Ferguson Latimer & Anor v Carney & Ors

This appeal is from the dismissal by the judge of an action brought by the landlords of certain premises against their former tenants for breach of the covenant to repair the premises. The landlords adduced no evidence at trial of the value of the premises at the date of the termination of the lease. The principal issue on this appeal is the application of the first limb of section 18 of the Landlord and Tenant Act 1927 (“the 1927 Act”). The first limb of section 18 of the 1927 Act imposes a limit or “cap” on the amount of damages that a landlord can obtain for breach of the covenant to repair. The second limb similarly prevents the recovery of damages where for instance the premises are to be pulled down. Section 18(1) provides:

Where a tenant is in breach of a covenant to repair the property demised to him, his landlord will have an action in damages against him at common law for breach of the repairing covenant. The general rule at common law is that the measure of damages is the cost of putting the premises into the state of repair required by the covenant ( Joyner v Weeks [1891] 2 QB 31).

In this case, the landlords (who were and are two individuals) obtained a report from a Mr D.H.Hughes, a chartered surveyor, as to the state of the premises on termination of the lease, and his report gave an estimate of what it would cost to make good the repairs. The landlords looked for a new tenant. In due course a tenant was found who applied for planning permission to make changes to the property. He obtained planning permission to turn what had been a carpet shop into a fast food outlet. The landlords refurbished the property to meet his needs. The work included works in addition to the work needed to effect the repairs. The new lease contained a full repairing covenant save that the new tenant was not required to maintain the roof of the premises in any better condition than it was at the date of the lease. At the trial of the action for damages for breach of covenant, the landlords failed to prove the actual cost of the repairs that the surveyor had identified as requiring to be done to remedy the items in his schedule of dilapidations which the judge held constituted breaches of the repair covenant.

To apply the cap in section 18 of the 1927 Act, the court has to make findings as to the value of the reversion at the date the term ends and as to its value if the breaches of covenant had not occurred. The parties to this appeal take different positions on the question whether, where the landlords cannot prove the actual cost of the repairs required to be done, the judge should ascertain the value of the reversion without expert evidence dealing explicitly with that value from the estimates set out in the schedule of dilapidations.

The judge, HHJ Gilliland QC sitting in the Salford County Court, gave a detailed and careful judgment. He found that there were several breaches of repairing covenant (judgment, para 24), and though he did not identify or aggregate the amounts estimated by Mr Hughes to remedy the breaches found by him, that is a matter which the parties, or in default, the judge below could do. The judge found that Mr Hughes did not have the lease covenants before him when he prepared his schedule of dilapidations but the covenant for repair in this case is not unusual. Provisionally I consider that apart from the roof the items identified by the judge would include the following items in Mr Hughes’ schedule of dilapidations: A1.2, A2.1, A3.3, A4.1, A5.1, A5.2, A6.1, A6.4, A7.1, B1.1, B1.2, B2.1, B3.1, B3.3, B3.5, B4.1, B4.3, B5.2, B6.1, C5.1, C5.2, C5.3, C6.1, C6.2, C7.1, C7.2, C7.3, C7.4, and C8.1. In this judgment I refer to the aggregate estimated costs of the items identified by the judge in paragraph 24 of his judgment, as the “the specific estimated costs of repair”. Reading the judgment as a whole, it is in my judgment clear that the judge found that the items described in the schedule of dilapidations, matching the items of disrepair identified by him in paragraph 24 of his judgment as breaches of the repairing covenant, constituted work required to be done to remedy the breaches he found.

Nonetheless the judge rejected the landlords’ claim. He held that the landlords had not proved either the actual cost of the repairs or the damage to the reversion resulting from the breaches of covenant. The landlords had failed to separate out the costs of repair from the costs of improvement. He accepted that in some cases it was sufficient to show the estimated cost of repairs but he held that that could not be done in this case for the following reasons:

The judge also held that, although the tenant had failed to redecorate the premises in accordance with covenant to redecorate at the end of the term, this also was a breach of a repairing covenant to which section 18(1) applied and accordingly the costs of redecoration could not be distinguished from the other costs of repair:

So the first and principal issue on this appeal is: (1) was the judge right to hold that the landlords had failed to prove damage to the reversion? (“the cap issue”). There are three subsidiary issues: (2) should the court grant the landlords permission to amend their notice of appeal so as to raise the argument that the judge should have inferred damage to the reversion from the difficulty in reletting the premises in their unrepaired state? (“the amendment issue”); (3) does a covenant for periodic decoration of the demised premises constitute a covenant to repair for the purposes of section 18 of the 1927 Act? (“the decoration issue”); and (4) was the decision of the judge as to costs in part plainly wrong because he ordered that the landlords alone should pay the costs of the third respondent in Part 20 proceedings between the tenants, even though the second respondent’s defence to the third respondent’s claim for an indemnity failed? (“the costs issue”).

