Cardtronics Europe Ltd & Ors v Syke & Ors (Valuation Officers) [2018] EWCA Civ 2472

[2018] EWCA Civ 2472
IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE UPPER TRIBUNAL (LANDS CHAMBER) MR MARTIN RODGER Q.C. AND MR A.J. TROTT F.R.I.C.S.
[2017] UKUT 0138 (LC)
Venue Royal Courts of Justice Strand, London, WC2A 2LLDate 9 November 2018
Before
Lady Justice Gloster, Vice President of the Court of Appeal, Civil Division Lady Justice KingLord Justice Lindblom
C3/2017/1590, C3/2017/1713, C3/2017/1714, C3/2017/1716 and C3/2017/1717
Between
Cardtronics Europe Ltd.Tesco Stores Ltd. and Tesco Personal Finance PlcSainsbury’s Supermarkets Ltd. and Sainsbury’s Bank PlcThe Co-operative Group Ltd.AppellantsChris Sykes and Others (Valuation Officers) Respondents
Mr Daniel Kolinsky Q.C. and Mr Luke Wilcox (instructed by DMH Stallard LLP) for First AppellantMr Timothy Mould Q.C. and Mr Guy Williams (instructed by Berwin Leighton Paisner LLP) for Second AppellantMr Richard Drabble Q.C. and Mr Christopher Lewsley (instructed by Dentons UKMEA LLP) for Third AppellantMr Timothy Mould Q.C. and Mr Christopher Lewsley (instructed by Dentons UKMEA LLP) for Fourth AppellantMr Timothy Morshead Q.C. and Ms Galina Ward (instructed by HMRC Solicitor’s Office and Legal Services) for RespondentsAnd between:C3/2017/1780Chris Sykes and Others (Valuation Officers) AppellantsTesco Stores Ltd. and Tesco Personal Finance Plc Respondents
Mr Timothy Morshead Q.C. and Ms Galina Ward (instructed by HMRC Solicitor’s Office and Legal Services) for AppellantsMr Timothy Mould Q.C. and Mr Guy Williams (instructed by Berwin Leighton PaisnerLLP) for Respondents
Hearing Hearing dates: 23 and 24 May 2018
Judgment Approved by the court for handing down(subject to editorial corrections)Lord Justice Lindblom:

Introduction

[1]What is the correct approach in law to the rating of the site of an automated teller machine (“ATM”) in a supermarket, shop, or petrol filling station? That is the basic question in these appeals. In 2015 there were about 70,000 ATMs in the United Kingdom, most of them in bank premises. But many are in supermarkets and other shops, where the operator of the ATM and the occupier of the premises are not the same company. How should the sites of these ATMs be rated? Are they capable in law of being separate hereditaments? And if so, are they in the rateable occupation of the banks that own the ATMs?[2]The appeals are against the decision of the Upper Tribunal (Lands Chamber) (Mr Martin Rodger Q.C., Deputy Chamber President, and Mr A.J. Trott F.R.I.C.S.) (“the Tribunal”), on 12 April 2017, allowing, in part, appeals against the decision of the Valuation Tribunal for England (“the VTE”), on 4 March 2016, to refuse to alter the 2010 rating list by deleting amendments made by Valuation Officers to create separate entries for the sites of ATMs. The sites are in supermarkets owned by Sainsbury’s Supermarkets Ltd. (“Sainsbury”), by Tesco Stores Ltd. (“Tesco”) and by The Co-operative Group Ltd. (“the Co-op”), and in premises where an ATM is operated by Cardtronics Europe Ltd. (“Cardtronics”). The ATMs in the supermarkets and superstores are operated by banks within the same corporate structure as the retailer. Cardtronics operates more than 16,000 of its own ATMs in shops and other premises.[3]Each of the disputed alterations to the rating list had the effect of including the site of the ATM as a separate hereditament with its own rateable value, but – in most cases – without any corresponding reduction in the rateable value of the shop or other premises in which it was located. Each site was the subject of an appeal to the VTE. Of those appeals, 11 were selected as lead cases. All 11 were dismissed by the VTE, which concluded, in each case, that the sites of the ATMs were separate hereditaments in rateable occupation by the bank or other ATM provider, not the owner of the host premises, and that the site in question should therefore remain the subject of a separate entry in the rating list. Nine of those 11 cases went on appeal to the Tribunal; the other two were stayed. The Tribunal allowed Tesco’s appeals for ATM sites inside its superstores in Nottingham, Walsall and Rugby, but dismissed the others. It subsequently granted each party permission to appeal to this court.

The main issues in the appeals

[4]The parties’ grounds of appeal raise a number of common issues, two of which are central:(1) Did the Tribunal err in its approach to the identification of a hereditament?(2) Did the Tribunal err in its approach to the rateable occupation of the ATM sites?

The ATM sites considered by the Tribunal

[5]The Tribunal provided (in paragraphs 21 to 75 of its decision) a very full and clear account of the relevant facts – including the relevant practical and contractual arrangements – which need not be repeated here.[6]The ATM in Sainsbury’s supermarket in Worcester is in an external wall, next to the main entrance door, and can be used 24 hours a day. It sits on a metal plinth, is chained to the floor of the cash room in the store, and is connected to the supermarket’s electricity supply. Sainsbury’s supermarket in Stroud has two ATMs sited next to each other in an external wall, within a secure cash room beside the main entrance door. Sainsbury is the leasehold owner of each of these stores, and is entitled to exclusive possession of all the floor space in them. There is no lease to Sainsbury’s Bank Plc. Under an agreement entered into between Sainsbury and the bank in February 2007 – known as the “Squadron” agreement – the bank was granted the right to install and operate the ATMs, “to be enjoyed in common with” Sainsbury. The bank has a licence to enter each ATM site. Responsibility for the day-to-day operation and management of the ATMs lies with Sainsbury. The cash dispensed by the ATMs is owned by the bank, but is kept in the security room of the store, under the control of Sainsbury’s staff, and at Sainsbury’s risk. Maintenance is carried out daily by Sainsbury’s staff, during the opening hours of the store.[7]Tesco’s superstore in Walsall has two internal ATMs. Its superstore in Nottingham has two ATMs in external walls, and one internal ATM. Its supermarket in Rugby has one internal ATM. The ATMs in those stores were installed under an ATM site licence entered into between Tesco and Tesco Personal Finance Plc on 29 November 2002. The licence conferred mutual obligations and benefits on the two parties for the installation and operation of the ATMs. Tesco Personal Finance Plc has the exclusive right to install and operate ATMs in Tesco’s stores. It is obliged to deliver agreed levels of service, which require the ATMs to be operational at specified times.[8]The Co-op’s supermarket in Newcastle-under-Lyme has an ATM in an external wall. In Huddersfield, it has a shop at a petrol filling station, with an ATM in an external wall. Its supermarket in Keighley has an ATM in an external wall. Under a licence agreement dated 1 January 2008, The Co-operative Bank Plc was permitted to install and operate ATMs on the sites created for that purpose by the Co-op, as owner and operator of the stores. The ATMs were maintained, and the cash in them replenished, under separate agreements between the bank and third parties. The ATMs themselves and the cash in them are owned by the bank, but the ATMs are maintained and operated by the Co-op’s staff. Electrical power is provided to them from the stores. The bank has access to them only during the stores’ opening hours.[9]Cardtronics operates an ATM in a Londis “convenience store” in Harefield, with about 60 square metres of floor space. The ATM is in an external wall, next to the entrance door. The ATM was placed in the store under a licence agreement with Londis, dated 26 March 2007, which makes provision for Cardtronics to gain access to it. It is owned, operated, maintained and loaded by Cardtronics. Maintenance and loading are undertaken within the store. Loading blocks an aisle, and the store is sometimes closed while it is being carried out.

The concept of a hereditament

[10]Non-domestic rates are a tax on individual units of property, known in England as hereditaments. Where a hereditament is wholly or partly occupied, rates are payable by the party who is in rateable occupation. While a hereditament may be occupied by more than one party, only one occupier may be in rateable occupation. Under section 42 of the Local Government Finance Act 1988 the rating list must show every relevant non-domestic hereditament. Under section 64(4)(a), a hereditament is a relevant hereditament if it consists of “lands”. Section 64(1) defines a “hereditament” as “anything which, by virtue of the definition of hereditament in section 115(1) of [the General Rate Act 1967], would have been a hereditament for the purposes of that Act had this Act not been passed”. The relevant definition in section 115(1) of the 1967 Act is “property which is or may become liable to a rate, being a unit of such property which is, or would fall to be, shown as a separate item in the valuation list”.[11]Whether a unit of property is capable of being a hereditament is determined by applying principles developed by the courts in a long series of cases. In Woolway (Valuation Officer) v Mazars LLP [2015] A.C. 1862, [2015] UKSC 53, the Supreme Court confirmed that the primary test of whether distinct spaces in common occupation were to be assessed for rates as a single hereditament was a “geographical” test – which, said Lord Sumption (in paragraph 6 of his judgment), “depends simply on whether the premises said to constitute a hereditament constitute a single unit on a plan”. He referred (in paragraph 11) to the Scottish rating decision in Burn Stewart Distillers Plc v Lanarkshire Valuation Joint Board [2001] R.A. 110, where it was emphasized by the court that “[the] underlying purpose is to provide a proper basis for a tax on property, not a tax on persons or businesses”. Lord Sumption recognized (in paragraph 12 of his judgment) three broad principles in the relevant authorities: “12. … First, the primary test is … geographical. It is based on visual or cartographic unity. Contiguous spaces will normally possess this characteristic, but unity is not simply a question of contiguity, as the second Bank of Scotland case [Bank of Scotland v Assessor for Edinburgh (1891) 18 R 936] illustrates. If adjoining houses in a terrace or vertically contiguous units in an office block do not intercommunicate and can be accessed only via other property (such as a public street or the common parts of the building) of which the common occupier is not in exclusive possession, this will be a strong indication that they are separate hereditaments. If direct communication were to be established, by piercing a door or a staircase, the occupier would usually be said to create a new and larger hereditament in place of the two which previously existed. Secondly, where in accordance with this principle two spaces are geographically distinct, a functional test may nevertheless enable them to be treated as a single hereditament, but only where the use of the one is necessary to the effectual enjoyment of the other. This last point may commonly be tested by asking whether the two sections could reasonably be let separately. Thirdly, the question whether the use of one section is necessary to the effectual enjoyment of the other depends not on the business needs of the ratepayer but on the objectively ascertainable character of the subjects. The application of these principles cannot be a mere mechanical exercise. They will commonly call for a factual judgment on the part of the valuer and the exercise of a large measure of professional common sense. But in my opinion they correctly summarise the relevant law. They are also rationally founded on the nature of a tax on individual properties. …”.[12]Lord Neuberger of Abbotsbury, Lord Carnwath, Lord Toulson and Lord Gill all agreed with Lord Sumption. Lord Neuberger said that “[normally] … both as a matter of ordinary legal language and as a matter of judicial observation, a hereditament is a self-contained piece of property (… property all parts of which are physically accessible from all other parts, without having to go onto other property), and a self-contained piece of property is a single hereditament” (paragraph 47). Lord Gill confirmed (at paragraph 35) that “[in] the law of Scotland, the identification of the valuation unit, or the unum quid, rests on a geographical test”.

