are situated (including drains gutters and external pipes) not amounting to the making good of structural defects;

(ii)

to make good any structural defect of whose existence the Corporation has notified the tenant in the notice served pursuant to [statutory requirements] which therein stated the Corporation’s estimate of the amount (at then current prices) which would be payable by the tenant towards the costs of making it good (such defects being listed in the Fourth Schedule hereto) or of which the Corporation does not become aware earlier than ten years after the grant hereof and

(iii)

to keep in repair any other property over or in respect of which the tenant has any deemed rights.”

As originally constructed, the building consisted of a concrete frame with the main east and west elevations largely clad in curtain wall glazing, contained by a framework of aluminium sections fixed to a timber sub-frame. That, in turn, was fixed to the edge of the floor slabs and ends of the cross walls of the main structure. The building has suffered from water penetration for many years. The Corporation has commissioned a number of expert reports; in particular, a report by Jenkins & Potter (who are structural engineers) produced in August 2002. That report concluded:

Further reports commissioned by the lessees confirmed these problems; and identified further ones.

The Corporation undertook a scheme of works, beginning in February 2016 and concluding in the summer of 2018, to address the problems. The works cost approximately £8 million; and consisted of:

It is those works that are the subject of the current dispute. If the Corporation is entitled to pass on the full cost of those works to the lessees, it will result in a potential bill of over £72,000 per flat.

In a nutshell, the Corporation’s argument is that whether works are or are not works of repair is a question of fact and degree. Works do not cease to be works of repair merely because they simultaneously eradicate a defect in the building that has been there from the time it was constructed, where that defect has caused damage to or deterioration in the subject matter of the covenant. On the other hand, if works eradicate a defect which has resulted in neither damage to nor deterioration in the subject matter of the covenant, then those works are not works of repair. If works are repair, properly so called, then they fall within paragraph (i) of the definition, and are properly chargeable to the lessees.

The Upper Tribunal Lands Chamber (Fancourt J, President) rejected the

It is common ground that the legal background is relevant to the interpretation of the definition. I begin with the common law.

Until the landmark decision of Forbes J in Ravenseft Properties Ltd v Davstone (Holdings) Ltd [1980] QB 12 it had been widely thought that a repairing covenant did not require the covenantor to make good what was described as an “inherent defect” in the subject-matter of the covenant. An “inherent defect” used in this sense was a fault in the original design of a building as opposed to faulty materials or workmanship: see [1980] QB at 18. Forbes J held that that was a misconception. Whether works amount to repair was a question of fact and degree; and the relevant inquiry was whether the carrying out of the works in question would involve giving back to the covenantee a wholly different thing from that which was demised. If a scheme of works did not amount to giving back to the covenantee a wholly different thing to that which was demised, they were works of repair. They did not cease to be works of repair merely because they also eradicated an inherent defect which had given rise to the need to repair in the first place.

This test was elaborated in McDougall v Easington DC (1989) 58 P & CR 201, 207 in which Mustill LJ said:

“It is sufficient to say that in my opinion three different tests may be discerned, which may be applied separately or concurrently as the circumstances of the individual case may demand, but all to be approached in the light of the nature and age of the premises, their condition when the tenant went into occupation, and the other express terms of the tenancy:

(i)

Whether the alterations went to the whole or substantially the whole of the structure or only to a subsidiary part;

(ii)

Whether the effect of the alterations was to produce a building of a wholly different character than that which had been let;

(iii)

What was the cost of the works in relation to the previous value of the building, and what was their effect on the value and lifespan of the building.”

Nevertheless, other cases introduced a further refinement. An obligation to repair was not triggered unless and until there had been some damage to the subject matter of the covenant: Quick v Taff Ely BC [1986] QB 809. As Lawton LJ put it in that case: “there must be disrepair before any question arises as to whether it would be reasonable to remedy a design fault when doing the repair”. It is only once some damage to the subject matter of the covenant has occurred, that the tests posed by Forbes J or Mustill LJ become relevant. This was taken a stage further in Post Office v Aquarius Properties Ltd (1986) 54 P & CR 61. In that case this court held that where defects in the building had existed since the date when it was constructed, but there had been no damage to or deterioration in the condition of the building, a repairing covenant did not require the defect to be eradicated. In that respect it did not matter whether the original defect “resulted from error in design, or in workmanship, or from deliberate parsimony or any other cause”.