For the reasons given below, in my judgment, the answers to the various issues are in summary as follows:

the decoration issue

The costs issue

By a lease dated 20 January 1989, James Ferguson Latimer and Jean Latimer, (to whom I refer together as “the landlords”) demised the premises known as 5 King’s Road, Prestwich, Manchester to the respondent for a period of six years from 25 December 1988 at an initial rent of £5,200. The permitted user was as a shop for the sale of carpets, bedroom furniture and general household furnishings, with living accommodation on the first floor. The covenants included the following:

The lease was extended for a further period of three years terminating on 24 December 1997 by which date the annual rent was £9,500. Subsequently to the grant of the extension of the lease, the respondents with the permission of the appellants assigned the lease to the second respondent alone. The second respondent remained possession but gave up possession at the end of the term. The judge found that the lease came to an end on 24 December 1997. Mr Hughes visited the premises and prepared his report in December 1997 though he produced an addendum in February 1998. His report was served on the second respondent.

Repairs to the roof were carried out in early April 1998. The landlords produced the invoice of £3,813 (including VAT) for this work. This showed that the cost of the work was the amount estimated (£3,008) and further amounts to be added or deducted, including an amount for taking down a concrete roof at the rear of the premises. The judge did not accept that it was necessary to remove that roof (judgment, para.12). Indeed, in the absence of some explanation, he did not accept that the amount of £3,008 was necessarily wholly related to the work identified by Mr Hughes since Mr Hughes had only estimated the sum of £775 (presumably exclusive of VAT) (see judgment, para. 29). Accordingly the judge’s finding as to the cost of remedying the breach of the repairing covenant in relation to the roof was by implication that the cost was £775.

The premises were put on the market for reletting in February 1998. A month or so later, a Mr Weizman expressed interest if satisfactory terms and substantial works and planning permission for a change of use were agreed. However, the final refurbishment was not carried out until October or November 1998 and Mr Weizman’s lease was not executed until 25 January 1999. It provides for a term of six years from 25 January 1999 with a peppercorn rent for the first month and an initial rent thereafter of £10,500. The lease contained full repairing covenants but save that the tenant was not required to keep the lease in any better condition than at the commencement of the lease.

The judge described the negotiations with Mr Weizman, who became the new tenant, in paras 10 to 16 of his judgment. It is clear from those paras. that Mr Weizman would not have been prepared to take the premises if the repairs advised by his own surveyor had not been carried out by the landlords. In the course of meeting these requirements, the judge found that “nearly all of the works in respect of which the claimants are claiming had been carried out” (judgment, para. 13). In making that finding the judge, was as I read it, there making no distinction between repairs done in the manner anticipated by the schedule of dilapidations and repairs done incorporating an improvement to the premises. The judge’s overall conclusion was in para. 16 of his judgment:

In 2002 the landlords began proceedings against all three former tenants for damages for breach of the covenants in the lease. However, Mr Carney was not served. The landlords claimed as damages for breach of the repairing covenant:

In their defence the respondents took the point that there had to be evidence that the cap imposed by section 18 was not exceeded. This court is only concerned with the claim for the costs shown in the schedule of dilapidations. The landlords’ claims for the actual costs of repair (£23,823.39) and for loss of rent (£11,104.11) and council tax (£1,508.23) failed at trial, and are not pursued on appeal.

The judge did not consider that the state and condition of the premises was the only reason for the delay in reletting. He held that the delay was in part due to Mr Weizman’s desire to obtain planning permission.

Mr Hughes gave evidence as to the estimated cost of repairs. The judge accepted that the costs in his schedule of dilapidations were “his estimates of the likely costs of the remedial works which he refers to.” (judgment, para. 20). The judge does not in terms hold that the costs were reasonably necessary, but save where otherwise stated by the judge, this must in my judgment follow by implication from his judgment. He certainly did not say that he had doubts about the estimates or that he was unable to form a view as to whether they represented amounts that were reasonably necessary for the purpose of remedying the breaches that he found. The judge’s concern was that Mr Hughes could not say what sums had actually been spent. In the circumstances the judge found that as there was no expert evidence as to why it was necessary to expend the sum claimed in the proceedings which was much greater than the amount estimated by Mr Hughes, and as Mr Hughes himself accepted in his evidence that the invoices for the work actually done were indicative of modernisation work having been done in addition to the repairs, the judge held that improvement works must have been included and that the landlords had failed to prove the actual cost of the repairs (judgment, paras. 27 and 28). The judge was satisfied that the sums that the landlords claimed had been spent on repairs had actually been expended (judgment, para. 33).

As to the claim for the estimated costs of repair, the judge appears by implication to have assumed that where the repair works were actually done he had to proceed on the basis of the actual costs of the works (judgment, para. 36, 37 and, in particular, para.39).

He rejected the claim for loss of rent and council tax on the basis that it was not possible to make any estimate of the time needed to do the repairs which he had held to be necessary because the work actually done to make good those repairs was not clearly ascertainable (judgment, para.40). He made no findings about the one month’s peppercorn rent, and the parties have not directed any of their argument to any inference that should have been drawn from it.

