England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE UPPER TRIBUNAL (LANDS CHAMBER)
MARTIN RODGER Q.C., DEPUTY PRESIDENT
[2021] UKUT 0085 (LC)

		Royal Courts of Justice Strand, London, WC2A 2LL
		13 April 2022

This judgment was handed down remotely by circulation to the parties' representatives by email and release to BAILII. The date and time for hand-down was deemed not before 4pm on 13 April 2022

The Senior President of Tribunals:

Introduction

1. What is the true interpretation of section 9(1A)(d) of the Leasehold Reform Act 1967 – a counter-factual deeming provision for the valuation of the freeholder's interest in a house and premises upon enfranchisement, requiring the price payable for that interest to be diminished by the extent to which its value has been increased by improvements carried out by the tenant, or his predecessor in title, at his own expense? That question is raised in a preliminary issue decided by the Upper Tribunal (Lands Chamber) in this case. Different interpretations of section 9(1A)(d) have been put forward on either side. In my view, the tribunal's was correct.
2. With permission granted by Lord Justice Lewison, the appellant, Cadogan Holdings Ltd., appeals against the decision of Martin Rodger Q.C., the Deputy President of the Upper Tribunal (Lands Chamber), on a preliminary issue to establish the scope of the assumption in section 9(1A)(d) and its effect on the price payable for the freeholder's interest in the house at 10 Cheyne Walk in Chelsea by the respondent, Mrs Fleur Alberti. That preliminary issue arose on an application under section 21(1)(a) of the 1967 Act for a determination of the price payable for the house and premises under section 9. The section 21 application was necessary because the parties could not agree on the valuation.
3. The dispute concerns works undertaken over 40 years ago by Mrs Alberti's predecessor in title, the cartoonist Gerald Scarfe, to convert the building from five flats to a single house. As those works were undoubtedly improvements, what was the effect of the assumption in section 9(1A)(d) on the valuation of the freeholder's interest? Did it require the freehold to be valued as if, on the valuation date, the building could lawfully be used only as five flats and not as a single house, or on the basis that it could lawfully be used as a single house?
4. The decision on the preliminary issue will have a significant effect on the price payable for the freehold. Mrs Alberti contends that if the improvements are assumed not to have been carried out, the use of the building as a single house on or after the valuation date would be development requiring planning permission, which would not now be granted, that the freehold would therefore only have been of interest to buyers seeking a profit from the refurbishment of the flats, and that its value was £2.6 million. Cadogan contends that, in reality, it was lawful on the valuation date to use the building as a single house; and that the value of the freehold, subject only to the unexpired term of the lease, was £11 million. Mrs Alberti maintains that if Cadogan's interpretation of section 9(1A)(d) were correct the value of the freehold would be £4.25 million.

The issue in the appeal

5. The issue in the appeal is the same as the preliminary issue itself, which was this:

"Does section 9(1A)(d) of [the 1967 Act] require the Tribunal to assume that it was unlawful as a matter of planning control to use 10 Cheyne Walk as a single house on the valuation date?"

6. The tribunal determined that issue in favour of Mrs Alberti. Cadogan contends that this determination was wrong.

The facts

7. The tribunal had before it the parties' statement of agreed facts, which it took as the basis for the determination of the preliminary issue. It set out a short summary of the "agreed facts" (in paragraphs 8 to 20 of the decision). We have been assisted by a further agreed statement, dated 15 February 2022, which records more of the history of relevant planning policy.
8. On 21 May 1971 a long lease of the building, then divided into five flats, was granted to Mr Scarfe. He undertook the works to convert it back into a single house in the 1970s. The works included the removal of internal partitions, bathrooms and kitchens. They were undertaken without planning permission, which the parties agree was not required. Under section 87 of the Town and Country Planning Act 1971, after four years had passed, no enforcement action could have been taken against the use of the building as a single dwellinghouse. It is not in dispute that this became the building's "established use". Mr Scarfe and his wife lived in it as their home.
9. In 1984 the building was listed, at grade II. On 2 June 1987 planning permission and listed building consent were granted for a rear studio extension at third floor level, which Mr Scarfe constructed and used as part of his studio.
10. Before August 2014 the development plan for the Royal Borough of Kensington and Chelsea contained a policy whose effect was that schemes of development involving the net loss of fewer than five residential units would not normally be regarded as a material change of use requiring a grant of planning permission. This was the policy prevailing when Mr Scarfe undertook the conversion works in the 1970s. But in August 2014 the relevant policies in the development plan changed, and now resist the grant of planning permission for works which would result in a net loss of one residential unit or more. Since then the local planning authority's position has been that all such "amalgamations" are development requiring planning permission. In the local plan adopted on 11 September 2019, Policy CH1 "Increasing Housing Supply" states that the local planning authority will "resist the loss of residential units through amalgamations of existing or new homes unless the amalgamation will result in the net loss of one unit only and the total floorspace of the new dwelling created will be less than or equal to 170sq m gross internal area …".
11. On 13 May 2019 Mr Scarfe gave notice to Cadogan of his intention to purchase the freehold under the 1967 Act. That is the valuation date. By then, less than two years of the term of the lease remained and the value of the freehold reversion was, on any view, considerable. It is common ground that the use of the building as a single house was not unlawful at that stage. Mrs Alberti purchased the unexpired term of the lease from Mr Scarfe in August 2019.
12. There is no dispute that the works carried out by Mr Scarfe – both the internal alterations and the rear extension – were "improvements" within the meaning of section 9(1A)(d). It is also agreed that if they had not been carried out before the valuation date, any purchaser of the freehold on the valuation date would have been advised that an application for planning permission and listed building consent to amalgamate the five flats into a single house would have had no chance of success. It appears also to be accepted on both sides that there was a good chance of planning permission and listed building consent being granted for the third-floor extension.