The concept of rateable occupation

[13]In John Laing & Son Ltd. v Kingswood Assessment Committee [1949] 1 K.B. 344, Tucker L.J. referred (at p.350) to “four necessary ingredients in rateable occupation”. The first was that “there must be actual occupation”; the second, that “it must be exclusive for the particular purposes of the possessor”; the third, that “the possession must be of some value or benefit to the possessor”; and the fourth, that “the possession must not be for too transient a period”. As Lord Russell of Killowen had said in Westminster Council v Southern Railway Company Ltd [1936] A.C. 511 (at p.533), “the crucial question must always be what in fact is the occupation in respect of which someone is alleged to be rateable, and it is immaterial whether the title to occupy is attributable to a lease, a licence, or an easement”.[14]In The Assessment Committee of the Holywell Union v Halkyn District Mines Drainage Company [1895] A.C. 117 (at p.126), Lord Herschell L.C. observed that “[where] a person already in possession has given to another possession of a part of his premises, if that possession be not exclusive he does not cease to be liable to the rate, nor does the other become so”. He gave the example of a landlord and his lodger, both of whom “are, in a sense, in occupation, but the occupation of the landlord is paramount, that of the lodger subordinate”.[15]The distinction between paramount and subordinate occupation was considered in Westminster Council v Southern Railway. That case concerned the status, for the purposes of rating, of various bookstalls and kiosks, a chemist’s shop and showcases used for advertising goods at Victoria railway station – all of them occupied by independent retailers under agreements with the railway company. In considering whether these units were capable of separate assessment from the station as a whole, Lord Russell concluded (on pp.529 and 530):
“… Occupation … is not synonymous with legal possession … . Rateable occupation … must include actual possession, and it must have some degree of permanence … . Where there is no rival claimant to the occupancy, no difficulty can arise; but in certain cases there may be a rival occupancy in some person who, to some extent, may have occupancy rights over the premises. The question in every case must be one of fact – namely, whose position in relation to occupation is paramount, and whose position in relation to occupation is subordinate; but, in my opinion, the question must be considered and answered in regard to the position and rights of the parties in respect of the premises in question, and in regard to the purpose of the occupation of those premises. In other words, in the present case, the question must be, not who is in paramount occupation of the station, within whose confines the premises in question are situate, but who is in paramount occupation of the particular premises in question.”
The critical consideration in ascertaining whose occupation was “paramount” was the exercise of “general control” over the premises in question. As Lord Russell put it (on p.530):
“… The general principle applicable to the cases where persons occupy parts of a larger hereditament seems to be that if the owner of the hereditament (being also in occupation by himself or his servants) retains to himself general control over the occupied parts, the owner will be treated as being in rateable occupation; if he retains to himself no control, the occupiers of the various parts will be treated as in rateable occupation of those parts.”
This he called “the landlord-control principle” (p.531). On the question of “control”, he said this (on p.532):
“In truth the effect of the alleged control upon the question of rateable occupation must depend upon the facts in every case; and in my opinion in each case the degree of control must be examined, and the examination must be directed to the extent to which its exercise would interfere with the enjoyment by the occupant of the premises in his possession for the purposes for which he occupies them, or would be inconsistent with his enjoyment of them to the substantial exclusion of all other persons.”
He concluded (on p.537) that “the sites of the bookstalls in question … are so let out as to be capable of separate assessment”.[16]Lord Wright M.R. reached similar conclusions. He acknowledged the restrictions imposed by the railway company for the proper working of the station as a whole. But the tenant was “still in sole occupation of his premises and [had] full use of them to carry on his business” (p.555). He could not see how the railway company could be said to be in occupation of the bookstalls. The “theory of the lodger” did not depend on the fact that the landlord still lives in the house, but on the fact that “he still retains control for purposes of his business of the whole house”. The railway company did not retain such control over the bookstalls, but let them out to W.H. Smith & Son “for purposes of their separate business” (p.561).[17]As for the requirement that possession or occupation must be exclusive for the purposes of the possessor, in Wimborne District Council v Brayne Construction Co. Ltd. [1985] R.A. 234, Lloyd L.J. said (on p.239) that the correct sense of the word “purpose” here was “the object of the activity in question, rather than the motive behind the activity” – what someone does rather than why he does it. The logical basis for the possibility that two occupiers of the same hereditament might both be in exclusive occupation was this (p.243):
“… [An] occupier, in order to qualify for rateable occupation, has only to be in exclusive occupation for his own particular purposes. This does not exclude others from occupying the same hereditament for their particular purposes. Paramountcy is a way of choosing between exclusive occupiers in that sense. The degree of control exercised by one occupier over the other, or by a third party, seems to be relevant to both questions – that is to say, to whether an occupier is in exclusive occupation for his own particular purposes, and also to which of two competing occupiers is in paramount occupation.”
Sir George Waller (on p.247) said that the word “particular” in the expression “the particular purposes of the possessor” was “to emphasise the work that the possessor is doing – selling newspapers and not running a railway – in the Southern Railway case; the carrying out of the contract in Laing’s case”.[18]Applying those principles to the question of whether a petrol filling station in a motorway service area was in the rateable occupation of the petroleum company, and not the company operating the motorway service area, the Lands Tribunal concluded, on the facts, that it was (Esso Petroleum Co. Ltd. v Walker (V.O.) [2013] R.A. 355, [2013] UKUT 052 (L.C.)). It accepted that “[an] essential fact of occupation is the relative position of the parties and the rights under which each party occupies” (paragraph 81). It recognized that “the essential factual test of paramountcy is control”, and that “[control] depends on the facts of the case” (paragraph 85). And it asked itself to what extent the “degree of control” of either company interfered with the occupation of the premises by the other for the purposes of that occupation (paragraphs 87 and 88).

The status of ATMs as items of machinery

[19]The disputed issues in these appeals concern not the ATMs themselves but the floor space on which they stand – the site of the ATM, not the machine itself.[20]The Valuation for Rating (Plant and Machinery) (England) Regulations 2000 were made under a power conferred by paragraph 2(8) of Schedule 6 to the 1988 Act, which is concerned solely with valuation. Regulation 2 of the 2000 regulations sets out the “[prescribed] assumptions as to plant and machinery”. It states that “[for] the purpose of determining the rateable value of a hereditament” for any day on or after 1April 2000, in applying the statutory valuation criteria, “(b) in relation to any … hereditament [other than one in or on which there is plant or machinery belonging to any of the classes in the Schedule], the prescribed assumption is that the value of any plant or machinery has no effect on the rent to be so estimated”. An ATM is not within any of the classes of plant and machinery in the Schedule. The effect of regulation 2(b), therefore, is that an ATM is an item of machinery the value of which must be assumed to have no effect on the rateable value of the hereditament on which it is sited. It may therefore be regarded as non-rateable.[21]But as Lord Neuberger acknowledged in Woolway v Mazars (in paragraph 49 of his judgment), it is “well established that premises are not merely liable to have their rateable value assessed, but also to have their status as a hereditament assessed, by reference to the machinery, plant and other structures which have been placed in or on them, whether by the occupier or someone else, sometimes even if the structure retains its character as a chattel – see per Lord Radcliffe in London County Council v Wilkins (Valuation Officer) [1957] AC 362, 378”.[22]In London County Council v Wilkins, Lord Radcliffe accepted (at p.378) that “the presence of chattels on land” could be “a relevant factor either in determining the assessment of the rateable value of a hereditament or in determining whether there is a rateable occupation or not”. He also acknowledged as “equally well established” that “a structure placed upon another person’s land can with it form a rateable hereditament, even though the structure remains in law a chattel and as such the property of the person who placed it there”. In that case the issue was whether huts on a building site, which were used as offices, stores and a canteen, should be entered on the rating list as rateable hereditaments, though the building site itself was not rated. The House of Lords accepted that the Lands Tribunal had made no error of law in holding that the occupation of the huts by the building contractors was rateable. Lord Radcliffe said (on p.380):
“In my opinion, the present case really centres round the question whether the sheds, erected on a building site by a building contractor for the purpose of his operations, involve such a degree of permanency in his occupation as to make it a rateable one. I have no doubt that, in considering this, it is at any rate relevant to ascertain to what extent and in what way these constructions have been made a fixed part of the site on which they stand, for the more casually they are attached the less likely it is that the occupation of them will be found to be a permanent one. In this sense it may be of some importance to inquire whether they are chattels or not. But to make the whole issue of rateability depend on the bare issue, for instance, whether a particular structure has or has not foundations in the ground which give it a measure of lateral as well as subjacent support would be to use a legal distinction for a purpose for which it was never intended.”
[23]In Vtesse Networks Ltd. v Bradford [2006] EWCA Civ 1339, the chattel in question was a fibre optic telecommunications network comprising pairs of extremely narrow fibres, leased to the ratepayer for use in the transmission of its own signals. The Court of Appeal concluded that the network of fibres was a unit of property forming a hereditament in rateable occupation (see the judgment of Lloyd L.J. at paragraph 33, and the judgment of Sedley L.J. at paragraphs 40 to 43). But as the Tribunal pointed out in this case (in paragraph 102 of its decision) “[the] builder’s hut in Wilkins and the network of optical fibres in Vtesse, though chattels, were rateable property in their own right, and the assessments were of the huts and cables and not simply of the space they occupied”, and “ATMs, in contrast, are non-rateable”.

