Compulsory purchase — Compensation — Valuation method — Development potential — Agricultural land — Landowners seeking to have development potential of land taken into account in valuation — No direct comparables sufficiently similar to land compulsorily acquired — Whether residual method of valuation applicable Whether spot valuation appropriate
The first respondent compulsorily acquired from the appellant an area of land in Mauritius for the purpose of building a hospital. The land formed part of a larger area in the appellant’s ownership that was cultivated for sugar cane. The appellant claimed compensation, which fell to be assessed as at the date of the statutory notice of acquisition in April 2000. At that time, the land was zoned for agricultural purposes. An issue arose as to whether the possibility that the land could be re-zoned for residential or other development in the foreseeable future could be taken into account when assessing compensation. At the relevant date, no application to re-zone the land had been made and no development permit had been obtained. A nearby tract of land was re-zoned for residential development in 2001 pursuant to a government support deal, which affected the zoning scheme of the entire area. However, expert evidence indicated that such a deal could not have been foreseen at the relevant valuation date.
The second respondent rejected the comparables put forward by the parties as being insufficiently similar to the subject land and accordingly rejected the direct comparison method of valuation. It valued the land according to the residual method, by reference to a hypothetical development, assuming that the land in question could be developed for ultimate sale to purchasers. That produced a figure of more than R39.743m. The Supreme Court allowed an appeal by the respondents and substituted a figure of R6.430m. It found that the development potential of the land was at best speculative and not one that could reasonably be expected to be achieved in the short term. Accordingly, it valued the land by comparison with other agricultural land but with an uplift for hope value. The appellant appealed.
Held (Baroness Hale and Sir Peter Gibson dissenting): The appeal was dismissed. The residual value method of valuing an interest in land that is compulsorily acquired should be applied only in exceptional cases, where a proposed development scheme has such prospects of success that the comparison method cannot give a realistic and reasonably assessable figure. It is materially more suitable for valuing land where variables such as the chance of obtaining planning permission are not large and the effect upon valuation of any contingencies can readily be assessed. It should not be applied where the open market value is otherwise ascertainable by a spot valuation, which can take into account the existence and amount of hope value. Hope value represents the premium over existing use value that a developer might be thought willing to pay in order to acquire the land in the hope of turning it to profitable account. Its assessment will depend upon an amalgam of factors, including the likelihood of the requisite planning permission being granted, the demand for the suggested development, the time that such a development would take and its projected costs. The appellant’s proposed scheme contained a number of variables that required assessment by the hypothetical developer, including the effect of various applicable taxes and the chances of obtaining the necessary permits. It was not possible to tell from the evidence what notional deduction a developer might make for such matters. There were serious difficulties in the way of accepting the residual method of valuation. The Supreme Court had been justified in adopting a valuation based upon existing use plus a modest addition for hope value. It had correctly assessed as low the possibility of a successful and profitable development taking place. Since there were no local sales with hope value, it was appropriate to apply a spot figure.
The following cases are referred to in this report.
Benmax v Austin Motor Co Ltd [1955] AC 370; [1955] 2 WLR 418; [1955] 1 All ER 326, HL
Gajapatiraju v Revenue Divisional Officer, Vizagapatam; sub nom Gajapatiraju v Vizagapatam (Revenue Divisional Officer) [1939] AC 302; [1939] 2 All ER 317; 55 TLR 563, PC
Lavender Garden Properties Ltd v Enfield London Borough Council [1968] 2 All ER 401; 66 LGR 544; (1968) 19 P&CR 480; [1968] RVR 268, CA, (1967) 18 P&CR 320; [1967] JPL 346, LT
Perkins v Middlesex County Council (1952) 2 P&CR 42, LT
Pointe Gourde Quarrying & Transport Co Ltd v Sub-Intendent of Crown Lands [1947] AC 565
Waters v Welsh Development Agency [2004] UKHL 19; [2004] 1 WLR 1304; [2004] 2 All ER 915; [2004] 2 EGLR 103
This was an appeal by the appellant, Mon Tresor & Mon Desert Ltd, from a decision of the Supreme Court in Mauritius allowing an appeal by the first respondent, the Ministry of Housing and Lands, from a decision of the second respondent, the Board of Assessment, assessing compensation for the compulsory acquisition of land.
Graham Stokes (instructed by Clifford Chance LLP) appeared for the appellant; Anthony Speaight QC (instructed by Carrington & Associates) represented the first respondent; the second respondent did not appear and was not represented.
Giving the joint majority opinion, Lord Scott of Foscote and Lord Carswell said:
[1] The issue on this appeal is whether the Supreme Court of Mauritius was right to reverse the decision of the Board of Assessment (the Board) to value the lands compulsorily purchased by the government of Mauritius by the residual value method and to accept instead the valuation propounded by the chief government |page:14| valuer, based upon comparisons with an added hope value. By an award dated 5 April 2004, the Board (Caunhye J, Mr Y Coret and Mr D Ramasawmy), having rejected the comparisons put forward, adopted the residual value basis and awarded the appellant the sum of R39,743,588. The Supreme Court (Matadeen and Domah JJ) allowed the respondents’ appeal and, in a written judgment given on 19 January 2006, amended the award by substituting the figure of R6.43m.