The legislative provisions

13. Section 9(1C) of the 1967 Act provides that the price to be paid for the freeholder's interest is the amount which, at the "relevant date", the freehold subject to the lease might be expected to realise if sold in the open market by a willing seller, on certain assumptions. The "relevant date" is the date of the service of the notice of claim. The assumption in issue here is the one specified in section 9(1A)(d):

"(d) … that the price be diminished by the extent to which the value of the house and premises has been increased by any improvement carried out by the tenant or his predecessors in title at their own expense".

Relevant authority on section 9(1A)(d)

14. In Shalson v Keepers and Governors of the Free School of John Lyon [2003] UKHL 32; [2004] 1 AC 802 the House of Lords considered the meaning of section 9(1A)(d) of the 1967 Act. Mr Shalson, the tenant, had reconverted a house consisting of five self-contained flats into a single dwelling. The question for the court was whether Mr Shalson's conversion of the house back into its original form as a single dwelling constituted an improvement which increased the value of the house and premises under the statute and the tenant was therefore entitled to a diminution by that extent to the price payable on his purchase of the freehold. The House of Lords unanimously held this to be so.
15. Lord Bingham of Cornhill said (in paragraph 3 of his speech) that the "fairness" of section 9(1A)(d) was "obvious". It would not be fair for the tenant to have to "pay an enhanced price to the extent that such enhancement was attributable to works done by him or his predecessors in title (probably voluntarily) at their own expense", because if that were so, "the tenant would in effect be paying twice", and "the owner would be reaping an adventitious gain as a result of works which he had no right to require". If the value of the house has been increased by any improvement carried out by the tenant or his predecessor in title, "the market price must be reduced so as to discount the increase attributable to that improvement …".
16. In Lord Hoffmann's view, the language of section 9(1A)(d) was clear. For the tenant to secure a reduction in accordance with the statutory assumption, he must do two things. First, he must identify improvements which he or his predecessors have carried out at their own expense. Secondly, he must "satisfy the tribunal that but for those improvements the house and premises would have been worth less" (Lord Hoffmann's speech, with which their Lordships all agreed, at paragraph 17). As for the first condition, an improvement, said Lord Hoffmann, is "a physical and not an economic concept"; it "refers to the works themselves and not to the effect, if any, which they have upon the value of the premises". The second condition deals with the effect on value. The tribunal was therefore right to regard both the original conversion of the building to flats and the subsequent reconversion as improvements, even though the original improvements had been stripped out before the valuation date, and, if they had remained intact, would have reduced the value of the house and premises, not increased it (paragraph 18).
17. Lord Hoffmann went on to say (in paragraphs 19 to 21):

"19. The issue in this appeal turns upon what I have called the second condition. What does it mean to say that the value of the house and premises has been increased by the improvement? In my opinion, it signifies a simple causal relationship: but for the improvement, the house and premises would have been worth less. The comparison is between the value of the house as it stands and what its value would have been if the improvement had not been made.

20. The hypothetical house envisaged by this comparison is in my opinion one which has all the features of the real house, including its history, save for one: that the improvement in question had not been made. … [If] the reconversion had not taken place, the 1947 improvements would still have existed and the house would have been worth less. To the extent that it was worth more, the tenant was entitled to a reduction in the open market value.

21. I can see no room in the statutory language for a comparative hypothesis which assumes, as the Court of Appeal did, that the improvement in question had not been done and also that there had been no earlier changes which the improvement reversed. In considering whether an improvement has added to the value of the house, the comparison is simply with the house as it would otherwise have been. This seems to me fair to both parties. If the tenant had not carried out the reconversion, the landlord's interest would have been the reversion on a house converted into five flats. The tenant was under no obligation to reinstate. If the tenant increases the value of the landlord's interest by expenditure on reconversion, it would not seem fair that he should have to pay a second time when the landlord's interest is valued for the purposes of a sale of the freehold."

