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 QC Deputy Chamber President and
Mrs Diane Martin MRICS FAAV
2021/032

		Royal Courts of Justice Strand, London, WC2A 2LL
		19/02/2021

Covid-19 Protocol: This judgment was handed down remotely by circulation to the parties' representatives by email, release to BAILII and publication on the Courts and Tribunals Judiciary website. The date and time for hand-down is deemed to be 10am on Friday 19th February 2021.

Lord Justice Lewison:

Introduction

1. The first issue on this appeal is whether, for the purposes of assessing compensation on compulsory acquisition, land at East Lane Farm, Morpeth, belonging to Leech Homes Ltd is to be treated for planning purposes as land to which green belt policies apply. The Upper Tribunal (Martin Rodger QC, Deputy President and Mrs Diane Martin MRICS, FAAV) ("the UT") held that it was. Their decision on that issue is at [2020] UKUT 150 (LC), [2021] RVR 21. The second issue is whether the UT had power to order the claimant to pay the acquiring authority's costs. The UT decided that it did. Their decision on that issue is at [2020] UKUT 328 (LC), [2021] RVR 34.
2. Leech Homes' land was compulsorily acquired for the purpose of constructing the Morpeth Northern By-pass. The first issue arose in the context of an application by Leech Homes for a certificate of appropriate alternative development (a "CAAD") in connection with the assessment of compensation. On such an application the local planning authority must decide whether on a hypothetical application for planning permission (made on certain assumptions) planning permission would have been granted for appropriate alternative development; and if so, for what kind of development. In the present case Leech Homes applied for a CAAD in respect of a residential development consisting of about 135 dwellings, together with associated infrastructure and landscaping. The Northumberland County Council ("the Council") decided that no planning permission would have been granted for development. Leech Homes unsuccessfully challenged that decision in the UT.
3. The crux of the decision was that the land in question was land to which green belt policies applied. It is now common ground that if the Council and the UT were correct in that conclusion, then its decision on the application for a CAAD was correct; but that if they were wrong in that conclusion, then planning permission would have been granted for residential development because the Council could not demonstrate a 5-year supply of housing.

The CAAD Appeal

4. An appeal against a CAAD lies to the UT under section 18 of the Land Compensation Act 1961. Section 18 (2) provides that on any appeal the UT must consider the matters to which the CAAD relates as if the application for a CAAD had been made to the UT in the first place.
5. The UT set out the legal framework laid down by the Land Compensation Act 1961 at [10] to [19] of its decision. It is common ground that its summary is accurate. For present purposes all that needs to be said is that the question whether planning permission would have been granted is determined in a "no scheme" world as at the "launch date" (9 August 2012).
6. It is also common ground that in determining the hypothetical applications for planning permission, the local planning authority (or on appeal the UT) must make its decision in accordance with the development plan unless material considerations indicate otherwise: Planning and Compulsory Purchase Act 2004 s. 38 (6). This reflects long-standing government policy that planning should be genuinely plan led. The development plan includes the regional strategy (if there is one): s. 38 (3) (a). But the decision-maker must correctly understand policies contained in the development plan. The interpretation of such policies is a question of law rather than one of planning judgment: Tesco Stores Ltd v Dundee City Council [2012] UKSC 13, [2012] PTSR 983. Material considerations include national planning policy.
7. The contents of the development plan were also common ground. The UT set them out at [26] to [30] in terms which I reproduce:

"[26] There is no disagreement in this appeal over the content of the development plan. At the valuation date it comprised the saved policies of the Castle Morpeth District Local Plan (2003) and a single saved policy from the Northumberland County and National Park Joint Structure Plan (2005).

[27] The Local Plan was originally intended to last from 1991 to 2006 but the adoption of a new plan was delayed by local government reorganisation. The Local Plan policies agreed to be of particular relevance to the Tribunal's determination were C1, which concerned settlement boundaries, C17 relating to the Green Belt, H16 on housing in the countryside, and MC1 which dealt specifically with the Morpeth settlement boundary.

[28] The Joint Structure Plan establishes the strategic framework for the preparation of local plans. The plan was largely replaced in 2008 and by the valuation date the only surviving component was Policy S5 which proposed an extension to the green belt. The interpretation of Policy S5 is critical to the issues in this appeal. It provided as follows:

"Policy S5 - Extension to the Green Belt

An extension to the Green Belt will extend from the existing boundary northwards to lie:

• [a list of 18 settlements and other locations covering a wide area incorporating Morpeth]

Precise boundaries, including those around settlements, should be defined in Local Plans having particular regard to the maintenance of the role of Morpeth as defined in Policy S7 and to the sequential approach in Policy S11."