[2] It was not in dispute that the appeal from the Board to the Supreme Court, under section 24 of the Land Acquisition Act 1982 (the 1982 Act), was a full appeal on both fact and law, as is the further appeal to the Privy Council. Such appeals are governed by the principles laid down by the House of Lords in Benmax v Austin Motor Co Ltd [1955] AC 370. An appellate tribunal ought to be slow to reject a finding of specific fact by a lower court or tribunal, especially one that is founded on the credibility or bearing of a witness. It can, however, form an independent opinion on the inferences to be drawn from, or evaluation to be made of, specific or primary facts so found, although it will naturally attach importance to the judgment of the trial judge or tribunal. On an appeal from a specialist tribunal such as the Board, the Supreme Court or the Privy Council should ordinarily be slow to reject its findings on matters of pure valuation, but if it considers that the tribunal has misapprehended material facts or that the primary facts established do not lead correctly to the inferences that it has drawn from them, it can and should reverse the decision of the tribunal.
[3] The subject land consists of a plot of land at Telfair, Moka, measuring 12 arpents and 86 perches (equivalent to 54,280m2 or 5.428ha). The land formed part of a larger area planted with sugar cane and owned by the appellant Mon Tresor & Mon Desert Ltd, which is part of the Lonhro group of companies. It was acquired by the Mauritian Ministry of Housing and Lands, under the 1982 Act, for the purpose of building a national children’s hospital and institute of cardiology and neurology. The statutory notice, under section 8 of the Act, was published on 8 April 2000; this forms the date upon which the land is to be valued.
[4] The land was surrounded by a large tract of prime agricultural land under sugar cane cultivation, owned on three sides by the appellant company. It lies approximately 200m from the Reduit-St Pierre public highway and 200m from an estate of public housing known as Cite Telfair. The site did not have electricity, water or foul drainage services, and access was by an untarred estate road. The land on the other side of the highway contained a substantial amount of development, including the University of Mauritius and the Mahatma Gandhi Institute. The subject land was zoned for agricultural purposes, and one of the issues in the appeal was the extent of the possibility that it might be re-zoned for residential or other development in the foreseeable future. No application had been made before 8 April 2000 to re-zone the site for planning purposes, and no development permit had been obtained for residential or other development. On the contrary, in correspondence with the Ministry of Housing and Lands in April 2000, the appellant resisted the compulsory acquisition on the ground of the value to it of the land for sugar cane production. In evidence before the Board, the company secretary stated that it had had difficulty in fulfilling all its commitments for producing sugar cane and wished to keep production as high as possible. There was accordingly no indication that, at the material date, it had any intention of parting with or developing the land. There was a significant amount of undeveloped agricultural land, some 100 arpents, on the other side of the highway that was within the area in which residential development could take place.
[5] In April 2001, things took an unexpected turn. The government brokered an arrangement referred to in evidence as the “Illovo deal”, described by the Supreme Court as a “very special support deal”, which affected the entire zoning scheme of the area. Under this scheme, a tract of land immediately surrounding the subject land was re-zoned for residential purposes in preparation for a major development. Mr Noor Dilmohamed, the chief government valuer, who gave expert evidence on behalf of the respondent government department, stated categorically in evidence that no reasonable man could have foreseen that such a deal would be forthcoming and that “at the relevant date no valuer could have unless he is a magician”.
[6] Both sides produced evidence of sales of land upon which they relied as comparables in order to establish the value of plots of land in the area. The appellant’s valuer, Mr Rhoy Ramlackhan, produced three comparables, but each of them was, as the Board pointed out, in a far better location, being proximate to developed areas with amenities. For this reason, both the Board and the Supreme Court, rightly in our view, declined to rely upon them for comparison. Mr Dilmohamed produced five comparables, all of which related to sales of plots of agricultural land in the district of Trianon, some 1.5km from the subject land. Each of these plots was in the middle of a large area of agricultural land under sugar cane production, well away from services or other development and with access only by estate roads. For this reason, the Board took the view that it did not have sufficient similar characteristics and that the direct comparison method should therefore be ruled out as unreliable. It accordingly resorted to the residual method of valuation. This method, deduced from a hypothetical development, assumes that the land in question can be developed for ultimate sale to purchasers. It is described in Johnson, Davies and Shapiro’s Modern Methods of Valuation of Land Houses and Buildings (9th ed) 2000, at p165, as follows:
The method works on the premise that the price which a purchaser can pay for such property is the surplus after he has met out of the proceeds from the sale or value of the finished development his costs of construction, his costs of purchase and sale, the cost of finance, and an allowance for profits required to carry out the project.
The Supreme Court, on the other hand, rejected the residual method on the ground that the development potential was “at best speculative and, in any event, not one that can reasonably be expected to be reached in the short term”. It accordingly accepted Mr Dilmohamed’s valuation, based upon the comparables of agricultural land, with an uplift for “hope value”.
[7] In our opinion, the following propositions may be deduced from the authorities:
(a) The value of an interest in land compulsorily acquired is the amount that that interest, if sold on the open market by a willing seller, might be expected to realise at the date of first publication of the statutory notice. This familiar principle is given statutory form in Mauritius by section 19(3) of the 1982 Act.
(b) In assessing this value, the best evidence is comparison with figures from other sales of comparable property.
(c) The land acquired must be valued not merely by reference to the use to which it is being put at the time at which its value has to be determined but also by reference to the uses to which it is reasonably capable of being put in the future: see Gajapatiraju v Revenue Divisional Officer, Vizagapatam [1939] AC 302.
(d) The use for which the land is being acquired must be disregarded in making this assessment: Pointe Gourde Quarrying & Transport Co Ltd v Sub-Intendent of Crown Lands [1947] AC 565; Waters v Welsh Development Agency [2004] UKHL 19; [2004] 1 WLR 1304*.
(e) Where there are no comparable sales, resort may be had to the residual value method. This should be reserved for exceptional cases and will not be applied where the open market value is otherwise ascertainable by assessments such as a spot valuation: see Cripps on Compulsory Acquisition of Land (11th ed) 1962, in para 4-200. As the Lands Tribunal stated in Perkins v Middlesex County Council (1952) 2 P&CR 42, at p47:
a spot valuation based upon experience of the market is more likely to be right than calculations which depend upon many assumptions and forecasts.
