United Kingdom VAT & Duties Tribunals Decisions

You are here: BAILII >> Databases >> United Kingdom VAT & Duties Tribunals Decisions >> Optika Ltd v Customs and Excise [2004] UKVAT V18627 (24 May 2004)
URL: http://www.bailii.org/uk/cases/UKVAT/2004/V18627.html
Cite as: [2004] UKVAT V18627

[New search] [Printable RTF version] [Help]

Optika Ltd v Customs and Excise [2004] UK V18627 (24 May 2004)

INPUT TAX – Partial exemption – Special method – Optician making taxable supplies of goods and exempt supplies of services – Principal business supplying prescription spectacles – Applications for approval of special methods of attribution of residual input tax – Whether special methods resulted in fair and reasonable recovery – Floor-space based special method – Whether "zoning" of floor space for rent and rating purposes applicable to attribution of residual tax – EC Sixth Dir, Arts 17.5, 19 – VAT Regs 1995, rr 101,102

JURISDICTION – Appeal relating to special methods of attribution – Whether Tribunal has appellate or supervisory jurisdiction

LONDON TRIBUNAL CENTRE

OPTIKA LIMITED Appellant

and –

THE COMMISSIONERS OF CUSTOMS AND EXCISE Respondents

Tribunal: ANGUS NICOL (Chairman)

KENNETH MANTERFIELD FCA

Sitting in public in London on 8-12 December 2003

Julian Ghosh and James Henderson, both of counsel, instructed by Grant Thornton, chartered accountants, for the Appellant

Owain Thomas, counsel, instructed by the Solicitor for the Customs and Excise, for the Respondents

© CROWN COPYRIGHT 2004

DECISION

This appeal was originally listed under the name The Optika Clulow Group, and reached this Tribunal under that name. We were informed at the hearing that the Appellant has changed its name, and is now known simply as Optika Ltd. We therefore direct that the title of these appeals be renamed accordingly.

There are two appeals. The decision of the Commissioners the subject of the first appeal was given on 25 October 2000. The Commissioners refused approval for the Appellant to use a partial exemption method, under regulation 102 of the Value Added Tax Regulations 1995 ("the Regulations") other than the standard partial exemption method prescribed by regulation 101. We will refer to this as "the First Special Method". The First Special Method was set out in a letter of proposal dated 22 September 1997 (see paragraph 24 below). The Appellant termed it "a purchase method". Summing it up very briefly at this point, the method included the apportionment of input tax by reference to the floor space of the shops, dividing the use of the shops into taxable, exempt, and mixed use; measurement of the floor space involved was capable of being measured accurately and could be banded in accordance with zoned rental valuations. The Commissioners took the view that the general reception areas of the shop could be used for dispensing and that the taxable area would be minimal, the majority being mixed or exempt. They therefore rejected the First Special Method, in a letter of 31 August 1999, and again on a number of later dates when the Appellant had endeavoured to convince the Commissioners that the First Special Method was fair and accurate, and finally in a letter dated 25 October 2000.

The second appeal relates to the rejection of a further method of apportioning input tax referable to property, described by the Appellant as a floor space method, which was proposed, and approval in principle sought in May 2001. The Commissioners replied that a complete new proposal would have to be made, and would be considered anew. Accordingly, "a separate and entirely different proposed special method ('the floorspace method')" was put forward in a letter of 19 June 2001 (see paragraph 36 below). The proposal was accompanied by extensive measurements and calculations carried out by a surveyor, shewing, the Appellant contended, that it was possible to calculate precisely an apportionment between taxable and exempt areas of the shops according to value. We refer to this as "the Second Special Method". Putting it very briefly again, the Commissioners rejected this method also, for the principal, but not only, reason that the retail area of the shops was used for making both taxable and exempt supplies, and that therefore only a relatively small part of the retail area was used exclusively for taxable activities.

It was conceded by the Commissioners that the sale of spectacle frames was an integral part of the overall supply of spectacles, and that the price of the spectacles included payment of the cost of dispensing services.

The Appellant's evidence

The first day of the appeal was spent in visiting the Appellant's premises at Bluewater, a vast and labyrinthine retail centre situated near Greenhithe in Kent. We spent some time there, inspecting the shop itself and hearing descriptions of the way in which the space was used. The shop is fairly long, and narrow, and was described in some detail by witnesses. Whilst we were there one or two customers entered the shop, but it was difficult to observe precisely what went on in respect of each of them, save that they spent some time looking at the frames exhibited for that purpose.

All the witnesses called by the Appellant had made witness statements, and these largely stood as their evidence-in-chief. The first to give evidence was Stephen Kane. He is a surveyor by profession, and principal of the firm that bears his name. Since 1971, he said, he has specialised in retail premises. For the last nine years he has been the sole retained property adviser to the Appellant, and is a non-executive director. He has also been involved in the Appellants' marketing and in the development of "the David Clulow concept". He said that the Bluewater shop was typical of Optika shops, with a few slight differences.

Mr Kane described the Bluewater shop. He said that it is rectangular and has an area of approximately 1,100 square feet, of which the front, retail area, that is from the front of the shop to the counter, was some 500 to 600 square feet. Generally the retail area of an Optika shop comprises between two thirds and three quarters of the floor space. In the Bluewater shop, the retail area contains displays of frames and sun-glasses. We observed that some of the frames are shewn as having been designed by well-known fashion designers. The frames are displayed in units all along the wall on both sides, and in one or more glass-topped tables. Inward of the retail area is the reception area which has a counter and the till, and which provides a natural break in the layout of the shop. Behind the reception area lie the consulting rooms, a customer waiting area, eye testing space, storage, and ancillary staff accommodation. Mr Kane said that some of the shops not in shopping centres have less regular shapes.

Mr Kane said that it was crucial that the retail area of any Optika shop should have a minimum size of 500 to 600 square feet, so that customers might have a reasonable amount of space in which to view the products in comfort and confidence. He said that the Appellant was less interested in selling eye tests than in selling spectacles. The critical thing was that the customer should be attracted to buying spectacle frames, the presentation of which must give the customer confidence that he is buying from high quality opticians.

Occupational costs, Mr Kane said, should not exceed about 20 per cent of gross turnover. Those costs included rent, rates, and service charges, of which rent and rates comprised about 90 per cent. Zoning was, he said, fundamental to the calculation of both rent and rates. He explained that the front of any shop was the most valuable part, and the value decreased as the distance from the front increased. The overwhelming proportion of the occupational costs is incurred on the retail area of an Optika shop. Zone A is the first 20 feet from the shop front. Zone B is the next 20 feet. If the value of zone A is £x, that of zone B is £x/2, and that of zone C is £x/4. Beyond zone C it was usually accepted that there is a remainder zone, whatever its area. For the purposes of rent, it was customary to express the floor area of any premises "in terms of zone A" (commonly abbreviated to ITZA); that figure multiplied by the zone A rate produced the basic rent. In nearly all the Optika shops the first two zones were used for the retail sales, the space behind that being used for testing and consulting (the "quasi medical purposes", as Mr Kane termed them). Mr Kane set out, by way of illustration, an example of a notional shop unit with a width of 20 feet and depth of 50 feet, which shewed that its physical area was 1,000 square feet, and ITZA 650 square feet. This approach to floor area was recommended by the Royal Institution of Chartered surveyors and is standard valuation practice, though Mr Kane added that when a lease is being acquired for the first time the overall rent will be negotiated rather than the zone A rate. However, the landlord would base the level of rent on zone A rates in order to remain consistent with the zone A rates for nearby premises.

Mr Kane described the normal course of events when a customer comes to purchase spectacles. When the customer enters the shop he should be approached by a member of the sales staff, and asked if he can be assisted. The customer may have an appointment with the optician, in which case he would be escorted to the counter, the appointment would be checked in the diary, and the customer would be handed over to the optician. If the customer wished to make an appointment for an eye test, he would again be escorted to the counter, and the appointment would be made. If the optician was not busy, the appointment might be made for there and then. If not, the customer would be asked to return at the appointment time, but might also be shewn a selection of frames in the retail area. In the case of customers who do not feel that they need a test, or who have already had a test and have a prescription, or just want to look at frames, the sales assistant is there to help and deal with the requirement. A customer might wish sun-glasses or some other accessory, in which case the sales assistant would be there to help. A customer might also have come simply to collect spectacles or lenses already ordered.

The eye test, Mr Kane said, was a means of selling spectacles, but it is not the way in which a retail optician makes its money. The retail area is designed to attract the maximum possible number of customers in to buy spectacles. It was necessary to drive retail sales in order to pay the rent. The landlord would not allow the Appellant to have, for example, a consulting room in the front of the shop: they were obliged to have a retail area. At Bluewater Zone A was very important to the developer, and was not allowed to drop below a certain level.

We were referred by Mr Kane to a standard text-book, Modern Methods of Valuation. On page 360 of that book, in the section headed "Shops" it establishes that the practice of zoning had been in use for many years, and is applicable equally to rating and rent. The purpose was to achieve uniformity. He said that there were no Optika shops which consisted of Zone A only: that would not be possible for an optician's shop.

Mr Colin Allen is the Finance Director of Optika, and has been since November 1997, prior to which he was chief accountant. His function is, principally, dealing with all the Appellant's financial affairs, but also overlaps with the other directors. He said that he visited all the shops regularly. There are 36 of these, of which 31 trade as David Clulow, four as Optika, and one as Harrods Opticians. Having confirmed the dimensions of the shop at Bluewater given by Mr Kane, he repeated that the purpose of having an open shop front was to encourage potential customers to look a the displays of spectacles and sun-glasses. The overall look of the front of the shop, he said, is crucial, since most of the Appellant's business came from people walking from the street with no prior appointment asking for a test. He said that in the retail area there were portable spinning racks of sun-glasses. The proportion of sales of sun-glasses to overall sales for the two previous years was between 10 and 15 per cent.

Mr Allen described the area beyond the counter as the "pre-screening area", and it contained a field scanner, for measuring peripheral vision, and an auto-refractor, which was used to measure a patient's prescription requirements. That part of the shop also included what he termed a "contact lens teach area", with a table and chairs, a looking-glass and a small sink. There were also chairs for customers waiting for eye tests. Beyond that again were two sight examination rooms; some shops have only one, and in some shops these are located in a basement or on the first floor, where, Mr Allen said, rents are even lower than at the back of the ground floor.

When the optometrist has concluded his tests, Mr Allen said, he writes a report and a suggested prescription, and he then hands the customer over to the sales staff. A sales assistant will take the customer to the retail area to select a frame. Mr Allen said that the Appellant's belief was that customers come to the shops because the Appellant is a seller of fashion frames. Once the frame has been selected, an order is written out, this usually being done at the seated area or perhaps at the table in the retail area. The customer is then taken to the counter to make payment. If the job was properly done, a member of staff should be with the customer until his departure, though very occasionally a customer did not wish any help. Occasionally a customer would buy a frame only: this was in about 5 to 13 per cent of cases. It was only rarely that there were queries by customers about lenses, because these would be dealt with by the optometrist. However, a customer might talk to sales staff about the look of the lenses. Mr Allen said that the optometrist has to be qualified. Of the shop managers, 13 out of the 29 were qualified dispensing opticians. All staff would have selling experience, but all undergo training sufficient to enable them to carry out the majority of sales on their own. No element of taxable supplies was provided by unqualified employees who are not under supervision. The staff are able to dispense, but they could always contact an optometrist, even on the telephone. The optician did not need to come out beyond the counter, though occasionally he might. Any query that the manager could not answer would be referred to the optometrist, but invariably the manager would go and ask him rather than the optometrist coming out. There were rare occasions when a customer had an eye test and then made no purchase, either because he needed no prescription or because he did not find anything which he wanted to buy.

The bulk of the Appellant's profits, about 70 per cent, came from the sale of spectacles, Mr Allen said. The front of the shop was intended to shew frames. Most customers, however, would not expect to leave the shop without lenses as well. Sight tests were offered as part of the package. The normal price of an eye test was £20, but the optometrists were very highly paid, and the Appellant would lose money if it only did eye tests; it was therefore necessary to make a profit on the frames. The frames on display were fitted with tags which caused an alarm to sound if one were taken out of the shop with the tag still on. Posters were used to entice customers in. The Christmas posters (which were up at the time of out visit) advertised free sight tests and reduced prices. More general posters were used to remind the public that eye tests were available, and also sun-glasses. Some posters were permanently stuck on the window at Bluewater. There were sale posters twice a year, and a general poster promoting a brand from time to time. These involved no change to the retail area or its function. The complete package, that is to say a complete pair of spectacles was what drew the customer in, but the Appellant hoped to sell high fashion frames. The function of the retail area was to attract customers in to buy "designer brand" spectacles or sun-glasses.

Mr Allen said that the price of spectacles was apportioned, 41 per cent being for the taxable supply of frames and lenses, and 59 per cent for the exempt supply of dispensing. In the case of contact lenses the ratio was 50.4 per cent taxable and 49.6 per cent exempt. Between 5 and 13 per cent bought only frames. Between 15 and 50 per cent (in summer) came in for sun-glasses, and between 15 and 20 during the rest of the year. Mr Allen was handed a document which shewed an analysis of the proportion of standard-rated sales in all the 29 shops during the six months to 31 June 2001. This was divided between sun-glasses and accessories. The percentage of sun-glasses varied between 5 per cent and 32 per cent, most of them being between 10 and 16 per cent, and only seven below 10 per cent. Sales of accessories were much lower: zero in nine of the shops, 2 per cent in one of them, and 1 per cent in the remainder. The analysis did not include those shops which only sold sun-glasses. That analysis, Mr Allen said, referred to two years ago, and there had been a considerable increase since then.