Policies S7 and S11, which are referred to in Policy S5, concerned the settlement boundary and the significance of Morpeth as a main town and focus for development. They were not saved when the Structure Plan expired but it was not suggested that their lapsing had any effect on the application of Policy S5 itself.

[29] Policy S5 was accompanied by two indicative diagrams. No reference to these appears in the text itself and Mr Pereira did not dissent from Mr Cairnes' suggestion that they are not to be regarded as part of the statutory development plan. The first diagram showed the County as a whole and indicated the approximate location of the existing adopted green belt boundary, the general extent of the proposed green belt extension, and other strategic designations. The second diagram was an inset of the area around Morpeth which showed the town surrounded by the proposed extension to the green belt. The diagrams did not purport to identify either the inner or the outer boundaries of the green belt extension with any precision, nor do they, or Policy S5 itself, say that all of the land shown within the general extent is Green Belt.

[30] Other material policy considerations are agreed to include the emerging Northumberland Local Plan Core Strategy which was published in draft in December 2014, and the consultation draft Morpeth Neighbourhood Plan published in January 2015. These were at an early stage of their progression through the plan making process and it is agreed that they should be given limited weight for that reason. The published evidence associated with these emerging plans is also agreed to be relevant. This material included a consultation on the methodology to be employed in reviewing the green belt boundaries and the Council's own Green Belt Settlements Assessment published in December 2014."

8. Policy S5 was accompanied by the two indicative diagrams to which the UT referred; but it is common ground that the diagrams are not part of the development plan itself. One of the diagrams showed Morpeth surrounded by a notional circle within which the green belt would be extended. But it did not purport to define either the inner or the outer boundaries of the ring.
9. As Mr Pereira QC correctly pointed out, land within the green belt (unlike, say, land within a conservation area) does not have separate legal status. It is, as he put it, a creature of policy. National green belt policy is contained in the National Planning Policy Framework ("the NPPF"). That describes the purposes of the green belt and the policies to be applied to it. In the version current at the valuation date it stated:

"79. The Government attaches great importance to Green Belts. The fundamental aim of Green Belt policy is to prevent urban sprawl by keeping land permanently open; the essential characteristics of Green Belts are their openness and their permanence.

80. Green Belt serves five purposes:

• to check the unrestricted sprawl of large built-up areas;

• to prevent neighbouring towns merging into one another;

• to assist in safeguarding the countryside from encroachment;

• to preserve the setting and special character of historic towns; and

• to assist in urban regeneration, by encouraging the recycling of derelict and other urban land. …

82. The general extent of Green Belts across the country is already established. New Green Belts should only be established in exceptional circumstances, for example when planning for larger scale development such as new settlements or major urban extensions. …

83. Local planning authorities with Green Belts in their area should establish Green Belt boundaries in their Local Plans which set the framework for Green Belt and settlement policy. Once established, Green Belt boundaries should only be altered in exceptional circumstances, through the preparation or review of the Local Plan. …

87. As with previous Green Belt policy, inappropriate development is, by definition, harmful to the Green Belt and should not be approved except in very special circumstances.

88. When considering any planning application, local planning authorities should ensure that substantial weight is given to any harm to the Green Belt. 'Very special circumstances' will not exist unless the potential harm to the Green Belt by reason of inappropriateness, and any other harm, is clearly outweighed by other considerations."