18. Lord Millett said (in paragraph 31) that the tenant's improvement had "increased the value of the property at the relevant time, in that the property would have been worth less if the work had not been carried out and the house had remained divided into flats". He observed (in paragraph 32) that section 9(1A)(d) is "designed to avoid the tenant having to pay a price which reflects a value in the property for which he has already paid", and that "[if] the tenant carries out alterations to the property which enhance its value he thereby increases the value of the landlord's reversionary interest which he afterwards claims to acquire". The assumption in section 9(1A)(d) "prevents his own expenditure resulting in an increase in the price he has to pay". On the approach to valuation, applying that assumption, Lord Millett said (at paragraph 40) that "[the] "extent to which the value of the house and premises has been increased" by an improvement is simply the difference between the value of the property with the improvement in question and the value of the property without it". And "[the landlord] gets the worst of both worlds; he receives a lower price if the tenant carries out alterations which reduce the value of the property, and does not receive the benefit if the tenant carries out alterations which increase it" (paragraph 42).
19. In Fattal v Keepers and Governors of the Free Grammar School of John Lyon [2003] EWCA Civ 1530; [2005] 1 WLR 803, this court had to consider whether, in valuing the property in question, its development potential, including in particular the value of any planning permission, was to be disregarded under section 9(1A)(d). The court held that it should not be.
20. It was common ground in that case, in the light of the speeches of Lord Hoffmann and Lord Millett in Shalson, that a comparison must be made, at the valuation date, "between the value of the property in its unimproved state and the value it would have had if it had not been improved" (the judgment of Sir Martin Nourse, with which Lord Justice Sedley and Lord Justice Buxton agreed, at paragraph 10). Sir Martin Nourse said (in paragraph 13):

"13. What assumption (d) requires is a calculation of the amount of the increase in value caused by the improvements. That necessarily involves a valuation of the property as it would have been on the valuation date if it had not been improved. Before the Lands Tribunal both valuers agreed that any potential for improvement would be included in the achieved sale prices of unimproved properties; in other words, that a valuation of an unimproved house and premises would include the value of any such potential. It follows that an increase in value caused by an actual improvement must be calculated as an excess over the unimproved valuation (including the value of the potential for improvement), notwithstanding that the potential is merged in or absorbed by the actual improvement. …".

And he added (in paragraph 15), that section 9(1A)(d) "does not credit the tenant with the value of the relevant improvements, but only with the increase in value they have caused".

21. Counsel for the tenants had submitted that the approach of the Lands Tribunal in valuing the property as if it had never been improved at all was erroneous. Rejecting that submission, Sir Martin Nourse said this (in paragraph 18):

"18. … Assumption (d) simply requires that the price be diminished by the extent stated. It does not impose any requirement that the house and premises shall be valued either from the top down or from the bottom up. The method adopted is a matter of valuation, not of law. Moreover, it appears that the method adopted here has been the standard method adopted by the Lands Tribunal ever since its decision in Norfolk v Master, Fellows and Scholars of Trinity College, Cambridge (1976) 32 P & CR 147 …".

As was stated in "Hague on Leasehold Enfranchisement", fourth edition (2003, at p. 225), "[the] manner in which the assumption is given effect is for the property to be valued (at all stages of the valuation – including the calculation of the marriage value) as if the improvements had not been made".

22. It was also submitted for the tenants in Fattal that the tribunal ought to have assumed not only that the improvements in question had not been carried out, but also that the planning permissions which enabled them to be carried out had never been obtained. In view of Lord Hoffmann's observation in Shalson that an improvement is a "physical concept", Sir Martin Nourse agreed (at paragraph 19) with the submission of leading counsel for the landlords that it is the increase in value caused by the physical works which has to be subtracted, and that "the existence or availability of planning permission is not part of those works". In support of that submission, leading counsel had cited the decision of Mr Justice Knox in Railstore Ltd. v Playdale Ltd. [1988] 2 EGLR 153, "where it was held that a rent review clause containing a direction to disregard any effect on rent of any improvements carried out by the tenant did not require the arbitrator to disregard the existence of a planning permission for the buildings which the tenant had built and was occupying".

The tribunal's decision

23. The tribunal observed that the preliminary issue was "concerned with the interplay between section 9(1A)(d) and the reality principle". It paraphrased the question as being "whether it is a necessary consequence of section 9(1A)(d) that, in its assumed condition, planning permission and listed building consent would need to be obtained by a purchaser … if they wished to carry out work to enable it to be occupied and used as a single house" (paragraph 33 of the decision). It said that "[the] difference in approach between the parties can be seen most clearly by comparing [Mr Mark Sefton Q.C.'s] suggestion [on behalf of Cadogan] that the alterations should be assumed notionally to be undone on the valuation date with [Mr Stephen Jourdan Q.C.'s] submission [on behalf of Mrs Alberti] that it should be assumed Mr Scarfe did not convert the building from five self-contained flats into a single house in the 1970s" (paragraph 44).
24. There was, in the tribunal's view, "no doubt about how section 9(1A)(d) is to be applied". It saw "authoritative guidance" in Lord Hoffmann's speech in Shalson, though the issue in that case was not the same as in this. To identify the extent to which the value of a house has been increased by an improvement, it is necessary to make a comparison "between the value of the house as it stands and what its value would have been if the improvement had not been made" – as Lord Hoffmann had put it (at paragraph 19). The same approach had been taken by Sir Martin Nourse in Fattal, where he had said (at paragraph 13) that what is necessarily involved is "a valuation of the property as it would have been on the valuation date if it had not been improved" (paragraph 45 of the tribunal's decision).
25. The tribunal went on to say that "[as] far as the works carried out in the 1970s are concerned, the only permitted departure from reality is therefore that it must be assumed that the house was not converted from a building divided into flats by works rendering it capable of single occupation". It rejected Mr Sefton's submission that the building must be assumed to be divided into flats on the valuation date alone, and that no assumption contrary to reality should be made about its condition at any other time. The exercise was "not limited to interfering with reality on the date of the valuation alone; the history of the house is to be re-written by treating the improvements as if they had not been made" (paragraph 46).
26. The core of the tribunal's reasoning is in the following three paragraphs of the decision (paragraphs 48 to 50):