Mr Allen was referred to a letter from Mr Laney, a senior officer of Customs and Excise, dated 25 August 1998, in which he said,

"I had suggested to you that I considered that as spectacle frames were separately priced at the branches of Optika, I saw no reason why VAT could not be accounted for them as a separate supply. Our H.Q. Supply of goods branch have accepted your argument that the frames are an integral part of the overall supply of spectacles the payment of which also covers the cost of dispensing."

Mr Allen agreed that that was an argument that the Appellant had put forward.

A letter from Grant Thornton, the Appellant's accountants, to Mr Laney dated 19 June 2001 referred to the writer, Mr Marcus Ward, and Mr Allen having visited a number of Optika shops and taken measurements, for the purpose of measuring accurately the areas of the shops which were used for specific purposes and allocating those areas to the appropriate zones. Mr Allen said that he had been to two shops, those at Bluewater and Cheapside. Measurements had been read from scale drawings, and had been checked with a tape. No other measurements had been taken.

Mr Allen was referred to a file note of a meeting at New King's Beam House on 16 July 2001 at which he had been present. A note to this document related that,

(b) Subsequently, [Mr Allen] stated that sunglass sales - by value, represented 20% and 45% of sales. We, and C&E, were not aware of this fact during the meeting and this could affect Customs' view."

Mr Allen agreed that on the list referred to (in paragraph 17 above) the largest was 32 per cent. The figures which he had stated at the meeting would have been from the top of his head and from his experience of the trade. He could not remember where he had got the figures, but, he said, it could only have been from the records at head office. He added that he did not think that the figures for sun-glasses included prescription sun-glasses, which was a very low proportion of sun-glasses.

The bulk of the Appellant's taxable inputs were purchases of raw materials. The main capital expenditure was on shop fitting, including machines for testing. These were mostly obtained on a lease or with a bank loan. Monthly profit and loss accounts were produced, and daily, weekly, and monthly sales analyses. The tills with electronic point of sale (EPOS) enabled these to be in considerable detail.

The first proposed special method

The evidence relating to the First Special Method was contained largely in the voluminous correspondence which passed between Grant Thornton, who acted for the Appellant, and a number of officers of Customs and Excise. Of those officers, two, Mr Keith Laney and Mr Philip Mattacks, also made witness statements and gave oral evidence.

Mr Laney said that he had been involved with the Appellant from about September 1997, in negotiations for a new method of apportioning sales between the standard rated supply of spectacles and the exempt supply of dispensing services. There was no agreement as to apportionment at that stage, and a method was imposed upon the Appellant based upon the full cost of making the supply of spectacles and of dispensing, which gave an apportionment of 41.19 per cent taxable supplies and 58.81 per cent exempt supplies. None of the latter related to contact lenses, for which the apportionment was 50.4 per cent taxable and 49.6 per cent exempt. That apportionment was not appealed.

Mr Laney produced three further documents. The first was a report of a visit by him to the Bluewater shop on 24 July 2001, the second was his conclusions on the floor-based method of apportionment proposed by the Appellant and the third the report of a meeting on 16 July 2001.

There was also a report of a visit by Mr Laney to the Ealing branch of the Appellant, which took place on 14 July 1998. He had walked round the whole of the premises, and had observed that spectacles were being sold in the retail area. He had spoken to the manager, and had asked to be taken through the normal routine for selling. The visit had lasted between 1½ and 2 hours. On that visit no agreement was reached as to the VAT treatment of sales of frames and lenses. Mr Laney asked if the Appellant had any written proposal for a special method. This had not yet been done. Mr Laney said that it had occurred to him afterwards "that what was a fair method before should be a fair method after, despite what result was concluded on the outputs. Clearly there is some dispute over what constitutes fair and reasonable which the trader appears to view as what gives them the best financial result."

The visit to the Bluewater shop was undertaken in order to confirm measurements, establish where possible, the exclusively taxable and exempt areas, and to record any other pertinent information. Mr Laney described the outside of the Bluewater shop, mentioning that it had an extra protruding display window, about 1½ feet deep by 3 feet wide, which displayed posters at the time of his visit which offered 50 per cent off the price of designer spectacle frames. He therefore considered that to a mixed area. As to the interior, his report ran thus:

"INTERIOR - MAIN RETAIL

Measured the zoned areas and confirmed that they were as indicated by the trader. In this particular store I noted that the flooring and lighting materials and style continued into zone B to a distance of 8 feet to the counter plus a further few feet to the side of the counter although beyond that the flooring material changed from wood laminate to a mixture of carpeting and linoleum (mainly carpet).

Of particular note was the reception desk which fell into zone B was a curved unit of wood and glass construction presumably a relatively high value item.

The public retail area is symmetrical in layout with in the central area two sets of standard display cabinet measuring 28" x 18" with a small wooden table abutting them table was the same size and had a chair either side which the customers were using whilst waiting to be seen etc. The two display units were displaying spectacles so were taxable areas but the two tables related to mixed."

Mr Laney gave the dimensions of the spectacle display racks which ran along the walls each side, and said that each did have some form of dedicated lighting either wall mounted or recessed in the ceiling. He mentioned also three other display cabinets displaying spectacle frames. He continued:

"COUNTER AREA

Clearly mixed. There are also drawer and cupboard units behind and to the side which contained paperwork, stationery etc. Recessed into the wall is an area where contact lens fluids are displayed but this is above the drawers so occupies the same floorspace and so cannot be considered exclusively taxable.

PRACTICE AREA

The area immediately behind the reception counter is used for:

A) waiting for eye examination

B) pre-test area - some basic eye examination checks and;

C) practice area for putting in contact lenses ie dispensing services.

To sum up in my view this whole area relates to exempt services. The area is carpeted and lighting 'appears' to be of a slightly less expensive type than that of the retail area."

Mr Laney gave a brief description of the eye examination rooms; the laboratory, as to which he was unsure whether lens storage there related to taxable or exempt supplies; a storage cupboard; a disabled w.c., of which he said that it could only be a mixed area; and a corridor. Under "Other Observations", he noted:

"A number of customers used the seats in the retail area whilst awaiting to arrange eye examination tests or discussing frames with the staff. Also saw on two occasions customers in consultation with staff standing literally in the area immediately next to the display units with the staff getting customers to try the frames to size etc."

In his oral evidence he added that he had seen people waiting in the retail area to arrange for eye tests. The manager had told him that. There were others who were buying spectacles. He said that he had not questioned the answers he was given, nor had he asked if what he had seen was typical. He had asked no questions about staff qualifications.

Mr Laney said that no difficulty arose in dealing with input tax directly attributed to either taxable or exempt supplies, apportionment was necessary for residual input tax, which generally related to costs which are overheads of the business. He added that since the decision in Customs and Excise Commissioners v Leightons Ltd [1995] STC 458 that included the costs incurred in refitting the retail area of an optician's premises. In the absence of a special method of apportionment, the standard method was required by statute to be used: regulation 101. Special methods, which may be approved by the Commissioners if the standard method does not give a fair and reasonable result, are usually devised by individual businesses to fit their own unique circumstances. The Commissioners may approve such a method only if they are satisfied that it provides a fair and reasonable recovery given the extent to which the residual input tax bearing costs are used to make taxable supplies.

The Appellant's First Special Method, Mr Laney said, sought to attribute the residual input tax incurred in relation to the shops on the basis of an analysis of the use of the floor space, the value of the floor space being weighted in favour of the front areas of the shops, calculated on the basis of rental valuations given by an independent valuer or surveyor. Mr Laney's concern was that the floor space in the retail areas of the shops was not exclusively attributable to either taxable or to exempt supplies, being related to both. It was the Mr Ward's letter of 26 October 1999, outlining proposals for a special method based on the use to which the properties were put, claiming that higher rents were paid for shop front areas, and mentioning the rental calculations that led Mr Laney to misunderstand the nature of the proposed special method, and to think that it was floor based.

In a letter to Mr Kirk of Customs and Excise, dated 23 October 1996, following earlier correspondence, Mr Ward said that the standard method for calculating exempt input tax, which had been used by the Appellant until that time, gave a distorted result. It was suggested that a floor space method might be used. The letter included the following:

"(ii) You assert that an output based method gives a fair and reasonable result. This does not, of course, preclude other methods being considered, and the very existence of special methods indicates that the standard method does not always provide fair and reasonable apportionment.

(iii) You may be aware that a floor space method of apportioning residual input tax is used by many taxpayers, not just opticians. It is not possible to identify space used exclusively for making taxable supplies and this was the reason why the floor space method was calculated in the way in which it was. It should be noted that whilst it is possible to identify floor space used solely for exempt purposes, of the residual floor space, the majority is used for taxable purposes, namely: displaying spectacles and contact lenses, storing spectacles and contact lenses, displaying and selling accessories, and advertising spectacles, contact lenses and accessories. The remainder of the residual floor space is used for both taxable and exempt purposes, namely: reception, record keeping and customer payments. As a consequence, the calculation of exempt floor space is actually biased towards the exempt side."

The letter also asked the Commissioners to accept that 61 per cent of sales of frames and lenses and 49.6 per cent of sales of contact lenses were exempt.

The response, dated 8 November 1996, to that letter came from Mr Donohoe, who said, inter alia,

"I have received your partial exemption proposals for recovery of the 'pot' items. I have no objection to the floor space method in principle as long as it is calculated on the basis of taxable floor area divided by total floor area. However, I agree with Colin Kirk that the floor space method that you propose is not a fair and a reasonable one because there would be a high amount of input tax directly attributable to the taxable supplies but very little directly attributable to the exempt supplies. The resulting calculation would unfairly allow input tax recovery of most of the 'pot' but in reality many of these items are used, at least in part, in making exempt supplies."

He expressed the view that the standard method did split the 'pot' items fairly and reasonably.

The First Special Method was set out in a letter from Mr Ward dated 22 September 1997 addressed to Mr Rishi at the Wembley VAT Office. The letter stated that the Appellant had apportioned its supplies in the percentages referred to in the previous letter, adding that that apportionment had been agreed retrospectively with the Commissioners. The special method is then set out in the following terms:

"Step 1

Input tax which is directly attributable to exempt supplies is identified (eg equipment used solely for eye tests). This input tax will not be claimed.

Step 2

Input tax which is directly attributable to taxable supplies is identified (eg purchases of sunglasses). This input tax will be reclaimed in full.

Step 3

Input tax which is directly attributable to properties (the shops) is identified (eg rent and shop fitting). This input tax will be apportioned by reference to the floor space of the shops. A split will be calculated between taxable/exempt/mixed use of the shop ie:

(a) exempt - eg eye examination rooms

(b) taxable - eg goods display/storage

(c) mixed - eg reception/waiting area

Measurement of the floor space can be made accurately and be 'banded' (ie zoned rental valuations attributed) by an independent surveyor/valuer.

As a consequence of this exercise expenditure on the shops may accurately be attributed to the taxable/exempt use of those shops even to the extent that the floor space measurement will be 'weighted' for the appropriate value. The calculation will thus be:

a = input tax not recoverable

b = input tax fully recoverable

c = mixed input tax to be apportioned as follows

b x c = d (recoverable element of c)

a + b

Total recovery = b + d

Step 4

The remaining non-attributable input tax will be apportioned using a sales based method, namely:

Taxable supplies x non-attributable input tax = recoverable element

Total supplies

I consider that this method is fair and reasonable and produces a result which is accurate, easily verifiable by your department, and is dependent upon independent third party valuations of Optika's properties."

In a letter of 31 August 1999, referring to an earlier letter dated 25 August 1998, Mr Laney referred to an issue relating to apportionment of spectacle supply, which was clearly still a matter of debate, and was separate from the matter of the special method. He stated the view that the general reception areas of the shop could be used for dispensing and therefore the taxable area would be minimal and the majority of the public area would be of mixed use or, as in the case of examination rooms, exempt. He asked Mr Ward to advise him within four weeks whether he wished to apply the special method proposed, and if so to say how it was intended to establish the shop areas which were taxable.

Mr Ward's detailed reply was dated 26 October 1999. Under the heading "Background", Mr Ward said,

"Whilst it is accepted that the standard method of apportioning input tax may produce a fair and reasonable result for an element of overhead input tax incurred by Optika, for input tax incurred in relation to property, this is not the case.

For input tax attributable to property, eg: rent and shopfitting, using the standard method produces a distorted result. This is because of the use made of the relevant properties. As you may be aware, the exempt use of the properties, the sight test rooms, are situated either in basements, rear rooms, or on higher floors, ie, not in the areas of the properties which are valued at a higher rate. Higher rents are payable for the shop front areas of properties where the (standard-rated) goods are displayed."

Mr Ward here appended a letter from Stephen Kane & Co explaining the zoning system, and calculating the area of the Bluewater shop in terms of zone A. The special method was then set out in the following terms:

"Method

In order to reflect the above facts, I propose the following partial exemption method:

Step 1 - Identify all input tax directly attributable to exempt supplies, eg sight test machines - This input tax would be irrecoverable

Step 2 - Identify all input tax which is directly attributable to taxable supplies, eg sunglasses and accessories - This input tax is recoverable in full

Step 3 - Unattributable input tax - This will be analysed into two 'pots' - that relating to property and the remainder

Step 4 - Input tax relating to the property, eg rent and refurbishment, will be apportioned as per appendix I. This recognises the actual use to which the properties are put

Step 5 - All other input tax will be apportioned as per the ratio between the taxable and exempt supplies made, ie the 'standard' method.

Appendix I

Apportionment of Input Tax Relating to Retail Outlets/Properties

I Identify the value of residual input tax incurred on supplies of, or in relation to retail properties (eg: rent, shop-fitting and refurbishment costs).

II The amount of such input tax that may be recovered should be determined by reference to the following calculation:

Value of input tax directly attributable to taxable supplies x 100

Value of input tax directly attributable to taxable and exempt supplies

The above calculation should be expressed as a percentage, calculated to two decimal places."