Previous cases on the rating of ATMs

[24]The correct approach to the rating of ATM sites has been considered in several cases, both in this jurisdiction and in Scotland.[25]In Stringer (V.O.) v J. Sainsbury Plc [1992] R.A.16, the only decision on the rating of ATM sites in this jurisdiction, the Lands Tribunal held that the site of six “hole in the wall” ATMs at Sainsbury’s supermarket in Leicester had been correctly assessed as part of the premises comprising the supermarket premises. The ATMs, each belonging to a different bank or building society, were located in a secure room and the connected security lobby, adjacent to the exit-entry lobby of a supermarket. The secure room and security lobby had been constructed at the same time as, and as an integral part of, the supermarket. Six apertures were incorporated into the external wall of the secure room, to house the part of the ATMs to which the general public had access. The “supermarket [was] open for trading Monday to Saturday inclusive for about 11.5 hours per day”, but the “cash points [were] accessible for use during the 24 hours of every day of the year” (p.21). The Lands Tribunal held (at p.29) that the site of the ATMs was not separately rateable. It tackled the issue of paramount, and thus rateable, occupation by having regard to Sainsbury’s “degree of control”. It concluded, among other things:
“1. The making available of bank notes at this location in Leicester serves the purpose of both J Sainsbury PLC and the financial institutions and some persons, but not necessarily all, would be customers of both parties. … … 3. The supply of the mutually beneficial service to customers is in practice a complementary joint venture has been acknowledged to be from the outset of the invitation to enter into an agreement. 4. Without the day to day management, oversight, services and control of the machines provided by Sainsbury the ATMs can bestow no benefit upon the financial institutions; their presence in the secure room becomes merely the storage of inoperative chattels. … 6. From the outset the parties to the agreements contemplated not only a shared objective and joint provision of the means to provide the facility but also that the on-site control would reside solely with Sainsbury. Financial institution personnel and nominated engineers would exceptionally, in the case of breakdown of an ATM, be afforded access to the secure areas but even then only so long as Sainsbury staff were present for security supervision purposes. 7. The degree of control thus exercised by J Sainsbury PLC makes its occupation clearly paramount and that of the financial institutions wholly subordinate. 8. The granting by the agreements of exclusive occupation to the financial institutions was never in the contemplation of the parties nor has it been afforded in practice”
. The secure room and security lobby were therefore in the rateable occupation of Sainsbury and no other party.[26]The first of the Scottish cases is Clydesdale Bank Plc v Lanarkshire Valuation Joint Board Assessor for Lanarkshire 2005 S.L.T. 167. There the Lands Tribunal for Scotland had to consider whether the sites of free-standing ATMs inside various supermarkets and shops, to which the public had access only from within the premises, should be entered in the valuation roll as separate units. The tribunal held not. It was “clear that no independently identifiable unit of lands and heritages existed before the use was made of a particular part of the shop as a site for the ATM” (pp.18 and 19). In many shops it would be “possible to identify “units” of heritable property by reference to use”, and “[subjects] such as “site for ice cream cabinet”; “site for photocopier” or “site for soft drinks machine” could expect to be identified in many convenience stores”. In the tribunal’s view it “would seem fanciful to start by treating these as separate units of lands and heritages requiring further investigation in each case”. It said that “[the] process of creating heritable units by reference to use inevitably makes that use the dominant characteristic of occupation”, and that “[this] cannot be regarded as a sound approach” (p.19), and that “the question of control must be seen as essentially subordinate to the broad question of purpose” (p.22). It found it “unnecessary to deal with dicta in Stringer”, but observed (ibid. and on p.23) that that case had been “concerned with heritable sites specially adapted to accommodate hole in the wall machines”. If the machines had been absent, the “special nature of the site” would remain “readily apparent”. The “subjects were not within the established hereditament, the store”. Whether they were to be regarded as part of the shop unit or as “separate subjects” was a question arising from the nature of “two distinct heritable units”. The question here was “quite a different one”.[27]The Lands Valuation Appeals Court upheld the tribunal’s decision. Lord Gill, the Lord Justice Clerk, described each of the ATMs as “a free-standing piece of moveable property” (paragraph 24). He did not accept the contention that the sites of the ATMs were in the sole, or at least predominant, occupation of the bank – a proposition based on cases in which the issue of rateable occupation “related to corporeal subjects within a larger site; for example, a bookstall in a railway station (Westminster Council v Southern Railway …) or a café, bar or shop in [an] airport terminal ([Renfrewshire Assessor v Old Consort Co. Ltd. 1960 S.C. 226]) or a contractor’s hut on a building site (London County Council v Wilkins)” (paragraph 25). In all those cases, the operator of the unit entered in the roll had a right of occupation of some kind in respect of it, usually conferred by a lease. Although it was possible for “an area of floor … used as the site for an item of moveable property” to be entered in the roll as “lands and heritages in separate rateable occupation”, this was “only … if the ratepayer has a right of occupation of it” (paragraph 26). Lord Gill continued (in paragraph 27):
“27. The flaw in these appeals lies in the contention that the agreements confer a right of occupation of the floor space upon the bank. In my opinion, neither agreement has that effect. Both agreements relate to the supply, use and control of an item of moveable property that the bank supplies to the retailer, whether or not the bank looks after it on a day to day basis, for use by the retailer as one of its retail attractions. Neither agreement expressly or impliedly confers any right of occupation of the site by the bank. The bank cannot be said to “occupy” the floor space in any real sense. It has at most a right of access to the machine.”
Those considerations were enough to indicate that the question of “rateable occupation” by the banks did not arise (paragraph 28). But if that question had truly arisen, any right of occupation that the agreements could be thought to confer on the bank would have been “subordinate to that of the retailer” (paragraph 29).[28]In Assessor for Central Scotland Joint Valuation Board v Bank of Ireland [2011] R.A. 195, the Lands Valuation Appeal Court considered the approach to be taken to the rating of the site of a “hole in the wall” ATM in an outside wall of a sub-post office. The ATM could be used when the sub-post office was closed, and was available at all times to bank customers and cardholders. Lord Gill distinguished the case on its facts from Clydesdale Bank. He said (in paragraphs 15 to 17):
“[15] In my opinion, the crucial difference in this case is that there is no direct link between the ATM site and the operation of the Sub Post Office. The ATM cannot reasonably be said to be one of the retail attractions provided in the Sub Post Office for its customers. Where an ATM is sited within a retail store, it is reasonable to infer that its primary purpose is to provide a facility for shoppers enabling them to access cash in-store in the course of shopping there. It is reasonable also to infer that few users will go to the store solely to obtain cash from the ATM. [16] In this case, however, although the ATM rests on the floor of the Sub Post Office, the operative part of it from the user’s point of view is accessible only from outside. The ATM is therefore not an in-store facility. Within the Sub Post Office the site of it is in effect dead space. The ATM is intentionally provided for the use of the general public. For that purpose the building has been altered and adapted by the opening of an aperture in the glass frontage of the building in virtue of a planning permission and a building warrant. Furthermore, the usage of the ATM is entirely unrelated to the opening hours of the Sub Post Office. [17] The sub-postmaster has no access to the ATM site save for re-filling and for simple first line maintenance, for all of which he receives a commission. Beyond that, he has no rights or duties in relation to the machine.”
[29]Lord Gill therefore concluded (in paragraph 18) that “on the facts of this case it is the Bank that is in rateable occupation of the ATM site”, and that “[the] site should therefore be entered separately in the Roll”. Lord Hardie agreed. In contrast to the ATMs in Clydesdale Bank, the ATM was, he said, “provided for the benefit of the general public passing the sub post office”, and could not, on any view, be described as “a facility primarily for the benefit of the sub-postmaster to assist him in the efficient operation of the sub post office” (paragraph 20).[30]Lord Malcolm referred (in paragraph 27) to the “very different circumstances” from those in Clydesdale Bank. He said (ibid. and in paragraph 28):
“[27] … I refer in particular to the 24 hour and seven days a week external usage of the ATM, and to it being fixed to the frontage of the building in a clearly defined position for at least eight years. Furthermore the premises were altered to accommodate the ATM. In my view the agreement in the present case relates primarily to the provision of banking services to those passing by the post office, though no doubt some of them will also enter the premises. The ATM is not an adjunct to services offered to customers of the post office. It is a separate service provided by the bank to a different customer base. The specific, albeit small site, is exclusively devoted to an ATM owned by the bank, controlled by the bank, and operated for the bank’s purposes. The bank pays the post office for the use of the site and for the maintenance services provided by post office personnel. In short, … the bank enjoys the beneficial value of the subjects. In my opinion there is sufficient disconnection between the usage and purposes of the ATM site and the rest of the premises to indicate independent uncontrolled occupation of the appeal site by the bank for the purposes of the bank’s business. A unum quid valuation would not be appropriate given the separate functions and purposes. In the Clydesdale Bank appeals the ATMs could be regarded as accessory to the retailers’ premises and purposes. In the present case it is not only realistic, but in my view correct to regard the appeal site as in the rateable occupation of the owner of the ATM. [28] While [counsel for the assessor] emphasised the services provided by post office staff in the running of the ATM, in my view these are wholly consistent with the above analysis. If the post office is to be seen as exercising some occupation or control over the appeal site, it does not interfere with the full and exclusive enjoyment of the site by the bank, rather it is paid for by the bank on the basis that it facilitates the bank’s operation of the ATM. If, contrary to the view reached above, there is an element of shared occupation, no rivalry is involved. The bank’s rights are paramount so far as the appeal site is concerned.” [28] While [counsel for the assessor] emphasised the services provided by post office staff in the running of the ATM, in my view these are wholly consistent with the above analysis. If the post office is to be seen as exercising some occupation or control over the appeal site, it does not interfere with the full and exclusive enjoyment of the site by the bank, rather it is paid for by the bank on the basis that it facilitates the bank’s operation of the ATM. If, contrary to the view reached above, there is an element of shared occupation, no rivalry is involved. The bank’s rights are paramount so far as the appeal site is concerned.”

Did the Tribunal err in its approach to the identification of a hereditament?