10. A shortage of housing is unlikely to outweigh harm to the green belt and is, therefore, unlikely to amount to "very special circumstances". In addition to land within the green belt, paragraph 85 of the NPPF discussed a category of land described as "safeguarded land". That is land between the urban area and the green belt (i.e. it is not within the green belt itself), but it may be required to meet longer term development needs beyond the plan period. The same paragraph makes clear, however, that safeguarded land is not allocated for development; and that planning permission for the development of safeguarded land should only be granted following a local plan review which proposes the development.
11. The problem arises because Policy S5 is contained in a high-level strategic plan which leaves the precise boundaries of the proposed extension to the green belt undefined. It is not possible to say with certainty, one way or the other, whether the land in question will or will not form part of the green belt once the detailed boundaries of the green belt have been fixed by a local plan.
12. The starting point for decision making must be policy S5, because that is part of the development plan. As the NPPF states at paragraph 82, the general extent of green belt across the country is already established. That statement leads to the conclusion that policy S5 establishes green belt, even though the detailed boundaries have not yet been fixed. It is also implicit in paragraph 83 that local planning authorities fixing detailed boundaries already have green belt in their area. That must be the consequence of the regional strategy. Policy S5 was therefore in operation and had the effect of generally extending the green belt. This appears to have been common ground in the UT (see paragraph [82] of the decision) and I did not understand Mr Cairnes QC to challenge that in this appeal.
13. The fixing of the detailed boundaries of the green belt is left by the NPPF to local planning authorities. They are to fix the boundaries of the green belt in local plans: NPPF paragraph 83. It is important to stress at this stage that a local planning authority may have many functions in connection with town and country planning. One of those functions is plan-making. Development control is a different and separate function. In exercising its development control function, a local planning authority is not plan-making.
14. The making of a plan is a different process from exercising powers of development control. It involves (at least) extensive consultation and independent examination before it is adopted. It is area wide rather than site specific. As paragraph 155 of the NPPF put it:

"Early and meaningful engagement and collaboration with neighbourhoods, local organisations and businesses is essential. A wide section of the community should be proactively engaged, so that Local Plans, as far as possible, reflect a collective vision and a set of agreed priorities for the sustainable development of the area, including those contained in any neighbourhood plans that have been made."

15. In view of the fact that plan-making and development control are different planning functions, a decision on an application for planning permission for a specific site should not, in principle, usurp the plan-making function. In relation to the applicability of green belt policies to land which is within the general extent of the green belt, an individual decision on an application for planning permission should not, as the UT put it at [85] "pre-empt the plan making process."
16. What, then, is the decision maker to do in the light of a general strategic policy such as Policy S5? Stuart-Smith J considered this question in Wedgewood v City of York Council [2020] EWHC 780 (Admin). (The paragraph numbers in that judgment differ between the print with which we were provided and the version published on Westlaw. I have used the latter.) That case concerned a challenge to the grant of planning permission for two extensions and additional parking to an existing rehabilitation centre. The challenge was based on the assertion that the site should have been treated as if it had been in the green belt. The judge noted that the boundaries of the green belt around the city of York had not been fixed. The regional spatial strategy ("the RSS") contained a policy which stated that the detailed inner boundaries of the green belt around York "should be defined" and a further policy that those boundaries should be defined by the City of York local development framework. An accompanying diagram showed the "general extent" of the green belt. As the judge said at [18]:

"The policies do not state or imply that every piece of land within the doughnut ring that is bounded by the inner and outer boundaries shall be Green Belt; nor do they say anything about whether all or some pieces of land within the doughnut ring shall not be Green Belt. No doubt, this lack of detail and precision is attributable to the fact that the RSS was and is a high-level strategic document. It leaves matters of practical detail to lower level plans and policies."

17. Planning policy in that case had, however, moved on since the publication of the regional spatial strategy. By the time of the challenged decision (in 2019) there was a draft local plan for York, published in 2005. It showed the site as being within the urban area, and outside the green belt. The accompanying proposals map showed the site as being outside the green belt. A later emerging plan also showed the site as being outside the green belt. The draft local plan proposals map accompanying that emerging plan (published in 2018) also showed the site as being outside the green belt. Consultation on that emerging plan elicited no suggestion that the site should be included in the green belt. Moreover, the draft local plan had been adopted as the basis for decision-making. It also showed the site as not being within the green belt. The judge described the formal position as follows at [21]:

"1) As a matter of planning principle, there is a Green Belt area around York.

2) The detailed inner boundaries and outer boundaries have not been defined by any formally adopted development plan.

3) Policy Y1C states that the detailed boundaries of the outstanding sections of the outer boundary shall be 'about six miles from York City centre,' and that the detailed boundaries of the inner boundary shall be defined in line with Policy YH9C.

4) There is no formally adopted Development Plan that identifies the site as being within the Green Belt, as opposed to being within the general extent of the Green Belt. The most that can be said is that the site falls within the area illustrated by the RSS key diagram as being, 'general extent of Green Belt (Policy YH9).'