(f) A spot valuation can take into account the existence and amount of hope value. Its assessment depends upon an amalgam of factors: the likelihood (ranging from complete certainty to a very slight possibility) of the requisite planning permission being granted; the demand for |page:15| the suggested development; the time that such development would take; and the projected costs. The resulting figure represents the premium over existing use value that a developer may be thought willing to pay in order to acquire the land in the hope of turning it to profitable account.
We accordingly consider that if a spot valuation based upon comparison plus an element of hope value can give a realistic figure for the amount that a speculative developer might be willing to pay for the land, it would be wrong to adopt the residual value method. In our opinion, that method should be adopted only where a proposed development scheme has such prospects of success that the comparison method cannot give such a realistic and reasonably assessable figure. It is materially more suitable for valuing land where variables such as the chance of obtaining planning permission are not large and the effect upon the valuation of any contingencies can be readily assessed: cf Lavender Garden Properties Ltd v Enfield London Borough Council (1967) 18 P&CR 320; affirmed [1968] RVR 268.
—————————————————————————
* Editor’s note: Also reported at [2004] 2 EGLR 103
—————————————————————————
[8] In the present case, we are of the opinion, for the reasons that we shall give, that the Supreme Court was right in assessing the possibility of a successful and profitable development taking place as low. We further consider that the comparables relied upon by Mr Dilmohamed gave an acceptable basis for assessing the value of the subject land, with an appropriate adjustment for a modest amount of hope value. There were no local sales with hope value, so the assessment has to be a spot figure. The comparables that he propounded varied in size between one arpent and 11 arpents, the sales took place between December 1996 and November 1997 and the price per arpent ranged between R300,179 and R409,636. The government’s figure of R500,000 per arpent for the subject land therefore contained a hope value premium somewhere between 18% and 40%.
[9] If this conclusion is correct cadit quaestio, but in order to determine whether it is correct it is necessary to consider, as the Supreme Court did, the prospects that a developer might calculate of being able to develop the land successfully for residential purposes.
[10] The appellant’s projected scheme contained a number of variables that required assessment by the hypothetical developer:
” the demand for the type of housing proposed, which affects both the price that the developer could charge on the sale of plots and the time that it would take to complete the sale of the plots;
” the projected costs of the infrastructure works;
” the developer’s projected profit margin;
” the effect of the several taxes chargeable;
” the chances of obtaining the necessary permits and the time required to do so if they can be obtained.
[11] The Board made findings on the first three of these, which were strongly disputed by the respondent ministry, whose valuer’s figures were less favourable to a developer. In the ordinary way, we would regard these matters as falling within the area in which an appellate court should be slow to interfere with the findings of a specialist tribunal. Although the figures accepted may have been somewhat generous to the appellant, one would not, on that ground alone, reject them on appeal. The Board was, however, plainly wrong on the road access costs and, on the evidence adduced, its estimate of the time that would elapse before completion of the project was quite unrealistic. These flaws would have entitled the Supreme Court to review the items concerned.
[12] The case really turns, however, on the last two factors: the treatment of which by the Board was inadequate or incorrect. It was established by reference to statute law at the hearing before the Privy Council that three permits would be required for the development to proceed, a development permit relating to zoning or planning, a morcellement permit and a land conversion permit. We shall deal with these separately and the evidence relating to them.
[13] The subject land was situated within a planning area within the meaning of the Town and Country Planning Act 1990 (the 1990 Act) and the outline scheme for the district zoned it for agricultural use. By virtue of section 14(3) of the 1990 Act, no authority is to pass or approve any plan for building or development that contravenes the scheme. Accordingly, the only way in which a developer could have obtained the development permit required under the 1990 Act was to submit an application under section 24 to have the scheme modified and the land re-zoned. That application would have had to go to the Town and Country Planning Board for its approval, which requires consultation with central government ministries and the local authority and, finally, to the president for the modification order to be made.
[14] Mr Ramlackhan, in his valuation, assumed that the necessary permits would all be granted and had not checked the chances of success in obtaining them. Mr Dilmohamed had made enquiries with the relevant bodies of the Ministry of Housing and concluded that it was unlikely that, in the normal course of things, the land would be re-zoned. His testimony on this point was not challenged in cross-examination. Mr Pubiswar Hemoo, principal town planner at the Ministry of Housing, expressed the opinion that because the site was far from the existing village it would be very difficult to have allowed the re-zoning and that there would be very little chance of obtaining the permit. He agreed in cross-examination that there was no reason why the subject land would have been excluded when the surrounding land was re-zoned. It is quite apparent, however, that he was speaking of the situation that appertained when the Illovo deal was in being and an application was made to re-zone the land surrounding the subject land as part of that deal. We do not accept that his answers at this point in his evidence negated the conclusion of the Supreme Court that there was only a “bleak and remote possibility” of obtaining a re-zoning at the material time. On the evidence presented to the Board, we do not see that any conclusion was open to it but that there was little chance of a developer being successful in a re-zoning application. The Board did not refer at any point in its decision to the difficulties involved, notwithstanding the evidence that it had received, and appears to have assumed that the lands would be re-zoned, the only question being the time it would take. In our view, the Board’s decision was unsustainable in this respect.