The letter then dealt with input tax incurred on advertising, proposing that that should be allowed in full, on the ground that although it appeared that advertising promoted the whole of Optika's business, opticians were specifically prohibited from advertising their services as better or cheaper than or different from those of their competitors.

Mr Laney replied on 17 December 1999, repeating the outline of the special method. He then continued:

"Although it is not clear in your letter I believe from our conversations and references made in relation to the rental valuations and zones in the surveyors report that item c) (your step 4) apportions input tax between exempt and taxable supplies on a floor space method but valuing this floor by zones ie areas at the front of the premises generally having a higher value than those at the rear.. . .

Item C. Whilst I have no problem in principle with the attribution of input tax on a floor based method I do not accept that only the consulting room space relates to an exempt supply as any waiting areas, reception areas and dispensing areas (which includes service counters and till areas) relate in part to the exempt supply of dispensing and therefore relate to a mixed supply.

My other reservation about this part of the calculation, if my understanding was correct, is the desire to give weighting on a zone basis where different areas of the premises are given different valuations dependent upon their location within the premises. Whilst I do not have sufficient knowledge about the valuation of premises for retail purposes to dispute what your Surveyor states I do not accept that this gives reasonable grounds to attribute your costs in this manner. My reasons for this area are that Optika will be rented the premises as a whole for a given price, not areas of the property at different prices. Furthermore any other cost in relation to the property will not be different for different areas eg if the premises were carpeted then the cost of that carpeting for any area would be in proportion of that area to the whole, it would not be less simply because it was at the rear of the property."

Once again, the proposed special method was rejected. In his evidence at the hearing, and in his statement, Mr Laney admitted that he had misunderstood the nature of the First Special Method, and had mistakenly thought that it was a floor-space based method. He said that his misunderstanding related only to the proposal in the letters of 26 October 1999 and 29 February 2000; he also said that Miss Nagarajah's review was based upon a correct understanding of the proposed method.

A later file memorandum made by Mr Ward, dated 28 February 2000, recorded a meeting held on that date, at which Mr Laney and Mr Rishi of Customs were present, as was Mr Allen. The meeting had been for the purpose of progressing the dispute over the method of apportioning sales between exempt and taxable supplies, though agreement was apparently not reached. The memorandum concluded by saying:

"We also discussed partial exemption. It was clear that [Mr Laney] had misunderstood my letter proposing a special method. He was still of the opinion that we were proposing a floor space based method. I explained that that was not the case, and that purely for input tax relating to property a purchase based method was used. On re-reading the letter he conceded that he had misunderstood the proposal. I will write to him tomorrow, addressing each of his concerns. I explained that now he understands the proposal, a lot of his objections will fall away."

The promised letter was written on 29 February. The essential parts were as follows:

"I should point out initially, that you appear to be under a misapprehension concerning the proposed method. Although we discussed the reasons why the standard method would be distortive when expenditure on property is considered (this is because the costs incurred on the property bear no relation to the use the property is put), I did not propose a 'floor-based method'. I proposed a straightforward purchase method.

Consequently, your comment and conclusion are based on an incorrect understanding. The purchase based method, more accurately reflects the actual use of the properties, and is simple to administer and for you to verify.

Although a floor based method has not been proposed, I address the observations you make as this may assist you in coming to a conclusion on the actual proposed partial exemption method.

i Optika is rented premises as a whole for a given price. However, that price is calculated by reference to the amount of floor space in each area which is valued at different prices. This is now the total price is arrived at. It is therefore clear that Optika pay different amounts for different areas. As an example Optika would pay a higher rent for a shop with more 'retail' area than another (all else being equal) with less.

ii Your point on carpeting is accepted. However, carpet along with say paint or wallpaper is usually all the money which is spent on the 'exempt areas'. This represents a very small fraction of the costs of a re-fit. As you will appreciate, the substantial costs relate to displays and storage areas, ie, the 'standard-rated' areas. Of course, the proposed method does apportion the 'overhead' part of costs relating to properties.

iii Regarding your comments on the costs of advertising, I am not contending that this is a special method, rather, these costs are directly attributable to standard-rated supplies and are therefore recoverable in full, as they would be under the standard method.

iv I cannot accept your point that a sale of spectacles is a mixed supply. It is clear from the Leightons and Eye-Tech case that there are two supplies: a) the standard-rated supply of goods, and b) the supply of exempt services. Since, as preciously explained, it is not possible to advertise an optician's services, it is only possible to advertise (standard-rated) goods. You use the example of sunglasses which are also standard-rated, but the principle is similar in the case of spectacles, ie, costs wholly attributable to the sale of standard-rated goods may only be attributed to that supply (and none other) and consequently the relevant input tax falls to be recoverable in full."

Mr Ward concluded by saying that the proposal for a special method remained the same as was proposed in the letter of 26 October 1999.

In his reply, dated 20 March 2000, Mr Laney dealt with the basis of the proposed method by saying:

"I note your comments with regard to your proposed method being one based on a purchase cost of the property by reference to valuations of areas within the property. When I responded to your prior proposal I believe that I did understand this point but have expressed it in a different manner."

He then went on to reject the proposed method once again, for the same reasons as before, adding the following:

"1. However, the purchase or rental value of one of the stores is calculated a single price is paid for the property as a while, they do not rent each area separately for separate considerations and I would presume in general that they do not have the option of renting only parts of the store.

[Dealt with costs relating to areas in the shops, e.g. carpeting.]

In my previous correspondence I have made reference to the retail area. By this I mean the normal sales area of the store other than the eye examination rooms. This 'retail' area is used for both the sales of goods and the dispensing services provided. Therefore this area could not be treated as a taxable only area. It does in fact relate to taxable and exempt sales.

[The Appellant's point as to advertising was also rejected.]"

It was the Appellant's case that in that letter Mr Laney demonstrated that he had continued to misunderstand the nature of the proposed special method. Mr Ward wrote to Mr Rishi on 18 April 2000, requesting a review of Mr Laney's decision, and dealing with the points raised by Mr Laney in the letter referred to in paragraph 29 above. He began by saying,

"...It would appear that, despite numerous attempts at explanation, Mr Laney has still not recognised that I am not advancing a method based on floor space. I think it would be useful if I set out how my proposal represents a fairer and more reasonable method than the standard method. As you will be aware, the purpose of a partial exemption method is to establish the amount of input tax recoverable on the basis of use. The method which most accurately achieves this is to be preferred. [Mr Ward's emphasis.]

. . ."

Mr Ward then described the proposed method as being a purchase based method, "ie the 'old' standard method [pre-1991]". The consideration of the use of various parts of the property was intended only to demonstrate that a fairer and more accurate method was needed. The letter continued:

"The proposed method is not one based on a 'purchase cost of the property by reference to valuation of areas within the property'. It is a straightforward purchase based method. No calculations will be used in respect of areas of the properties. [Mr Ward's emphasis.]

Although a single price is paid for a property (as one would expect in a normal commercial transaction) that single price is calculated by reference to valuations of the quantum of zones occupied in that property. Clearly a property with more 'retail' area would be valued higher than one with less (all else being equal).... I should repeat, the purpose of analysing the zones is to demonstrate that for areas where more 'taxable activities' take place, eg: the display of goods, more rent is paid than for the zones where 'fully exempt' activities take place, eg: sight test rooms which are sited in zones valued at less than the retail area. This analysis is not the basis of the proposed method."

Mr Ward accepted that such costs as those of paint and carpeting would not vary according to the value of the zone in which they were applied, but "they would, of course, be apportioned by virtue of the purchase based method". He pointed out that such costs in fact amounted to less than 1 per cent of overall property costs, and that shop-fitting, the major expense, related to the taxable elements of the properties, "eg: display areas for the (taxable) goods, and storage for goods".

The matter was referred by Mr Rishi to Miss Nagarajah, the partial exemption specialist at Wembley VAT office. In a letter dated 7 July 2000, she said,

"In considering your application to use a special method you need to demonstrate to Customs that the standard method is not fair and reasonable.... You go to state that the vast majority of expenditure relates to 'taxable elements' of the properties, e.g. Display areas for the taxable goods and storage of goods. Whilst I can accept that the shop front or the retail aspect of the shop may be the most cost intensive I do not think that the expenditure relates to the taxable elements of the property as this relates to both taxable and exempt element, as there will be dispensing services undertaken in the retail area too. This is a point that must first be addressed.

[Having referred to the appendix to the letter of 26 October 1999 (see paragraph 27 above), Miss Nagarajah continued:] Hence the formula ignores the input tax that is not directly attributable to either taxable or exempt supplies from the denominator. Without figures to show this definitively, I think it would be reasonable to assume that there would be input tax directly attributable to taxable supplies, such as frames, consumables etc. But I cannot think of the input tax that would be directly attributable to the exempt supplies (apart from the purchase of eyetest equipment or prescription forms which would not be purchased on a frequent basis). Hence, surely the apportionment would result in a very high percentage recovery rate in most periods, if not 100%. This apportionment would not give a fair and reasonable recovery rate to reflect the exempt supplies that are generated from the usage of the properties."

Miss Nagarajah referred to the question of advertising, and pointed out that the advertising generated both taxable and exempt income, it having been accepted that the frames were an integral part of the overall supply of spectacles, the cost of which included that of dispensing. She continued:

"Therefore in the vast majority of cases, the Leightons decision applies and it is accepted that in these cases there is a supply of spectacles and dispensing. To me, the supply of dispensing is an integral part of the supply of spectacles, and it is not unreasonable for this to be assumed by the end user, i.e. your clients customers, who would expect the highest quality of dispensing service to be provided, at a standard charge, with the sale of frames. So although promotional material will concentrate on the sale of spectacles, as the services given will be standard to most opticians we must not forget that the adverts generate both taxable and exempt income and thus the advertising costs must be seen to be residual expenditure."

In his reply, Mr Ward replied to the points made by Miss Nagarajah in terms similar to those in his previous letters to Mr Laney and Mr Rishi. He conceded that the proposed method ignored input tax which was not directly attributable to either taxable or exempt supplies from the denominator in the calculation set out in the appendix to the letter of 26 October 1999 (paragraph 27 above). He continued:

"Nevertheless, in order to reach agreement in this case I should like to propose the following:

The standard method, as previously set out, to be used for appointing input tax other than that referable to properties.

Input tax incurred in relation to property to be apportioned as follows:

taxable input tax x 100 = recoverable percentage

all input tax

The purpose of a partial exemption method is to establish the amount of input tax attributable to taxable supplies, and therefore recoverable, on the basis of use. One may see that this method most accurately achieves this, as there is no issue (as I understand it) that there is a greater demand on (or use of) the properties in making taxable supplies than the business overall."

Miss Nagarajah replied on 18 September 2000, to say that the "new proposal" had been passed to Mr Laney, who would consider it.

Mr Laney wrote on 25 October 2000, saying that Miss Nagarajah had regarded Mr Ward's letter as containing the proposal of a new method, and had declined to deal with it. Mr Laney said:

"Thank you for the information you gave me by phone on the 23 October. You advised me that the 'taxable input tax' in your suggested formula relates to VAT on shop overheads etc where these overheads can be directly attributed to taxable supplies. In your view such overheads would relate to items such as shelf fittings used in the sale of glasses. You also stated that the 'total input tax' part of the formula includes all shop overheads where VAT is incurred. This would include input tax on the purchase or lease of a property where exemption has been waived."

That was Mr Laney's understanding of the formula in Mr Ward's letter after amplification in a telephone conversation.

"An apportionment method for partial exemption must accurately reflect the ultimate use to which a purchase of a supply is put ie: the use for making a taxable or exempt supply. An inputs or input tax based method if agreed would also be required to reflect how the purchases are to be used in the making of taxable or exempt supplies.

The method proposed for your client does not have a direct link to the supplies made and does not reflect the ultimate use to which the purchases are put. In other words in my view it does not provide a reasonable apportionment of input tax to the ultimate use of the supplies. I believe the following examples best illustrate my point. If for instance the taxable element of your equation increased because of a bulk purchase of new display units then recovery on all non attributable input tax for the shops would increase. However there would not be an increase in the taxable usage of the non attributable items.

A more extreme example would be to examine what effect the inclusion of any VAT on property purchase, lease or rental would have on other overheads. The input tax on property (if an option to tax existed) would only be included in the bottom half of the equation as VAT on property would be non-attributable. This would mean if your client were charged VAT on all his properties he would have a lower recovery on all property overheads than he would have if no properties had VAT charged...."

In cross-examination, Mr Laney agreed that this might also be true in respect of the standard method; so also in the case of an option to tax property. The letter continued:

"...Again whether or not a property is exempt has no bearing on the use of other overheads for the property ie the VAT status of the property as an expense should not effect [sic] the apportionment of input tax on say heat and light but in your proposed method it would."

The proposed special method was rejected. That was the letter containing the decision the subject of the first appeal. It was Mr Laney's view, as he said in his oral evidence, that the method would not result in a fair and reasonable recovery, there being no clear connection between the ratio or recovery rate and the use to which the non-attributable input tax incurred in relation to the shops was put, and he set out an example in his statement. He also considered that under the method proposed an increase in the number of spectacles purchased would result in a higher proportion of non-attributable tax being recovered. If the usage of a store did not change, Mr Laney said at the hearing, the recovery rate should not change. He also said that it had never been suggested to him that the supplies of dispensing in any of the shops are made by unqualified staff not under qualified supervision.

The second proposed special method

On 22 May 2001, Mr Ward wrote to Miss Nicola Joad at the Solicitors office of Customs and Excise. Having reasserted that the standard method of apportionment failed to deal fairly with input tax relating to property, as agreed with the Commissioners, Mr Ward continued:

"Despite misunderstandings in previous correspondence, a floor-based apportionment method had not been proposed. One of the reasons for this was that the purchase based method was simpler to calculate and to check. However, in order to progress this issue, I now propose, without prejudice, the following method to apportion input tax referable to property.