The judge held that the third respondent was not released by the terms in which the landlords had given consent to the assignment to the second respondent and he accordingly made an order for nominal damages for breach of the repairing covenant against both the second and third respondents.

The third respondent brought a Part 20 claim against the second respondent and the first respondent for an indemnity against any monies for which he was found liable to the landlords. The third respondent claimed that the third respondent’s signature on the assignment to the second respondent was forged. The judge accepted the evidence on this point but held that the third respondent was entitled to an indemnity under the terms of an agreement dissolving the partnership between them. There is no appeal from this part of the judge’s order.

The judge made various orders as to costs. The landlords were ordered to pay the second respondent’s costs of the action and third defendant’s costs both of the action and of the Part 20 claim which he had had to bring against the second respondent for an indemnity.

The basic measure of damages for breach of the covenant to repair is the reasonable costs of executing the repairs required to fulfil the covenant (see Hanson v Newman [1934] Ch 298). This general rule is subject to the statutory cap in section 18(1) of the 1927 Act, set out above. It is also subject to general principles of law, including the principle established in Ruxley v Forsyth [1996] AC 344. In that case the House of Lords held that, where the expenditure required to be done to an asset to remedy a breach of contract is out of all proportion to the benefit to be obtained, the appropriate measure of damages will be the diminution in value of the asset rather than the expenditure. In Ruxley, the owner of the asset had no intention of expending the money required to remedy the defect and this point raises the question of the extent to which the subjective intention of the claimant is relevant. However, it is unnecessary to deal with that point in this case as the landlords have executed the repairs, that is, they have done works which are or supersede the repairs the respondents were bound to effect. Although courts are not normally concerned with what a claimant does with his damages, a landlord’s conduct in taking steps or not taking steps to remedy a breach of the covenant to repair may throw light on the question whether the repairs are reasonably necessary, and thus on the question whether there was any diminution in value of the reversion as a result of the disrepair.

The effect of section 18 is, in any case where its application is in issue between the parties, to require the court to find the amount of the damage to the value of the reversion of the premises caused by the failure to repair. To do this the court has to find the difference between the value of the premises in disrepair on the open market and the value that the premises would have had if there had been no breach of the covenant to repair. It need not do more than find that this difference was at least as great as the amount claimed against the tenant. In an ideal world, the parties will agree the relevant values or alternatively they will have produced the evidence of a single expert as to those values, or, if the court has given permission for more than one expert, they will produce their experts’ evidence together with a joint report from them identifying the differences between their views and the reasons for such differences so that the judge can come to a conclusion as to which expert evidence he prefers. But the failure to adduce expert evidence does not preclude a finding as to those values by other means because in many cases it will be obvious that the disrepair must have caused some damage to the value of the reversion and that the cost of doing the repairs is a reliable guide to the amount of that damage.

In this case, there was no expert valuation evidence and that complicated a trial in which the judge already had to make findings of a detailed nature on the state of the premises and the repairs to them that were needed. On the face of it, this was unfortunate. The judge’s judgments on liability and costs reveal a deep sense of frustration at the failure of the landlords to take obvious steps to present their case properly. It would appear that a great deal of cost and delay could have been saved if the landlords had produced the evidence needed and the parties had co-operated and tried to reduce the areas of dispute between them.

However, in cases where the amount in dispute is not large, courts regularly have to do their best on less than ideal material. That endeavour is consistent with the approach of the Civil Procedure Rules 1998, which provide that their overriding objective is “enabling the court to deal with cases justly.” (CPR 1.1(1)). The overriding objective states a fundamental value of any properly run system for the administration of justice. “As medicine is the art of healing, so is law the art of seeing that justice is done.” (Honoré, Ulpian, 2ed (2002) page 78). Justice and the rule of law of course also involve deciding cases in accordance with the law.

The result of the judge’s ruling in this case, however, is that the landlords recovered no damages, even though there were significant breaches of the covenant to repair and the landlords had clearly expended substantial sums on repairs. That result is on the face of it surprising and invites scrutiny. In Crewe Services and Investment Corporation v Silk [1998] 35 EG 81, this court, however, gave guidance on what the court should do in the circumstances where the material is less than ideal. That is one of the cases on which the appellants relied most heavily and I will be examining it below after I have explained the type of circumstances on which courts have relied to draw inferences as to the diminution in value of the reversion for the purposes of section 18.