"48. In my judgment the inevitable consequence of treating the works as if they had never been done is that any occupation of the house between the date on which the works were carried out and the valuation date must be assumed to have been of the building in its unimproved condition. It follows that the prospective purchaser of the unimproved house on the valuation date would not be advised that, although the building was divided into five flats, it nevertheless had the benefit of an established planning use which would render it lawful, without planning consent, to occupy it as a single house.

49. I accept Mr Jourdan's submission that, in principle, the best proxy for the value of the unimproved house would be a house next door which had been divided into flats on the date the lease was granted and remained in that condition on the valuation date. The planning status of the two properties would be the same and Cadogan should therefore expect to receive the same price on a notional sale of both properties. To credit Cadogan in the assessment of the price of 10 Cheyne Walk with the benefit of a planning status which was a consequence of the occupation of the property as a single house, when that style of occupation was enabled only by the improvements carried out by Gerald Scarfe, would not be to diminish the price by the extent to which the value of the house had been increased by those improvements.

50. Mr Sefton submitted that Fattal was binding authority that the planning status of the building should be taken to be as it actually was on the valuation date. I disagree. In Fattal the tenant's improvements did not cause the planning permission to be granted; the planning permission was not itself an improvement nor was it an inevitable consequence of the improvements (on the contrary, it pre-dated the improvements). In this case the planning status of the building is not an improvement, but it is a direct consequence of the improvements and would not have been enjoyed without them. That requires that the building must be assumed to have had the same planning status on the valuation date as it would inevitably have had if the improvements had not been carried out."

27. For those reasons, the tribunal determined the preliminary issue in favour of Mrs Alberti, concluding (in paragraph 52) that "[section] 9(1A)(d) does require the Tribunal to assume that it would have been unlawful as a matter of planning control to use 10 Cheyne Walk as a single house on the valuation date".

How should the preliminary issue be determined?