5) In accordance with NPPF paragraph 48, to which I have referred, the draft local plan can be afforded weight according to the stage of preparation of the emerging plan, the extent to which there are unresolved objections to it, and the degree of consistency of the relevant policies in the emerging plan to the policies of the NPPF. The draft local plan has been submitted to the inspector, who has raised certain issues. It is now progressing towards phase two of the inspector's involvement as it approaches adoption.

6) The 2005 Draft Local Plan has been adopted only to be used as a basis for decision making. It shows the site as not being in the Green Belt."

18. Having set out the formal position, the judge described his preferred approach as follows:

"[32] This appeal seems to me to raise a question that is novel and difficult for the court, though it is not novel for the city of York. At bottom lies the question whether the adoption of a high-level strategic plan such as the RSS is, of itself, sufficient to constitute and define what is Green Belt land. If one adopts a binary approach, each alternative is unpalatable. If it is held that more is required in order to create the Green Belt than the RSS, then York has no Green Belt land unless and until a further plan, probably a Local Development Plan, defines the detail of its scope. On the other hand, if it is held that a high-level strategic plan such as the RSS converts everything to which it refers into Green Belt, the restrictions which that would impose on developing land that has none of the characteristics normally associated with Green Belt land would be unsatisfactory from a number of different perspectives.

[33] In my judgement, the solution to this binary conundrum is to adopt a more nuanced approach, as suggested by the defendant. It must be acknowledged that the RSS, as a high-level strategic document, establishes that, in principle and as a matter of policy, there is a Green Belt within the doughnut ring. That policy must be implemented by the defendant, but the policy does not state that all land that is (as a matter of high-level policy) within the inner and outer boundaries is Green Belt land.

[34] In the absence of a defining Local Development Plan that specifies what is and is not Green Belt, the defendant must apply the high-level policy rationally in order to determine what land within the doughnut ring is and is not to be treated as Green Belt land. In doing so, it may have regard to the 2005 draft local plan incorporating the full set of changes, as it has previously taken a policy step by resolving to take it into account for development and management purposes. It may take into account the emerging Local Plan, provided it has due regard to the guidance at paragraph 48 of the NPPF. Furthermore, it may and should take into account site specific features that may tend to treating the site as Green Belt or not."

19. On the particular facts of that case Stuart-Smith J held that the local planning authority had not acted unlawfully in declining to apply green belt policies to the particular parcel of land under consideration. It is to be particularly noted that at the time of the challenged decision the site had been treated as being outside the green belt in every draft plan for the best part of 15 years. But I note that he did not say that if the local planning authority had applied green belt policies that would have been unlawful.
20. Mr Cairnes placed particular reliance on Stuart-Smith J's observation at [5] that whether the site should have been treated as being within the green belt was "a planning judgment within the exclusive province of the local planning authority". That is no doubt correct, as far as it goes, but that planning judgment must be exercised upon a proper understanding of the effect of the policies in the development plan. I do not consider that the judge meant to say that the planning judgment was at large.
21. The principal bone of contention between the parties is whether the UT adopted the approach set out in Wedgewood. Before going any further, however, I would point out some differences between the issue facing Stuart-Smith J and the issue facing the UT. First, the issue in Wedgewood was whether the local planning authority had acted lawfully in reaching their decision to grant planning permission. That question was to be decided on ordinary public law principles. The issue before the UT was not whether the Council had acted lawfully in reaching its decision. Under section 18 (2) (a) of the 1961 Act the UT was required to consider the question as if the application for the CAAD had been made to it in the first place. In other words, it was exercising an original jurisdiction rather than a review jurisdiction. Second, in Wedgewood, the issue was to be determined by reference to what the local planning authority itself actually did. By contrast, the issue before the UT was, as they correctly said at [22]:

"… when considering under s.14(4)(b) whether planning permission for the appellant's scheme could reasonably have been expected to be granted at the valuation date, or later, the Tribunal is not required to ask itself how Northumberland County Council is likely to have determined the notional application for consent. The Tribunal must put itself in the position of a reasonable decision maker, properly applying the law. It follows that, if at the statutory valuation date the County Council's officers and members had a particular understanding of the meaning of a relevant planning policy, the Tribunal is not required to adopt that understanding or to interpret the policy in the same way, but must decide for itself what the policy means, and apply it correctly."