It is not possible to give a definitive list of the matters which, in the absence of valuation evidence, might enable the court to make a finding as to the diminution in value of premises due to disrepair or as to the circumstances in which such matters would be a reliable guide as to such diminution. Cases have, however, arisen in which the courts had been prepared to infer the amount of the diminution in value of the reversion from circumstantial evidence. The leading case is Jones v Herxheimer [1950] 2QB 106 (relied on by the appellants and referred to in this judgment below). But there have been many other instances, as where the court has inferred a diminution in value from the fact that the new tenant was given a rent free period or was paid a reverse premium to take the premises (see Shortland Investments Ltd v Cargill plc [1995] EGLR 51), or from the fact that the premises were sold shortly after the term date at a price below the market value of the property if sold in the state of repair the premises would have had but for the breaches of the repairing covenants (Culworth Estates Ltd v Society of Licensed Victuallers (1991) 62 P&CR 211). Whether the circumstances can bear the inference that the repairs led to a diminution must depend on the particular facts of the case.

The valuation exercises required by section 18 fall to be performed at the date of determination of the lease containing the covenants to repair (the term date). But on general principle (see the authorities cited by me in Gorne v Scales [2006] EWCA Civ 311) subsequent events can be taken into account if they relate to the bases of valuation and thus throw light upon it. Such events would include refurbishment or sale of the premises after the term date.

In arguing this appeal the parties referred the court to a number of authorities but on the cap issue the principal authorities were Jones v Herxheimer, Firle Investments Ltd v Datapoint International Ltd ( Mr Colin Reese QC sitting as a deputy judge of the Queen’s Bench Division, Technology and Construction Court, 8 May 2000, unreported) and Crewe v Silk. I will deal with those authorities in turn.

Jones v Herxheimer is the leading authority for the proposition that the court may infer that the amount by which the reversion has diminished in value as a result of disrepair from the amount of the costs to remedy the defects, and that valuation evidence may be unnecessary. In Jones v Herxheimer, premises consisting of four rooms on the first floor and all the ground floor of a house were let to the defendant for one year. The tenant covenanted to keep, and at the end of the tenancy deliver up, the interior the premises in good tenantable repair. When the tenant gave up possession in 1949, the landlord found the rooms in a bad state of decorative repair, and he had them redecorated throughout, and then relet them. The landlord claimed the cost of those decorations. The judge having given judgment for the landlord, the tenant appealed contending that was no evidence of damage to the reversion for the purposes of section 18 of the 1927 Act, not have been awarded more than nominal damages. His appeal was dismissed. The Court of Appeal held that there was evidence on which the judge could hold that the cost of executing the repairs was a measure of the damage to the reversion. There had been no change of user or alteration of the premises. Jenkins LJ, giving the judgment of the court, held:

The respondents drew particular attention to the words “small and simple case” in this passage. However, Jenkins LJ uses those words as an example of the type of case where valuation evidence is not necessary. Accordingly the fact that this case is factually complicated because of the modernisation works which the judge found that the landlord had carried out does not mean that this is not capable of being a case in which the court could infer diminution in value from estimates of the repair costs if no valuations were available.

The next case is the Firle case. This again was an action by a landlord to recover damages for breach of the covenant to repair. The landlord claimed inter alia the estimated costs of repair as set out in a schedule of dilapidations. The case differs from the Jones case because it deals with the situation which arises where the premises have potential for redevelopment. Indeed the landlord planned to redevelop the premises as soon as the lease came to an end. The tenants relied on section 18(1) of the 1927 Act and contended that the premises could not be relet without substantial modernisation and that the works would have rendered the repairs done by the tenant worthless. By the date of the trial, the landlord had in fact modernised the premises. It contended unsuccessfully that it had done so only because of the disrepair for which the tenant was responsible. Reference was made to Dowding and Reynolds on Dilapidations (now 3^rd ed. 2004). That distinguishes two broad types of case. The first is where the notional purchaser would simply demolish the premises or alter them so substantially as to make the existing state of repair irrelevant to the value of the premises. In that sort of case, the court might well find that there was no damage to the reversion caused by the disrepair and that the landlord could recover no damages. The second broad type of case is where the purchaser would be likely to upgrade the premises in such a way that the pre-existing state of repair was relevant only to a limited extent. In such a case only some of the repair works would survive the refurbishment. Dowding and Reynolds state that in such a case:

Mr Colin Reese QC agreed with that statement but went on as follows:

As it seems to me there is a difference between the proposition in Dowding and Reynolds, which was approved by Mr Reese, and the proposition in which Mr Reese expressed the law in his own words. He has made the proposition far firmer and more absolute than Dowding and Reynolds. That work goes on to give an example of a case where the premises did have redevelopment potential but the court held that a purchaser would still pay more if the premises were in a good state of repair (Craven Builders Ltd v Secretary of State for Health [2000] 1 EGLR 128). I consider, therefore, that the judge’s reformulation must be interpreted as not departing from Dowding and Reynolds. On this basis, what Mr Colin Reese holds is that it is likely that there is no diminution in value when repair works are superseded by works of refurbishment that would be undertaken by the purchaser. Later passages in the judge’s judgment support a more nuanced approach (see paras. 80 and 81).