[15] Morcellement is the division of a plot of land into two or more plots and, under the Morcellement Act 1990, requires a permit from the Morcellement Board. By virtue of the provisions of that Act, an applicant developer is required to submit details of infrastructural work, comprising such matters as roads, access and road connections and sewerage. A morcellement fee is payable, which, at R6 psm would amount to R325,680 for the subject lands. There was no suggestion in the evidence that obtaining a morcellement permit would cause particular difficulty or delay, but it is one more hurdle to be surmounted before a development could proceed. The need to satisfy the conditions of the Ministry of Public Infrastructure meant that an access road would have to be constructed on the line and to the specification laid down by it, with a consequential effect upon the costs, which the Board failed to acknowledge in its costing of the infrastructural works.
[16] The third permit, discussion of which formed a considerable part of the argument before us, is the land conversion permit. Mr Ramlackhan did not take this into account at all, assuming, incorrectly as it was established, that it would not be required if the land were re-zoned. The Board accordingly left it out of account in reaching its conclusions, whereas the issue of whether a permit could be obtained at all was very significant, as was the effect of the land conversion tax. This factor alone casts a considerable shadow over the validity of the Board’s conclusions.
[17] Land conversion is dealt with under the Sugar Industry Efficiency Act 1988 (the 1988 Act), which is aimed at regulating the conversion of agricultural land, especially land under sugar cane cultivation, to non-agricultural use. Section 5 provides that no agricultural land shall be put to a non-agricultural use except: (i) where the prescribed conditions are satisfied; (ii) with the prior written authority of the minister; and (iii) upon payment of the land conversion tax. The minister is advised by a Land Conversion Committee and he has to have regard, inter alia, to: the necessity of ensuring that the level of production of sugar is sufficient; preserving agricultural land; optimising agricultural production; preventing speculation in agricultural land; and respecting outline schemes and planning and development directives. |page:16|
[18] The land conversion tax, based upon area, was calculated at R18.998m. Section 5(7) of the 1988 Act specifies a number of situations in which land conversion tax will not be payable. The one material to the present case is set out in section 5(7)(f):
in respect of land… where… the applicant undertakes
(i) to sell to the Government at nominal rates, within a period of 6 months after the application is granted, 25 per cent of the agricultural land to be converted;
(ii) to plough back at least 60 per cent of the proceeds arising from the conversion, of which at least half to sugar production, or diversification within sugar in Mauritius, in the schemes specified in the Fifth Schedule, and the remainder to any other economic activity in Mauritius.
The Fifth Schedule sets out a range of schemes in which the proceeds could be invested, including both agricultural and industrial projects. It was represented on behalf of the appellant that it could readily satisfy the conditions, but it is less clear that the hypothetical developer would find it so straightforward. Moreover, Part III of the Sixth Schedule provides for a further restriction on the developer:
Where an authorisation for conversion granted under section 5 is in respect of land to which the rates applicable are the rates specified in Category I of Part I, and where the land converted is in excess of 5 hectares, the applicant shall, within a period of two years
(a) plough back at least 50 per cent of the proceeds arising out of the conversion to sugar production at field or factory level or diversification within sugar;
(b) fully compensate the loss in agricultural production computed by the committee by generating an equivalent amount of such production for at least one crop cycle of eight years by
(i) putting under cane cultivate other land belonging to the applicant; or
(ii) implementing projects relating to water and energy saving irrigation methods.
This restriction was not the subject of discussion in the Board’s decision or that of the Supreme Court, and we were not informed whether there are any avenues of escape from this requirement, but prima facie it appears to be a significant restriction on a developer’s freedom of movement and a deterrent to the conversion of agricultural land. It would therefore appear very likely that the hypothetical developer would be unable to take advantage of the exemption and would be liable for the tax.
[19] The appellant’s valuer had not taken the issue of obtaining a land conversion permit into account or checked the chances of success in obtaining a permit. Nor did Mr Dilmohamed deal with the prospects of success in the course of his evidence. The Supreme Court stated in its judgment (see record, at p214) that “the unchallenged evidence of the Town Planner was that at that time there would have been very little chance of obtaining the land conversion permit under the Sugar Industry Efficiency Act”. Mr Hemoo’s evidence appears, however, to have been directed entirely to the possibility of re-zoning under the planning legislation, save for an unresolved point concerning the time that it would take to obtain a land conversion permit. We are left to speculate about the issue, which we are reluctant to do, and the most we can say is that the minister would have had to approve the conversion, having regard to the factors in section 5(5) of the 1988 Act that may constitute contrary factors, and that the possibility of obtaining exemption from the land conversion tax appears very problematical.
[20] One further fact emerged in evidence that did not receive any attention in the decisions of the lower courts, but which seems to us to have some significance. In his cross-examination, at p80 of the record, Mr Ramlackhan stated that, in the previous 10 to 15 years, developers had not been buying properties to convert them into residential properties. What they had been doing was developing other people’s land at a fixed fee, without any risks on their part. Although the residual value method presupposes a hypothetical developer, it is part of an exercise designed to ascertain what the land would have fetched in the open market. If there were in fact no buyers in the open market for development, this would tend to show that the residual value method in the present case will not give a realistic figure for the true value of the land.
[21] In our opinion, the decision of the Board contained a number of defects. In the first place, it was too ready to depart from the comparison method of valuation of the land and to adopt the residual method. Second, its calculation based upon the residual method was flawed, in that:
(a) it assumed the existence of a hypothetical developer and disregarded the evidence of the absence of purchases for development;
(b) the estimate of the infrastructural costs, in particular the access road, was, on the evidence, too low;
(c) the estimate of the time that the project would require was substantially too low;
(d) it left out of account the issue of obtaining a land conversion permit and the effect upon the project of land conversion tax; and
(e) no allowance was made for the risk of failing to obtain the necessary permits, ignoring the evidence adduced by the respondent ministry.