After research, it has been established that it is possible to accurately apportion the floor areas of Optika Clulow Group's properties to reflect taxable, exempt and mixed use. I am able to confirm that the majority of the floor space is capable of being analysed into taxable or exempt usage with a substantially smaller element representing mixed use, (this is, of course, subject to agreement with the local office). The benefit of using a floor space method is that it accurately reflects the actual usage of the input tax incurred in relation to it."

Mr Ward sought the Commissioners' agreement in principle to such a method, subject to agreement of the precise details with the local VAT office. The letter was passed by the Solicitor's office to Mr Laney. (We observe, at this point, that no accurate apportionment into taxable, exempt, and mixed use areas was put forward by the Appellant.)

Mr Laney replied to Mr Ward's letter on 26 May 2001. He pointed out first that the Commissioners had not agreed that the standard method did not give a fair and reasonable result when applied to input tax in respect of property. He also said that he had understood the first proposed special method and had explained why he found it unacceptable. He explained that agreement in principle could not be given, since all new proposals had to be considered on their merits, and he suggested that the new proposal be submitted in full detail. He added that there was no indication as to how apportionment of deductible and non-deductible input tax for the mixed areas would be dealt with, although he had raised the point several times before. Nor did he accept that the mixed area was substantially smaller than the wholly taxable or wholly exempt areas, and expressed the view that the whole of the retail area would be of mixed use.

Replying on 5 June 2001, Mr Ward said that the proposal was an attempt to resolve the issue before the Tribunal hearing, and that the Appellant was still prepared to seek a compromise. He said yet again that the standard method was not appropriate for input tax related to property, and referred to a meeting on 16 April 1997, chaired by Mr Topping of the Solicitor's office, at which it was agreed between the parties that a special method was appropriate for VAT bearing costs incurred in respect of property.

In a further letter dated 19 June 2001, Mr Ward provided more detailed information as to the proposed floor-space method, stating that this was a separate and entirely different method from that already under appeal. He said that for this purpose he had visited a number of shops to take detailed measurements. He mentioned that, in a letter of 8 November 1996, Mr Donohoe had said that he had no objection in principle to a floor-space based method. Mr Ward repeated his views relating to the use of property and the expenditure incurred on various parts of any property. He said that the Appellant had selected a representative sample of Optika branches which reflected differences in location, size and layout, as well as social mix of customers, rent, and number of sight-test rooms, which was considered to be an accurate cross-section of their shops. (The results of this survey, relating to the Bluewater branch, which we take as a representative sample, are set out in Appendix I to this decision.)

"The areas we have defined as taxable are those which have no connection to services and relate solely to the display, storage and advertising of goods. This floorspace is occupied with, inter alia, wall and window displays, free standing displays, seats and mirrors for trying on spectacles (as distinct from sight testing or pre-sight test examination) security devices and floorspace for customers to stand before the abovementioned displays (clearly this is necessary in order to view the goods and promotions of goods, and to examine and try on the goods).

The areas which were identified as exempt are: sight test rooms (apart from areas used for storing goods). Areas where pre-eye test checks are made, eg: pupil measurement etc, and any waiting area which relates specifically to sight test rooms.

Mixed areas are the counter/till and a corresponding area in front of the counter where customers stand to pay for both goods and services, certain walkways and corridors, a 'pathway' through zone A through which customers pass to reach the exempt areas, staff lavatories and an area which is used to carry out work on goods (spectacle lenses).

One may see that the mixed element represents less than a third of floorspace according to value. The apportionment of the mixed areas has been calculated by reference to the ratio represented by taxable versus exempt areas.

It is intended to apply the percentage calculated from the representative shops to all unattributable input tax incurred in respect of property."

The calculation relating to Bluewater gave the taxable element as 85.86 per cent, and the exempt as 14.14 per cent.

A file note by Mr Ward dated 26 June 2001, relating to a telephone conversation with Mr Laney stated that Mr Laney had apologised for his admitted misunderstanding of the First Special Method, though he considered that Miss Nagarajah had addressed the issues correctly in her letter of 7 July 2001. Mr Laney had asked for figures for the First Special Method for comparison with the standard method. He also stated that he considered, with respect to the Second Special Method, that some of the floor space allocated to taxable input tax was used for dispensing (exempt).

On 16 July 2001 a meeting took place at New King's Beam House to discuss the Second Special Method. Mr Laney and Mr Laurence James attended for Customs, Mr Allen for the Appellant, and Mr Ward and Mr Frank Hartley (VAT partner) from Grant Thornton. Mr Ward provided a note of the meeting. The note records that agreement was reached that the input tax to which any special method would apply would be that on rent (where there was an option to tax), shop-fitting, repairs, and maintenance. Mr James agreed that any agreement as to the special method would be backdated to 1 April 1997, possibly subject to the capping provision (of section 73(6) of the 1994 Act). Mr Laney apologised again for his previous misunderstanding, but stressed again that it related only to the proposal set out in the letters dated 26 October 1999 and 29 February 2000.

Customs expressed certain concerns about the Second Special Method. (We observe that some of these appear to contradict the agreement as to the input tax mentioned above.) First, Customs did not consider that the method should be applicable to expenditure on shop-fitting, because, the note stated, "they considered it possible, and acceptable, to analyse the supply from the shop-fitter and recover the taxable elements". Mr Allen would obtain a breakdown of shop-fitting supplies and that would be considered. Secondly, Customs were unwilling to accept that a breaking down of rental supplies was possible: he considered it a single supply which was not capable of being divided into zones. The way in which the Appellant proposed to do this in the method was not acceptable. Thirdly, Customs would not agree upon the allocation of the whole of zone A, less a mixed walkway, to taxable supplies. They contended that the whole area was mixed, since customers came in to purchase both eye tests and spectacles. The Appellant maintained that the whole of the front of the shop was intended to attract customers in by displaying standard-rated goods. Customs accepted that the actual display areas, plus a floor area in front of or round the displays, were wholly taxable. It was at that meeting that Mr Allen stated that the sales of sun-glasses represented between 20 and 45 per cent of sales.

Mr Laney wrote to Mr Ward on 12 September 2001 with reference to Mr Ward's letter of 19 June 2001 and the meeting referred to above, and also to a visit that he had made to Bluewater on 24 July 2001, following both of which he had come to the conclusion that the Second Special Method did not produce a fair and reasonable recovery. He said in his oral evidence that he had experience of visiting opticians' shops and seeing how they operated. He would, he said, have been aware from his visit to the Ealing branch of the use made of the retail area. He said that there was no dispute with the Commissioners as to that part of the proposal which was, that the standard method be adopted for all input tax which was either fully attributable or which fell outwith the costs incurred in connexion with the properties. The dispute was only as to the treatment of the input tax relating to the properties. Mr Laney made the following comments upon that proposal:

"1) The method is reliant upon the six shops proposed as being representative. Whilst I cannot conclusively dispute this I do have some doubts as the calculated restriction on the sample shops shows a result variance of over 12%."

At the hearing, in cross-examination, Mr Laney said that he would have expected some variation between shops, within a range. But he had not been sure what range he should have expected, which was why he had expressed a doubt. He had not sought further information, nor had he asked for drawings of the shops. The letter went on to say:

"2) Using Bluewater as an example I sought to verify your measurements and assumptions through a visit to the store. My own examination suggested that not all areas highlighted in your calculations as taxable were in fact so and an area highlighted as mixed was in fact exempt:

a) Retail area - see separate comment at para 3.

b) Solutions behind till - they are actually in an area sited above

the drawers containing paperwork and stationery relating to both

taxable and exempt supplies.

c) An area exists behind the till area used for teaching customers

to put contact lenses in. In my view this is purely a dispensing area

which would therefore be a wholly exempt one. In the calculations

it has been treated as a mixed area. NB this is an area separate to the

pre sight test area.

d) Storage of standard rated items in test room 2. My checks did

confirm the presence of boxes of spectacle cases but the assistant

branch manager told me that they shouldn't be there and they would be

removed shortly.

3) Your proposed method assumes that the majority of the retail area of the shops is fully taxable and in relation to the mixed area provides for a narrow walkway to the exempt areas of the shop and to the till. In effect this assumes that the majority of that retail area is used exclusively for taxable purposes. From my inspection of the Bluewater shop I cannot accept this a reflection of the use of that space. In fact the majority of that area is used for making both taxable and exempt supplies and is enjoyed by customers receiving both taxable and exempt supplies. During my visit I noted the front retail area being used by customers awaiting eye tests, waiting for the help with contact lenses and receiving dispensing services from staff with regards to spectacles. This activity took place at all parts of the retail area including in areas immediately adjacent to the display and storage cabinets. Accordingly, it is my opinion that the only part of the retail area that is used for exclusively taxable activities is the spectacle display and storage units, which cover a relatively small area.

In his oral evidence, Mr Laney said that the exempt use of the retail area was not fleeting or ephemeral. Customers were there for some minutes trying frames, seeing that they fitted properly, and discussing them with staff. The counter was used for making appointments and paying. When customers walked out of the shop with spectacles they would have received exempt supplies of dispensing. The letter continued:

"4) Even if the Commissioners were to accept the six sample stores as being representative, because of differences in use of the areas between taxable, exempt or mixed all six sites would potentially require a visit by Customs to check the details - this does not in my view constitute an easy method of verification. However it may be that we could reach an agreement to resolve this aspect."

Mr Laney also said that he did not believe that the method gave a fair attribution in respect of sales to customers who only received standard-rated supplies, and then continued:

"I appreciate your arguments that when a total refurbishment takes place a single price may be negotiated for the whole store whilst the front areas of the store potentially absorb a significant element of that cost. However because my view is that the majority of that shop front area is used to make taxable and exempt supplies it follows that your proposal would allow a distorted recovery for such costs. The one area that I accept that your client is not achieving sufficient recovery is in relation to the spectacle display racks. I understand that these costs were included in a total refurbishment cost and therefore I.T. recovery was restricted. As we indicated at our meeting we do not dispute that the spectacle display and storage units relate to an exclusively taxable supply and subject to any other normal VAT rules have no objection if they were treated as such if they were itemised on the suppliers invoice."

A second letter of the same date informed Mr Ward that the Appellant remained on the standard method for partial exemption, since the latest proposed method had not been approved.

Mr Hartley replied to that in a very long letter dated 16 October 2001. That letter states that Mr Laney's of 12 September is not a rejection of the proposed special method, for this reason:

"...Because you have failed to consider that which you had undertaken to consider and have purported to reject something entirely different, we say that your letter is a nullity so far as consideration of the proposals is concerned and that you have therefore not succeeded in rejecting them.

Your letter continues to be relevant, however, as evidence of continuing failure properly to consider the relevant issued."

The rest of that part of the letter appears to establish that the First Special Method is referred to, not the second.

The second part of that letter is headed "The distance between the parties". Mr Hartley stated that he did not think that there were any points of substance relating to the matters raised in Mr Laney's letter except with regard to the treatment of input tax on the rent. He then considered each of the points in Mr Laney's letter in turn.

Dealing with sampling, Mr Hartley said,

"You note that the figures coming from the six shops indicate recovery rates varying between 85.37% and 97.98%. In our view this tends to suggest that the sample chosen is very likely to be representative of the variations across the business.

I cannot understand how you have a concern about the existence of a variation. In fact most observers would have been concerned if there had not been a variation to reflect differences of location and age of the shops. Also, as you appear to acknowledge, you are really not in a position to dispute that the sample is representative whereas Mr Allen is in a position to devise a representative sample on the basis of his experience and knowledge of the business.

Any purported conclusion by Customs on this point, therefore, is not based on evidence and is not a conclusion, which a reasonable officer could reach."

Mr Hartley's letter went on to deal with "the points at indent 2 of your letter". We suppose this to be a reference to the paragraphs numbered 2) set out in paragraph 40 above. Mr Hartley said that the Commissioners' conclusion on these matters was not accepted, but offered a number of concessions with a view to disposing of the matter. These were:

"Solutions behind the till

The area is negligible. For the sake of agreement my client would be prepared to treat this area as mixed.

Teaching area

Again this has no appreciable significance and for the sake of agreement my client would be prepared to treat this area as exempt.

Storage in test room 2

This is the same as the other two. For the sake of agreement my client would be prepared to treat this as exempt."

Turning to the matter of the retail area, Mr Hartley summarised the Appellant's arguments on this subject, as follows:

"In your letter you concede at least that the spectacle display and storage units are used for exclusively taxable purposes. In fact the space around such areas is necessarily used in the same way; after all a customer cannot actually stand inside the display unit. My client would therefore be prepared to agree that the general retail area should be treated as mixed provided you agree that display units (including the window and shop front) and the two feet immediately around each of them (or in the case of the shop front, behind it) is treated as wholly taxable.

Lest we have to go to the next stage, I should like to make it clear that we view your analysis as being so strained as to be indefensible. In fact if it were applied consistently, areas we had originally treated as exempt would become mixed. The mere fact that customers can be seen waiting in this area before they have their tests does not mean in any real sense of the words that the area is used for testing. The reference to customers receiving dispensing services is similarly strained. If a member of staff fits frames on a customer in this area that activity is more readily characterised as retail than dispensing.

Mr Hartley went on to say that those were not his prime objections, and stated that any method based on floor space would treat as wholly exempt or wholly taxable areas which had a fleeting use for the other purpose, that being entirely insignificant. Thus, he said, consulting rooms had been treated as exempt although the opticians were actively engaged in selling standard-rated goods. Such a method "gives a trade off that arrives at a fair and reasonable result," he concluded.

As to verification, Mr Hartley expressed the view that there would be differences remaining between the parties as to the use of areas in the various shops, and he reminded the Commissioners that Mr Laney had suggested the likelihood of reaching agreement on the point. He took it that the Commissioners would agree that that point alone could not be an impediment to agreeing a special method.