The Firle case is about the supercession of repairs by modernisation works. In this case the landlords have not purported to prove the diminution in value by reference to a formal valuation. In the absence of indications to the contrary (and the judge does not refer to any) the likelihood is that the price a purchaser would have been prepared to pay in this case would have taken into account the state of repair of the premises. The only case which the landlords make is that the diminution in value can be inferred from the estimates, coupled with the fact that when a new tenant came along he insisted that the landlords carry out substantial repairs and that he was only liable for the existing condition of the roof. The authorities, such as Jones v Herxheimer, make it clear that a valuation is not the only means of proving the value of the reversion.

I would, however, accept that where the repair works which the tenant in breach of a covenant to repair is bound to do will be overtaken by refurbishments which the landlord or a purchaser of the property proposes to do, that indicates that the reversion has a latent development value. The landlord would have to show that the repairs caused damage to the reversion, and this may in the circumstances be difficult. Alternatively he will have to show that the refurbishment would incorporate some of the repairs the former tenant should have carried out, ie that specific repairs would “survive” the refurbishment. The question whether the repairs to be done by the former tenant in any particular case survive may raise difficult issues of fact and judgment, but no specific examples of difficulty have been suggested to us.

The third principal authority in this appeal is Crewe v Silk, and this authority is most helpful on the question how the court might apply section 18 when the evidence about values is sparse in order to achieve the objective of dealing with the case justly, which, as I have explained is an objective of general application. It is an example of the “art”, referred to by Ulpian, of deciding cases justly in the sense I have given. I proceed on the basis that Parliament certainly did not intend that section 18 should render it impossible for a landlord to obtain proper recompense for breaches of the repair covenant without undue expense and delay. Moreover, as I have explained, this court in Jones v Herxheimer did not consider that it was necessary for there to be formal valuations done by experts in every case, and in addition expressly envisaged that the judge might not accept the evidence of an expert valuer and thus by implication reach a valuation on some other basis (at pages 118 to 119).

In Crewe Services & Investment Corporation v Silk, the landlord had brought proceedings against his agricultural tenant for failure to keep his tenanted farm in a good state of repair. The judge below awarded the landlord the cost of doing the repairs himself, making no discount for the possibility that the tenant might in fact remedy the breaches before the end of the tenancy. The tenant appealed to this court: Lord Woolf MR, Millett and Robert Walker LJJ. This court held that since the tenant might decide to repair himself during the term and there was no evidence before the trial judge of the Court of Appeal that the landlord intended to carry out any works of repair at all, the costs of the repairs might be regarded as being a starting point. The court discounted these for the uncertainties as to whether the work would be done.

This is a case therefore where the breaches had occurred during the currency of the lease and where there was evidence as to the cost of repair and no repair had yet been carried out by the landlord. This court held that as the breaches had occurred during the currency of the lease the correct measure of damages was the extent of the damage to the reversion and that section 18(1) did not affect the position. But as I see it those are not differences of principle. There is no uncertainty in this case as to whether the landlords will carry out the repairs. The judge found that they had nearly all been done. But there is uncertainty as to whether the repair works which were done made the repairs which the respondents were bound to do unnecessary.

Robert Walker LJ referred to the dictum of Lord Goddard LCJ in Bonham-Carter v Hyde Park Hotel (1948) 64 TLR177 at 178 that parties had to prove the damage. Robert Walker LJ, giving the judgment of the court, continued:

This court decided that on the facts the judge had been in error in taking the undiscounted cost of the repairs. The court held that nominal damages were not appropriate and that the case should not be remitted to the county court to ascertain the amount of the diminution in value of the reversion but rather that this court should fix the discount to be applied. The court held that as the landlord was responsible for the meagre evidence on diminution in value the rate of discount should be severe. It determined that the appropriate discount was 75%.

It is to be noted that in Crewe v Silk there was no expert evidence as to the diminution in value of the reversion but this court held that such evidence was not in that case necessary. It was obvious that the breaches of covenant would affect the price the purchaser was prepared to pay for the farm. Likewise in this case, it is in my judgment axiomatic that the defects as found, for instance the repair needed to the roof and the decoration, were likely to affect the value of the premises to a purchaser. A purchaser is likely to want to let the premises or use them himself. There is no suggestion that they were to be pulled down or rebuilt. Mr Weizman’s reasons for substantial alterations to the premises were no doubt related to the change of use. This may not have been anticipated by a purchaser of the reversion in the open market as at the term date but it would be realistic to suppose at that date in the circumstances of this case that any new tenant would have required a degree of modernisation and that that modernisation would have rendered unnecessary some of the repairs which the respondents were bound to do. So, in my judgment, the court has to make some allowance for the possibility that the extent of the damage to the reversion caused by the tenants’ breach of their repairing obligations would be less than the cost of the repairs. Evidence as to the extent that the works required by Mr Weizman in the event rendered the repairs for which the respondents were liable unnecessary might have assisted the judge, but the fact that such evidence was unavailable does not mean that the judge could not make a discount for these matters, doing the best he could on the evidence that he had. I will deal with the discount below but first I turn to deal with some of the points made by counsel and the judge.