In the process, the Board accepted with too little question the evidence of Mr Ramlackhan, which was deficient in a number of material respects, particularly in respect of land conversion and its cost and to the risk of failing to obtain the permits for the development. In the result, the board failed to give proper consideration to the issue of whether the hypothetical development would have been viable and whether any developer would make an offer at all or be prepared to pay more than agricultural value with a modest hope value in addition.
[22] It may be seen from the foregoing that there were serious difficulties in the way of accepting the residual method of valuation, in particular the effect of land conversion tax and the very substantial possibility that the necessary permits could not be obtained at all. Several calculations of the costs were put forward at various times in an attempt to furnish a value of the land based upon the residual method. The appellant’s figure, contained in Mr Ramlackhan’s written valuation of 8 August 2002 was scaled down by the Board as being excessively high. The Board’s own assessment suffers from the defects to which we have alluded. Mr Dilmohamed’s assessment based upon the residual method (see record, at p269) concluded that the project would not be viable, but it did not bring the deduction of land conversion tax into account in the correct part of the calculation and required adjustment. An attempt was made to provide that adjustment by the production during the hearing of the appeal of a revised assessment. This reworked figure showed a value per arpent that was very little more than the government’s offered figure and made assumptions about cost based upon the government’s own figures and not those accepted by the Board. It also made a deduction of 50% for the risk, which is a purely arbitrary assessment. The best conclusion that one could reach on these figures is that if the risk factor were ignored and the Board’s assumptions about costs accepted, there could be enough profit to justify an offer price materially higher than the government’s figure of R500,000 per arpent; perhaps two or three times that figure. We do not propose to attempt to rework the calculation, which is a difficult exercise requiring a valuer’s professional skills and, in any event, would be highly speculative.
[23] In our view, it is impossible to tell from the evidence what notional deduction a developer might then make for the risk of failing to get the permits, the time factor involved and the doubts concerning the extent of demand for housing assuming any developer could be found that would be interested in such a project. This leads us to the conclusion that resort to the residual method is an inappropriate means of assessing the value of the subject land. We consider accordingly that the Supreme Court was right to reverse the decision of the Board and reject a valuation based upon that method. It was right to accept a valuation based upon existing use value plus a modest addition for hope value. The only figure that it had before it on this basis was that of Mr Dilmohamed and, in our view, the Supreme Court was justified in adopting it.
[24] We would therefore dismiss the appeal with costs.
Giving his concurring opinion, Lord Brown of Eaton-under-Heywood said: |page:17|
[25] I have had the advantage of reading in draft the joint opinions respectively of Lord Scott of Foscote and Lord Carswell, who favour dismissing this appeal, and of Baroness Hale of Richmond and Sir Peter Gibson, who favour allowing it. In common with Lord Scott and Lord Carswell, I too would dismiss it but, in the light of what will be the minority opinion, rather than simply subscribe to Lord Scott and Lord Carswell’s opinion I prefer to explain my decision in my own words. I recognise that this is an unusual course to take, but I see no objection to it. Not merely is it the conventional course taken in comparable final appeals to the Appellate Committee of the House of Lords but it is the course taken by the board itself in Scottish devolution appeals.
[26] I gratefully take the detailed facts from the other opinions; they are most fully set out in that of Lord Scott and Lord Carswell. As they make it clear, the appeal concerns an island of land some 5.5ha in area (roughly equivalent to 240m2) within the appellant’s extensive sugar cane estate in Mauritius. The question raised on its compulsory acquisition by the government was as to its value on 8 April 2000. Section 19(3) of the 1982 Act provides that:
The value of any interest in the land should be the amount which that interest if sold on the open market by a willing seller, might be expected to realise at the date of the first publication of the notice under section 8.
[27] Elementarily, the price that the land might reasonably have been expected to fetch on the open market on 8 April 2000 would have been expected to reflect whatever development potential the land had. As stated by the Privy Council in Gajapatiraju v Revenue Divisional Officer, Vizagapatam [1939] AC 302, at p313:
[T]he land is not to be valued merely by reference to the use to which it is being put at the time at which its value has to be determined… but also by reference to the uses to which it is reasonably capable of being put in the future… No one can suppose in the case of land which is certain, or even likely, to be used in the immediate or reasonably near future for building purposes, but which at the valuation date is waste land or is being used for agricultural purposes, that the owner, however willing a vendor, will be content to sell the land for its value as waste or agricultural land… [T]he possibility of its being used for building purposes would have to be taken into account.
[28] The foundation of the Board’s decision was that “it is beyond dispute… that as at 8 April 2000 the subject property had a real and obvious potential for higher development”. On that basis, and “in the absence of any appropriate comparables with sufficient similar characteristics for residential development”, it adopted the residual method of valuation, namely a calculation of the net profit a developer might reasonably have expected to achieve from the residential development of the land.
[29] The Supreme Court on appeal took a very different view of the evidence. On its reading of it, “there was only a bleak and remote possibility of obtaining a re-zoning at the material time”; the evidence suggested “that the development potential was at best speculative and in any event not one which can reasonably be expected to be reached in the short-term”. In these circumstances, the Supreme Court thought that the residual method of valuation was inappropriate and substituted for it the rival approach contended for by the minister: “the direct market comparison approach together with an enhancement for a slight hope value”.