Mr Hartley then turned to the use of the front area. He said:

"You have already been given evidence of he commercial reality of how rents are arrived at and how the calculation of rents is weighted so that a premium is paid for the front area. The premium is paid because of the importance of ensuring 'foot fall'. This term is used to describe the effect in attracting customers into the shop. This expenditure is used to attract customers to buy the standard rated items. You need only have regard to the design of shops to appreciate the truth of this. To say that the use of parts of the shop is mixed does not lead to the conclusion that the front (which displays standard rated goods and which should therefore be treated in the same way as display units inside the shop) is used to attract eye test business. Such business is passive: an optician will get it anyway and no optician actively solicits it - indeed there are rules restricting the ability of opticians actively to solicit such business even if they wanted to.

The approach based on floor space zones and the commercial practice in regard to the build up of rents is not predicated on splitting up the rent. The rent is seen as consideration for a single supply. We say that the standard method would not give a fair or reasonable result and we take it from the course of correspondence, and in particular from the meeting of 16 July, that this is accepted by Customs, such that the basic point at issue is only the appropriate special method, and not the need for one. Our suggestion on floor space is calculated to give an apportionment based on the factual use, not to treat the rent as being composed of separate supplies in respect of discrete zones.

It should also be noted that in the light of this letter (in particular the concession offered in the next paragraph) the floor space approach would, if our proposal were accepted, apply only to the rent."

That next paragraph, headed "Non-rent costs", said,

"For the sake of agreement, my client is prepared to apportion such costs (other than refurbishment costs) on the basis of the standard method."

As to refurbishment costs, Mr Hartley said,

"I note your agreement that treating refurbishment costs as a single purchase for mixed use would give a distorted recovery and that my client may in future arrange for split invoices. For the past Optika will carry out a reasonable apportionment."

The next part of the letter consisted on criticism of the way in which the Commissioners had dealt with the application for a special method, delay due to their misunderstanding of the proposal. Mr Hartley stated that he was not in this letter seeking a reconsideration but was stating that Customs should use its own procedures to weed out unmeritorious appeals before they reached the Tribunal.

The remainder of the letter contained a suggested basis for settling the matter, expressed as follows:

• "the floor space method as put to you in our letter of 22 May 2001 amended as indicated in this letter (for convenience the amendments proposed are set out below)

• the method to be retrospective to 1 April 1997

• the area behind the till to be treated as mixed

• the area in which customers are taught to put in contact lenses to be treated as exempt

• storage area in test rooms to be treated as exempt

• the display areas (including shop fronts) and storage units and the floor space for two feet immediately contiguous to such areas to be treated as fully taxable

• all other retail areas to be treated as mixed

• save for refurbishment, all non-rent overhead costs to be apportioned on the basis of the standard method

• for the future the elements of refurbishment costs relating to display and storage areas to be invoiced separately and treated as fully taxable, the balance to be treated as mixed

• no issues arise as to the operation of capping rules because the letter of 22 September 1997 falls to be regarded as a claim"

The offer was termed an "exploding offer", in that the Appellant reserved the right to reargue any of the points set out in the offer if the matter had to be appealed.

In a letter of 26 November 2001, Mr Richard Dowling, who had apparently taken over the case upon Mr Laney having been transferred to another office, replied. He first said that it was not accepted that a composite proposal containing two alternatives had been made, since the proposals were fundamentally different and therefore fell to be considered separately. Mr Laney, having admitted an early misunderstanding of the inputs method, had said that he had understood the letter of 18 September 2000. Mr Dowling said that he did not accept the contention that the inputs method, the First Special Method, had not been properly considered by the Commissioners. He also rejected the Second Special Method on the ground that, even with the concessions contained in the letter of 16 October it did not give a fair and reasonable apportionment. His reasons were as follows:

"1. The method allocates too much floor space to taxable supplies, namely the areas in front of the display cabinets and the area surrounding the shop window, ignoring the fact that exempt dispensing activities take place in these areas.

The manner in which the mixed floor space is apportioned is unrepresentative of the taxable usage, i.e. it places taxable areas over exempt areas rather than taxable areas over total area.

The precise method of apportionment of the refurbishment costs is not specified.

The fact that a single supply of rent is received by the appellant means that the weighting of the rent into zones is an artificial breakdown of the cost incurred. Furthermore, the weighting is not representative of the actual usage of the space.

In arriving at my decision that the proposed floor based method is not acceptable, I have also given consideration to the recovery that would be involved if the method were used without the weighting of the rent and the recovery given by the standard method. The detailed calculations are set out in the schedule attached. Suffice to say that it can be seen that the proposed method (including weighting of the rent) would allow an unjustifiably higher recovery rate than either the standard method or the proposed method without weighting. On the other hand, it can be seen that the proposed method without the weighting would give a very similar recovery rate to the standard method. This reinforces the putative fairness of the standard method.

Accordingly, the proposed floor based method is rejected."

In cross-examination, Mr Laney said that he was not sure why it was necessary to look at the standard method to see if a special method was fair. The fairness of the standard method had not entered his mind. Regulation 102 said that he had to see whether a proposed special method was fair. On 25 October 2000 he was under a misapprehension as to the First Special Method. He had looked at the letter from Mr Ward of 8 October, and still had that misunderstanding when he wrote his letter of 25 October. When he looked at the letter of 8 October he looked at it as a new method, which broke the mental link. When he rejected the proposed method in that letter, he said, all that he considered was that letter and what he knew about the business. He said that he would have been aware of the previous correspondence, but he did not go back and review it. He had spoken to Mr Ward. In the letter of 18 September 2000 the method proposed was the same as the previous method but for the concessions. By that time he had seen the shop at Ealing: the Ealing visit, however, was to do with output tax, sales, not input tax. There was no question of zoning.

The letter also dealt with a point relating to the spectacle display units, which we include here since it was raised later by the Appellant:

"...As [the display] units are used for exclusively taxable purposes, we have no objection if the tax incurred on the purchase of these units were to be attributed to taxable supplies if this tax were clearly itemised on the suppliers' invoices. However, this could only be done as part of an approved special method. No special method for the display cabinets alone is acceptable."

(We include the schedule of comparative recovery rates as Appendix II to this decision. The schedule shews that the recovery rate under the standard method would be 54 per cent, and under the proposed special method it would be 55 per cent neglecting the zone weighting. With the weighting, the recovery rate would vary between 85.86 per cent (Bluewater) and 97.98 per cent (Cambridge).)

In a letter dated 30 April 2002, Mr Paddy Behan, senior manager of Grant Thornton, replied to Mr Dowling's letter. Having quoted the paragraph from the letter of 26 November 2001 set out at the end of paragraph 51 above, a voluntary disclosure was enclosed claiming a repayment of £33,294, consisting of a letter of 13 February 2002 and schedules. It was mentioned that past invoices for shop-fitting did not list the display units separately, but a fair apportionment had been calculated using proportions derived from the suppliers' quotations.

Mr Behan's letter then requested a review of the decision in the letter of 26 November 2001 relating to the floor space method, and observed that the reasons were apparently those contained in the first four indents on page 2 of that letter, and an observation on the alleged relevance of the standard method. Before dealing with the four indents, Mr Behan set out a passage from page 2 of the letter (set out in paragraph 51 above, the first two sentences after the numbered paragraph 4). Mr Behan said that that statement was fundamentally misconceived. He continued:

"...Any request for a special method is predicated on the standard method's not giving a fair and reasonable result. This question must be considered in the light of all the relevant facts but it is not legitimate to use the standard method as a control mechanism."

To do so, Mr Behan said, vitiates the decision. Mr Behan then set out the proportions of the taxable areas given in the Schedule to the Commissioners' letter of 26 November (see Appendix II) as varied by the concessions offered in Grant Thornton's letter of 16 October. The result was as follows:

Shop % taxable % taxable

(Customs) after concessions

Bluewater 85.96 80.49

Kingston 92.89 89.17

Cambridge 97.98 96.16

King's Road 95.87 94.58

Mr Behan added, "I am instructed that purely for the sake of agreement, Optika would be prepared to settle on the basis of an 80 % taxable proportion."

Turning then to indent 1 (see paragraph 51 above), Mr Behan said,

"We dispute this as a matter of fact. This point was considered in our letter of 16 October 2001. It will be remembered that, in our view, it could not sensibly be stated that any dispensing activity took place in the retail area. Nevertheless, we were prepared to concede that the general retail area could be treated as being for mixed use. However, once it is conceded that the display units are used for wholly taxable purposes, we say that an area immediately around the units is similarly so used and this point applies to the shop window. An area immediately around a unit is, as it were, part of the curtilage so as to form a unity with it in terms of use."

Mr Behan then repeated the points made in Grant Thornton's letter of 16 October 2001 (see paragraph 45 above).

As to indent 2, having stated the Commissioners' view (see paragraph 51 above), Mr Behan said,

"The method needs to apportion mixed-use areas. If the calculation were carried out as Customs suggests that which must be apportioned would have to be included in the apportionment. The use of taxable over exempt has the virtue of logic in that it uses identified values and is, in our view, the more likely to produce a fair and reasonable result."

Indent 3 was said to have been dealt with in relation to the voluntary disclosure. Having set out indent 4 (see paragraph 51 above), Mr Behan continued,

"It is not disputed that Optika pays rent for a single supply. The process of breaking down the rent is not an attempt to treat it as a number of different supplies. Rather the reference to zoning reflects how rents are actually arrived at and demonstrates that apportionment on the basis of the standard method would not yield a fair and reasonable result because it would not reflect what the retailer is actually paying for and, hence, the real use.

The analysis of the rent by zones is not artificial; it is based on actual commercial practice...."

A letter from Stephen Kane & Company was produced, which explains the method of zoning, and included by way of illustration eight rent review analyses of Optika shops. Mr Behan continued,

"The weighting certainly reflects the commercial reality throughout the retail sector. The VAT question is then whether it reflects actual use. In this connection the commercial reality must be a good starting point for VAT purposes and the points already made in correspondence together with the various concessions we have offered for the sake of agreement mean that the end result is fair and reasonable."

On 10 May 2002 Mr Mattacks, of Customs and Excise, wrote to Mr Behan confirming the decision to refuse approval for the Second Special Method, which, he said, did not achieve a fair and reasonable apportionment. He then gave the following reasons for the decision:

"The basic assumption behind the method is that the main sales of the business are taxable spectacle sales and exempt eye tests (making little recognition of dispensing). Another assumption is that an attractive shop front will attract customers to buy taxable spectacles only. Your correspondence specifically states 'If a member of staff fits frames on a customer in this area (the shop) that activity is more readily characterised as retail than dispensing'. This ignores the fact that a sale of prescription spectacles (the principle trading activity of the business) is a mixed supply of taxable goods and exempt dispensing, with a considerable portion of the single consideration payable relating to the exempt part. [He referred to Southport Visionplus and Leightons, and continued:] Thus, for example, an attractive shop front must be viewed as bringing customers in to enjoy the whole range of products offered by the business, including the exempt dispensing element of spectacle sales, and cannot be directly attributed to taxable sales. By this logic very few areas of the premises can be directly linked to either taxable or exempt sales, possibly as few as the examining rooms (used for exempt eye tests) and the display cabinets (displaying the taxable goods). If very few areas of the premises can be directly linked to taxable or exempt their relative sizes will be very unlikely to give a fair attribution of the rent between taxable and exempt.

Another point that leads me to conclude that floor space is not fair and reasonable is the high levels of assumption needed to quantify areas. For instance your argument that the display cabinets area flows out into a two foot viewing corridor, and the width and very existence of mixed use walkways leading through to exempt areas behind. Clearly the higher the levels of assumption required, the less accurate the final answer will be. The high level of disagreements over the measuring of the one shop where an officer of this department has reviewed this illustrates this point and also raises the further concerns over whether this method is fully checkable. It would certainly be unreasonable if six plus site visits were needed, with protracted further correspondence, every time that the Commissioners wished to review the correct working of the method.

Lastly I would view the weighting of the premises by zones as arbitrary and artificial. That, of two equal floor area retail units, that with the greater frontage will command the greater rental is understandable. That some surveyors may employ a zoned calculation in setting rents to reflect this is possible. It remains a single supply of a licence to occupy land however. You cannot rent just the front or merely the middle third of a unit so the supply is a single one covering all areas under a single consideration."

Mr Mattacks said in his statement that he had seen all the relevant correspondence, and agreed with Mr Laney's letter of 12 September 2001, that the Commissioners did not accept that the standard method was not fair and reasonable for Optika, and that all proposed special methods had been considered with an open mind as to whether they were fairer than the standard method. He also said that in considering whether a special method was fair it was reasonable to compare it with something, and the standard method was a readily available comparator. The case officer was therefore entitled to make that comparison.

Mr Mattacks also said that the fundamental logic underpinning the Second Special Method was that the sales made from the retail area were largely taxable, but that that was not usually the case with opticians generally, and with the Appellant in particular, who viewed their principal trading activity, that of selling prescription spectacles, as a mixed supply. Mr Mattacks's view was that that mixed nature of the majority of the Appellant's supplies made a floorspace based method of partial exemption attribution very unlikely to achieve a fair and reasonable result.