22. Third, whereas Wedgewood was concerned with decision making in the real world, the UT had to make their decision in the "no scheme" world.
23. In this case, the UT described the evolution of planning policy since the promulgation of Policy S5. It is important to stress that those emerging policies were all policies applicable in the real world, rather than in the "no scheme" world that the UT were required to assume. The UT referred to a number of planning documents that were published in 2014. They said:

"[45] … In [December 2014] the Council published its Green Belt Settlement Assessment. The Assessment recognised that the inner boundaries of the green belt which were to be identified for main towns and service centres (of which Morpeth was one) needed to provide sufficient capacity to accommodate housing and economic development needs, both for the emerging plan period and subsequently.

[46] The appellant's land was one of the sites assessed by the Council against four of the five purposes of the green belt identified in paragraph 80, NPPF (see paragraph 28 above). The site was considered overall to make a medium contribution to green belt purposes, with the assessor noting its importance in safeguarding the countryside from encroachment and in achieving a satisfactory transition between urban and rural landscapes. The context in which those judgments were made assumed the presence of the forthcoming bypass.

[47] The draft Northumberland Local Plan Core Strategy, also published in December 2014, was intended to cover the period to 2031. It built on the Green Belt Settlement Assessment and indicated that the preferred option for Morpeth was to allow most development to take place to the north of the town, including the previously developed sites at the St George's and Northgate Hospitals. Once again, this preference was expressed in the context of the proposed new bypass. The preferred inner green belt boundary was shown on a plan which included the appellant's land in an area straddling the bypass and extending west to the A1. The whole of this area was shown as lying within the inner boundary (i.e. outside the green belt extension) and as safeguarded for future development after the period covered by the draft Core Strategy document. The proposed green belt boundary was to the north of the appellant's land, for the most part beyond Northgate Hospital, and to the west, beyond the line of the A1. The plan showed the line of the proposed new bypass crossing the appellant's land and then forming the boundary of the green belt along part of its northern edge, before continuing southeast through land intended to be designated as part of the green belt.

[48] The third of the December 2014 documents was the Council's Strategic Housing Land Availability Assessment, which considered potential development sites around Morpeth over a time scale of 0-15 years. The appellant's land, and other land adjoining it between the A1 and the settlement boundary, was designated as "uncertain", which we take to reflect the proposal in the draft Core Strategy that the same area should be safeguarded for development after 2031."

24. The UT approached the question by first setting out some of the legal principles that they proposed to apply. At [82] they recorded that it was common ground that Policy S5 was already operative; and that it already had the effect of generally extending the green belt. At [83] they observed that it was agreed that a site was either currently within the green belt or outside it; and that there was no "indeterminate category where the application of green belt is a matter of discretion". They regarded the effect of Policy S5 as a question of law: see [88].
25. Dealing with the two points, the UT said at [84]:

"These are both important points. The decision maker is required by s.38(6) of the Planning and Compulsory Purchase Act 2004, to determine a planning application in accordance with the statutory development plan unless material considerations indicate otherwise. If the effect of Policy S5 is that land outside the settlement boundaries of Morpeth is green belt, the decision maker is not free to disapply green belt policies, and must give effect to the presumption against development unless very special circumstances outweigh the harm to the green belt which would result."

26. The key parts of the UT's decision on this point are in paragraphs [85] and [86]:

"[85] Policy S5 distinguishes between green belt and settlements, and provides that precise boundaries between them is to be defined in local plans. Until those boundaries are fixed, it is not possible to know with any assurance whether a particular site is within the green belt or not. To proceed, in that state of uncertainty, on the basis that green belt policies do not apply, would in a sense be to pre-empt the plan making process. That is the justification for the "precautionary approach" taken by the Secretary of State in the Germany Beck decision and the other York appeals.

[86] The precautionary approach is a response to this state of uncertainty. It amounts to a presumption against granting consent unless consent would be granted if the site was known to be within the green belt."

27. Nevertheless, although the UT referred to a "presumption" in paragraph [86], elsewhere in the decision they referred to an "assumption". To call it a "presumption" may be putting it too high. But even if there was a presumption, it is clear that the UT considered that the presumption was rebuttable. Thus at [90] they said:

"Once a site is included within the description of the general extent of the green belt in the relevant policy, the mere fact that the precise boundaries of the extension have not yet been defined is not a reason for treating the land as if green belt policies do not apply to it. The only safe assumption, in the absence of some good reason for concluding that the site is not within the green belt, is that it is green belt land." (Emphasis added)

28. They returned to the point at [91]:

"There are aspects of judgment in determining whether, despite being within the general extent of the extension, there is a sufficient reason to conclude that a particular site is not green belt land but, in general, the proper application of s.38(6) of the 2004 Act requires it to be assumed that green belt policies apply unless and until precise boundaries are defined through the local plan process."