[31]Before the Tribunal, Tesco, the Co-op and Cardtronics contended that the boundaries of a hereditament could not be defined merely by the presence of a piece of machinery – such as an ATM – not itself liable to be rated, and that if the ATMs themselves were ignored there was nothing to define the extent of the putative hereditaments, and therefore the “geographical test” in Woolway v Mazars could not be satisfied. They relied on the decisions of the House of Lords in Townley Mill Company (1919) Ltd. v Oldham Assessment Committee [1937] A.C. 419 and Kennet District Council v British Telecommunications [1983] R.A. 43. The Valuation Officers maintained that although regulation 2(b) of the 2000 regulations had the effect of requiring an ATM to be disregarded for the purposes of valuation, it did not require the presence of an ATM to be disregarded for the purposes of identifying the boundaries of a hereditament, and that, in accordance with the approach indicated by the Supreme Court in Woolway v Mazars, the sites of the ATMs were therefore liable to be treated as being in the rateable occupation of the ATM operator – the bank.[32]In Kennet District Council a main issue was whether BT was in rateable occupation of two telephone exchanges that were not yet in use, but in which telephone equipment was being installed. The equipment fell within the statutory provision that was the predecessor to regulation 2 of the 2000 regulations. The House of Lords restored the hereditaments to the rating list on the basis that the justices had been entitled to find that BT’s business purpose was the housing of telephone equipment and that the premises were being used for that purpose once the equipment was moved in. Lord Keith of Kinkel – with whom Lord Diplock, Lord Wilberforce, Lord Roskill and Lord Brightman agreed – acknowledged (on p.45) that the hereditament “does not include any plant or machinery within the building”. He went on to say (on p.46): “[Section 21(1)(b) of the 1967 Act] provides that it is for the purposes of valuation that plant and machinery within para (b) is to be left out of account, but it must, I think, follow that it is impossible to treat such plant and machinery as part of the hereditament for any rating purpose, even though it be so fixed or attached that it would have fallen to be valued as part of the hereditament under the law prevailing before the statutory ancestor of [section] 21 was enacted … . Nothing can be rated which is not capable of being valued for the purposes of rating, and nothing which is not so capable can be the subject of rateable occupation. So it was rightly conceded … that the hereditament … was land with the bare shell of the building on it, excluding all of the equipment therein. …”.[33]Relying on that passage in the speech of Lord Keith, Tesco, the Co-op and Cardtronics argued that non-rateable machinery must be ignored for “all rating purposes”, including the question of whether the site on which the machinery was placed was a separate hereditament. The Valuation Officers argued that their Lordships’ decision supported the proposition that such machinery was not to be assumed to be absent, but simply to have no effect on value.[34]As the Tribunal said, in this case, as in Edmondson v Teesside Textiles Ltd. (1984) 83 L.G.R. 317 there was no suggestion that the machinery in question – here the ATMs – formed part of the hereditament. In its view Kennet District Council could be distinguished on that basis. The authorities all concerned buildings whose boundaries clearly defined the extent of the relevant hereditament, the existence of which was not in doubt. None of them assisted “in determining whether a unit comprising only the space occupied by an item of non-rateable plant or machinery should be recognised as a hereditament” (paragraph 124 of the decision). It concluded (in paragraph 126):
“126. In principle, … we consider that the presence of an item of non-rateable machinery, such as an ATM, should not be ignored when determining whether a separate hereditament exists. The statutory assumption applies only for the purpose of valuation and may not legitimately be applied in answering the logically prior question of whether there is or is not a hereditament which needs to be valued.”
[35]Tesco, the Co-op and Cardtronics also argued before the Tribunal that no separate hereditament could be found to exist in the circumstances here, for two reasons: first, because the site of an ATM could not be regarded as “self-contained” in the sense referred to by Lord Neuberger in Woolway v Mazars, and secondly, because the ATM sites were all functionally dependent on, and inseparable from, the “host” store, and thus satisfied the second of the three requirements identified by Lord Sumption. The Valuation Officers opposed this argument, contending that the ATM sites were akin to shop units in a shopping centre, each of which can be a separate hereditament. The Tribunal concluded (in paragraph 130):
“130. … Once a machine has been installed there should … be no difficulty in defining the boundaries of a fixed ATM site with sufficient precision to satisfy the geographic test of self-containment. In cases involving more mobile equipment it may additionally be necessary to consider the nature of the occupancy and whether the Bank’s right was a right of occupation of a specific unit of property or simply a right of access to a machine wherever it happened to be located. But assuming the occupier of the site had a sufficient right of occupation, identifying the unit of occupation would not be problematic.”
[36]Where the ATM was operated by the owner of a “host” store, there would be, in the Tribunal’s view, “no question of the ATM site being a separate hereditament”. Where it was owned and operated by a third party, a separate entry in the list was unlikely until the machine itself was in place and the hereditament capable of being defined (paragraph 132). Even in the case of a “hole in the wall” machine, the extent of the hereditament “cannot be fully ascertained until a specific model [was] installed” (paragraph 133). The distinction drawn by the court in Clydesdale Bank, between a “right of occupation” of a unit of property and a right of access to a moveable piece of equipment, not itself rateable, was important. It reflected the fact that hereditaments are units of property, and rates are a tax on property (paragraphs 134 and 135). The Tribunal contrasted the circumstances here with the type of arrangement in Clydesdale Bank – where the bank was given “a right of access to a more or less [free-standing] moveable machine placed in a location chosen by the Store from which it could be readily moved … when the operational requirements of the Store so required”. In a case of that kind there was “nothing to identify a unit of property at all before the machine has been put in position, and little to indicate once it is there whether the current site is a distinct unit of property without examination of the rights granted”. So “no separate hereditament can be said to have been created and no question of rateable occupation by the Bank which supplies the machine arises for consideration” (paragraph 135).[37]The Tribunal therefore concluded (in paragraph 136):
“136. Arrangements of that nature can be distinguished from the different factual circumstances with which we are concerned in most of these appeals. Where premises have been designed or adapted to receive an ATM by the construction of a separate enclosed space to house the machine or by the creation of an opening in an external wall (or both), the physical circumstances will make the identification of a unit of property with the potential to be recognised as a separate hereditament more realistic. The deliberate creation of a specific space in a fixed and apparently permanent location, visibly different from the generality of the host store and clearly intended for a particular use, is sufficient to differentiate most of these cases from the arrangements considered in Clydesdale. In such cases enhanced security, visibility and permanence all contribute to the separation and identification of the unit.”
[38]This analysis the Tribunal saw as consistent with the approach taken in the Scottish cases, and, in particular, the reliance placed on the “physical characteristics” of a site by Lord Gill and Lord Malcolm in Bank of Ireland (paragraph 137). There would be a “comparable degree of certainty” over the extent of the hereditament in cases where the site of the ATM was “fixed by the agreement”, but that had not been done here (paragraph 138).[39]The Tribunal considered each of the ATM sites individually (in paragraphs 139 to 150). It concluded that all but one of them were capable of forming a separate hereditament. The sole exception was the ATM on the first floor of the Tesco store in Nottingham. That ATM stood in a space that had not been designed or adapted to accommodate it, and there was nothing to prevent it being moved elsewhere in the store. The “essential qualities” of this arrangement were, in the Tribunal’s view, “impermanence and mobility”. The space occupied by the machine “from time to time” was not a “unit of property separate from the remainder of the Store” (paragraph 143). The Tribunal therefore concluded (in paragraph 151):
“151. We are therefore satisfied that each of the appeal sites, with the exception of the first floor site at Tesco’s Nottingham store (where the machine is free standing), is capable of being the subject of a separate entry in the rating list. With that single exception, each site is more than just an indistinguishable space on the shop floor which happens to be occupied by an ATM; in each case the site has either been designed or adapted to accommodate such a machine. We are satisfied that the physical arrangements of a site, rather than incidental details of access or servicing arrangements, justify treating it as a potential hereditament. There are inevitably borderline cases ([Cardtronics’] machine at Harefield, and Tesco’s at Walsall being closest to the boundary), but a clear distinction can be drawn between the space occupied by free standing machines on the one hand and specific sites which have been designed or adapted for the purpose on the other. That distinction is practical and appropriate to a tax on property, it is consistent with the Scottish jurisprudence and it provides a clear answer to the first issue for each of the appeal sites.”
[40]For Cardtronics, Mr Daniel Kolinsky Q.C. submitted to us that the Tribunal’s approach was incorrect. It was wrong to conclude that the presence of an ATM was relevant in determining whether the site of the ATM was a separate hereditament. There is no reason to think that in Kennet District Council Lord Keith did not consider the identification of the hereditament to be a relevant “rating purpose”, for which the presence of non-rateable plant – such as an ATM – must be disregarded. The identification of a hereditament is prior to the question of occupation, and must be undertaken without considering occupation. An ATM is simply the means by which the putative hereditament is occupied. If the putative hereditament can only be defined by its being occupied, it cannot be a “unit of property” with coherent physical boundaries, capable of being separately rated. London County Council v Wilkins was distinguishable because in that case the builder’s huts were themselves rateable property, which became part of the hereditament when placed on the land.[41]Mr Kolinsky submitted that the site of the ATM in the Londis store in Harefield was not materially different from those considered in Clydesdale Bank, and was incapable of being a separate hereditament. In finding that the site met the “geographical” test in Woolway v Mazars, the Tribunal had wrongly assumed the presence of the ATM itself. The hole in the wall was not enough on its own to enable the boundaries of a hereditament to be defined. Even if the presence of an ATM could be taken into account in identifying a hereditament, the Tribunal had failed to consider whether the ATM site in the Londis store had coherent physical boundaries. And it did not. The putative hereditament could only be defined by occupation. Its size and extent were defined only by the floor space occupied by the machine itself – and liable to change with a different one.[42]For the Valuation Officers, Mr Timothy Morshead Q.C. submitted that the Tribunal was right to reject the argument that the presence of non-rateable machinery should be ignored when determining whether a separate hereditament exists, but wrong to hold that a hereditament could not exist in the absence of adaptations to the “host” store or a right to occupy a defined area, and to ignore the presence of ATMs when considering whether a hereditament was physically identifiable. There was no principle of law requiring the identification of a hereditament to depend on the existence of what Mr Kolinsky described as a “physically coherent unit of separate property” before the hereditament is brought into being by separate occupation. In a case such as this it is simply the separate occupation by a third party that brings the new hereditament into being. The “geographical” testis satisfied at the moment when it needs to be considered. This is shown, for example, by the treatment of the builders’ huts in London County Council v Wilkins, in the “wayfarer” cases considered in Reeves v Northrop [2013] 1 W.L.R. 2867, which concerned the creation of a hereditament where an area of land or river is occupied, respectively, by a caravan or a boat – and also in Westminster Council v Southern Railway, where the sites occupied by the operators of the kiosks and display cabinets had no separate physical identity until the railway company parted with control of them. The crucial question, Mr Morshead submitted, is this. What does the occupier occupy, once he starts to occupy it? Each of these ATM sites, including the one on the first floor of Tesco’s store in Nottingham, was readily identifiable on the ground and capable of being a separate hereditament if it was in separate rateable occupation from the “host” store.