28. The competing interpretations of section 9(1A)(d) presented to the tribunal were repeated and amplified before us.
29. For Cadogan, Mr Mark Sefton Q.C. submitted, as he did before the tribunal, that section 9(1A)(d) is a counter-factual deeming provision, whose effect is to deem the house in question to be in a different physical state on the valuation date from its physical state in reality. It is concerned only with the physical state of the building on the valuation date. It simply obliges the tribunal to ascertain whether, on the valuation date, there was a difference between the value of the house with the improvements which had been carried out and its value without those improvements. This is the full extent of the statutory assumption. Mr Sefton cited Lord Millett's speech in Shalson (at paragraph 40) as authority for that proposition. The notion that section 9(1A)(d) also deems the lawful use of the building, under the planning legislation, to be different from what it is in reality is mistaken. The tribunal therefore fell into error when it accepted that the provision not only deems the building to be in a different physical state on the valuation date, but also, in effect, deems the lawful use to be as it would have been if the building had not been used in the way that the improvements made possible. Here, therefore, the tribunal was wrong to hold that section 9(1A)(d) deemed it to be unlawful under the planning regime to use the building as a single house, given that such use was lawful in reality and, indeed, had been the actual use of the building continuously for some four decades before the valuation date. The "reality principle", Mr Sefton submitted, dictates that where there are competing interpretations of valuation provisions, the court must choose the one which involves the most limited departure from reality. In this case, that interpretation is Cadogan's.
30. Mr Sefton contended that the "enormous profit" which would result for Mrs Alberti if her interpretation of the statutory assumption were correct, would be "out of all proportion" to the works Mr Scarfe had carried out. It was not those works which had produced the "discount" in the value of the freehold, but the change in planning policy. Cadogan would be "hugely undercompensated" on Mrs Alberti's approach, which was "capricious". Neither Mr Scarfe nor Mrs Alberti had paid anything at all to secure the lawful use of the building as a single house. That had resulted from the then planning policy.
31. It would be arbitrary and anomalous, submitted Mr Sefton, if a planning permission, whoever had applied for it, were treated as existing for the purposes of the valuation, but not a lawful use of the building under the statutory scheme for planning control. The lawfulness of the use of the building, which exists in reality, should be treated as existing for valuation purposes, no matter whether it derives from a grant of planning permission or from a change of use which became lawful without requiring permission to be granted. In support of this submission, Mr Sefton cited Sir Martin Nourse's judgment in Fattal.
32. For Mrs Alberti, Mr Stephen Jourdan Q.C. submitted that the tribunal was right to decide the preliminary issue as it did. The ordinary meaning of the language used in section 9(1A)(d) required a valuation on the assumption that the improvements carried out by Mr Scarfe had never been made. This is how the provision had been interpreted in all the relevant cases. This understanding also reflects the purpose of the assumption, which is to ensure that the landlord does not get a higher price for the freehold than he would have done if the tenant had not undertaken the improvements. The "reality principle" does not compel a different construction. Nor do the cases on valuation provisions in other statutory contexts. Moreover, the interpretation urged by Mr Sefton would have the unsatisfactory consequence that a comparable property, unimproved, which would provide the best evidence of value here, could not be used in the valuation exercise.
33. In my view, as Mr Jourdan submitted, the tribunal's interpretation of section 9(1A)(d) was correct. It accords with the ordinary and natural meaning of the statutory words – "the extent to which the value of the house and premises has been increased by any improvement carried out by the tenant or his predecessors in title at their own expense". It finds support in the authorities in which those words have been construed and applied. It accurately reflects the purpose underlying the assumption. It is not at odds with the "reality principle". And it is conducive to a realistic approach to valuation, in which the value of the unimproved building can be derived by looking at the prices paid in the open market for similar buildings sold unimproved.
34. Interpreting the assumption in section 9(1A)(d) should not be difficult. The starting point must be that Parliament phrased that assumption no more broadly or narrowly than the language it actually used. The ordinary and natural meaning of the statutory words is, I think, plain. It prescribes an exercise in which the valuer establishes by how much the value of the house and premises "has been increased by any improvement" carried out by the tenant, or his predecessor, and reduces the price payable for the freehold interest to that extent. The price payable for the freehold must leave out of account any increase in value which is a consequence of improvements made at the tenant's expense – that is, any increase in value which would not have happened but for the improvements.
35. An ample explanation of that concept has already been provided in authority at the highest level, and also in this court. In Shalson, Lord Bingham spoke of an enhancement to the price of the freehold "attributable to works done by [the tenant] or his predecessors in title", and the requirement to reduce the market price "so as to discount the increase attributable to that improvement …" (paragraph 3). Lord Hoffmann described the concept of the value of the house and premises being increased by an improvement as "[signifying] a simple causal relationship: but for the improvement, the house and premises would have been worth less". So the comparison required is a "comparison … between the value of the house as it stands and what its value would have been if the improvement had not been made" (paragraph 19). Likewise, Lord Millett construed the statutory phrase "the extent to which the value of the house and premises has been increased" as connoting "the difference between the value of the property with the improvement in question and the value of the property without it" (paragraph 40). In Fattal, Sir Martin Nourse saw in those observations of their Lordships in Shalson the need for a comparison to be made "between the value of the property in its unimproved state and the value it would have had if it had not been improved" (paragraph 10). The statutory assumption calls for the amount of the "increase in value caused by the improvements" to be calculated, which "necessarily involves a valuation of the property as it would have been on the valuation date if it had not been improved" (paragraph 13).
36. These observations on the meaning of section 9(1A)(d) are, in my view, authoritative. Their gist is captured well in Hague's reference to the practice of valuing the property "as if the improvements had not been made".