29. Contrary to Mr Cairnes QC's submission, this is not a "blanket presumption" that all land within the ambit of Policy S5 is green belt land. Rather it entails a two stage process. Since the land falls within the ambit of Policy S5, it is capable of being green belt land. To disapply green belt planning policies at this stage would be to remove options for designation as green belt that the local planning authority currently has. From that starting point, the second stage is to decide whether there is "sufficient reason" for concluding that when the boundaries of the green belt are finally determined by a local development plan (in the "no scheme" world) the particular land in issue would be excluded from the green belt. In this connection I do not think that there is any material difference between a "good" reason (paragraph [90]) and a "sufficient" reason (paragraph [91]). In other words the decision-maker must have a degree of confidence that, ultimately, the land in issue would fall outside the boundaries of the green belt, once they have been definitively fixed. In Wedgewood it was not surprising on the facts that the local planning authority had that confidence.
30. Nor is it the case, again contrary to Mr Cairnes' submission, that the UT rejected any appeal to planning judgment. They rejected his submission that the applicability of green policy was "simply" a matter of planning judgment. In my judgment they were correct to do so, because they recognised that Policy S5 was in operation and effective to extend the green belt. But they said explicitly at [91] that the exercise of planning judgment is required in order to decide whether there is sufficient reason for excluding the land from the green belt; and hence from the application of green belt policies.
31. Having posed the question in that way, the UT then went on to answer it. They said at [92]:

"Planning judgment is required in the assessment of a particular site against the five purposes of the green belt identified in NPPF paragraph 81. But, in agreement with Mr Bolton who said that he had considered this issue "for the sake of robustness", we prefer to regard that analysis as serving a subordinate role. It will be relevant in cases where there is said to be a good reason for disapplying green belt policies, despite a site being within the general extent of the extension to the green belt. We have therefore asked ourselves whether there is any reason not to apply green belt policy for the time being, and whether the site contributed to any of the purposes of the green belt."

32. Mr Cairnes criticised the UT's use of the adjective "subordinate" in describing the function of planning judgment. But that, to my mind, is not a fair criticism. What the UT clearly meant was that the exercise of planning judgment was relevant in considering whether there was "any reason" for not applying green belt policies. That was part of the evaluative judgment required at stage two in the UT's analysis. It was subordinate only in the sense that whether land was or was not green belt land was not an open question, but had to take Policy S5 as its starting point.
33. The UT then proceeded to evaluate the site against the five purposes of the green belt laid down by the NPPF. They concluded at [94]:

"Approaching the status of the appeal site as a matter of planning judgment and assuming only a medium contribution to the achievement of green belt purposes, we can nevertheless see no reason not to apply green belt policies to the appellant's land. Nothing in the evidence or submissions identified any reason not to do so." (Emphasis added)

34. They also said that in the "no scheme" world they were required to consider the land "might have attracted a higher rating" in terms of its fulfilment of green belt policies.
35. Clearly, then, their ultimate conclusion was one that they reached as a matter of planning judgment. It is worthy of note that Leech Homes' own planning witness agreed that if the UT were satisfied that the land was "in the general extent of the green belt by reason of Policy S5" then "green belt policies should apply to it".
36. Having reached that conclusion, they decided that the Council had been correct in its determination of the application for a CAAD.
37. In my judgment, the method that the UT adopted is a sound way of dealing with land within the general areas of a proposed green belt whose boundaries are not yet definitively fixed. It may be better not to describe it as a "precautionary principle" because that principle has a particular meaning in EU law. In short, the decision-maker must start with the policy which shows that the site is capable of being included within the green belt, and then consider whether there is sufficient reason not to apply green belt policies to it.
38. I do not detect any clear difference between the approach of the UT in this case and that of Stuart-Smith J in Wedgewood. In Wedgewood the planning officer (whose advice the planning authority followed) concluded by reference to the planning history of green belt policies and the particular nature of the site that there were good reasons for not applying green belt policies. One important consideration in that case (which is not present in this case) was the draft local plan and the emerging local plan which excluded the site in question from the green belt. As Stuart-Smith J said at [38]:

"Although short and concise, it appears to me to be a correct approach and to involve the exercise of planning judgement on the basis of the RSS, the emerging local plan and the 2005 proposals, following by site-specific consideration. That was a correct approach in principle, and it has not been shown to be irrational in practice or application."