[43]Mr Timothy Mould Q.C., for Tesco, supported the Tribunal’s conclusions on the ATM on the first floor of the Nottingham store (in paragraphs 135 and 143 of its decision) as correct in principle, and consistent with the decision of the House of Lords in London County Council v Wilkins and the Lands Valuation Appeal Court in Clydesdale Bank.[44]Two questions arise. The first is whether items of plant or machinery, not themselves rateable, can, in principle, be relevant in the exercise of determining the existence and extent of a separate hereditament. The second is whether, assuming their potential relevance in that exercise, the presence of such items of non-rateable plant or machinery is capable of establishing the existence and extent of a separate hereditament without distinct physical arrangements being made within the premises to accommodate them. It has never been suggested – nor of course could it be – that the ATMs themselves constituted hereditaments in their own right.[45]On the first question, I cannot accept Mr Kolinsky’s submission that the presence of an ATM – as an item of non-rateable plant or machinery – within a supermarket or shop, whether sited inside the premises or in an external wall, must, in principle, be disregarded as being of no relevance to the question of whether a separate hereditament has come into existence when the ATM is located in the store. That is not required by any statutory provision. Nor can one find in the relevant case law, either in this jurisdiction or in Scotland, any support for an approach to identifying a hereditament in which the presence of non-rateable plant must simply be ignored. That is not, in my view, a correct understanding of what Lord Keith said in his speech in Kennet District Council (on p.46), nor a proper inference to draw from it. And I see no basis for it in principle.[46]As the Tribunal recognized, the statutory assumption in regulation 2 of the 2000 regulations is specifically and solely “[for] the purpose of determining the rateable value of a hereditament”, in applying the statutory valuation criteria. It does not extend to the prior exercise of ascertaining whether or not there exists a hereditament to be valued. In the passage of his speech in Kennet District Council on which Mr Kolinsky sought to rely, Lord Keith did not say that the presence of non-rateable plant and machinery on a site can have no part to play, or can carry no evidential weight, in establishing whether a hereditament exists. The context for what he said was the assumption in section 21 of the 1967 Act, the precursor to the one in regulation 2. The question he was considering was whether it was possible to treat “plant and machinery” within section 21(1)(b) as “part of the hereditament for any rating purpose, even though it be so fixed or attached that it would have fallen to be valued as part of the hereditament under the law prevailing before the statutory ancestor of [section] 21 was enacted”. He held not. He did so because, as he said, it is impossible to rate something that is itself incapable of being valued for the purposes of rating and hence incapable of being the subject of rateable occupation.[47]That, however, is a quite different question from the one we are considering now, which is not whether items of non-rateable plant and machinery can be, in themselves, a constituent part of a putative hereditament, but simply whether it is legally possible to take them into consideration in determining, in the first place, whether a hereditament may have come into existence, and, if so, its extent. Such a determination is not itself, in my view, a “rating purpose” in the sense in which Lord Keith used the expression “as part of the hereditament for any rating purpose” – meaning, as I understand it, for any rating purpose that flows from the hereditament being the “unit of property” comprised in it. The idea that Lord Keith had in mind the very process of identifying a hereditament as a relevant “rating purpose” is, in my view, mistaken. He was concentrating on the rating consequences of the plant or machinery in question not being part of the hereditament once identified, rather than on the relevance of the plant or machinery being present on the site when one was considering whether a hereditament could be identified. It does not follow from what he said that plant and machinery within regulation 2(b) of the 2000 regulations is irrelevant in that exercise.[48]There is nothing illogical or contradictory in this. It is possible, in principle, for an item of non-rateable plant to be a relevant factor in identifying a hereditament, yet not treated as part of that hereditament once it has been identified. As the Tribunal recognized, none of the authorities points to any different analysis. In none of them did the court have to deal with the specific question of the relevance of non-rateable plant in assessing the existence and extent of a hereditament in circumstances parallel to those of this case, because, as the Tribunal said, each of them concerned a site whose boundaries clearly defined the extent of the relevant hereditament, and the existence of the hereditament was not itself in doubt. None of them suggests that non-rateable property is not only to be taken as having no effect on the rateable value of the hereditament on which it is sited, but must also necessarily be assumed to be absent when the existence and extent of that hereditament are being determined.[49]Such an assumption would, I think, be artificial and unrealistic. It could exclude, for no good reason, a consideration that may go to the status of the site in question as a “self-contained piece of property” – the concept to which Lord Neuberger referred in Woolway v Mazars. Citing Lord Radcliffe’s speech in London County Council v Wilkins, Lord Neuberger also acknowledged the principle that premises are liable to have their “status as a hereditament” assessed by having regard to, among other things, the “machinery, plant and other structures … placed in or on them …”. An example of the principle in practice, perhaps extreme in its facts, is to be seen in Vtesse Networks. Although the chattels in London County Council v Wilkins and Vtesse Networks were themselves rateable and ATMs are non-rateable, I can see no reason, in the light of the relevant authorities, to limit the principle to property that is itself rateable. The presence of non-rateable plant or machinery on a site may be a manifestation of occupation, and may indicate the existence of a hereditament. I stress “may”, because it is possible, in a given case, that this will not be so. Whether it is so will depend on the circumstances (see Lord Radcliffe’s speech in London County Council v Wilkins, at p.377, and the judgment of Sir Alan Ward in Reeves v Northrop, at paragraph 19).[50]I therefore reject Mr Kolinsky’s argument on the first question, and accept Mr Morshead’s. In my view the Tribunal did not err in law in having regard to the presence of the ATMs when considering whether each of the sites it was considering was capable of being regarded as a separate hereditament. It was right to conclude (in paragraph 126) that “the presence of an item of non-rateable machinery, such as an ATM, should not be ignored when determining whether a separate hereditament exists”.[51]On the second question, I would not accept as a correct statement of the law the proposition that a separate hereditament can only ever come into existence through the presence of nonrateable plant or machinery on land, or in a building, if the site on which it is located has been specifically designed or physically adapted to accommodate that plant or machinery. The authorities do not produce such a principle.[52]What is necessary, as Lord Sumption explained in Woolway v Mazars when describing the “geographical” test, is that the putative hereditament can be represented as a “single unit on a plan” and has the quality of “visual or cartographic unity”. In many cases this quality may be apparent in a physical differentiation between one hereditament and another. The boundaries of the hereditament may be sharply defined on the ground. But as Lord Sumption emphasized, the principles involved in the “geographical” test will often require “a factual judgment” and the exercise of “professional common sense”. The facts will vary from case to case. In some cases, there will already exist a “physically coherent unit of separate property” – as Mr Kolinsky described the concept – before a separate hereditament is created. There will be others where the physical occupation of the site by the act of placing some structure or item of plant or machinery upon it may lead to a hereditament being formed, subject always to the requirements of the “geographical” test being fully met.[53]None of this, however, detracts from the need for sufficient certainty, both on the existence of the hereditament and on its extent. One of the essential attributes here, as the Tribunal acknowledged, is self-containment. As Lord Neuberger said in Woolway v Mazars, a hereditament is a “self-contained piece of property”. Unless the site is capable of being identified as a unit of property sufficiently defined by its own boundaries to be regarded as “self-contained”, it will not be capable of constituting a hereditament.[54]The Tribunal recognized this. It concentrated on the question of self-containment, both physical and in terms of a “right of occupation”. It was, I think, right to say (in paragraph 130) that in the case of what it called “a fixed ATM site”, once the machine itself has been installed there ought to be no difficulty in defining the boundaries of the site with “sufficient precision to satisfy the geographic test of self-containment”. As a general statement, this seems correct. In cases of “more mobile equipment”, however, the Tribunal distinguished – again rightly in my view – between circumstances in which the bank operating the ATM has a “right of occupation of a specific unit of property” and those in which it has nothing more than a “right of access to a machine wherever it [happens] to be located”. What was required in such a case was that the occupier of the site had a “sufficient right of occupation”. If it did, there would be no difficulty in identifying the “unit of occupation”. In my view this conclusion is also sound. It reflects the requirement that to be a hereditament a site must not be inchoate or ephemeral, but identifiable as a self-contained “unit of property”. This is consistent with the underlying scheme of non-domestic rating as a tax on property. As a tax on property, it depends on the relevant property being clearly defined.[55]The Tribunal’s general reasoning (in paragraphs 134 to 138), and its conclusions on the individual ATM sites (in paragraphs 139 to 151) were guided by the approach of the Lands Valuation Appeal Court in Clydesdale Bank, including the principle expressed by Lord Gill (in paragraph 26 of his judgment) that “within larger subjects an area of floor only that is used as the site for an item of moveable property can be entered in the roll as lands and heritages in separate rateable occupation; but only … if the ratepayer has a right of occupation of it”. This was echoed in Bank of Ireland, both by Lord Gill himself (in paragraph 16) and by Lord Malcolm (in paragraph 27).[56]I would accept Lord Gill’s formulation as consistent with the “geographical” test in the way in which it was explained by Lord Sumption in Woolway v Mazars. The Tribunal clearly did not regard it as incompatible with that test. Rightly in my view, it saw a significant distinction between all but one of the ATMs with which it had to deal in this case and the kind of arrangement considered in Clydesdale Bank, where the bank had a right of access to a “more or less free standing moveable machine” located in a part of the store chosen by the retailer “from which it could be readily moved … and the right of access diverted … when [the retailer’s] operational requirements … so required” (paragraph 135). In a case such as that, as the Tribunal recognized, it is difficult to see how the “geographical” test could ever be met. There would be no sufficiently coherent and settled “unit of property” capable of forming a separate hereditament, only an undifferentiated area of floor space within the hereditament comprising the retail premises as a whole. By contrast, in a case where the ATM is not free-standing and liable to be moved from one place in the store to another to suit the retailer’s requirements for the time being, but on a site deliberately designed or adapted to house it, that site will generally be capable of being a hereditament.[57]In that category the Tribunal was able to place all the ATM sites with which it was concerned, with the single exception of the ATM on the first floor of Tesco’s store in Nottingham – because the space occupied by that machine lacked the requisite definition and permanence to qualify as a hereditament (paragraph 143). These were, in the circumstances, findings and conclusions that one might expect. I do not think they can be criticized in law. The Tribunal also acknowledged that the sites of the ATMs in Tesco’s superstore in Walsall and the ATM operated by Cardtronics in the Londis store in Harefield were “borderline cases”. But it found, on the facts, that they were each capable of forming a separate hereditament (paragraphs 145 and 149 to 151) – and it was therefore “necessary to consider in the case of each ATM site capable of being a separate hereditament whether the Bank or the Store is in rateable occupation” (paragraph 152) (see, for example, the judgment of Lloyd L.J. in Vtesse Networks, at paragraph 28). This too was an unsurprising conclusion, and not, I think, legally flawed.[58]In my view, therefore, the second question is also resolved by endorsing the Tribunal’s conclusions. There was no error of law in the approach it took, which was orthodox, in conformity with the authorities on the concept of a hereditament, and faithful to the “geographical” test in Woolway v Mazars. Its relevant findings of fact are unimpeachable. It was entitled, and I think right, to conclude as it did. On this point, therefore, both the appeal of Cardtronics and that of the Valuation Officers must fail.