37. Valuing the property "as if the improvements had not been made" does not equate the increase in value attributable to the tenant's improvements with the cost to him of undertaking those works. That is not what Lord Millett meant in Shalson when he said that the assumption in section 9(1A)(d) was designed "to avoid the tenant having to pay a price which reflects a value in the property for which he has already paid". The exercise is concerned not with the cost or value of the improvements as such, but rather with the diminution in the value of the property itself if the increase attributable to those improvements is disregarded. The whole increase in value which can be attributed to the improvements must be discounted – nothing more and nothing less.
38. If one is to establish the real extent to which the value of the property "has been increased" by the tenant's improvements, one does not merely assume the improvements to be absent on the valuation date, ignoring their consequences for the value of the property. Those consequences must be built into the assumption. Otherwise, one will fail to make a realistic comparison "between the value of the house as it stands and what its value would have been if the improvement had not been made", and it will be impossible to arrive at "a valuation of the property as it would have been on the valuation date if it had not been improved".
39. As Mr Jourdan submitted, the purpose of the assumption in section 9(1A)(d) is to ensure that the price paid by the tenant to buy the freehold does not exceed what it would have been had the tenant or his predecessors in title not made the improvements, instead leaving the property as it was when the lease was granted. The intended outcome is twofold: that the tenant does not have to pay twice and that the landlord does not receive a higher price because of works which he was not entitled to require the tenant to undertake. The provision is intended to favour the tenant. Lord Millett pointed out in Shalson (at paragraph 42) that the landlord gets a lower price if the tenant makes alterations whose effect is to lessen the value of the property, but does not benefit if the tenant's alterations increase that value.
40. This understanding of the statutory assumption is founded on its proper construction. It embodies the concept of discounting an increase in value attributable to a tenant's improvement, whatever that increase might be. It is not vulnerable to the level of increase in value for which the statutory assumption is responsible. The increase in value, in itself, must be attributable to the improvements carried out by the tenant or his predecessor in title. The scale of that increase may be small or, as in this case, very great.
41. Here, the increase in value attributable to the improvements, which would not have occurred but for those improvements, arises because the value of 10 Cheyne Walk as a single house is greater than its value in its unimproved state as five flats. The scale of that increase in value is largely the consequence of two things in combination: first, the significant evolution in development plan policy between the carrying out of the works and the valuation date, so that relevant policy now presents an obstacle to planning permission being granted for the conversion of buildings from flats to single houses; and secondly, the dynamics of the housing market, which have caused a vast rise in value for large single houses in prime locations in certain parts of London, including Chelsea. Neither of these two things was the responsibility of Mrs Alberti or Mr Scarfe, or of Cadogan. Both were, in that sense, "external factors" – as Mr Jourdan described them – though not wholly unpredicted. In the upshot, a substantial gain will accrue to both parties, landlord and tenant, largely because of events for which neither was responsible. One of them, of course, may get a much greater profit than the other. Which party enjoys that success, however, will depend on the true interpretation and lawful application of section 9(1A)(d). To suggest that in these circumstances "fairness" points to an outcome favouring Cadogan is, in my view, mistaken.
42. The counter-factual assumption required under section 9(1A)(d), as the tribunal held, is that the house had not been converted in the 1970s – or re-converted – from a building divided into five flats to a building which could be occupied as a single house. As the tribunal recognised, if one is to ascertain the extent to which the value of the property has been increased by the improvements carried out by Mr Scarfe, one must not shut one's eyes to what was achieved by those improvements, namely, under the planning legislation, the emergence of an established use of the building as a single house after its occupation in that use for the requisite period – which in this case, under section 87(3) of the 1971 Act, was four years (see the judgment of Lord Mance in Welwyn Hatfield Borough Council v Secretary of State for Communities and Local Government [2011] UKSC 15; [2011] 2 AC 304, approving Impey v Secretary of State for the Environment (1980) 47 P. & C.R. 157). There is nothing arbitrary or anomalous here. The hypothesis generated by the assumption that Parliament inserted in section 9(1A)(d) is thus given its intended effect, in full. Giving it only partial effect would distort the valuation exercise required by the statute.
43. The artificiality of the approach advocated by Mr Sefton is that it assumes a diminution in value to the extent of ignoring only the improvements themselves on the valuation date, and not the uplift in value brought about by the use of the building as a single house – an uplift in value which resulted from, and would not have happened but for, those very improvements. This would involve two suppositions which in my view are hard to reconcile. The first would be that the improvements undertaken by Mr Scarfe did not exist on the valuation date. But the second would be that the planning status of the building which was the consequence of those improvements had somehow come into existence without them and now subsisted. This, I think, would not produce a true comparison between "the value of the house as it stands" and "what its value would have been if the improvement had not been made". It would not lead one to the value of the property "as it would have been on the valuation date if it had not been improved".
44. I therefore agree with the tribunal that an inevitable consequence of treating Mr Scarfe's improvements as if they had never been carried out is that any occupation of the house that took place after those works were done and before the valuation date would have been occupation of the building in its unimproved state, as five flats. But for those improvements, the use of the building as a single house would not have come about. Any would-be purchaser of the house on the valuation date would have been told that planning permission, and listed building consent, would now be required if such use, and the works necessary to make it possible, were to be lawful.