It is common ground that, because the leases were granted pursuant to the right to buy, the legislative background is also an aid to interpretation of the covenants. At the time when the first of the leases was granted, the relevant legislation was contained in the Housing Act 1980. Despite changes in the legislation, the leases have remained in the same form throughout. It is not suggested that leases granted later in time should be interpreted differently from those granted earlier in time.

Schedule 2 paragraph 13 of the Housing Act 1980 imposed on the landlord a statutory obligation:

“(a)

to keep in repair the structure and exterior of the dwellinghouse and of the building in which it is situated (including drains, gutters and external pipes) and to make good any defect affecting that structure;

(b)

to keep in repair any other property over or in respect of which the tenant has any rights… ”

Because that obligation is imposed by statute, the lease itself contains no express obligation to like effect on the part of the landlord.

Paragraph 15 of that Schedule provided that:

“Any provision of the lease or of any agreement collateral to it shall be void insofar as it purports –

(a)

…

(b)

to enable the landlord to recover from the tenant any part of the costs incurred by the landlord in discharging or insuring against his obligations under paragraph 13(1)(a) or 13(1)(b) above,…

but subject to section 19 of this Act and paragraph 16 below.”

Paragraph 16 provided:

“A provision is not void by virtue of paragraph 15 above insofar as it requires the tenant to bear a reasonable part of the costs of carrying out repairs not amounting to the making good of structural defects or of the costs of making good any structural defects falling within paragraph 17 below or of insuring against risks involving such repairs or the making good of such defects.”

Finally, paragraph 17 provided:

“A structural defect falls within this paragraph if—

(a)

the landlord has notified the tenant of its existence before the lease was granted; or

(b)

the landlord does not become aware of it earlier than 10 years after the lease was granted.”

These paragraphs are clearly meant to be read together. Paragraph 13 requires the landlord to make good defects. The defects covered by the obligation are “defects affecting that structure”; that is to say the structure and exterior of the dwelling-house and of the building in which it is situated. When the later provisions refer to “structural defects” it is to those defects that they are referring. The defects to which these paragraphs refer are not confined to “inherent defects” as that expression was used in the law of landlord and tenant, before the landmark decision in Ravenseft.

The second limb of paragraph 13 (1) (a) (making good structural defects) is a freestanding obligation which applies to any structural defect, whether it has caused damage or not.

Paragraph 15 of Schedule 2 to the Housing Act 1980 starts by imposing a blanket ban on recovery from the tenants of any part of the costs of the landlord’s compliance with paragraph 13. The words “any part” are important. But paragraph 16 creates a carve-out from that prohibition. The first part of that carve-out does not maintain any distinction between the eradication of a structural defect in the course of effecting repairs, where that defect has caused damage; and the eradication of structural defect where it has not. On the contrary it explicitly recognises that some repairs may eradicate structural defects which have caused damage, otherwise they would fall outside the scope of the first part of the carve-out in paragraph 16. Whether or not a structural defect has caused damage, the cost of eradicating it cannot be recovered from the tenants, unless it falls within the second part of the carve-out. The second part of paragraph 16 refers to the making good of structural defects even if that work goes beyond repair; but that cost is recoverable only where the tenant has been notified of the defect before entry into the lease, or where the landlord did not become aware of it within 10 years from the date of grant.

These provisions have been through various iterations some of which have modified them; but not so as to affect the issue arising on this appeal. But the statutory scheme was significantly modified by the Housing and Planning Act 1986 which inserted new paragraphs 16A to 16D into Schedule 6 of the Housing Act 1985 (which had, by then, become the principal statute governing the right to buy).

Paragraph 16B now relevantly provides:

“(1)

Where a lease of a flat requires the tenant to pay service charges in respect of repairs (including works for the making good of structural defects), his liability in respect of costs incurred in the initial period of the lease is restricted as follows.