39. It is to be noted that in that paragraph in Wedgewood Stuart-Smith J commended an approach which took the regional strategic policy as its starting point, supplemented that by reference to the emerging local plan, both of which were then followed by a site-specific assessment. That, to my mind, is exactly the approach that the UT took; and, as in Wedgewood, it has not been shown to be irrational.
40. Mr Cairnes' second ground of appeal is that the UT was wrong to say that consistency in decision-making had no part to play in its determination; and that consequently it failed to take into account highly material considerations. Those material considerations included the Council's own assessment of the site when applying for the development consent order that authorised the by-pass scheme; and the decisions reached in relation to other sites within the general extension of the green belt.
41. The UT held that Mr Cairnes' reliance on previous decisions was misplaced, because those decisions did not correctly determine the legal effect of Policy S5. At [86] they held that their own conclusion on the legal effect of Policy S5 was "the antithesis" of the approach taken by the Council and the Secretary of State in previous decisions. As they said at [88]: "The law is not different in York and Morpeth". That reinforces the point that the UT were dealing with the legal effect of a policy like Policy S5. They returned to this point at [90] in which they said:

"The approach taken in the Morpeth cases appears to give no weight at all to Policy S5 (even in cases where it was recognised as being applicable, such as Stobhill and Fairmoor) which does not seem to us to be a defensible position."

42. I do not consider that it can be plausibly asserted that a decision based upon a legal premise which has been held to be erroneous is a material consideration in the context of planning decision-making. As the UT put it at [81] a decision maker is not required or permitted to repeat the mistakes of his predecessor.
43. We were taken to a number of the decisions on which Mr Cairnes relied. In most of them there was either no or scant mention of Policy S5; and none of them considered in any detail what the effect of that policy might be. In addition, those decisions were taken in the real world, rather than the "no scheme" world which the UT were required to consider. Moreover, his reliance on these decisions seemed to me to be little more than an attack on the UT's planning judgment, which is not permissible in the absence of a public law error.
44. If, as I consider, the UT adopted the correct legal approach to Policy S5, they were entitled to disregard previous decisions which had applied the wrong legal approach.
45. I would dismiss this appeal.

The Costs Appeal

46. The second appeal concerns costs. Put shortly, the UT decided that Leech Homes, having lost its appeal, should pay the Council's costs of the appeal on the standard basis. It did so in purported exercised of the power contained in rule 10 of the Tribunal Procedure (Upper Tribunal) (Lands Chamber) Rules 2010 ("the Procedure Rules"). That rule permits the UT to make an order for costs in proceedings "for compensation for compulsory purchase".
47. The argument that the UT accepted is that a challenge to a CAAD is part of the statutory code for the assessment of compensation on compulsory acquisition; and that it is artificial to try to divide up that code into discrete parts. The sort of issues that arise on an appeal against a CAAD may arise in the context of any dispute about compensation. The kinds of development that would be permitted in a "no scheme" world are always relevant to value, whether the landowner has applied for a CAAD or not. The UT's jurisdiction to award costs cannot depend upon such procedural niceties.
48. The UT's general power to award costs is contained in section 29 of the Tribunals Courts and Enforcement Act 2007, But section 29 (3) provides that the power to award costs is "subject to tribunal procedure rules". Rule 10 of the Procedure Rules provides (so far as material):

"(1) The Tribunal may make an order for costs on an application or on its own initiative.

(2) Any order under paragraph (1)—

(a) may only be made in accordance with the conditions or in the circumstances referred to in paragraphs (3) to (6);

(b) must, in a case to which section 4 of the 1961 Act applies, be in accordance with the provisions of that section.

…

(6) The Tribunal may make an order for costs in proceedings—

(a) for compensation for compulsory purchase;…

(8) In proceedings to which paragraph (6) applies, the Tribunal must have regard to the size and nature of the matters in dispute."