As I have explained, the appellants’ case is that the judge should have found that the breaches of the repairing covenant had diminished the value of the reversion on the basis of Mr Hughes’ evidence as to the estimated costs of the repairs. The respondents’ case is that, if the work of repair has been done, the tenant has to know the actual cost of the repairs. It is not enough to find an estimated figure. The landlords ought to adduce valuation evidence to show the damage to the reversion. I consider that that submission goes further than the proposition relied on by the judge for, as the passage from para. 39 of his judgment cited above shows, his reason for rejecting the estimates of the costs of repair was as follows:

It is clear from the Jones case that expert evidence is not required and that in an appropriate case the court can infer that there has been a diminution in the value of the reversion, and the amount of such diminution, from the fact that the landlords have had to carry out certain repairs. But this is not the only circumstance in which diminution will be found. It may be found where for instance the landlord has not carried out the repairs through lack of funds. In my judgment the judge was in error in thinking that if the repairs had actually been carried out he could not infer that damage to the reversion had occurred from the fact of those repairs unless he had the actual cost of those repairs. In my judgment he could make that inference from an estimate of those costs if he was satisfied the estimated costs were reasonable. In an ideal world every landlord will come to court with the appropriate valuations but where this does not happen the landlord should not fail to make any recovery if it is clear that breaches occurred, that the repair work had to be done and that damage to the reversion as a result of the breaches is proved in other ways. As Denning LJ held in Haviland v Long [1952] 2 QB 80 at 84, the fact that repairs are necessary is a fact from which damage to the reversion can be inferred.

The judge’s view that, once the repairs had actually been carried out, damage to the reversion could only be inferred from the costs of the repairs if the actual cost was known and also it was clear what work had been carried out may even have arisen from a concern about two things: first the possibility that the repairs for which the respondents were liable had not in fact been carried out and secondly the possibility that they had cost less than the estimated costs.

The failure to carry out the repairs would clearly be an indication that the repairs were not necessary as the landlords claimed. Put another way, whether sums were actually spent on doing repairs is relevant to the question whether the repairs were necessary or not. If they were not necessary, damage to the reversion could not be inferred from them. But even where the repairs had not been carried out there could be other explanations for the failure that could satisfy the judge that the indication was not well-founded, as where the landlord decides not to repair the property himself but proceeds to sell it at a lower price than he could have obtained if the repairs had been remedied. In this case, there was, as I read the judgment, a finding that the repairs identified by Mr Hughes had substantially been made good and so far as I can see the judge did not suggest that the repairs should have been done for less than the amounts estimated by Mr Hughes. In some cases, Mr Hughes inserted a provisional sum in his report but that fact does not mean that a claim cannot be made for that amount.

Underlying the judge’s conclusion may be an assumption that if more was actually spent by the landlords they had to claim that amount. The fact that more was spent than is claimed does not of course mean that what was estimated did not relate to a necessary repair or that the amount claimed is not a reasonable amount. The fact that the incoming tenant wanted more extensive works of modernisation than were required to be done by the outgoing tenant does not necessarily mean that the latter tenant is not liable for breach of the covenant of repair. Simply because the landlords did more extensive work than was necessary does not mean that there was no damage to the reversion at the term date or that the landlords did not suffer any loss as a result of the outgoing tenants’ breaches of covenant.

So far as the roof is concerned, I can see no basis for concluding that the repair was not necessary or that it was not carried out or that the sum estimated by Mr Hughes (£775) was not reasonable. Accordingly, in my judgment, the judge was in error in not holding that this sum was evidence from which he could infer damage to the reversion. I then move to the other sums estimated by Mr Hughes. There were a number of possibilities. The first was a possibility that even if the repairs identified by Mr Hughes were carried out, the property could still not have been let. If this were the case, then the failure to do the repairs did not cause any damage to the reversion: the premises were unlettable. This possibility is not realistic. The premises had been used as a shop and there is no suggestion that if it had been reinstated as a shop it could not have been let as a shop. Another possibility was that the landlords had all along proposed to modernise the premises and the fact that the outgoing tenants had not done the repairs they were due to do was of no consequence, because the work would have been completely superseded by the works of modernisation. In my judgment, this too was an unlikely scenario, since save for repairing the roof to make it wind and watertight the landlord took no steps to repair the premises until he had the prospect of a new tenant. He then sought to make the new tenant do the repairs but his surveyor advised against that. The bargaining situation was such that Mr Weizman was able to persuade the landlords that they should do the repairs. He also persuaded the landlords that they should do some improvements at the same time. For instance, in relation to the windows that were rotten, Mr Weizman was able to persuade the landlords to replace the windows with UPVC windows. This went beyond merely repairing rotten woodwork. On the other hand, there is nothing to suggest that the landlords would have undertaken this change themselves, had it not been for the poor state of repair in which the premises were left by the former tenants. As I see it, the matter has to be approached on the judge's findings on the basis that the landlords had no option but to spend the money they in fact spent on repairing the premises. Otherwise, Mr Weizman would not have been prepared to take the premises. There was evidence that there was no other tenant in the offing.