Did the Tribunal err in its approach to the rateable occupation of the ATM sites?

[59]For Sainsbury, Mr Richard Drabble Q.C. submitted to the Tribunal, as did Mr Mould for Tesco and the Co-op, that the VTE had been wrong to regard the sites of the ATMs as being in the sole occupation of the banks operating the ATMs. In fact, in each case the retailer had exclusive possession of the whole retail premises under its lease, and had not parted with possession of any part of it to the bank. The ATMs were being used by the retailer to provide cash and other banking services as part of the store’s “offer” – the provision of goods and services to the public. This purpose ought to have been taken into account, as should the dayto-day control of the sites by the retailer. Had this been done, the conclusion should have been that the ATM sites were occupied both by the retailer and by the bank, with the joint purpose of providing the same service to the public, and for their mutual benefit. This case was materially different from Westminster Council v Southern Railway, especially in the degree of control exercised by the retailers over the ATM sites. It was indistinguishable from Stringer. Here, the retailers’ occupation of the ATM sites was paramount. There was no separate rateable occupation by the banks. Mr Kolinsky submitted, in the same way, that Londis was in paramount occupation of the ATM site in the Harefield store. Cardtronics’ occupation was concurrent with the retailer’s, and complementary to the retailer’s commercial purpose.[60]Mr Morshead submitted to the Tribunal that the test of rateable occupation must be applied only to the premises in question, not to any larger premises. Occupation must be exclusive for the purpose of the occupier, and it is the occupier’s use of the premises that is relevant. The only relevant use here was the provision of ATM services by the banks. They had contractual obligations to the retailers to provide ATM services, and the retailers were obliged to support their operation of the ATMs. There was no parallel between an ATM and a “lodger”. If a retailer contracted with a bank to give the bank the exclusive right to provide and operate an ATM on the retailer’s premises, the only purpose relevant in determining the rateability of the ATM site would be the purpose of the bank’s occupation – to provide banking services to its own customers. The role of the retailer was simply to provide the site. Its purpose as a retailer occupying the store would not bear at all on question of the separate rateability of the ATM site. In having regard to that purpose, the decisions in the Scottish cases were inconsistent with the decision of the Court of Appeal in Wimborne District Council v Brayne Construction.[61]In considering those arguments, the Tribunal said (in paragraph 161):
“161. Where a distinct space has been created to receive a machine and the question arises whether there are rival candidates for rateable occupation of that space, the Scottish cases have focussed on whether, objectively assessed, the dominant purpose of the arrangement is to provide a facility or “retail attraction” for the host’s customers, or whether that purpose is insufficiently connected to the host’s own business to enable the host’s occupation (if any) to be regarded as paramount. That was the crucial factual difference between Clydesdale, where the use of the ATM was “part of the business of the shop” …, and Bank of Ireland, where to a degree the businesses of the Bank and the sub post office were in competition with each other and the site of the ATM was exclusively dedicated to serving the business of the Bank.”
[62]It noted that in Bank of Ireland “the degree of separation between the use of the site and the business of the post office was sufficient for the Bank to be seen as in sole occupation” (paragraph 163). Referring to the decision of the tribunal in Clydesdale Bank, it said (in paragraph 167):
“167. Far from disregarding Southern Railway the Tribunal in Clydesdale explained … that in cases of concurrent occupation “the question of rateable occupancy depends upon the question of whose possession is paramount and that is to be considered in regard to the purpose of occupation.”
The Tribunal’s sole modification of the approach taken in Southern Railway concerned the value of “control” as a means of resolving the issue of paramount occupation. In this very different factual context, where “occupiers are not truly rivals but are both deriving a direct benefit from the same use of the subjects” it regarded “the question of control … as essentially subordinate to the broad question of purpose”. That seems to us to be an unobjectionable refinement of the approach to paramount occupation where the circumstances do not justify treating concurrent occupiers as deriving different benefits from the use of the same unit of occupation. On the facts the Tribunal considered (and the Lands Valuation Appeal Court agreed) that the use of the ATM site was “for a purpose integral to the basic operation of the store” so that “although physically occupied by the machine owned by the bank, the substantive purpose of use of the site can properly be described in terms of being to facilitate ready access to cash for customers of the company.” That view of the facts led to the conclusion that the Store was in rateable occupation; if it is assumed (contrary to the approach on appeal in Clydesdale) that a moveable machine is capable of being a separate hereditament, we would have reached the same conclusion on those facts.” “control” as a means of resolving the issue of paramount occupation. In this very different factual context, where “occupiers are not truly rivals but are both deriving a direct benefit from the same use of the subjects” it regarded “the question of control … as essentially subordinate to the broad question of purpose”. That seems to us to be an unobjectionable refinement of the approach to paramount occupation where the circumstances do not justify treating concurrent occupiers as deriving different benefits from the use of the same unit of occupation. On the facts the Tribunal considered (and the Lands Valuation Appeal Court agreed) that the use of the ATM site was “for a purpose integral to the basic operation of the store” so that “although physically occupied by the machine owned by the bank, the substantive purpose of use of the site can properly be described in terms of being to facilitate ready access to cash for customers of the company.” That view of the facts led to the conclusion that the Store was in rateable occupation; if it is assumed (contrary to the approach on appeal in Clydesdale) that a moveable machine is capable of being a separate hereditament, we would have reached the same conclusion on those facts.”[63]The Tribunal therefore rejected the suggestion that the approach taken in the Scottish cases was either “defective or unreliable as a guide to the resolution of these appeals” (paragraph 168). It concluded (in paragraph 169):
“169. We also differ from the VTE in that we agree with the appellants that the floor space on which an ATM stands may be regarded as occupied both by the Store and by the Bank. The Store has not, in any of these cases, parted with possession of the site of the ATM, but it has agreed to confer rights on the Bank which substantially restrict the Store’s use of that small part of its premises which comprises the ATM site. The Store has agreed to that restriction because the presence of the ATM furthers its own general business purposes and because the operation of the ATM by the Bank provides the Store with an income.” of the site of the ATM, but it has agreed to confer rights on the Bank which substantially restrict the Store’s use of that small part of its premises which comprises the ATM site. The Store has agreed to that restriction because the presence of the ATM furthers its own general business purposes and because the operation of the ATM by the Bank provides the Store with an income.”
[64]To the extent that the appeals before it involved “fixed sites”, it was, said the Tribunal, “necessary to address the question of rateable occupation in the conventional way, by examining the degree to which there is concurrent occupation of the ATM sites and considering which party’s possession is paramount and which subordinate” (paragraph 170). It then said this (in paragraph 171):
“171. When applying that approach to external hole in the wall ATMs, the sites of which are identifiable as separate units of property, and which are accessible by the public at large without entering the store, there is no need to consider the occupation of the store as a whole. What matters is the purpose for which the ATM has been installed on the external wall of the building.”
[65]Agreeing with a submission made on behalf of the Valuation Officers, the Tribunal said that in its view “[the] relevant purpose for which a site is being used cannot depend on the subjective intentions of the occupiers of the site (“the motive behind the activity” as Lloyd LJ put it [in Wimborne District Council v Brayne Construction]), but must be objectively ascertainable”. It recognized that the “availability of ATMs in almost all larger supermarkets and in many smaller stores demonstrates their attraction to customers, who clearly wish to have access to the services provided by ATMs and find it convenient to do so in the location where they shop”, and that “[retailers] naturally wish to satisfy that expectation to attract and retain their customers” (paragraph 173). But a retailer’s “choice of an external ATM is also a choice to make the service separately accessible to the public at large, irrespective of the usage of the other services provided within the store” (paragraph 174). The survey evidence suggested that, “as would be expected, the location of an ATM outside a store, even where it is accessible only by coming into the car park, significantly widens the pool of users of the ATM service beyond the customers of the Store itself” (paragraph 175). The Tribunal then said this (in paragraph 176):
“176. We do not consider that it is generally helpful to characterise the Store and the Bank as rivals in their occupation of the site of an ATM. Both parties derive a direct benefit from the use of the site for the same purpose, and share the economic fruits of the specific activity for which the space is used. … .”
[66]It went on to say that “[when] considering rateable occupation in the context of a complementary activity like the provision of an ATM”, it did “not regard control or interference as particularly relevant considerations …” (paragraph 177). Nor were the arrangements for servicing ATMs “of much significance”. The “manner in which a particular occupier chooses to arrange for servicing, cleaning and maintenance does not change the physical character of the property or the purpose of its occupation”. The payment of a fee for servicing made it “more difficult to regard the servicing provided by the Store as an aspect of its occupation of the site of the ATM” (paragraph 178). In the Tribunal’s view it was “more helpful to consider the purpose of the occupation of the site in the light of the decisions the parties make about the manner in which the space dedicated to ATMs will be used” (paragraph 179). It continued (ibid.):
“179. … We regard it as significant that, by design, the target market of an external ATM is much broader than the retail customers of the store. An external ATM is not only physically remote from the generality of the “retail offer of the store”, … but its purpose is also distinctively different. It is to reach as wide a market for ATM services as possible, rather than to restrict usage to those who have entered the Store to make use of facilities only available to customers of the Store.”
[67]It did not consider the proportion of ATM users who spent some of the cash they had withdrawn in the store as a significant factor, in view of Cardtronics’ evidence that in its stores the total average spend by ATM users was only about £13 (paragraph 180).[68]Those who use external ATMs were, it said, more appropriately described “as customers of the Bank using a service which happens to be available at their local supermarket or convenience store, rather than as customers of the Store making use of an “in-store facility”” (paragraph 181). It concluded (ibid.):
“181. … We therefore agree with the approach taken in Scotland, exemplified by the Lands Valuation Appeal Court’s decision in Bank of Ireland, that external ATMs available to the public at large should not be regarded as an in-store facility.”
[69]Was a different approach required for an ATM in a smaller store such as the Londis supermarket in Harefield? The Tribunal thought not. It accepted that in a very confined space the “minimal segregation” between the two uses “creates a degree of rivalry or interference … absent from larger stores where there is the opportunity to accommodate the ATM in a secure room or housing of some description” (paragraph 182). In its view, the fact that “[the] ATM cabinet is surrounded by goods offered for sale and occupies space which would otherwise be fully used for retail purposes” and the fact that “[the] whole of the aisle is blocked when the machine is opened for servicing, which sometimes requires that the store close temporarily” made it “more realistic to regard the arrangement as involving an element of rivalry or interference” (paragraph 183). It concluded (in paragraph 184):
“184. The physical segregation of an external ATM site from the general retail activities of a store (even if it is within a secure room which is also used for other purposes) contributes to the recognition of the Bank as being in paramount occupation of the site because the activities of the Store in and around the site are restricted. Where physical segregation cannot be achieved we have nevertheless concluded that the case for regarding the Bank as being in paramount occupation of a fixed ATM site is just as strong. The presence of the ATM imposes to a greater degree on the practical operation of a very small store, even to the extent on some (relatively rare) occasions of requiring that the store be closed temporarily to enable cash to be replenished. In our judgment this imposition on the use and management of the store points to the paramountcy of the Bank’s occupation of the ATM site. On the other hand, the fact that this level of interference is tolerated by the Store confirms the benefits which the Store must see in having the ATM available, which might be said to enhance the case for regarding the machine principally as part of the retail offer of the Store. We do not think that this conclusion would be justified as the interference consequent on re-provisioning the machine would be no different if the ATM was internal and served only the Store’s customers and not the public at large. An external ATM is a facility available to all, whether or not they are customers of the Store, and in our judgment that justifies treating the Bank as the party in rateable occupation.”
And its general conclusion on the external ATM sites was this (in paragraph 185):
“185. Having regard to the broad customer base at which the service of an external ATM is targeted, the distinct character and branding of the space and the security arrangements associated with its use, the practical impossibility of the Store making any different use of the same space while it is occupied by the Bank’s ATM, and the inconvenience and impracticality of the machine being removed to a different location, we consider it to be realistic and workable to regard the Bank as being in paramount occupation. Although obviously the Bank and the Store have a mutual interest in providing ATM services, and both derive a benefit from the presence of the machines, where the parties have chosen to make the service available to all, and at all times, and have physically separated the ATM from the facilities offered within the Store, we consider it is right to treat the primary purpose of the occupation of the site of the machine as being a purpose of the Bank. The Bank’s occupation for that purpose is exclusive: only one machine can be accommodated on the site and in each case the arrangements between the Store and the Bank provide that only the Bank is to have the right to locate such a machine in the Store.” customers of the Store, and in our judgment that justifies treating the Bank as the party in rateable occupation.”