In his oral evidence, Mr Mattacks said that half of the sale of a pair of spectacles was the sale of a prescription lens, and was therefore both taxable and exempt. It was not a fair assumption that the shop front was only for the purposes of selling spectacles, since it related to both frames and lenses. He said that he had not visited the Appellant's shops, but he had visited other opticians and they sell spectacles in their shops. There was no need to visit a shop to know that. When he spoke, in his decision letter, of "direct" attribution of spectacle sales to taxable supplies, he meant "exclusive" attribution. Referring to the two-foot area round display shelves and cabinets, he said that he could not agree that such areas were capable of precise measurement. People will walk where they like in a shop; if there was a designated walkway in a shop, how wide was it? He maintained that dispensing took place all over the retail area. Whether the six shops surveyed were representative or not did not concern him; what did concern him was that a large amount of time and effort would be required by both sides in reviewing the method and in carrying out measurements and calculations. He had been previously engaged in output apportionment methods, and had visited a large number of opticians' shops and went through their entire sales process. In every one that he had visited, dispensing was carried out in a comfortable area in the front of the shop, not in the less comfortable area at the back

Mr Mattacks said that he did not agree that it was illegitimate to compare the special method with the standard method; even if the standard method was not fair it was still reasonable to compare the special method with it, to see in what respect the unfairness was imposed. He had not considered whether the special method was fairer than the standard method. There was nothing in regulation 101 or 102 to say that a special method can only be requested if the standard method is not fair. The review was concerned with the question whether the special method was fair, not whether the standard method was not fair: but if the standard method was not fair it was up to the taxpayer to put up a special method that was. Attribution of input tax involved looking at what the premises are used for, looking at the building not merely the retail area, and the attribution of input tax on rent, and the rent is for the building not for one area.

Mr Mattacks agreed that you could have spectacles with no eye test and an eye test without selling spectacles. But if there were no eye test there could be no sale of prescription spectacles. He repeated that the attribution of input tax involved looking to see what the premises are used for. The process of dispensing was getting the right lens in the right frame for the customer. If he was not buying prescription spectacles, he was not receiving dispensing services. Mr Mattacks said that he had not asked the Appellant about the eleven points (in paragraph 7 of the amended statement of case). He had not asked where the discussion of suitable frames would take place (point (ii)). He considered that the examination of existing spectacles and discussion of associated problems (point (i)) would take place mostly in the shop, according to his experience. Administration (point (x)) would take place where the records were kept, or at the counter, not in the retail area. That, he said, was part of the activity of dispensing. Customers would wait in the waiting area (point (xi)). Mr Laney had also seen customers waiting in the retail area. Mr Mattacks agreed that if there were questions which needed to be answered, then the review was flawed.

Asked about the use of the premises, Mr Mattacks said it was possible to look at spectacles from two feet or three feet, or any distance. But the method postulated a precise measurement. If one were looking from two feet away one could also be just discussing frames. The Second Special Method was based upon an assumption that the retail area was all taxable and that the supply of spectacles was a single taxable supply. The shop window was there to attract customers in to make taxable and exempt purchases: therefore the shop window cannot be taxable. The Appellant was selling prescription glasses, and sought to draw people in so that they could sell them. Sun-glasses were fully taxable. Mr Mattacks said that he had not asked about the numbers coming in solely to buy sun-glasses. He had rejected the floor-space method because if there were very little exempt use of the area then there would be a very large recovery, and if a negligible use there would be virtually 100 per cent recovery. The Second Special Method required the assumption that the retail area was used to make taxable and exempt supplies. A floor-space method can give a fair and reasonable result, if large areas can be identified as purely taxable or purely exempt. If, for instance, half the premises were used purely for taxable and half for exempt supplies, each bearing its own costs, a floor-space method would be fair.

Mr Mattacks expressed the view that zoning was artificial, though it might not be if there were more than one lease covering different parts of the premises. But zoning was not relevant to the VAT status of the premises, because the Appellant had not paid one price for one part and a lesser price for another; they have paid a single rent for a single lease. It would make no difference, he considered, if the rent in the lease were apportioned between the zones, since there would still be a single supply. He said that he had not asked to see the lease.

Referred to the Customs and Excise Manuals, volume 6, chapter 2, in paragraph 7.9, headed "Directions", Mr Mattacks pointed out the passage which said,

"To direct the termination of a special method is in effect to direct the trader to use the standard method. In some cases this could produce a less fair result than the special method previously used."

That could also be true, he added, of a proposed special method; and it was possible that a proposed special method might be even less fair and reasonable. In a later passage the Manual said,

"Where you consider that a trader's method does not produce a fair and reasonable attribution of input tax to supplies with a right to deduct, and you have taken all reasonable steps towards persuading the trader to change to a more suitable method, you should first report the relevant facts of the case to your local Partial Exemption Liaison Officer (PELO). These will include [inter alia] details of a suggested method that would produce a fair and reasonable result compared to the existing method."

Mr Mattacks said that where the Commissioners imposed something on a taxpayer they should have reasonable grounds for justifying it, but it was not the same if the special method had been requested. Mr Mattacks was also referred to the following passage, in paragraph 7.10:

"Every case needs to be considered on the basis of all the available information; do not assume that a method that has been approved for an apparently similar trader will automatically be acceptable for your trader. There may be small but significant differences between the two traders or the person who approved the method for the other trader could have overlooked something relevant...."

Mr Mattacks did not consider that that sat uncomfortably with his not having visited any of the premises or not asking questions. The proposed special method was not fair and reasonable on the face of the shop front and retail area.

The last witness, called by the Commissioners, was Professor Mohammed Jalie, called by the Commissioners as an expert witness. He is a Fellow of the Association of British Dispensing Opticians (FABDO), has the Senior Award of the Spectacle Makers Company, and is an honorary Fellow of the College of Optometry. Currently he is Visiting Professor in Optometry at the University of Ulster. He defined those qualifications which were relevant to the provision of an optician's services. These were: an ophthalmologist, who is a medical doctor with a post-graduate qualification, probably a diploma in ophthalmology, and may be a Fellow of the Royal College of Surgeons or of Ophthalmologists. An optometrist is a person qualified to conduct eye tests, and to dispense spectacles or contact lenses resulting from the tests. A dispensing optician is qualified to dispense prescriptions from ophthalmologists or optometrists. Since changes in the law in 1984, a dispensing optician is not obliged to be a member of the ABDO nor to be registered with the General Optical Council; but the full range of dispensing services can only be provided by someone who is so registered, and in order to register it is necessary to be a member of ABDO.

Professor Jalie set out in his statement and in evidence to define "dispensing". He said that a dispensing optician has to consider what he termed "the Three Ps" of dispensing: the physiological fit, the physical fit, and the psychological fit of the spectacles. The physiological fit involved the power of the lenses and the form and centration of them. Also a suitable lens type must be selected, whether half lenses, or large lenses, or to cater for someone who needed to look through the top of the lenses. The physical fit involved the comfort of the frame and the appropriate frame for the type of lens. The psychological fit involved ensuring that the customer was satisfied that the spectacles meet the desired cosmetic and practical requirements. The activities which Professor Jalie listed as being included within the definition of dispensing were: the taking of accurate measurements by the dispensing optician; ensuring that the positions of the optical centre of the lenses are correct and the lens type suitable for the prescription; correct positioning, in the case of multifocal lenses, of the segments and fitting cross; inclusion, in the case of medium to high strength lenses, of the vertex distance as part of the prescription; choosing the lens shape to ensure minimum thickness and weight; determining suitable eye size, bridge size, side length, angle of sides, and head width. Professor Jalie said that all the activities listed in paragraph 7(i) to (xi) of the amended statement of case came within the definition of dispensing. We include paragraph 7 herewith as Appendix III.

In his oral evidence, Professor Jalie said that the sale of prescription spectacles cannot be divorced from dispensing. He had visited one of the Appellant's shops, and had formed the view that the tables in the retail area were dispensing tables. They were narrow tables so that the customer and an assistant could sit opposite and discuss whatever was necessary. He observed in the shop the selection of frames and the writing of order forms going on, both of which he would say were dispensing activities. He expressed the view that dispensing starts as soon as a person undertaking dispensing receives a prescription. When a customer comes in and gets an eye test and is brought back and handed over to the sales staff, that is when dispensing begins. An optometrist cannot give advice on lenses until he knows into what mount the lenses will be put. Not all frames would be suitable for a given lens. The sales assistant will be there to help the customer choose an appropriate frame. This will depend upon the use to which the spectacles will be put, and the preferred colour and shape, there being many different types of frame. Professor Jalie said that you cannot just choose a frame and assume that it is apt for the lens and the customer, otherwise there would be no need for dispensing services. He considered that fitting frames, selecting the appropriate frames, writing the order, discussing the lens with the customer, reaching agreement as to the recommended lens, discussion of lens coating, measurements, and discussing availability of a thick or thin lens all come within dispensing activities. Dispensing takes place, he said, wherever the person providing the dispensing services happens to be. The dispensing which he had seen took place at a dispensing table, some 6 feet from the front door.

In cross-examination, Professor Jalie said that of the items (i) to (xi), (i) to (vi) were part of the dispensing function and (vii) to (xi) were ancillary to dispensing. He agreed that (vii) to (xi) were not within the definition of dispensing in Southport Visionplus, which was why he considered them to be ancillary aspects of the dispensing service.

Grounds of appeal

That letter of 10 May 2002 from Mr Mattacks was the one which contained the Commissioners' decision as to the Second Special Method. It was followed on 31 May 2002 by the notice of appeal, which set out the following grounds of appeal:

"We wish to appeal on the grounds that the decision is irrational or unreasonable within the meaning ascribed to those terms in the decision in Associated Picture Houses Ltd [sic] v Wednesbury Corporation Ltd [sic] [1948]1 KB 223 in that the officer failed to consider relevant factors and failed to exclude from consideration irrelevant factors, and otherwise erred in fact and in law in that in particular he:

1. concluded that the service of dispensing is supplied in the sales areas of the Appellant's shops, notwithstanding that all the evidence in this case is to the contrary

2. assumed that the decision of the High Court in Customs and Excise Commissioners v Leightons Ltd [1995] STC 458 and the Tribunal in Southport Visionplus Ltd LON/01/0505 established as a matter of law that the shop front of an optician's shop is of necessity used to promote both the sale of goods and the supply of the services of dispensing and failed to consider the evidence on this point in the instant case

3. concluded that the method proposed rests on assumptions, notwithstanding that the evidence is to the contrary

4. characterised disagreements as to the use of the areas in one of the Appellant's shops as disagreement on measurements and thereby mistakenly concluded that the proposed method would not be readily susceptible to checking by HM Customs & Excise for assurance purposes

5. characterised the weighting of zones in the proposed method as 'arbitrary and artificial' for this purpose disregarding expert evidence that this methodology is standard practice in the retail sector

6. formed the view that only some surveyors adopted the methodology referred to at 5 above notwithstanding that the only evidence on this point was the expert evidence referred to at 5 above, thereby substituting an assumption not based on evidence for the said expert evidence

7. failed to acknowledge that the method proposed treats rent paid by the Appellant in respect of the shops in question as consideration for a single supply but rather formed the view that the method sought to apportion the rent

8. failed to give sufficient or any weight to the evidence of commercial practice in regard to rents of retail units and how this reflects actual use such that it demonstrates the unfairness of the standard partial exemption method

9. used the standard partial exemption method as a comparator

10. reached a conclusion that so conflicts with the evidence and the weight of the evidence that it could not have been reached without either considering irrelevant factors or failing to exclude from consideration irrelevant matters

The law

The law relevant to these appeals is contained in regulations 101 and 102 of the 1995 Regulations, which respectively reflect the provisions of Articles 19 and 17.5 of the EC Sixth Directive. So far as bears upon these appeals, those regulations provide:

"101-(1) Subject to regulation 102, the amount of input tax which a taxable person shall be entitled to deduct provisionally shall be that amount which is attributable to taxable supplies in accordance with this regulation.

(2) In respect of each accounting period—

(a) . . .

(b) there shall be attributed to taxable supplies the whole of the input tax on such of those goods or services as are used or to be used by him exclusively in making taxable supplies,

(c) no part of the input tax on such of those goods or services as are used or to be used by him exclusively in making exempt supplies, or in carrying on any activity other than the making of taxable supplies, shall be attributed to taxable supplies, and

(d) there shall be attributed to taxable supplies such proportion of the input tax on such of those goods or services as are used or to be used by him in making both taxable and exempt supplies as bears the same ratio to the total of such input tax as the value of taxable supplies made by him bears to the value of all supplies made by him in the period.

. . .

102-(1) Subject to paragraph (2) below ... the Commissioners may approve or direct the use by a taxable person of a method other than that specified in regulation 101...."

Paragraph (2) of regulation 102 has no application in these appeals. Since reference was made to it, we also set out the relevant parts of Group 7 of Schedule 9 to the 1994 Act:

"Item No

The supply of services by a person registered or enrolled in any of the following—

(a) . . .

(b) either of the registers of ophthalmic opticians or the register of dispensing opticians kept under the Opticians Act 1989 or either of the lists kept under section 9 of that Act of bodies corporate carrying on business as ophthalmic opticians or dispensing opticians

. . .

NOTES

(1) . . .

(2) Paragraphs (a) to (d) of item 1 ... include supplies of services made by a person who is not registered or enrolled in any of the registers or rolls specified in those paragraphs where the services are wholly performed or directly supervised by a person who is so registered or enrolled."

The Appellant's contentions

Mr Julian Ghosh and Mr James Henderson appeared for the Appellant, and provided us with a skeleton argument. Mr Ghosh began by saying that what the Commissioners should be looking at in these appeals is the basis of the actual use of the premises as they are used to generate the tax. The Appellant has suggested two different special methods, both of which were refused. It is the Appellant's contention that the Commissioners have acted unreasonably in refusing to accept the special methods, in that they took into account irrelevant factors. The special methods were confined to the rent of the properties: all else is dealt with under the standard method. The appeals are concerned only with the method employed by the Commissioners in making their decisions. The review officer would have had to look at the correspondence in this case in order to see if the standard method was fair or the special method fairer. All else, Mr Ghosh said, was irrelevant. There was no relevance in the Commissioners' supplemental bundle of documents (though there was no objection to its being put in). There was no relevance in Professor Jalie's evidence as to the meaning of "dispense". The appeal was not to do with a notional view of "dispensing", it was concerned with the use to which the Appellant put its properties, and whether the Commissioners have exercised their discretion reasonably in the Wednesbury sense. It was all a matter of fact, and nothing to do with expert evidence.