(2)

He is not required to pay in respect of works itemised in the estimates contained in the landlord's notice under section 125 any more than the amount shown as his estimated contribution in respect of that item, together with an inflation allowance.

(3)

He is not required to pay in respect of works not so itemised at a rate exceeding—

(a)

as regards parts of the initial period falling within the reference period for the purposes of the estimates contained in the landlord's notice under section 125, the estimated annual average amount shown in the estimates;

(b)

as regards parts of the initial period not falling within that reference period, the average rate produced by averaging over the reference period all works for which estimates are contained in the notice; together, in each case, with an inflation allowance.

(4)

The initial period of the lease for the purposes of this paragraph begins with the grant of the lease and ends five years after the grant…”

Thus under the current legislation, the tenant’s liability to pay is only capped for the period of five years from the date of grant of the lease; and thereafter he may be required to pay for the making good of structural defects. It is not suggested that this legislative change affects the issues on this appeal.

It is, nevertheless, clear that the immediate legal background to paragraphs (i) and (ii) of the definition of “specified repairs” in the leases was paragraphs 16 and 17 of the Housing Act 1980.

Payne v Barnet LBC (1997) 30 HLR 295 concerned the right to buy provisions. The appeal arose in consequence of an order striking out the claim. The main issue in the case was whether the prospective landlord had a duty to disclose costs and potential costs to the prospective lessee over and above the duties imposed by the legislation relating to the right to buy. This court held that it did not. Having quoted paragraph 13 (1)(a) of Schedule 2, Brooke LJ said at 301:

“We will call the repairs referred to in these two paragraphs “ordinary external repairs” as distinct from making good structural defects.

We make this distinction because it appears to us that the draftsman of this schedule was well aware of the vexed problem in landlord and tenant law of distinguishing between a liability to repair and a liability to make good an inherent defect in the property demised (see Woodfall on Landlord and Tenant, Volume 1, paras 13.029–13.037 and the well-known cases there cited). In Post Office v Aquarius Properties Ltd … for instance, this court held that a covenant by a tenant to keep demised premises in good and substantial repair did not impose any obligation on him to remedy a defect in the structure of the premises, whether that defect resulted from faulty design or workmanship, if it had been present from the time the building was constructed and had caused no damage to it. In the Housing Act scheme the landlord is fixed not only with the liability to keep the dwellinghouse's structure and exterior in repair, but also with the liability to make good any defect affecting that structure. However, the requirements he must fulfil if he is to be able to pass on to the tenant any of the expense he may incur in meeting these liabilities are different in each case.” (Emphasis added)

Mr Straker QC relies on this passage as establishing a “bright-line distinction” between two mutually exclusive concepts: repair on the one hand and works to make good inherent defects on the other. It is this submission that is the foundation of his argument. He also submitted that Payne is binding on us. So it is, but only for what it decides as ratio. Brooke LJ’s discussion of the service charge provisions of the legislation was not, in my judgment, part of the ratio of the case. Moreover, since the actual decision of the court was that the action should proceed to trial, that, too, is a further restriction on its precedential value.

But in any event, that is not how I read the passage on which Mr Straker relies. At this point in his judgment, what Brooke LJ was discussing was the extent of the landlord’s obligation: not the extent of the tenant’s obligation to contribute. The landlord’s implied covenant has two separate limbs (joined by the word “and”). In that context, the way in which paragraph 13 was drafted made it clear that not only was the landlord responsible for repairs (where there had been damage to the subject-matter of the covenant) but also for making good structural defects, even where they had not caused damage. Hence the draftsman’s use of the phrase “keep in repair” in limb 1; and “make good” in limb 2. Limb 2 also refers to making good a “defect”; not to making good “damage”. That is also why, in my judgment, Brooke LJ referred to the Post Office case as an illustration of the wide reach of paragraph 13 (1) (a). I do not consider that he intended to cast any doubt on the cited passage in Woodfall which stated in terms that works could be works of repair even though their effect was to eradicate an inherent (or structural) defect, provided that the defect had caused damage and the “fact and degree” test was met. What Brooke LJ described as

“ordinary repairs” was not the totality of works described in paragraph 13 (1) (a): but “the repairs referred to in [those] two paragraphs.