The position therefore, is this. The landlords were obliged to do the repairs. But they went further than simply doing the repairs. They carried out improvements at the same time. In fairness to the outgoing tenants, there is the possibility that the landlords’ work rendered some of the work required to remedy the breaches of covenant by the outgoing tenants futile and that some of the improvements rendered some of the work of repair unnecessary. For example, if the outgoing tenants were required to replace a window, which, as part of the work done for Mr Weizman, was to be blocked up, the landlord would have suffered no loss as a result of the breach of covenant and moreover it could not be said that the failure to repair the window constituted any evidence from which damage to the reversion could be inferred. On the other hand, it does not seem to me that the landlords should necessarily be deprived of their remedy simply because they performed the repairs to a higher standard than the outgoing tenants were required to do.

In all the circumstances, the right course in my judgment was for the judge to have inferred diminution in value to the reversion from the estimated costs of any repairs required to be done by the outgoing tenant which the landlord could actually show they had done. This would include damage to the roof. In the case of other repairs, the judge should have applied a discount to take account of the possibilities referred to above. The amount of such a discount was for the judge doing the best he could on the material available to him. The parties have not invited us to refer the matter back to the judge for the purpose of making this discount. In my judgment, it would be disproportionate to take that step. This court should formulate its best judgment on the material available to it, and in all the circumstances of this case it should be generous to the outgoing tenants. In my judgment, the discount should be 60%. This is not of course a case where it is suggested that the estimates were excessive and that when the work was actually done it was found that the costs of doing the work were less than what had been estimated. The discount should be applied to the specific estimated costs of repair as defined above

It may seem unjust that the landlords should recover so little to reimburse them for the costs of repairing the premises when the respondents had caused the damage and had undertaken to make good items of disrepair. But the landlords also on the judge’s findings took the opportunity to upgrade the premises and thus would be able to charge more for the premises. In so far as the repairs were not works of modernisation the appellants have themselves to blame for failing to foresee that the court would not have the raw material from which more precisely to draw inferences. Perhaps the appellants thought that they would be able to take a short cut. If that was the case it is worth recalling the advice in Dowding and Reynolds on Dilapidations at para.29-62:

The appellants have a subsidiary argument (skeleton argument, para. 8) that it is unclear why the judge failed to accept that the defects identified in specified paras of Mr Hughes’ report in respect of windows in the interior and the walls and exterior goods in the exterior were items of disrepair. I would reject this argument. The windows were replaced with UPVC windows as part of the refurbishment of the premises (see the judgment at para. 28). This work clearly went further than repair of the windows. The other items referred to by the appellants in para. 8 of counsel’s skeleton argument were minor items of repair which could well have been swept up in any scheme of refurbishment (see the judgment at para. 29). In those circumstances there is no basis for going behind the judge’s finding as to what constituted breaches of the covenant to repair.

Subsequently to the hearing of this appeal and at the invitation of the court, the appellants made a formal application to amend their notice of appeal to reflect their argument before us that the judge should have inferred damage to the reversion from the fact that there was delay in reletting. As formulated in the amended notice of appeal the appellants’ argument is that the judge failed to take into account the fact that although the premises were advertised as available for letting in February 1998, they were not in fact relet until January 1999, the fact that Mr Weizman required the landlords to spend about £20,000 on repairs and the fact that Mr Weizman’s obligations to keep the roof in repair were qualified. The second and third respondents oppose this application. They submit that this point was not raised at the trial and should be rejected for that reason. I agree that this application should be rejected. It is a new point and it is not a pure issue of law. The judge was unable to make a finding as to the period required to do the repairs which he found the respondents bound to do (judgment, para. 40). To adjudicate on the draft amended ground of appeal this court would need to go into the matter to see precisely how much of the delay should be attributed to the repair work, ie make new findings of fact. Moreover the court would need evidence as to the state of the market so that it could make a finding as to the likely date for reletting if the breaches of the repair covenants had not occurred. Thus to deal with the new ground of appeal the court would have to deal with factual matters which the respondents had not had the opportunity of dealing with at trial. Even if the findings could be made on the material in the judge’s judgment, it would not, in those circumstances be fair to the respondents to allow this new ground to be raised.

The appellants have also formulated further draft grounds of appeal which invite the court to make an inference as to the diminution in value of the reversion by discounting the actual costs of the repairs, and raising an entirely new point, namely that the judge’s should have found that the respondents were liable to pay all the professional costs of the schedule of dilapidations.

The fundamental difficulty with the first of these grounds is that identified by the judge, namely that the court cannot be satisfied as to the extent to which the actual costs relate to repairs, as opposed to works of improvement. So there is no starting point, or evidential basis, for discounting in this way. The point about costs raises an issue of fact which should have been raised earlier. It is too late to raise it now. I would therefore dismiss the whole of this application to amend.