Mr Ghosh submitted that there were five features to the concept of reasonableness. First, that the Commissioners must consider whether the current method of apportionment was a fair and reasonable attribution of input tax. The correspondence contained no letter which dealt with that point. Secondly, the Commissioners must then consider whether it was reasonable to change to a fairer method: that was a comparison which the Commissioners must make. Thirdly, the decision must take account of all relevant matters and not of irrelevant matters. Mr Ghosh referred to BMW (GB) Ltd v Customs and Excise Commissioners (1997) (Decision 14823) at paragraph 68. Fourthly, the special method must be linked to an assumed or estimated use (see Article 17), be reasonable to operate, and be capable of being checked: see Merchant Navy Officers Pension Fund Trustees Ltd v Customs and Excise Commissioners (1996) (Decision 14262) at paragraph 25. Both BMW and Merchant Navy suggest a three-stage approach: first, look to see if the standard method is fair; next, look to see if it is reasonable to change to a special method; and third, change the method using the proper procedure. If the special method is to do with use of the premises, how could the Commissioners fail to look to see if the standard method was operating as it should when answering the question whether the special method was fair and reasonable? To say that it is not necessary to look at the standard method, Mr Ghosh contended, is wrong. The critical thing, Mr Ghosh said, is the actual use of the premises: that actual use cannot be ascertained without looking at the premises. Mr Laney did not visit the premises, nor did the review officers, Mr Dowling and Mr Mattacks. The fifth feature is that the review is a de novo exercise. The Appellant's case was that the Commissioners had not used the right method in considering the two special methods put to them. In De Smith, Woolf and Jowell (7th edition, page 347) it was said to be enough to shew that a decision-maker was materially or substantially influenced by irrelevant considerations, and need not be shewn that those considerations were the only ones taken into account. Nor was it necessary to shew that the decision would have been different if the irrelevant factors had not been considered.

The First Special Method was set out in the letter of 26 October 1999. The proposition was: if you actually took at the use to which the Appellant put its premises, given that the Appellant paid more rent for certain parts, that is a distortion and is unfair. The first thing that Mr Laney should have done was to find out whether the standard method was unfair: that is the first requirement of regulation 102. There was no letter from Mr Laney which dealt with the unfairness of the standard method, his letters deal only with the special method. Step 4 of the method (see paragraph 25 above) links the attribution of input tax to the use of the premises. Mr Laney's letter of 17 December 1999 (see first quoted paragraph in paragraph 26 above) was simply wrong, he having misunderstood. But step 4 and the Appendix to that letter made no reference to floor space, and clearly used the input tax directly attributable to taxable supplies and the input tax directly attributable to exempt supplies as the only variables in determining the recoverable proportion.

In the same letter, Mr Laney did not accept that only the consulting room related to an exempt supply, all waiting, reception, and dispensing areas relating to mixed supplies. But, Mr Ghosh pointed out, Mr Laney had never visited the shop and had no means of knowing. The special method was based upon use; in order to make such an assertion Mr Laney ought to have found out if it were accurate, or at leased asked. Although Mr Laney admitted, in that letter, knowing little about valuation of premises for retail purposes, he had none the less rejected attribution of costs on that basis. The reference to carpeting further illustrated his misunderstanding. None of Mr Laney's arguments in that letter was relevant to the inputs method proposed by the Appellant, which involved no division of property-related input tax according to floor space. It involved, Mr Ghosh said, determination of the proportion of property-related input tax that may be recovered according to the ratio of other (non-property-related) input tax directly attributable to taxable supplies to other (non- property-related) input tax directly attributable to exempt supplies. There was a meeting on 28 February 2000 at which Mr Ward explained to Mr Laney that he had misunderstood, and that an inputs method, not a floor-space method, was being proposed. At that meeting this appeared to have been understood.

The matter of the single rent being paid for a single supply, Mr Ghosh said, was a recurrent point. There was only one lease, so how could it be relevant that the floor space was zoned? Whether money was spent on a single supply of a lease or more than one was not relevant. Zoning was a standard practice in quantifying rent. But if a taxpayer spends more on more expensive parts of the premises, in order to attract customers in to buy frames, and less money on dispensing supplies, that is a very relevant matter. But the method proposed concerned the use to which the property was put. If a lessee pays more for one part of the premises than for another, there is still a single lease, but that is not the point. In theory it would be possible to have more than one lease for different parts of the premises. But different zones are valued in different ways, and that leads to an unfair attribution in the standard method. Mr Laney's letter of 20 March 2000 (see paragraph 29 above) he misses the point in the first paragraph and in the paragraph numbered 1. In paragraph 3 of that letter Mr Laney made an assertion of fact as to something which he did not know. At best, without a visit, it was an educated guess from previous experience. The fact was that the front of the shop was used to sell spectacles and the back of the shop was used to sell eye tests. Mr Laney had made no attempt to find out what part of the dispensing process took place in the retail area, nor did he address the point as to whether the standard method was unfair or the special method fairer. In his letter of 18 April 2000 to Mr Rishi (paragraph 30 above) Mr Ward stressed that, and also suggested a visit to an optician's shop, since none had been made up to that point. Miss Nagarajah made a similar assertion of a fact that she could not have known in her letter of 7 July 2000. The reply by Mr Ward on 8 September 2000 (paragraph 32), conceded the point that the proposed method ignored input tax which was not directly attributable to either taxable or exempt supplies, adding that that was no reason why the method itself was flawed. It was a question of fact whether dispensing took place in the retail area, not a matter of expert opinion, and is a matter of fact for the Tribunal. In Southport Visionplus the Tribunal, at paragraph 14, adopted expert evidence that the skill of dispensing optical products is "to convert an optical prescription into a pair of spectacles, or other optical appliance, appropriate to an individual patient's needs". The Tribunal also considered paragraph 14 of the ABDO's "Advice and guidelines on Professional Conduct for Dispensing Opticians", which contained a list of ten matters which the dispensing optician should normally ensure.

The matter of supervision of those who were not registered within Item 1(b) of Group 7 was also a matter of fact. Expert evidence of that was rejected by the Tribunal in Land. Some of the activities carried on in an optician's shop may not be under supervision because they do not need to be, for instance items vii, viii, and ix in the amended statement of case. But Mr Laney should have made inquiries about that as well, and failed to do so.

Referring to the review letter (see paragraph 33 above), Mr Ghosh said that the reasoning was indefensible, and that there was no evidence to shew that previous misunderstanding of the method had been reversed. The letter asserted that

"The method proposed for [the Appellant] does not have a direct link to the supplies made and does not reflect the ultimate use to which the purchases are put. In other words in my view it does not provide a reasonable apportionment of input tax to the ultimate use of the supplies."

Again, an assertion of fact which he could not have known without a visit: if that is not irrelevant, Mr Ghosh asked rhetorically, what is? But Mr Laney said that he did not heed a visit. But he made no inquiry as to purchase methods, or as to purchases of standard-rated goods. A hypothetical example should not have been present in his mind; he should have been asking what actually went on, not what might go on. Again, in the review letter, there was no consideration of whether the standard method was fair. The Commissioners are obliged to use the comparison to find the fairest method, and no such comparison was made here. In their amended statement of case, the Commissioners adopt what Mr Laney said in his letter of 25 October 2000, and do not accept that the bulk of shop overheads can be directly attributable solely to taxable supplies, which is irrelevant. The Appellant accepted the assertion of the Commissioners in paragraph 19 of their skeleton argument, that the Appellant's assertion that the decision in the letter of 25 October 2000 was influenced by Mr Laney's earlier misunderstanding, was based not on the letter itself but on the assertion that the contents of earlier letters had not been retracted. The contents of the previous letters which were based upon the misunderstanding should have been retracted.

The decision, Mr Ghosh contended, was infected both in its terms and by Mr Laney's misunderstanding of all the correspondence up to 18 September 2000. In evidence Mr Laney had said that he had recollected some of the facts, when asked if he had read the correspondence. Either he did at least in part, base the decision on the previous misunderstood correspondence, and if so whether he recollected it or read it does not matter since he was looking at facts which were clothed with his misunderstanding, or, if he did not so base the decision, there was simply an immediate rejection of the special method without more. The reasons in that letter were improper reasons, even if taken in isolation. Those reasons concerned with bulk buying and comparison of recovery between opted and not opted property were misconceived They were both reasons for rejecting the special method because recovery of input tax could vary, even though the use of the premises did not. Bulk buying was purely a hypothetical example. The Commissioners made no inquiry as to purchasing, nor did they suggest averaging. Mr Laney said that he had relied on his visit to Ealing, in which he was considering sales and not attribution of input tax. But only a fraction of the time spent at Ealing might have been spent looking at matters which were relevant to the use of the premises. Visiting is important: the Manual tells officers to visit, for good reasons, because different traders have small but significant differences.

The Commissioners were contending that the proposed method was unconnected with the differing use which is made of the property, and that the inputs method posited no connexion between the recovery rate and the use to which non-attributable tax was put. They said that there is no reason to suppose that the property related non-attributable input tax would be consumed in the sale proportion as the quotient of taxable input tax. Mr Ghosh said that it was not a question of "supposing". The standard method was unfair, but the inputs method, which was the pre-1991 standard method, was fairer.

In summary, as to the first appeal,

(1) Not a word was said as to whether the standard method was fair, let alone fairer than the First Special Method.

(2) There was a clear misunderstanding of the proposal which was never expressly retracted by the same person who made the decision.

(3) According to the decision in BMW and Merchant Navy, one should look at the actual or estimated use of the property. That cannot be done without visiting the property, or at least making proper inquiries as to representative shops (it was not suggested that all shops should be visited).

(4) The express grounds of refusal in the letter of 25 October 2000:

(i) the assertion of fact as to linkage of use to the special method with no knowledge of the facts;

(ii) the hypothesis of what might be done, not what actually was done;

(iii) that rent might fall if the Appellant paid VAT on rent, which was not sensible.

Any one of those would be enough for the first appeal to succeed. The Appellant invited findings of fact on all their complaints which were complaints as to the facts, all of which were contained in the correspondence.

Mr Ghosh said that the rejection of the Second Special Method was even more infected that that of the first. Again, no consideration had been given to whether the standard method was fair; Mr Mattacks had made no visit to any premises, but had relied on his experience of other shops, and said that he was not obliged to visit; and the special method was rejected because of assumptions made by the Commissioners, and with little recognition of what was involved in dispensing. What was dispensing was not only a matter of fact for the Tribunal, but was a matter of fact for Mr Mattacks as well. In Mr Behan's letter of 30 April 2002, it was conceded, for the sake of agreement, that the general retail area could be treated as being for mixed use; but once it was conceded that the display units were for wholly taxable purposes an area immediately round those units must similarly be used. Mr Mattacks had asked no questions relating to dispensing, invited no debate or discussion, made no visit, made no inquiry about sales of sun-glasses, made no inquiry about items i to xi, all of which he should have done. He referred to a "high level of assumption", when all he had to do was look or ask.

Mr Ghosh said that the measurements that the Appellant had taken reflected what the Appellant actually did with the properties. As Mr Ward had said in his letter of 22 May 2001 (see paragraph 42 above), the floor space was capable of analysis into taxable or exempt use; the benefit of a floor-space method being that it did accurately reflect the actual use of the property. Detailed measurements were made in a number of shops. The floor-space approach was not an all or nothing point. If a small amount of exempt business was carried on in the retail area, it is a matter of attribution of the amount of use for exempt supplies. It was no part of the Appellant's case that one has to postulate a particular area as, for instance, a walkway. That was simply a concession made by Grant Thornton. In his letter of 12 September 2001, Mr Laney had based his rejection of the Second Special Method only on his unaccompanied visit to Bluewater, and not on any visit to any other shop. He had had no other visit in mind.

As to the note of the meeting on 17 July 2001, Mr Ghosh said that he was asking for a finding of fact on point iii of that note: Is it the case that the front of the shop is designed to attract customers to buy frames or to attract customers also to buy dispensing services?

The reasons for rejection given by Mr Laney in his letter of 12 September 2001 (see paragraph 49 above) were not acceptable reasons. In the first, he expressed doubts as to the calculations relating to the representative shops. That was not a proper basis for rejection: he should either have asked for more information or abandoned that as a reason for rejection. In reasons 2 and 3, the point is missed by asking whether an area is used exclusively for taxable purposes: the real point is, what is the area actually used for, as a matter of fact? His decision was based on one unaccompanied visit, during which Mr Laney did not ask if what he saw was typical. The evidence shewed that bookings were made at the counter, and that the waiting area was the other side of the counter from the front. The walkway was simply a notion which the Appellant had invented as a concession for the purpose of obtaining agreement. The expenditure on the retail area was heavily weighted towards the sale of frames and spectacles. For eye tests the customer has to go to the back of the shop. Reason 4 did not appear to be a ground of objection. In any case, it was not accepted that it would be onerous for the Commissioners to visit the sites. If, as Mr Laney said, that was a matter upon which agreement might be reached, it could hardly be an objection. In reason 5, again Mr Laney had missed the point. Very few, if any, customers would buy frames only. The relevant question is, is the retail area (and the money spent upon it) used to attract customers to buy frames or to buy exempt optician's services?

Mr Ghosh repeated and stressed the points made by Mr Hartley in the letter of 16 October 2001. On the matter of the retail area, he added that the display cases were obviously a taxable use. The idea of the two-foot margin for looking at them was arbitrary, but if one were in that two-foot space one could hardly do anything but look at the frames. That two-foot area must therefore be used for taxable purposes. Even if it were also used for dispensing, one must ask what its primary use is: for what purpose is money spent on that area? It is necessary to find out what the area is used for and what, and how much, the ephemeral use is. Under the heading of "Use of the front area", Mr Ghosh contended that the point had again been missed. That part of the shop attracts customers to buy frames, not eye tests. There are spectacles all the way up to the counter. A customer buys an eye test at the counter, and waits for the test the other side of the partition.