Brooke LJ went on to discuss the extent of the permitted service charge. He said:

“(b)

Ordinary External Repairs

These costs would be included among the service charge headings, but the landlord of a flat would only be entitled to require the tenant to bear a reasonable part of these costs.

(c)

Making good structural defects

There was no obligation to say anything about these in the notice. However, if the landlord wished to pass on to the tenant any part of the cost of making good any structural defect of which he had become aware at any time earlier than 10 years after the lease was granted, he could only require the tenant to bear a reasonable part of such costs if the section 10 notice not only informed the tenant of the existence of the relevant defects, but also stated the landlord's estimate of the amount which the tenant would be liable to pay towards the cost of making them good. (Emphasis added)”

His observations on passing on the cost of “ordinary repairs” is consistent with his earlier statement that that label attached only to “repairs” within the ambit of paragraph 13 (1) (a); and in this passage also, Brooke LJ included all structural defects under the heading “making good structural defects”; not merely those which were eradicated in the course of carrying out repairs. Nor did he limit structural defects to “inherent defects”.

In the Upper Tribunal Fancourt J said at [40]:

“A structural defect is not confined to a so-called inherent defect but must be something that arises from the design or construction (or possibly modification) of the structure of the Building. It is to be contrasted with damage or deterioration that has occurred over time, or as a result of some supervening event, where what is being remedied is the damage or deterioration. That is repair and is not in the nature of work to remedy a structural defect, even if it is a part of the structure that has deteriorated. As a simple example, mastic sealant is part of the structure of a modern building. The replacement of degraded sealant – even with a more modern and better type of sealant – is repair, not the making good of a structural defect; similarly, repointing a flank wall, or replacing spalled brickwork. These are simple examples.”

Mr Straker criticised this as being inconsistent with Payne. He relied in particular on the headnote in the Housing Law Reports which states:

“(3)

“Structural defects” are defects affecting the structure which require making good, as opposed to ordinary items of repair or maintenance; in the context of right to buy applications, structural defects are limited to the narrow category of inherent defects;” (Emphasis added)

I do not accept this submission, for three reasons. First, the legislation does not use the phrase “inherent defects;” and therefore does not contain the suggested limitation. Second, I do not consider that Brooke LJ anywhere decided that the statutory phrase was so confined. In my judgment, therefore, the headnote is inaccurate. Third, if that is what Brooke LJ meant (even though he did not say so) it was an obiter observation, and I would decline to follow it.

Although the legislative background is undoubtedly relevant to the interpretation of a contract, it is not necessarily determinative. In Floe Telecom Ltd v Office of Communications [2009] EWCA Civ 47, [2009] Bus LR 1116 this court considered the interpretation of a licence to operate a mobile telephone network. The legislative background included EU Directives. Mummery LJ said:

“[109] Floe's alternative argument on construction relied on the EC Directives as aids to construction of the licence. Of course, the language in which the licence is expressed must be construed in context. The context may include EC or domestic legislation. For example, a licence may use technical terms without defining them, but against the background of legislation, including EC legislation, in which they are defined. The terms as defined in the related legislation would be aids to the interpretation of the licence.

[110] In this case, however, the licence defines its own terms. It is they and not the contents of the Directives which control the meaning of the licence. The licence authorises the use of radio equipment. The definition of radio equipment does not include GSM gateways, which fall within the definition of different apparatus, “user stations”.”

This is consistent with the general principle of interpretation that where a contract defines its own terms, the court will give effect to the agreed definitions: see, most recently, Rockliffe Hall Ltd v Travelers Insurance Company Ltd [2021] EWHC 412