[70]It was therefore satisfied that “the sites of externally accessed ATMs should be entered as separate hereditaments in the occupation of the Bank” (paragraph 186). This conclusion applied to the ATM sites in Sainsbury’s Stroud and Worcester stores, those at the entrance to Tesco’s store in Nottingham, those in all three of the Co-op stores, and the site of the ATM at the Londis store in Harefield. It applied to the ATM site at the Co-op store in Newcastle under Lyme, even though the ATM was available only during trading hours, because its purpose was “to provide ATM services to all who wish to use them, whether Co-op customers or not …” (paragraph 187).[71]As the Tribunal acknowledged, these conclusions were inconsistent with the decision of the Lands Tribunal in Stringer, which, as it said, “concerned a very similar arrangement for the provision of external, hole in the wall ATMs at Sainsbury’s store in Leicester”. But it was “not unduly disturbed” by that. Stringer was a decision on the facts applicable to a proposal made in March 1987. The passage of time had had “brought changes in payment technology … relevant to the assessment of the purpose of the occupation”. More weight had been given in that case to the role played by Sainsbury’s employees in servicing the ATMs, which in this case the Tribunal did “not regard … as a factor to which much significance should be given”. And Stringer had not been regarded as authoritative in the Scottish cases (paragraph 188).[72]Turning to the ATMs located inside the retail premises – the two ATMs in Tesco’s Walsall store and the single ATM in its Rugby store – all of “which face onto the shop floor and are accessible only by those who have entered the store and can therefore be assumed (in the great majority of cases) to be customers of the store” (paragraph 189), the Tribunal concluded (in paragraphs 190 and 191):
“190. We consider that the sites of these internal ATMs are in the paramount occupation of the Store, and not the Bank. The service is primarily offered to shoppers in the store, and is not aimed at attracting passing trade (although no doubt there will be occasions when someone who wishes to use an ATM and is aware of its presence inside a store may make an incidental purchase). The purpose of the Bank’s occupation of the site is to provide a service to the Store’s customers, which is also the purpose of the Store’s occupation of the whole of the premises including the site. By its control of the opening hours of the premises the Store limits the use which may be made of the ATM by the Bank. We do not think it is appropriate to make any distinction between the normal arrangement where access for all purposes is from within the store and the arrangement at Tesco’s store in Rugby where, for servicing, access to the room which houses the machine itself is from outside the store. 191. An internal site, even one which has been designed or adapted to house an ATM, is likely to be more easily relocated elsewhere in the store than an external hole in the wall site. The space vacated by an internal ATM is also likely to be more readily usable for an alternative purpose (the recess in which the machines at Walsall are housed could equally accommodate the vending machines, display cabinets or recycling bins seen in other photographs of Tesco stores).”
In the Tribunal’s view, those considerations were “sufficient … to justify treating the Store as the party in paramount occupation of the site of an internal ATM” (paragraph 192). The outcome, therefore, was that the appeals relating to the ATM site on the first floor of Tesco’s Nottingham store, where there was “no separate hereditament in the suggested location”, and to the sites of the internal ATMs in its Walsall and Rugby stores were allowed, and the others all dismissed (paragraph 193).[73]Mr Drabble submitted to us that the Tribunal was wrong to conclude that the bank was in paramount occupation of the sites of the external ATMs at the Worcester and Stroud stores. It ought to have concluded that Sainsbury was in rateable occupation of the ATM sites, and that each of these sites was part of the hereditament of the store itself, not a separate hereditament. As it had accepted, Sainsbury occupied the sites, and had not parted with possession of them. The secure room in which the ATMs are housed was controlled by Sainsbury. The operation of the ATMs was governed by reciprocal obligations under the “Squadron” agreement, and the day-to-day activity undertaken by Sainsbury’s staff. These arrangements had been put in place for Sainsbury’s own purposes. An ATM is a normal part of the “retail offer” of its stores such as these. Both Sainsbury and the bank had a commercial interest in providing ATM services, and there was mutual benefit in their doing so. Their purposes were complementary. The Tribunal’s adoption of the concept of “primary” or “dominant” purpose finds no support in Westminster Council v Southern Railway. And in any event the purpose here was “common”. So the question of which party was in “paramount occupation” did not arise. On a correct application of the relevant legal principles, it should have concluded that no separate hereditaments had been carved out of the hereditaments formed by the stores themselves. The facts here were materially different from those in Westminster Council v Southern Railway, where the railway company had had no role, for example, in running the book-selling business of W.H. Smith. The Tribunal should have adopted the approach taken in Stringer – where the contractual and practical arrangements were similar – and with the same result. This would have been consistent with the Scottish cases, including Bank of Ireland, where a different result emerged on different facts.[74]Mr Mould’s submissions for Tesco and the Co-op were to the same effect. The Tribunal had erred in concluding that the banks were in paramount occupation of the external ATMs, and that the sites of external ATMs should be entered as separate hereditaments in the rateable occupation of the banks. The approach indicated in Westminster Council v Southern Railway should have led, on the agreed evidence, to the conclusion that Tesco was the paramount occupier of all the ATM sites in its stores – a conclusion consistent with Stringer. The Tribunal had misdirected itself in applying a test of “primary or dominant purpose”. It was also wrong to put to one side “the occupation of the store as a whole”, wrong to discount the evidence of Tesco’s intentions and actions, and wrong to distinguish between the purpose for which the ATM sites were occupied and Tesco’s “retail purpose”. Tesco had given up neither possession nor actual occupation of the ATM sites. The purpose for which those sites are occupied – the siting and operation of an ATM – was of direct benefit to Tesco, and in each store Tesco retained control over the site to realize that benefit. And the link between Tesco’s occupation and that benefit was demonstrated. In those circumstances, on the approach indicated in Westminster Council v Southern Railway, the question of “paramountcy” was to be resolved by a straightforward application of the principle of “general control”. There was undisputed evidence before the Tribunal that ATMs, including external ATMs, were an essential part of Tesco’s “retail offer”, and that this was reflected in the design of its stores. On the evidence it accepted, it should have concluded that Tesco had retained sufficient control in the relevant sense, and was in rateable occupation of the ATM sites. The Scottish cases do not suggest otherwise.[75]As for the Co-op, the crucial question again, Mr Mould submitted, was whether it had retained sufficient control over the ATM sites. It had. It was the paramount occupier. Even if the Tribunal had been right to apply a test of “primary or dominant purpose” – which it was not – that “primary” purpose, in the light of unchallenged evidence, was the Co-op’s purpose as a retailer, not the Co-operative Bank Plc’s as a bank. ATMs, including those in the external walls of its stores, were part of the Co-op’s “retail offer”.[76]Mr Kolinsky adopted the legal arguments of Mr Drabble and Mr Mould. He submitted that the Tribunal’s analysis in Cardtronics’ appeal was wrong. Cardtronics was not in paramount occupation of the ATM site in the Londis store in Harefield. Physically and functionally, there was no separation between the ATM site and the store. Nor was this a case of separate or competing purposes. Cardtronics’ commercial motive may not have been the same as Londis’, but their purpose was common – to provide ATM services on these premises. Cardtronics was not a “rival” occupier of the ATM site. Closing the store to permit the reloading and servicing of the ATM was not in conflict with Londis’ purpose as a retailer in occupying the ATM site. It demonstrated that purpose. The Tribunal’s approach to “paramountcy” was flawed by its reliance on the fact that some users of the ATM were not also customers in the store, and by its not having considered whether there was nevertheless a “direct link” of the kind referred to in Bank of Ireland.[77]Mr Morshead submitted that in none of these cases was the retailer in paramount occupation of the ATM site. The Tribunal did not follow the approach indicated in Westminster Council v Southern Railway (and exemplified in Old Consort – see the judgment of Lord Patrick at pp.234 to 236, the judgment of Lord Sorn at pp.237 and 238, and the judgment of Lord Guest at p.241). It had substituted a test of “primary purpose” for the test of “control” and “interference”. Indeed, it had rejected “control” and “interference” as relevant considerations. Rather than asking itself whether the retailer’s occupation of the store interfered with the bank’s operation of the ATM site, it had concentrated on the legally irrelevant question of whether operation of the ATM site interfered with the operation of the store. It had adopted the same erroneous understanding of the concept of “rivalry” as the Scottish court in Clydesdale Bank. What matters is the purpose for which the ATM was installed and operated by the bank, because it is the degree of control over – and interference with – this purpose that determines whether the retailer remains in rateable occupation. Only the bank had the right to operate the ATM, for its own business. Its purpose in operating internal ATMs was the same as for those in external walls – to provide the banking services that an ATM does. The distinction between “the object of the activity in question” and “the motive behind the activity”, stressed by Lloyd L.J. in Wimborne District Council v Brayne Construction was critical. To describe the retailer as having a relevant purpose served by the ATM in its store was to confuse the retailer’s motive in permitting the bank to operate the ATM with the purpose for which the ATM is operated by the bank. The Tribunal had not focused on what was being done on the ATM site, but on why it was being done.[78]The Tribunal’s conclusion (in paragraph 185) that “the arrangements between the Store and the Bank provide that only the Bank is to have the right to locate such a machine in the Store”, was, said Mr Morshead, enough to resolve the question required under the simple approach in Westminster Council v Southern Railway. The ATM sites were indistinguishable from the bookstalls and kiosks in that case. Here there was no retention of “general control” by the retailer, or any interference with the purpose for which the site was occupied by the bank. That the retailer’s activity and the bank were symbiotic or “complementary” was irrelevant. Only the banks’ business was being carried on in the ATM sites, facilitated by the retailers. The banks were in exclusive occupation of them, for their own purposes. In every case the bank should have been found to be in rateable occupation. The retailers’ argument, based on the Scottish cases, was hostile to the principle that liability to pay rates, like any other tax, should be decided by the application of “black letter” rules, not by value judgments on questions such as the extent to which one company’s business complements another’s.[79]Mr Morshead also submitted that, in any event, the Tribunal erred in holding that the retailers were in actual occupation of any of the ATM sites in a legally relevant sense. This is a question of law (see the judgment of Buckley L.J. in Case (V.O.) v British Railways Board [1972] R.A. 97, at p.119). The only purpose for which each ATM site can be used is the provision of banking services. The retailer does not provide such services, is not by law permitted to do so, and does not occupy the site for that purpose. It merely provides services to the bank, which are not acts of occupation. Once again, the Tribunal had confused the retailer’s motive for allowing the bank into occupation of the ATM site with the question of whether the retailer remains in occupation for the purpose for which the site is used.[80]In my view, the main submissions made on this issue by Mr Drabble, Mr Mould and Mr Kolinsky are essentially correct, and I prefer them where Mr Morshead’s differ. I am unable to accept that the Tribunal correctly followed the approach prescribed in Westminster Council v Southern Railway, at least in its application to the sites of the ATMs in the external walls of the buildings. And in my view this error of approach led it to conclusions on the question of rateable occupation with which we are entitled to interfere, and should.[81]I see no reason to doubt that, as the Tribunal concluded (in paragraph 169), in each of the cases before it both the retailer and the bank could be regarded as being in actual occupation of the ATM sites, and also that in none of them had the retailer parted with possession of the site. The retailer had only agreed to confer rights on the bank over that small part of the premises because the operation of the ATM furthered its, the retailer’s, own commercial purposes and also provided it with an income. To this extent the Tribunal’s general conclusions were, it seems to me, amply justified by the evidence before it. And they were clearly intended to apply no less to an ATM site located in an external wall of a store than to a site inside the premises. The Tribunal took care to say that the retailer had not “in any of these cases” parted with possession of the site of the ATM. And this seems obviously right. There was no basis for distinguishing between internal and external ATM sites in these conclusions, and the Tribunal did not do so.[82]Those general conclusions on actual occupation and possession are not only sound in law, but also significant for the application of the legal principles relevant to the question of which party was in rateable occupation of the ATM site. So too are the Tribunal’s conclusions (in paragraph 176), that the retailer and the bank should not be regarded as “rivals” in their respective occupation of an ATM site, since they both gain from using the site “for the same purpose” and also share between them the proceeds of doing so, and (in paragraph 177) that the provision of an ATM in a store is “complementary” to the retailer’s purpose.[83]The relevant principles of law are well established, familiar and complete. They do not require to be expanded, qualified or adapted before being put to use in determining the rateable occupation of the sites of ATMs in retail stores. They can be used as they stand. The most basic and enduring among them was expressed by Lord Herschell in Halkyn District Mines Drainage Co. (at p.126) – that where the person in possession of premises has given another person possession of part of those premises he nevertheless remains in rateable possession of that part of the premises unless the other person has exclusive possession. This concept informed the principles stated by Lord Russell of Killowen and Lord Wright M.R. in Westminster City Council v Southern Railway, which remains the guiding authority on the concept of “paramount” occupation. Where actual occupation of land is shared between two persons, the question of who is in rateable occupation makes it necessary to establish which of those two occupiers is in paramount occupation. And in that exercise the parties’ respective rights and purposes in occupying the site are relevant. As this court held in John Laing & Sons and the House of Lords confirmed in London County Council v Wilkins, in ascertaining who is in rateable occupation, it is necessary to discern the “particular purposes of the possessor”. Purpose and motive must not be confused, as Lloyd L.J. emphasized in Wimborne District Council v Brayne Construction (p.239).