[30] That the Board’s decision could not stand so that the Supreme Court had no alternative but to allow the appeal from it is agreed by all members of this board. As Lord Scott and Lord Carswell point out in [12] of their opinion, for residential development to proceed three permits would be required: a development permit relating to zoning or planning; a morcellement permit; and a land conversion permit. Paragraphs [13] to [18] of that opinion detail the many obstacles and uncertainties that would have been faced in obtaining all these permits. Astonishingly, however, the Board took no account whatsoever of the risk that residential development might not be permitted and, indeed, valued the land on the basis that the entire development process would be completed within just two years from 8 April 2000. Lady Hale and Sir Peter Gibson make plain, in [14] of their opinion, that they too regard the Board as having been clearly in error. These errors, indeed, seem to me to have been so egregious as to deny the Board’s views the entitlement to such substantial degree of respect as is ordinarily due to an expert valuation tribunal. Lady Hale and Sir Peter Gibson suggest, in [7] of their opinion, that “the transcript of evidence shows that the Judge who chaired the Board subjected the evidence on both sides to a proper level of scrutiny”. Be that as it may, the transcript of evidence certainly belies the Board’s all-important conclusion that the land’s “real and obvious potential for higher development” was “beyond dispute” on the contrary, it was hotly disputed.
[31] The appeal from the Board’s determination therefore had to be allowed. What divides your lordships is: with what result? Was the Supreme Court entitled, as it did, simply to substitute the minister’s contended-for valuation for the Board’s unsustainable assessment or should it have remitted the case to the Board for reconsideration? Lady Hale and Sir Peter Gibson favour the latter course, and I confess that at one time I too leaned towards it. In the end, however, I have come to the contrary conclusion and now think the Supreme Court right to have disposed of the appeal as it did.
[32] Critically, of course, the question here is whether really this was a case for the residual method of valuation at all. Lady Hale and Sir Peter Gibson, in [15] of their opinion, quote from Johnson, Davies and Shapiro’s Modern Methods of Valuation of Land Houses and Buildings (9th ed) 2000, a passage at pp279-280, that I regard as going to the heart of the matter. For convenience, I repeat it, adding the final sentence that completes the paragraph in the text:
A valuation to determine hope value is often impossible other than by adopting an instinctive approach, particularly in the stages when the hope of permission is remote; it can only be a guesstimate of the money a speculator would be prepared to pay. As the hope crystallises into reasonable certainty of a permission at some stage, a valuation can be attempted based on the potential development value deferred for the anticipated period until permission will be forthcoming, but with some end deduction to reflect the lack of certainty. Indeed, since most developers will buy only when permission is certain (preferring an option to buy or a contract conditional on the grant of permission before certainty has been reached) any sale in the period of uncertainty will probably require a significant discount on what might otherwise appear to be the full hope value.
[33] In this case, the fundamental uncertainties as to whether ever, and if so when, it might be possible to acquire all three necessary permits and then successfully complete the residential development of this land (put aside the further uncertainties as to the likely costs of such a development, including any land conversion tax) were to my mind such as to rule out the residual method of valuation in this case. No doubt, it made sense to adopt this method if all these uncertainties were to be ignored (as they were by the Board). But not otherwise. In determining the hope value as at 8 April 2000, all that could sensibly be achieved was “a guesstimate of the money a speculator would be prepared to pay”. By no means had the stage been reached when “the hope crystallises into reasonable certainty of permission” when “a valuation can be attempted based on the potential development value deferred for the anticipated period until permission will be forthcoming with some end deduction to reflect the lack of certainty” (that is, the residual method of valuation). There was no such “reasonable certainty” here, still less the absolute certainty upon which the Board based its own calculations. Whether or not the Supreme Court was correct in characterising the prospect as “bleak and remote” matters little; it was certainly justified in describing it as “at best speculative” and unlikely “in the short-term.”
[34] Chapter 11 of Modern Methods of Valuation of Land Houses and Buildings, entitled “Residual Method of Valuation”, appears to me to support the view that the development prospects of this plot were altogether too speculative to justify use of the residual method of valuation. Lady Hale and Sir Peter Gibson, in [10] and [11] of their opinion, quote from chapter 11. However, the “uncertainty” produced by “a large number of variables” as discussed in that chapter is as nothing compared to the yet more fundamental uncertainties as to whether, and if so when, and at what cost (including the likelihood and extent of land |page:18| conversion tax payable) the three permits would have been obtained surrounding the possible future development of this land.
[35] Lady Hale and Sir Peter Gibson would remit the case for reconsideration in the light of their opinion. However, how should the Board factor in all these many uncertainties that initially it quite simply overlooked. Moreover, how confident could the respondent ministry be that it was now doing so with complete objectivity? Or would it be necessary to have a complete rehearing before a freshly constituted Board?
[36] Had the appellant, as a willing seller on the open market, advertised for sale this island of land, I find it difficult to suppose that any bids forthcoming from property developers would have been calculated by reference to the residual method of valuation. (I refer to it as an island of land simply to emphasise how matters stood before the Illovo deal which, of course, has to be ignored so dramatically altered the development landscape.) The most the appellants could have expected (and the final sentence from the above cited passage from Modern Methods of Valuation of Land Houses and Buildings is of some significance in this regard) would have been a bid that included a premium over the basic agricultural land value.
[37] For my part, I readily acknowledge that that premium might well have exceeded the very modest amount (18-40% of the basic agricultural value as calculated by Lord Scott and Lord Carswell in [8] of their opinion) included in the minister’s offer (although, as Lady Hale and Sir Peter Gibson point out, in [4] and [16] of their opinion, the offer had in fact first been made solely by reference to the land’s agricultural value). The fact is, however, that no alternative case was ever advanced by the appellant contending for a higher uplift on basic value: the contest was at all times simply between the residual method of valuation and agricultural land comparables with a small hope value premium. If, as I believe, the residual method of valuation is fundamentally inappropriate in a case of this sort, the valuation dispute ought now to be regarded as finally at an end and the litigation concluded.
[38] It is in these circumstances and for these reasons that I too would dismiss this appeal.