The definition of “specified repairs” in the lease to a large extent follows the statutory provisions. Part (i) of the definition tracks paragraph 16; and paragraph (ii) tracks paragraph 17. I do not therefore accept Mr Straker’s submission that there is in the lease a “bright-line” distinction between repairs on the one hand and making good structural defects on the other. Works of repair may or may not also eradicate structural defects. If they do, then the cost of making good the structural defects is not recoverable, unless the tenant was notified of them before entry into the lease or the landlord was unaware of them until 10 years after the date of grant. In one respect, however, the scope of the service charge is narrower than that which the statutory scheme allows. That is because the whole of the definition is prefaced by the words “repairs in order to”. It follows, therefore, that unless works qualify as repairs both under the “fact and degree” test, and under the test that the works are carried out in order to repair deterioration or damage, they are outside the scope of the service charge. The point is reinforced by the label used for the term defined. Mr Straker accepted that this was so. It follows that, under the terms of the lease, the cost of eradicating a structural defect is not recoverable from the lessees under paragraph (ii) of the definition unless that defect has caused damage or deterioration in the subjectmatter of the covenant, so as to qualify as “disrepair” within the established meaning of that word; and the resulting scheme of work is properly described as work of repair.

In my judgment Mr Straker’s interpretation gives no real content to the carve-out in paragraph (i) of the definition.

That may well mean that the cost of some works that are within the landlord’s obligation to “make good structural defects” cannot be passed on to the lessees. But that is a consequence of the way in which the definition is drafted; and there is no presumption that the cost of all works that the landlord is obliged to carry out can be passed on to the lessees: Campbell v Daejan Properties Ltd [2012] EWCA Civ 1503, [2013] 1 P & CR 14 at [56].

The second argument focuses on the opening phrase in the definition: repairs carried out “in order” to. Mr Straker submits that what this means is that the purpose of carrying out the works is the determining factor. If the purpose of carrying out the work is to repair the structure or exterior of the building, then the fact that the work also makes good a structural defect does not exclude that work from paragraph (i) of the definition. I agree with this submission, up to a point. I accept that if a scheme of works can fairly be called a scheme of repair, it does not cease to be a scheme of repair, merely because the scheme incidentally involves the making good of a structural defect. But this argument ignores a critical part of the definition. The governing phrase is not “in order to;” but the phrase “not amounting to the making good of a structural defect” which introduces the carve out from paragraph (i) of the definition.

To put the point another way, works fall within paragraph (i) of the definition if:

In short, I do not consider that the phrase “in order to” will bear the weight that Mr Straker would have it bear. Quite apart from anything else, the statutory provisions on which the lease definition was based do not contain any reference to the purpose of carrying out work. Paragraph 16 avoids a service charge in so far as it “requires the tenant to bear a reasonable part of the costs of carrying out repairs not amounting to the making good of structural defects,” irrespective of the purpose for which those works are carried out. If and to the extent that Mr Straker’s interpretation goes further than paragraph 16, the service charge provision would have been, to that extent, void. It is well established that, if there are two realistic interpretations of a contractual provision, one of which means that the provision is invalid, and the other of which means that the provision is valid, the court should adopt the interpretation that results in the validity of the contractual provision: Tillman v Egon Zehnder Ltd [2019] UKSC 32, [2020] AC 154. The application of that principle to this case means, in my judgment, that we should interpret the lease so that it conforms with the applicable legislative scheme.

The Upper Tribunal dealt with this argument at [34] as follows:

“the words starting “not amounting to ….” operate by way of exception or proviso to what has gone before, one would expect the language of the exception to define what is excluded. What is excluded is repairs that amount to the making good of structural defects. The natural reading of these words is that they are works, which in law are properly characterised as repairs, but which amount to the making good of a structural defect. The words “amounting to” seem to me to be used in the sense of what the works in substance do or achieve.”

I agree. Putting the obverse of Mr Straker’s point: merely because works are works of repair according to the “fact and degree” test, does not mean that they are not also works amounting to the making good of structural defects, if that is in fact what they do. I agree with Mr Baker that they may be both repairs in the conventional sense, and also works for the making good of structural defects. To the extent that they do amount to the making good of structural defects, they are excluded from paragraph (i) of the definition.

That leads to the question at the heart of the appeal: what is the meaning of the phrase “not amounting to the making good of a structural defect”? Mr Straker’s first answer was that it was all a question of classification or description. Works of repair fall within paragraph (i) of the definition whether or not they make good a structural

The scope of the exclusion cannot be dependent on whether or not the defect has caused damage. In the first place, as I have said, the overarching requirement that rechargeable works must be works of repair necessarily means that there has been deterioration or damage. Second, if there were a distinction between rectifying structural defects which caused damage (paragraph (i)) and rectifying those which had not (paragraph (ii)) there would be what the FTT described at [90] as a “perverse incentive to landlords to wait until structural defects had caused disrepair before carrying out works”.