45. As Mr Jourdan submitted, this interpretation of section 9(1A)(d) makes possible a practical and reliable approach to valuation. The tribunal rightly concluded that the best comparable – or "proxy" – for establishing the value of the house as if it were unimproved in the sense of the assumption in section 9(1A)(d) "would be a house next door which had been divided into flats on the date the lease was granted and remained in that condition on the valuation date". Planning permission and listed building consent would therefore be needed for its conversion into a single house. This would be a dependable basis for establishing the value of the property "as it would have been on the valuation date if it had not been improved". The alternative, based on Cadogan's interpretation, would be false. It would not exist in the real world at all, but only in the realm of make-believe. In that scenario, no genuine comparable could ever be found.
46. Crucially, I think, if one is to diminish the price of the freehold by the extent to which the value of the house had been increased by the improvements undertaken by Mr Scarfe, it would be misconceived to credit Cadogan with the benefit of the planning status gained by the building through its occupation as a single house when, on the facts here, that planning status would not have been obtained but for those improvements. This point was central to the tribunal's analysis (also in paragraph 49), and in my view it is sound.
47. Neither the House of Lords' decision in Shalson nor this court's in Fattal supports a different treatment of the planning status of the property at the valuation date from that favoured by the tribunal. Sir Martin Nourse in Fattal did not say, or imply, that there is a general principle that the planning status of the property must always be taken as it actually was on the valuation date. Rather, he recognised that what was to be subtracted in the valuation was any "increase in value caused by the physical works", and that "the existence or availability of planning permission is not part of those works". His conclusion was confined, as it had to be, to the particular facts of the case in hand. He reached that conclusion on the orthodox basis – as Lord Hoffmann had stressed in Shalson – that an improvement is a physical thing, and that the assumption in section 9(1A)(d) calls for the subtraction of increases in value attributable to such physical works. The grant of planning permission by the local planning authority in Fattal was not part of the physical works; nor was it a result of them. This reasoning reflects what Lord Hoffmann referred to in Shalson (at paragraph 19) as a "simple causal relationship: but for the improvement, the house and premises would have been worth less". And it was this "causal relationship", fundamental to the statutory assumption, that the tribunal had in mind in determining the preliminary issue as it did.
48. In Fattal, as the tribunal said in this case, the improvements carried out by the tenants did not cause the local planning authority's grant of planning permission. The planning permission was neither an improvement itself nor a consequence of the improvements undertaken by the tenants. Indeed, it had been granted before those improvements were carried out. Here, by contrast, the planning status of 10 Cheyne Walk on the valuation date did not rest on a grant of planning permission for the works undertaken by Mr Scarfe in the 1970s. Neither Mr Scarfe nor Cadogan had ever applied for planning permission for those works, though it would have been possible for this to be done, either before the works were carried out or afterwards. As the tribunal said, therefore, the planning status of the building was "a direct consequence of the improvements" and "would not have been enjoyed without them". The necessary "causal relationship" between the improvements and the increase in value was present. So, for the purposes of section 9(1A)(d), it had to be assumed that, on the valuation date, the building had the same planning status as it would have had if the improvements had not been carried out. This conclusion seems to me both conventional and correct, and to accord with the intention of Parliament in framing the statutory assumption as it did.
49. Contrary to Mr Sefton's submission, the opposite conclusion is not justified by Knox J.'s decision in Railstore v Playdale. That case did not relate to the valuation provisions in the 1967 Act. It concerned a rent review clause in a lease, which required the open market rent to be ascertained "disregarding any effect on rent of any improvement carried out by the lessees otherwise than in pursuance of the obligation to the board herein contained". This assumption was different from that in section 9(1A)(d) of the 1967 Act. It was directed at "any effect on rent" of an improvement made by the tenant, whereas the section 9(1A)(d) assumption goes to the amount by which the value of the house and premises "has been increased" by any improvement.
50. Knox J. followed the approach to rent review cases described in this court's decision in Basingstoke and Deane Borough Council v Host Group Ltd. [1988] 1 W.L.R. 348, which includes, as he put it (at p.155F), "the strong presumption against either the landlord or the tenant obtaining an advantage which is referable to a factor which has no existence as between the actual landlord and the actual tenant". This "presumption of reality" – later elucidated in Chancebutton v Compass Services UK and Ireland Ltd. [2005] 1 P. & C.R. 8 – embodies the basic commercial purpose of a rent review. That commercial purpose is not the same as the statutory purpose of the valuation assumptions in section 9 of the 1967 Act, which are designed to promote the interests of the tenant even where this is to the landlord's disadvantage. As Lord Salmon said in United Scientific Holdings Ltd. v Burnley Borough Council [1978] A.C. 904 (at p.948), in a classic statement explaining the nature of rent review provisions in leases, "[it] is totally unrealistic to regard [rent review] clauses as conferring a privilege upon the landlord or as imposing a burden upon the tenant"; they are "plainly for the benefit of them both".
51. It must be said that the planning circumstances in Railstore v Playdale are not entirely clear from the law report. It is not apparent whether the planning status of the demised land was the consequence of planning permissions granted before the works of construction were undertaken or came into being only after such works were done, and as a direct result of those works. Knox J. referred to "the existing planning permission on what are called buildings G and H" (p.155F), having earlier said that those two buildings "were put up by the tenant, and there is no doubt but that that was a perfectly lawful operation" (p.154E to F). He went on to say that if "at the period when the lease was granted there was an expectation of planning permission which ripened into fact so that a building was erected and thereafter the planning law either went through a legislative or a policy change at some ex hypothesi wholly uncertain date in the future, this … quite fortuitous event, which might fall on either side of a rent review date, could, if the tenants are right in their argument, have the effect of radically altering the rent which it would be proper for the tenant to pay in respect of a totally undisturbed and unaffected occupation pursuant to the lawful planning permission that the tenant earlier obtained". He did not find himself driven to that result in the case before him (p.155G to H).