Giving their joint dissenting opinion, Baroness Hale of Richmond and Sir Peter Gibson said:
[39] It is unusual in valuation cases for either side to be completely right. The government of Mauritius compulsorily acquired a plot of land, 12 arpents and 86 perches (54,280m2) in area, in order to build a national children’s hospital and institute of neurology and cardiology. The government valuer assessed its value at R500,000 per arpent, giving a total of R6.43m for the entire plot. The owner’s valuer proposed a value for the whole plot of R74,36m. The Board decided that it was worth R39,743,588. The Supreme Court allowed the government’s appeal and substituted the government valuer’s figure of R6.43m. However, it does not follow from the fact that the decision of the Board was open to criticism that the government’s figure had necessarily to be accepted as correct. In our view, both were wrong.
[40] Section 19(3) of the 1982 Act simply provides that:
The value of any interest in the land shall be the amount which that interest if sold on the open market by a willing seller, might be expected to realise at the date of the first publication of the notice under section 8.
There are no provisions comparable to those in the UK’s Land Compensation Act 1961 relating to the assumptions that are to be made concerning the grant of planning permission for the development of the land. Nevertheless, it is common ground that the principle stated in Gajapatiraju v Revenue Divisional Officer, Vizagapatam [1939] AC 302, at p313, applies:
For it has been established by numerous authorities that the land is not to be valued merely by reference to the use to which it is being put at the time at which its value has to be determined… but also by reference to the uses to which it is reasonably capable of being put in the future… No one can suppose in the case of land which is certain, or even likely, to be used in the immediate or reasonably near future for building purposes, but which at the valuation date is waste land or is being used for agricultural purposes, that the owner, however willing a vendor, will be content to sell the land for its value as waste or agricultural land… the possibility of its being used for building purposes would have to be taken into account.
[41] There were two issues in this case. The land in question was in agricultural use as part of a sugar plantation. The first issue was whether the possibility of developing the land for “higher uses” in future should be taken into account. The second issue was as to the correct method of calculating its value. The government valuer had relied solely upon sales of agricultural land that he regarded as comparable. The Board and the owner’s valuer adopted the “residual method”, calculating what might eventually be realised if the site were developed for sale as residential building plots and then deducting the costs of that development.
[42] In his written report, dated 13 March 2002, Mr Dilmohamed, the deputy chief government valuer, did not take into account the possibility of future development at all. He stated that “the highest and best use of the property is agricultural at the time of the acquisition and will remain unchanged in the foreseeable future considering its location outside the limits of permitted development as more fully shown in the Outline Scheme” for the particular area. Hence: “The land has been assessed on the basis of its current use, that is agricultural land.” He put a comparatively high value on it, “bearing in mind the location, accessibility and existing irrigation improvements”. In his oral evidence to the Board, however, he said that his figure of R500,000 per arpent “will take into consideration [a] slight hope value of around 10-15 per cent because on a purely agricultural basis it would not exceed 350,000 to 400,000 an arpent but that 100,000 as a surplus I have granted it as a hope value potentiality in the long term”. If, contrary to his written evidence, that was what he was doing, he must have regarded the possibility as very slight indeed because there was evidence that residential plots in the vicinity were selling at a rate of approximately R11.1m per arpent.
[43] The Board was of course aware that, in the “Illova deal” in 2001, an area of land around the site had been re-zoned and the appellant permitted to sell it for residential building purposes exempt from land conversion tax. It was careful to remind itself that it should not take account of evidence that was not available on the valuation date. Nevertheless, it concluded:
Yet, it is beyond dispute, independently of any evidence which came to light afterwards, that as at 8 April 2000 the subject property had a real and obvious potential for higher development although it was currently in an agricultural zone. This is mainly due to its location. It is located at about 200 metres from the Telfair Housing Estate and quite proximate to the Motorway at Reduit and near substantial institutional development like the University of Mauritius and the Mahatma Gandhi Institute.
[44] The Supreme Court disagreed. Taking into account the evidence “that there was only a bleak and remote possibility of obtaining a rezoning at the material time”, the need to apply for land conversion and to pay land conversion tax, the costs of providing services and infrastructure and the fact that all the major institutional development was on the other side of the main road, they concluded that “these factors, when looked at objectively, tend to suggest that the development potential was at best speculative and in any event not one which can reasonably be expected to be reached in the short term”.
[45] In our view, the Supreme Court should not have overturned the finding of the Board on this issue. The question was whether the land had a reasonable possibility of development that a willing buyer and a willing seller would take into account when negotiating a purchase price. There was evidence each way on the prospects of development and members of the Board were also entitled to take its own expert opinions into account. The location of the land was very close to a junction between the major trunk road, described as a motorway, going north to south on the island, and a main road to the east. It was easy to get to from all over the island. Hence, there had already been major institutional development close to the road junction, which had quite recently been improved. That development brought with it increased demand for housing. There had already been some residential development on this side of the main road. The site was close to this development and to the main road. Zoning and other obstacles were not |page:19| insurmountable. All of this was apparent to the Board from the evidence of the witnesses and of its own eyes. As an expert valuation tribunal, it was better placed to make the necessary judgments and predictions than any other party. The transcript of evidence shows that the judge who chaired the Board subjected the evidence on both sides to a proper level of scrutiny.
[46] The real issue, in our view, is how the land, as agricultural land with a real possibility of development for residential use in the foreseeable if not immediate future, should have been valued. The government valuer adopted the “direct capital comparison” approach. He looked at recent sales of plots of agricultural land from all over the island, some of them quite close to the subject land. The problem with this approach, as the Board pointed out, was that none of the plots chosen was directly comparable. Those that were in the same area of the island were not close to the main transport hub, indeed not close to the roads at all, or to another built-up area. As Johnson, Davies and Shapiro point out in Modern Methods of Valuation of Land Houses and Buildings (9th ed) 2000, at p14: “Property can never be absolutely identical, so that the use of this method is limited to the simplest cases.”