Mr Baker’s answer to the question is that what is to be considered is the effect of the works. If their effect is to make good a structural defect, then their cost cannot be passed on to the lessee under paragraph (i) of the definition. Whether that is the effect of the whole or part of a scheme of works is a question of fact, to be determined on the evidence (including expert evidence where necessary). Some parts of a scheme of work may amount to the making good of a structural defect, while other parts may not.

Mr Straker objects that this is an uncertain test. The service charge should be as easy to operate as is possible. In my judgment there are at least two answers to that argument. First, whether a scheme of works falls within the scope of a service charge often gives rise to uncertainty at the edges. The FTT is a specialist tribunal well used to resolving such difficulties; and the powers conferred on that tribunal extend to deciding whether works would fall within the scope of a service charge, whether wholly or partly, even before the works are carried out. So any uncertainty can be cured before the landlord is irrevocably committed to any particular scheme of works. Second, this test is no more uncertain than a test which depends on ascertaining the “purpose” of the works (which must mean more than the landlord’s subjective purpose).

I do not consider that the disadvantages of that interpretation (at least from the perspective of the landlord) which Mr Straker urged upon us outweigh the twin considerations of the natural meaning of the words, and the statutory limitation on the recovery of costs in force at the time that the leases were first granted.

Mr Straker had another argument based on the statutory provisions introduced by the Local Government and Housing Act 1989 which require a local housing authority to maintain a housing revenue account and keep it in balance. I did not consider that this argument shed any light on the question which we have to consider. First, by the time that these provisions came into force, the service charge regime applicable to leases granted under the right to buy had radically changed. In the light of those changes the Corporation could have changed the form of the leases; but chose not to do so. Second, earlier legislation is not usually interpreted by reference to later legislation. Third, whether a housing authority may have to raise rents across the whole of its

Nor do I consider that the interpretation that the Upper Tribunal adopted is in any way contrary to the legislative purpose of the right to buy. The purpose of the legislation was to encourage home ownership by council tenants; often persons of modest means. That was achieved not only by giving them the right to buy at all, but allowing them to buy at a substantial discount from market value. Although it is true (as Mr Straker said in his skeleton argument) that many people were attracted by the investment value of such acquisitions, the legislation deterred quick turnover by providing for clawback of the discount in the case of early sales. It has long been the common law that (except in the case of the sale or lease by a builder of a dwelling in the course of construction) a seller (or lessor) gives no implied warranty that the dwelling has been properly constructed; or, indeed, that it is fit for habitation. But it is by no means surprising that Parliament gave a measure of consumer protection to persons to whom the right to buy was given by partially insulating them from liability to contribute towards the cost of rectifying structural defects in the property in question. The case is all the stronger in the case of a building like Great Arthur House; because it cannot be supposed that the purchaser of single flat in a large block would commission a structural survey of the whole building before committing himself to acquiring a long lease.

The Upper Tribunal concluded at [39]:

“In my judgment, therefore, works of repair of the structure and exterior of the Building do not fall within para (i) of the definition of “specified repairs” if the effect of the works is to make good a structural defect. The costs of works that do have the effect of making good a structural defect are only recoverable if they fall within para (ii). Thus, if works have the effect of making good a structural defect, it makes no difference that the works also remedy deterioration that has occurred over the time that the defect existed.”

Again, I agree.

Moreover, I consider that Mr Straker overstates the severity of the outcome on the landlord. It is not as though the lessees have complete immunity. They can be required to contribute to the cost of making good structural defects of which they were made aware before taking up the lease. They may also be made liable to contribute to the cost of making good structural defects of which the landlord only became aware ten years or more after the grant of the lease. That achieves a fair balance between the interests of landlord and tenant.

In short, in my judgment the decision of the Upper Tribunal was correct, substantially for the reasons given by Fancourt J. I would dismiss the appeal.

I agree.

I also agree.