In para. 37 of his judgment (quoted above) the judge enunciated and applied the following proposition:

The issue for this court is whether that proposition is correct as a matter of law. The appellants’ case is that the covenant to paint the premises in the third and final year of the term was not part of the covenants to repair and therefore is not caught by section 18. The respondents seek to uphold the judge’s judgment on this. They referred to sections 69 and 70(1) of the Landlord and Tenant Act 1954 Act. Section 69 states that the word “repairs” “includes any work of …decoration” and section 70(1) provides that the 1954 Act and the 1927 Act may be cited together as the Landlord and Tenant Acts 1927 to 1954. (However, it does not go on to provide that they shall be construed together). I do not consider that it is necessary to express a view on the true interpretation of the word “repairs” in this case, though the 1954 Act and 1927 Act are clearly in pari materia.

In reality, the question for this court is whether to interpret section 18(1) literally or purposively. If the former approach is taken then, a covenant for periodic decoration is not a covenant to repair because it will have to be performed even if the property is not in poor decorative repair (see for example Gemmell v Goldsworthy [1942] SASR 55,57 on which counsel for the appellants, Mr Michael Mulholland, relied) However, where the breach of the covenant for periodic decoration coincides with a breach of the repairing covenant, the landlord’s real complaint is that there has been a failure to repair. More importantly, Parliament enacted the cap in section 18(1) to meet the rigour of the measure of damages for breach of the repair covenant at common law. It may be that the courts would not apply the common law measure of damages in all cases today: I would accept the argument of counsel for the second respondent (Mr Matthew Hall) that, if the common law measure alone were relevant to a landlord’s claim, the courts today might in an appropriate case adopt the measure of damages in section 18(1) in preference to that which has previously been held to be the measure at common law ( see generally Ruxley v Forsyth).

In all the circumstances I consider that this court should treat a failure to repair the decorative state of the premises as a breach of the covenant to repair for the purposes of the first limb of section 18(1) of the 1927 Act even if that failure also constitutes a breach of a covenant for periodic decoration in the same lease. On that basis the judge’s finding that there was a breach of the covenant to paint the premises at the last year of the term (see clause 2(5) set out above) does not assist the appellants.

In the light of this conclusion, it is unnecessary to deal with the point raised by counsel for the third respondent (Mr Nigel Bird) that substantial damages would not be awarded at common law for breach of the covenant to redecorate because it would not survive the new tenant’s refurbishment works.

On this issue it is not enough for the appellants to show that this court might reasonably have made some other decision. They must show that the decision of the judge with respect to the exercise of his discretion as to costs was so plainly wrong that the only legitimate conclusion was that he had erred in the exercise of his discretion. The second respondent submits that the order which the judge made for the third respondent’s costs of the Part 20 claim was an order he was entitled to make in the exercise of his discretion as to costs.

The effect of the judge’s decision was that the second respondent was not liable for any of the costs of defending the Part 20 claim brought against her by the third respondent. The appellants submit that the judge should at least have ordered that the second respondent was jointly and severally liable with the landlords for the costs of the third respondent. The Part 20 claim was brought because of a claim for an indemnity. This had nothing to do with the landlords. If the landlords had sued only the third respondent they would not have had to incur the costs of any proceedings by the third respondent against the second respondent. The second respondent had brought the costs on herself. She had denied the obligation to indemnify the third respondent and relied on the terms of the assignment. The judge rejected her defence to the Part 20 claim.

In my judgment, all but an insignificant part of the costs incurred by the third respondent in relation to the Part 20 claim were due to the second respondent’s decision to defend the Part 20 claim. The judge in his extended judgment on costs considered that despite this the appellants should bear all the costs of the third respondent because of their conduct of the case. While there would have been no Part 20 claim if the landlords had not started proceedings, their commencement of those proceedings was not the sole cause of the costs that were in fact incurred. The conduct of the case against the respondents was a matter which was certainly relevant in determining the order for costs to be made in the action, but their conduct of their case in the main action did not result in the second respondent having to defend the third respondent’s Part 20 claim. Those costs were increased because of her own decision to defend them on grounds which could not be substantiated. In those circumstances it is in my judgment wrong in principle for the whole of the costs of the third respondent to be borne by the appellants.

One way of dealing with this would have been to make the order proposed by the appellants. Other forms of order could have been made which would have been less generous to the second respondent. The judge’s exercise of his discretion must be set aside. It is unnecessary to decide what order the judge should have made as in my judgment the judge proceeded on a wrong basis namely that the landlords’ claim (other than in purely nominal terms) wholly failed. If the parties cannot agree on the appropriate order there should be written submissions in writing to be filed when the parties make submissions to the court as to the form of order arising out of the disposition of this appeal. The court can rule on those submissions on paper.

For the reasons given above, I would allow the appeal in the manner and to the extent described above.

I agree.