49. Mr Cairnes submits that the UT had no power to award costs, because the appeal before it did not amount to proceedings "for" compensation for compulsory purchase. His essential point is that an appeal against a CAAD is a separate, and often preliminary, step towards the ultimate goal of the assessment and payment of compensation. But it is nevertheless a separate step. It does not actually result in the payment of compensation. It is the subject of its own statutory code, and is designed to replicate the outcome of an application for planning permission in the context of which costs are not usually awarded.
50. Disputes over the amount of compensation payable are assigned to the UT under section 1 of the Land Compensation Act 1961. Section 4 of that Act allows parties to make sealed offers which, if beaten, will have a decisive effect on an award of costs. In its original form, section 4 was quite generally phrased. But as the result of amendments made by the Transfer of Functions (Lands Tribunal and Miscellaneous Amendments) Order 2009, it is now common ground that section 4 (A1) limits the applicability of section 4 to proceedings on a question referred to the UT "under section 1." Section 4, therefore, does not apply to an appeal under section 18 (which is in a different part of the 1961 Act altogether).
51. Until changes introduced by the Localism Act 2011 (with effect from 6 April 2012) appeals against CAADs were assigned to the Minister. The Minister would appoint a planning inspector to conduct an inquiry. Although the Minister (now the Secretary of State) has wide statutory powers to make orders for costs under section 250 (5) of the Local Government Act 1972 and sections 322 and 322A of the Town and Country Planning Act 1990, it has been the long-standing practice to let costs lie where they fall except where one party has been guilty of unreasonable behaviour.
52. The Planning and Compensation Act 1991 amended section 17 by inserting a new sub-section (9A) which provided:

"In assessing any compensation payable to any person in respect of any compulsory acquisition, there must be taken into account any expenses reasonably incurred by the person in connection with the issue of a certificate under this section (including expenses incurred in connection with an appeal under section 18 where any of the issues are determined in the person's favour)."

53. The express inclusion of expenses incurred in connection with an appeal is limited to those of the expropriated landowner. That must have been intended to carve out an exception from the Secretary of State's general policy in relation to the award of costs. I would infer that the carve out is limited to the costs of the expropriated landowner precisely because it is he who is being expropriated.
54. Section 17 in its current form was inserted by the Localism Act 2011. What had been section 17 (9A) is now found in section 17 (10) but its form is unchanged.
55. The genesis of rule 10 of the Procedure Rules is a report on costs published by the Senior President of Tribunals entitled "Costs in Tribunals" in December 2011. The main relevant recommendation in the report was that the UT (Lands Chamber) should depart from its general costs shifting practice towards a practice of not awarding costs against an unsuccessful party if they had behaved reasonably and incurred reasonable costs in exercising their rights. The new rule 10 was intended to implement that recommendation. But the Explanatory Memorandum accompanying the new rules went on to say that there would be occasions and individual cases in which an element of costs shifting would remain. Those cases included "cases relating to compulsory purchase". What was to be encompassed in that phrase was not further discussed.
56. It is to be noted, however, that at the time of the SPT's report appeals in CAAD cases remained assigned to the Secretary of State, whose practice remained that of not awarding costs except in cases of unreasonable behaviour; and section 17 (9A) entitled the expropriated landowner to his reasonable costs. Paragraph 7.14 of the Explanatory Memorandum explained that they were made in response to that report.
57. It must, in my judgment, follow that rule 10 (6) (a) cannot have been intended to capture CAAD appeals, because at the date of the SPT's report, which the new rules were intended to implement, they were not within the jurisdiction of the UT. It was not until 6 April 2012 that jurisdiction in CAAD appeals was conferred on the UT.
58. Even when that change was made, it could not have resulted in completely discretionary costs shifting because section 17 (10) remained in force. Nor was rule 10 (6) (a) amended in order to widen its previous scope. On the contrary, in the meantime section 4 of the 1961 Act had been amended as I have explained to narrow its scope.
59. It would, in my judgment, have been surprising if the mere fact that the route of appeal against a CAAD had been changed was intended to effect a silent but radical change in the long-established practice of awarding costs in CAAD appeals. If that had been intended the Explanatory Memorandum would surely have said so. Rule 10 (6) (a) is well capable of being interpreted as being limited to disputes referred to the UT under section 1 of the 1961 Act, all the more so because of the narrowing of the scope of section 4 which would give either party costs protection in such circumstances. In my judgment, it should be so interpreted.
60. For these reasons, I conclude that the power of the UT to make a costs order (absent unreasonable behaviour) was not enlarged when CAAD appeals were transferred to it; and that in consequence the UT did not have the power to make the costs order that it did. It is a matter for the Tribunal Procedure Rules Committee to consider whether any change in the Procedure Rules is desirable.
61. I would allow the costs appeal.

Lord Justice Peter Jackson:

62. I agree

Lord Justice Nugee:

63. I also agree.