[47] The appellant’s valuer, on the other hand, had adopted the residual approach. He had calculated what the land would realise if parcelled out into building plots with appropriate roads and services, deducted the costs of doing this, originally arriving at the sum of R74.36m but later revising this to R66.5m. The government valuer, although not accepting that the approach was valid, had also done a residual calculation, with a view to demonstrating that residential development of this agricultural land was not feasible. After deducting morcellement tax and capital gains tax from the gross profits, he arrived at the sum of R28,140,915 or R2,188,250 per arpent. He reduced that figure by 50% to R1,094,125 for the risk that permission for the residential development would not be obtained. He then referred to the purchaser’s liability for land conversion tax at the rate of more than R1.4m per arpent.
[48] The variables between the two valuers’ calculations of the gross profits included the realisable price of the plots, the amount of the land to be devoted to infrastructure and landscaping, the costs of providing the various items of infrastructure and, in particular, the cost of providing an access road, and the delay in realisation. As Johnson, Davies and Shapiro, at p176, comment:
Given a calculation based on a large number of variables, the actual range of answers which can be produced is wide. This uncertainty is the method’s weakness but it is one which is acceptable so long as the estimates are prepared with as much information as is available to narrow possible errors.
[49] Johnson, Davies and Shapiro also comment that, in the UK, the residual method “is disliked by the Lands Tribunal in compensation cases… because it is not tested by ‘haggling in the market’. In the open market, however, the residual method will continue to be the main cornerstone of many opinions of value, particularly those involving land for development or redevelopment.” The UK compensation scheme is, of course, more complex than that in Mauritius, not least because it involves statutory assumptions concerning the grant of planning permission.
[50] The Board concluded that, “in the absence of any appropriate comparables with sufficient similar characteristics for residential development”, the direct comparison method should be ruled out and the residual method adopted. It then went through the various variables and, in general, adopted a middle course somewhere between those suggested by the appellant and those suggested by the government. The figure at which it arrived R39,743,588 was closer to that proposed by the government valuer in his residual value calculation than to that proposed by the appellant’s valuer.
[51] The Supreme Court concluded that because, on its view, the land fell to be valued as agricultural land only, the Board had been wrong to reject the direct comparison method. Furthermore, even if the residual method could be adopted, the Board had erred in not taking into account the land conversion tax payable under the 1988 Act, exemption from which was not automatic. It had also erred in not taking into account the risk of failing to obtain the necessary planning permits and the length of time that all this might take.
[52] In our view, the Board was clearly in error in failing to take into account the possibility that the necessary permits might not be obtained, the various possible permutations under which land conversion tax might or might not become payable and the length of time that all this might take. Its calculations appear to have been made upon the basis that the development would definitely be permitted and the benefit realised within two years. It took into account capital gains tax but not land conversion tax. The incidence of the latter was mentioned but not fully explored before them. On Mr Dilmohamed’s approach to the calculation, the incidence of the tax would render virtually all residential development of agricultural land unviable unless the scheme could be exempted. In our view, he had overstated its effect. It would be a deduction from gross profits in the same way as infrastructure costs and would thus reduce the profits to which capital gains tax applied. There are also circumstances in which such development can be exempted. That much at least it is permissible to conclude from the “Illova deal” in 2001. However, if nothing else the possible incidence of the tax is one of the uncertainties that must be factored into any residual method calculation.
[53] At the end of the day, where there is a reasonable prospect of development in the future, some method has to be found of assessing the “hope value” in the property. As Johnson, Davies and Shapiro, at pp279-280, candidly admit:
A valuation to determine hope value is often impossible other than by adopting an instinctive approach, particularly in the stages when the hope of permission is remote; it can only be a guesstimate of the money a speculator would be prepared to pay. As the hope crystallises into reasonable certainty of a permission at some stage, a valuation can be attempted based on the potential development value deferred for the anticipated period until permission will be forthcoming, but with some end deduction to reflect the lack of certainty.
In other words, there will be a sliding scale from a “comparables plus” approach to a “residual value minus” approach. A hypothetical developer, purchasing land for its “bank”, would be bound to do some calculation of how much it might eventually make from the development, as well as the risk that it might not be permitted to do it. It would do this even if there were truly comparable sales, although it would also look at these to make sure that it was not proposing to pay too much. That is no doubt why Johnson, Davies and Shapiro comment that the residual method is the main cornerstone for many opinions of value in the open market. It is certainly more scientific than a so-called “spot” valuation, which is not a term of art and was used in quite a different legal and factual context in Perkins v Middlesex County Council (1952) 2 P&CR 42.
[54] In our view, therefore, both the Supreme Court and the Board fell into error. The Supreme Court erred in leaving out of account altogether the undoubted development potential of this land and thus adopting an approach that had been premised on purely agricultural comparables. It cannot have taken the evidence of the government valuer as indicating that he had made a serious attempt to assess the hope value of the land. He had put the same value on it as purely agricultural land in his written report. The modest increase to which he referred in his oral evidence bore no relationship to the enormous disparity between the price of residential plots and the price of agricultural land. On the other hand, the Board also fell into error in adopting the residual approach without discounting for the risk that the development might never happen, or might not happen soon, and at least considering the possible incidence of land conversion tax. The truth, as always, must lie somewhere between the two.
[55] We would have allowed the appeal and remitted the case to the Board for reconsideration in the light of this opinion.
Appeal dismissed.