[84]The basis of the “landlord-control principle”, as described in Westminster Council v Southern Railway, is the concept of the owner of the premises retaining sufficient control of the part of the premises occupied by another party, as well as by himself, to be treated as being in rateable occupation of that part. The degree of control required is not absolute. Lord Russell referred (at p.530) to “general control” over the occupied parts, which is plainly not the same thing as total control. If the landlord retains “no control”, he will not be treated as being in rateable occupation (ibid.). Lord Wright referred (at p.561) to the landlord retaining “control for purposes of his business of the whole house”. Throughout the reasoning in both speeches the theme of “control” prevails. How much control is required will, of course, differ from one case to the next. As Lord Russell said, this will always depend on the particular facts of the case in hand. In every case “the degree of control must be examined” (p.532). It seems clear, however, that if less than total control can suffice, it will be possible, in a case of concurrent occupation by two occupiers, for both to retain a degree of autonomy in their occupation without the owner of the premises being deprived of “general control”.[85]Given the Tribunal’s conclusions that in each instance here the retailer had not parted with possession of the ATM site and had remained in occupation of it, sharing actual occupation with the bank, that retailer and bank were not “rivals in their occupation” of the site, and that they were using it “for the same purpose”, one must consider whether, on a proper application of the relevant principles, there was any justification for concluding that the bank rather than the retailer was in rateable occupation of it as the paramount occupier. I am not persuaded that there was.[86]I am unable to reconcile the Tribunal’s approach, at least to the external ATMs, with that adopted by the Scottish court – albeit in different factual situations, which may, in a particular case, justify a different outcome on the question of rateable occupation. In particular, I cannot agree with the Tribunal’s conclusion (in paragraph 167) that in the Scottish cases one can discern, on the part of the court, an intentional modification, or refinement, to the approach indicated in Westminster Council v Southern Railway, to address the situation in which the two occupiers are “not truly rivals but are both deriving a direct benefit from the same use” of the site in question. In such a case it may not be realistic to refer to an “interference” by the “owner” with the “enjoyment by the occupant of the premises in his possession for the purposes for which he occupies them” or an “inconsistency” with his “enjoyment of them to the substantial exclusion of all other persons” – as Lord Russell contemplated in Westminster Council v Southern Railway (at p.532). But the absence of such “interference” or “inconsistency” does not have the effect of disapplying the principle of “general control”, or call for it to be modified or refined by the introduction of a test of “dominant” or “primary” purpose such as the Tribunal seems to have applied here. Nor can I accept the Tribunal’s conclusion (in paragraph 171) that, when applying the approach in Westminster Council v Southern Railway to external ATM sites, “there is no need to consider the occupation of the store as a whole”. This does not seem a true reflection of relevant principle.[87]I think there is force in the submission of Mr Drabble, Mr Mould and Mr Kolinsky that where the “owner” has given up neither possession nor actual occupation of the site in question, where the purpose for which that site is occupied – in this instance, the operation of an ATM – is a common purpose with that of the other party in occupation and is of direct benefit to the “owner”, and where the “owner” retains physical or contractual control over the site to realize that benefit and this can be demonstrated by objective evidence, the principle of “general control” applies, in the normal way. Rateable occupation is not resolved in such a case by weighing one party’s “purpose” against another’s. “General control” remains the decisive factor in establishing who is in rateable occupation of the site. There is no need for a further test to be imposed to gauge which of two purposes is the “dominant” or “primary” purpose, or for the “general control” principle to be subordinated or made subject to such an enquiry. Such a test is not prescribed in the jurisprudence. And in my view it is neither necessary nor appropriate to resort to it as a means of resolving the question of rateable occupation. This is not to relegate the consideration of “purpose” to irrelevance. It is simply to recognize that in the approach indicated by the House of Lords in Westminster Council v Southern Railway, the critical question is ultimately, in every case, the question of “general control”, not one of “dominant” or “primary” purpose. Does the “owner” retain “general control” over the site, or not? That is the question to be answered. Here, therefore, on the Tribunal’s findings of fact on largely uncontroversial evidence, it had to consider, in each of these cases, whether the retailer as “owner” had retained sufficient “control” over the ATM site to be treated as being in rateable occupation.[88]On a straightforward application of the “general control” principle, in the light of the facts the Tribunal accepted, the correct answer seems to me to have been that the retailer, as “owner”, had in all these cases – both internal and external ATM sites – retained sufficient control of the site, in contractual, physical and functional terms, to be regarded as being in rateable occupation of it. This conclusion would not be defeated even if the respective purposes of the retailer and the bank in occupying the ATM site were not identical in terms of the “target market”. In applying its test of “primary purpose”, the Tribunal drew a distinction between the retailer’s purpose in operating the store and the bank’s in operating the ATM. It did so (in paragraph 179) by having regard to the “target market” of an external ATM, which, it said, was “much broader than the retail customers of the store”, the fact that the location of the site was, it said, “physically remote” from the “retail offer of the store”, and the “distinctively different” purpose of such an ATM, which, it said, was “to reach as wide a market for ATM services as possible”, and not merely members of the public who were customers of the store. Those commercial objectives, assuming they were correctly identified by the Tribunal, do not reflect any difference in the essential purpose of siting an ATM in a retail store, as between one that is located inside the store and one that is placed in an outside wall. This much is implicit in the Tribunal’s own conclusions (in paragraph 176) that retailer and bank use the ATM site “for the same purpose”, and (in paragraph 185) that the bank and the retailer “have a mutual interest in providing ATM services, and both derive a benefit from the presence of the machines”. None of this serves to demonstrate that the retailers had ceded “general control” of the ATM sites to the banks.[89]I do not see any tension between that analysis and the approach adopted by the Scottish courts in the ATM cases. In Bank of Ireland the court concentrated on the question of whether the post office retained control over the ATM site. This is clear in the judgment of Lord Malcolm (in particular, in paragraph 27). He did not merely accept that the ATM in that case was a “separate service” from those offered to customers of the post office, and provided to a “different customer base”. Crucially, he also saw a “sufficient disconnection” between the use of the ATM site and the post office premises to indicate “independent uncontrolled occupation” of it by the bank for the purposes of its business. Lord Gill (at paragraph 15) emphasized the absence of any “direct link” between the ATM site and the operation of the post office, as well as the fact that it was not “one of the retail attractions” provided for the customers of the post office. Those observations, as I read them, show that the court was applying – not departing from, or adjusting – the principles in Westminster Council v Southern Railway. It was, in effect, concentrating on the question of “general control” as the decisive consideration. The conclusions it reached on that question were based on materially different factual findings from those made by the Tribunal in this case, including those in paragraphs 169 and 176 of its decision.[90]In my view, therefore, the approach of the court in the Scottish cases was not in error, as Mr Morshead submitted, and does not justify either the approach taken by the Tribunal here, or its conclusions on the rateable occupation of the external ATMs. The Scottish court did not depart from the principles established in previous authority, including the “general control” principle in Westminster Council v Southern Railway. It reached the conclusions it did in Clydesdale Bank and Bank of Ireland by applying those principles to the particular facts of the cases before it.[91]The same may be said of the Lands Tribunal’s decision in Stringer, from which the Tribunal consciously departed (in paragraph 188), but which also seems to me to be consistent in its approach with the principles in the relevant authorities, and not materially distinguishable on its facts from the cases now before us.[92]Here, the undisputed evidence before the Tribunal, which it appears to have accepted, was that the ATMs the retailers had chosen to have sited in their stores, whether inside the store or in an external wall, enhanced their store’s “retail offer” by adding to the range of services available at the store; that some at least of the stores had been either designed to accommodate an ATM or physically adapted to accommodate it; that access to the ATM for regular servicing, maintenance and loading could only be achieved from within the store, and with the retailer’s co-operation or consent; and that in some, if not all cases, the retailer’s staff were involved in those tasks. This was, in every case, more than merely a symbiotic relationship between retailer and bank. Not only was the ATM site physically incorporated into the store premises, it was also functionally integrated with the operation of the store. Both with the internal and with the external ATM sites, the bank relied on the active assistance of the retailer in operating its ATM from that site. This was so even in the case of an external ATM available for use by the general public outside the store’s opening hours. Without that assistance, the bank would not be able to operate an ATM on the site, for the purposes of its own business. Taken together, these considerations are, in my opinion, strong indicia of the retailer having retained, in the relevant sense, “general control” over the ATM site, whether located inside the store or in an outside wall. They were all demonstrated by objective evidence, not bare subjective assertion.[93]The Tribunal accepted, rightly in my view, that “[both] parties” – bank and retailer – “derive a direct benefit from the use of the [ATM] site for the same purpose, and share the economic fruits of the specific activity for which the space is used” (paragraph 176). That conclusion is, I think, unassailable. The purpose of both bank and retailer, properly distinguished from motive, is aptly described as a “common purpose” – to provide ATM services to the public at a retail store. That purpose, if fulfilled, is to their mutual advantage. They had co-operated to provide an ATM on each of these sites. The bank’s shared occupation of the ATM site with the retailer, the retailer’s continued possession of that site, and the fact that the retailer had, as the Tribunal also found (in paragraph 169), agreed to restrict its own use of the site “because the presence of the ATM furthers its own general business purposes …”, are in stark contrast with the “independent uncontrolled occupation … by the bank for the purposes of the bank’s business” and the absence of a “direct link”, as was found to be so on different facts in Bank of Ireland. There, on the evidence, the synergy one sees in this case was lacking. Here the retailers remained in occupation and possession of the ATM sites in their stores, the banks had not been given exclusive possession for their own purposes, the relevant purpose of the banks and the retailers was the same, and the retailers had retained “general control” over those sites in the relevant sense. That analysis is a reflection of the approach in Westminster Council v Southern Railway applied to the relevant facts, not a departure from it.[94]For all those reasons, I cannot agree with the Tribunal’s reasoning towards its conclusion (in paragraphs 185 and 186) that the sites of the external ATMs ought to be entered as separate hereditaments in the occupation of the bank, and its consequent conclusions (in paragraph 187) on the individual sites. Its approach was not, in my view, consistent with relevant authority, including the House of Lords’ decision in Westminster Council v Southern Railway.[95]The Tribunal considered the issue of the rateable occupation of the internally located ATM sites (in paragraphs 190 to 192) without referring to the question of “dominant” or “primary” purpose. It did refer to the “purpose” of the bank’s occupation of the internal sites being “to provide a service to the Store’s customers”, which it equated to the “purpose” of the retailer’s occupation of the whole of the store premises, including the ATM site. On this basis it was able to conclude that in these cases the retailer, not the bank, was in paramount occupation. It is not entirely clear whether in coming to that conclusion it applied the “general control” principle, unmodified, and without recourse to any different or additional test. However, its conclusion is consistent with the application of that approach, and does not seem to have turned on the question of which occupier’s purpose was “dominant” or “primary”. Among the considerations it mentioned, it clearly gave some weight to the retailer’s “control of the opening hours of the premises”, by which it “limits the use which may be made of the ATM by the Bank”. This will of course be a further practical restriction on the bank’s occupation of the site, in addition to those affecting its occupation of an external site. But the Tribunal did not say it was a consideration that made a decisive difference to its conclusion on paramount occupation in the case of an internal site, and I do not think it would be right to assume this was the view it took.[96]Viewed on its own, I do not think this part of the Tribunal’s reasoning is open to the same criticism as its treatment of the external ATM sites. Its conclusions are not inconsistent with the court’s approach and the outcome in either of the Scottish cases or the decision of the Lands Tribunal in Stringer. They do not, however, validate the approach it took to the external sites. They do not justify the conclusion that, where the retailer’s and the bank’s purposes in providing an ATM in a store are either the same or closely aligned, the retailer remains in occupation and possession of the ATM site, and the contractual, physical and functional arrangements are as they were here, an internal ATM site is in the paramount occupation of the retailer, but an external site is in the paramount occupation of the bank. In short, it is not clear from the Tribunal’s decision how the application of the principles established in the authorities, including the principle of “general control” in Westminster Council v Southern Railway, could properly lead to that result.[97]It seems to me, therefore, the Tribunal did err in its approach to the issue of rateable occupation. Had it not done so, I cannot see how it could have concluded, on the evidence before it, that in the case of any of these ATM sites, internal or external, the bank was in rateable occupation of the site as paramount occupier, the retailer having failed to retain sufficient control over the site to prevent a separate hereditament being formed. It follows that, in my view, none of the alterations to the rating list should have been made by the Valuation Officers, or sustained on appeal to the VTE and, in turn, the Tribunal.

Conclusion

[98]For the reasons I have given, I would allow the appeals of Sainsbury, Tesco, the Co-op and Cardtronics, and dismiss that of the Valuation Officers. Lady Justice King[99]I agree. Lady Justice Gloster, Vice President of the Court of Appeal, Civil Division[100]I also agree.

Cited in 5 later judgments