52. The obscurity of the facts in Railstore v Playdale did not prevent Sir Martin Nourse concluding as he did in Fattal (at paragraph 19): Knox J. had held that the rent review clause did not require "the existence of a planning permission for the buildings which the tenant had built and was occupying" to be disregarded. This is also the understanding of the editors of "Hill & Redman's Law of Landlord and Tenant" (at paragraph A[2066]), and the authors of the "Handbook of Rent Review" (at paragraph 6.3.7).
53. Whether the thrust of Knox J.'s reasoning was as Sir Martin Nourse took it to be, I do not think it undermines the tribunal's analysis here. It certainly should not be taken as authority for a different interpretation of section 9(1A)(d) of the 1967 Act. It is not inconsistent with the concept that the statutory assumption in that provision operates to exclude any uplift in value attributable to the tenant's improvements, and that this exclusion necessarily extends to increases in value caused by those physical works, but not to increases in value attributable to events which are not, on the facts, the result of any physical works of improvement. As is clear in the relevant authorities, this will, of course, always depend on the particular facts of the case under consideration.
54. Mr Jourdan also drew our attention to the decision of the New Zealand Court of Appeal in S&M Property Holdings v Waterloo [1999] 3 NZLR 189, where it was held that a direction in a rent review clause that an improvement to the demised land had to be disregarded also had the effect of requiring the disregard of a regulatory restriction on the use of the land, which applied because the improvement was made.
55. Some support for the logic of the tribunal's analysis, though under a different statutory scheme, may be drawn from the Privy Council's decision in Toohey's Ltd. v The Valuer-General [1925] AC 439. The relevant statutory provision – section 6 of the Valuation of Land Act, No.2 of 1916 (N.S.W.) – directed that the "unimproved value of the land" was to be ascertained "assuming that the improvements, if any, thereon or appertaining thereto, and made or acquired by the owner or his predecessor in title, had not been made". The property in that case was a hotel in Sydney with a liquor licence. On the agreed valuations, the difference between the value of the licensed premises and the value of the premises without a licence was due entirely to the fact that in one case the premises were licensed, and in the other not. The Privy Council held that the value of the liquor licence had to be discounted. If the existence of the hotel had to be disregarded, so did the licence – because a licence could only be granted for a building. Lord Dunedin, in emphatic terms, declared that the valuer had to consider "what the land would fetch as at the date of the valuation if the improvements had not been made". The "improvements were to be left entirely out of view". They were "to be taken, not only as non-existent, but as if they had never existed" (p.443).
56. Finally, I see no difficulty in reconciling Mr Jourdan's argument with the "reality principle". Where a statutory assumption requires a valuation to be conducted on a different basis from reality, the language and purpose of the statute must be heeded. Here, both language and purpose are plain. The statutory assumption itself is not ambiguous. It is in clear terms. It was formulated by Parliament to concentrate the valuer's mind on increases in value attributable to improvements carried out by the tenant or his predecessors in title, and not more broadly than that. And there is no plausible interpretation of it competing with Mrs Alberti's. In these circumstances, as Lewison L.J. said in Harbinger Capital Partners v Caldwell [2013] EWCA Civ 492 (at paragraph 23), whilst the valuer must not depart from the real world further than the hypothesis compels, it is necessary to give that hypothesis "full effect", and any inevitable consequence of it must also be assumed. In the same case Lord Justice Mummery observed (at paragraph 122) that "the principle does not determine or limit what the statute commands us to assume contrary to reality". As he emphasised, "[the] statute determines that". The "reality principle", he said, "is about what is not covered by the statutory assumption".
57. Submissions were made to us in the light of decisions on valuation provisions in various statutory contexts, including Trocette Property Co. Ltd. v Greater London Council (1972) 28 P. & C.R. 408, Sharp v Cadogan (1998) (unreported) LRA/33 & 35/97, Hoare v National Trust (1998) 77 P. & C.R. 366, Secretary of State for Transport v Curzon Park Ltd. [2021] EWCA Civ 651; [2021] PTSR 1560, Mundy v Sloane Stanley Estate [2018] EWCA Civ 35; [2018] 1 WLR 4751 and Whitehall Court London Ltd. v Crown Estate Commissioners [2018] EWCA Civ 1704; [2019] 1 WLR 2319. Most pertinent here, in my view, is the fact that the "reality principle" generally requires effect to be given to the inevitable consequences of the assumed counter-factual state of affairs.
58. In the recent decision of the Supreme Court in Fowler v Revenue and Customs Commissioners [2020] UKSC 22; [2020] 1 WLR 2227 Lord Briggs set out (in paragraph 27 of his judgment) five principles applicable to the interpretation and application of statutory deeming provisions. First, "[the] extent of the fiction created by a deeming provision is primarily a matter of construction of the statute in which it appears". Secondly, "[for] that purpose the court should ascertain, if it can, the purposes for which and the persons between whom the statutory fiction is to be resorted to, and then apply the deeming provision that far, but not where it would produce effects clearly outside those purposes". Thirdly, however, "those purposes may be difficult to ascertain, and Parliament may not find it easy to prescribe with precision the intended limits of the artificial assumption which the deeming provision requires to be made". Fourthly, "[a] deeming provision should not be applied so far as to produce unjust, absurd or anomalous results, unless the court is compelled to do so by clear language". But fifthly, "the court should not shrink from applying the fiction created by the deeming provision to the consequences which would inevitably flow from the fiction being real …" (see also the speech of Lord Browne-Wilkinson in Marshall (Inspector of Taxes) v Kerr [1998] A.C. 148 (at p. 164), and the judgment of Lord Walker in DCC Holdings (UK) Ltd. v Revenue and Customs Commissioners [2010] UKSC 58; [2011] 1 WLR 44, at paragraphs 36 to 39). To emphasise this last principle, Lord Briggs also cited the salutary words of Lord Asquith in East End Dwellings v Finsbury Borough Council [1952] A.C. 109, at p. 133: "[the] statute says that you must imagine a certain state of affairs; it does not say that having done so, you must cause or permit your imagination to boggle when it comes to the inevitable corollaries of that state of affairs".
59. In my view, the tribunal's reasoning and conclusion comply well with those principles, in particular the fourth and the fifth.

Conclusion

60. For the reasons I have given, I would dismiss the appeal.

Lord Justice Newey:

61. I agree.

Lady Justice King:

62. I also agree.