Exemption applies to the supply of services by a person enrolled in either the register of ophthalmic opticians or of dispensing opticians, or on the list of those carrying on business as ophthalmic opticians or dispensing opticians under section 9 of the Opticians Act 1989: Item 1(b) of Group 7 of Schedule 9 to the 1994 Act. Note (2) to that Group provides that exemption will also apply to the services supplied by a person who is not so enrolled but whose services "are wholly performed or directly supervised by a person who is so registered or enrolled". The Commissioners, Mr Ghosh contended, should take into account the fact, if it be so, that in some shops the services were supplied by those who were under supervision and in some shops not.

Mr Dowling's letter of 26 November 2001 to Mr Hartley (see paragraph 60 above), again begs the question, Mr Ghosh contended. The first paragraph of the reasons for rejection, once again, made an assertion of fact without inquiry having been made. Mr Dowling referred to no visit to any Optika shop. In the second reason, the fraction was too big. There was no reason to include the exempt area in the denominator of the fraction. It was not clear why this was an objection. The third reason also begged the question, by merely stating that the precise method of apportionment of the refurbishment costs was not specified. As to the fourth reason, zoning was the standard method of valuation for rent; it cannot, therefore, be called artificial. Here again, Mr Dowling makes an assertion of a fact of which he has no knowledge, that the weighting is not representative of the actual usage f the space. It was improper to say that "it can be seen that the proposed method (including weighting of the rent) would allow an unjustifiably higher recovery rate than either the standard method or the proposed method without the weighting". Again, the point is, what is fairer, not whether there is an unjustifiably higher recovery rate. It is not relevant that you might recover more than under the standard method. Mr Ghosh referred to and adopted the letter in reply from Mr Behan of 30 April 2002, and described Mr Dowling's letter as a misdirection as to each and every ground, in that they addressed the wrong questions.

The further review letter by Mr Dowling, dated 10 May 2002, also contained, under the heading "Logic behind my conclusion", assertions of fact without any basis in knowledge. Further, the decisions in Southport Visionplus and Leightons had nothing to do with the question of what use was made of the premises. The question was, has the Appellant spent its money to make people buy spectacles or eye tests? Eye tests are bought at the counter. Quantifying the areas is not a matter of assumption. Under "Response to specific points", all proposed special methods should be fully considered with an open mind as to whether the standard method is fair and whether the proposals are fairer, as Mr Mattacks said, but it is not enough simply to say so. Mr Ghosh took issue with the assertion that it was necessary to compare the special method proposed with the standard method. The test is whether the standard method was fair, which depends upon the actual facts, and what was in fact done: Merchant Navy. There was nothing to be said for comparing the special method with the standard method: that was an obvious misdirection. Mr Mattacks should have gone to look at the shops, and should have asked if the shops were representative and whether the staff were qualified or supervised. He did none of those. He simply compared the standard method with the special method, which can lead nowhere. The review was infected by all five of the complaints (see paragraph 83 above), and a sixth, that the review officer did not go and look when he should have, or at the very least should have asked.

The Commissioner's contentions

Mr Owain Thomas, for the Commissioners, also produced a skeleton argument. He said, first, that regulation 101(2)(d) was in two parts: first, that which is to be apportioned, and secondly the ratio of apportionment. That is the standard method. It is subject to regulation 102, according to which either a special method may be requested for approval, or it may be the subject of a direction. A special method has to have some link to use: Article 17. Regulation 102 contains no condition upon which the Commissioners' discretion is to be exercised, so that the discretion is a wide one. Section 23(3) of the 1994 Act provides that the Commissioners have to fulfil an overriding condition in making regulations for special methods, namely that of securing a fair and reasonable attribution of input tax to supplies under subsection (2). The question for the Commissioners on an application for a special method is, whether the special method provides fair and reasonable attribution. They are not required to find that the standard method is unfair; it may be fair, and any number of special methods may also be fair. To say that the Commissioners are obliged to decide upon the fairness of the standard method is therefore wrong, and failure to do so cannot render the decision unreasonable in the Wednesbury sense. Determining whether a special method is fair and reasonable can be done without looking at the standard method, which is not linked to use, whereas a special method is linked to use.

A taxpayer is entitled to apply for a special method if it can be shewn that the standard method is unfair, but that is not a necessary prerequisite. The concept of the special method being "fairer" than the standard method is not to be found in the legislation. If there should be two unfair methods, the Commissioners are not entitled to approve a special method which is unfair. The test is to ask at each stage whether the special method secures a fair and reasonable attribution of input tax.

In BMW (GB) Ltd v Customs and Excise Commissioners (1997) (Decision 14823), the Tribunal considered that its task was to examine whether the statutory condition was satisfied, and whether the Commissioners decision was reasonable, not whether the special method concerned was fair and reasonable. That was a case in which a special method had been imposed upon the taxpayer. In the present case the only condition for exercise of the Commissioners' discretion was that the special method should secure a fair and reasonable attribution. If that condition was satisfied, the next consideration was whether the method was easy to operate and verify. The Commissioners were entitled to look at the previous method, but were not required to do so. They would be entitled, but not required, to approve a special method if it were fair and reasonable, and if it were easy to operate and verify; there must also be a demonstrable link between the method and its use. The Commissioners must consider properly all material put before them. There was no requirement that they should visit the premises concerned, and it is not correct to say that they cannot consider the use of a special method without a visit.

Rejection of a special method can properly be based upon a sound understanding of the nature of the taxpayer's business, informed by knowledge of the kind of business and the nature of the supplies made. Here the bulk of the supplies made were prescription spectacles, and included eye tests, sun-glasses, and other supplies. The Commissioners can compare that with the use to which it is claimed that the input tax is put, and form a judgment as to whether the input tax is used to a greater degree for taxable supplies than is allowed. The Tribunal's jurisdiction is to decide whether the decision was Wednesbury unreasonable; if it is decided in favour of the Appellant, the appropriate remedy is to make a direction that the Commissioners make a proper decision. It would be inappropriate to make a finding of fact on the basis of the evidence, including the site visit and the evidence of Mr Allen, as to whether the retail area was primarily used by customers to buy frames as opposed to supplies of dispensing services. That the whole of the retail area is used for dispensing is correct. A conclusion to the contrary would not be properly made on the basis of the site visit and Mr Allen's evidence. There would be no sense in finding that the retail area was used exclusively or dominantly or mainly for taxable supplies. The retail area is used for both taxable and exempt supplies on each occasion when the supplies are brought together by the dispensing services. An attempt to divorce supplies of prescription spectacles from the use of the premises is foredoomed to failure, Mr Thomas contended. The supply of spectacles takes place in the shop; the shop is used to make those supplies. Also the exempt supplies exist. Sales assistants work in the retail area, and therefore supplies are made in that area; the same is true of the dispensing opticians. This has nothing to do with the eye tests, which are a complete red herring in both appeals., since it is agreed that those areas are used exclusively for exempt supplies.

Once you know what the dispensing services are, you can decide what is an automatic part and what is ancillary. Whether the member of staff dispensing needs to be supervised or not is not relevant: all dispensing is exempt. Mr Thomas referred to the facts of and various passages in the Tribunal decision in Land v Customs and Excise Commissioners (Decision 15547).

Mr Thomas referred to a letter of 25 August 1998 from Mr Laney to Mr Ward, in which he said,

"I had suggested to you that I considered that as spectacle frames were separately priced at the branches of Optika, I saw no reason why VAT could not be accounted for them as a separate supply. Our HQ supply of goods branch have accepted your arguments that the frames are an integral part of the overall supply of spectacles, the payment of which also covers the cost of dispensing."

The frames being an integral part of the spectacles, they have to be dispensed, Mr Thomas contended. That establishes what is happening in the shops to the frames, from removal from the rack to delivery to the customer. That can only happen in the retail area. Professor Jalie's evidence was that dispensing starts when you hand a prescription to the optician or to the sales assistant; that is from after the eye test, on hand-over, until the customer receives the spectacles and leaves. The spectacles are supplied by the dispensing optician or by someone under the supervision of the dispensing optician, working in the retail part of the shop.

In the letter of 31 August 1999, Mr Laney was answering the application for a special method, he expressed the view that the reception area of the shop could be used for dispensing, and the taxable area would therefore be minimal, and the majority of the public area would be mixed. Mr Thomas pointed out that in his answer Mr Ward failed to address that point, but just made an assertion, that the exempt use of the properties, namely the sight test rooms, were sited away from the areas which were valued at a higher rate. He said that no evidence had ever been submitted that the use of the property could be analysed as to where dispensing takes place. The Commissioners analysed the special methods on the basis of the evidence before them. No account was given in the statement of the method or the appendix to it as to how it was said to reflect the use of the property, only a mathematical formula. The resulting recovery would be higher than the standard method, but was not connected to how the input tax was used.

As to advertising, Mr Thomas contended that the Appellant's advertising was not restricted to taxable goods. The advertisements for spectacles applied to prescription lenses only. There was an unfounded assumption that the front of the shop was attributable to taxable services. In fact its purpose was to attract customers in to buy the whole package of dispensed spectacles. An advertisement for the sale of spectacles was designed to generate both taxable and exempt supplies. It would be artificial to construe the advertisements as encouraging the purchase of only one particular aspect of the supplies.

The letter of 17 December 1999 was clearly written under the misunderstanding that Mr Laney thought that the special method was based upon floor space. The meeting on 28 February 2000 recorded that Mr Ward had explained to Mr Laney that he had misunderstood and that a purchase based method was proposed. Mr Laney acknowledged that he had misunderstood. It appeared to be accepted by Mr Ward that at the end of the meeting Mr Laney correctly understood the proposal. In his explanatory letter of 29 February 2000, Mr Ward said that the purchase based method "more accurately reflects the actual use of the properties". Mr Thomas contended that it did not. Miss Nagarajah's letter of 7 July 2000 contained no assertion that the proposed special method was floor based. That letter is not affected by Mr Laney's misunderstanding. Miss Nagarajah said that expenditure on the retail area was related to both taxable and exempt supplies, since dispensing would take place in that area. This was a point which Mr Laney had wished the Appellant to address, but the Appellant had not done so. Miss Nagarajah also observed that the proposed method ignored the input tax which was "not directly attributable to either taxable or exempt supplies from the denominator". Both those reasons were relevant and entitled the Commissioners to reject the proposal. Miss Nagarajah also said, and it was the Commissioners' case, that in her view the supply of dispensing was an integral part of the supply of spectacles: although promotional material would concentrate on the sale of spectacles, the advertisements generate both taxable and exempt income. Advertising costs must therefore be residual expenditure.

Mr Thomas contended that the infection of the first review letter by Mr Laney's misunderstanding of the nature of the proposed method was not borne out by the contents of the letter itself: on the contrary, it was clear that that misunderstanding was not a part of the consideration. There being no reference to previous correspondence, the misunderstanding could not come into it. Mr Laney did not say that he had read all the previous correspondence and taken it into consideration. He had looked at the letter of 8 September 2000 quite separately. He had not disputed that he had had a misunderstanding about the earlier proposals, but he looked at that of 8 September 2000 as a new proposed method, and had not reviewed the earlier correspondence. There was, therefore, no evidence that he had taken the previous correspondence into consideration. His letter of 25 October 2000 makes this clear: the first paragraph refers specifically to the proposal of a new method, and the second paragraph raised questions which could only relate to a new method and not to the old. In the third paragraph he describes the new method as an input tax based method, which he would not do if his mind were affected by the misunderstanding that the first method had been floor-based. Nothing is mentioned about floor space.

The formula for the new proposed method (given in Mr Ward's letter of 8 September 2000 (see paragraph 40 above) gives, Mr Thomas contended, a relatively high rate of recovery. The question which Mr Laney was asking himself was whether it was a proper reflexion of the use to which the premises were put. He did know what supplies were made by the business, and whether they were taxable or exempt, or both (in relation to spectacles). He had a general knowledge of the business of an optician, and specific knowledge of the Appellant's business in negotiating the floor cost method on supplies. The element of dispensing involved the optician and sales staff in dispensing. Mr Laney knew where the supplies took place, as a matter of his experience and from the visit to Ealing, i.e. in the Appellant's shops. Those shops are used to sell spectacles. Mr Kane's evidence had been that the process of selling spectacles was similar in all opticians' shops. The Ealing visit had been for the purpose of looking at outputs, and Mr Laney had viewed the process of buying a pair of spectacles. But even if he was considering outputs, he was still able to see what went on. In relation to property costs, Mr Laney had knowledge of the way that property was used. On the basis of his experience, the proposed formula did not reflect the use to which the property was put. That was his conclusion. The method did not link recovery to use, and nothing in the proposal of 8 September 2000 shewed that it did. That link was fundamental.

The Commissioners were, Mr Thomas contended, entitled to consider the fairness of the standard method, but were not obliged to do so. It was not Wednesbury unreasonable not to.

Mr Laney's examples of the consequences of bulk buying, in his letter of 25 October 2000, do shew that the proposed method produces a greater recovery, which would not be based on or linked to any increase in the taxable use of the non-attributable items. There was no evidence that the Appellant could not or would not make bulk purchases. This was not, Mr Thomas said, a fatal objection, but it was relevant. The second example given was not hypothetical. Some of the properties were opted to tax. The question to be answered was not whether a proposed method secures the greatest recovery of input tax, but whether the method were fair to both sides. It was not accepted that the standard method can result in a higher recovery in spite of there being no change in use. If the proportion sales of (non-prescription) sun-glasses to total sales increased, the Appellant would recover input tax on the purchases, and would also be able to recover a greater proportion of non-attributable input tax. That is a matter that was not within the Appellant's control, because it is seasonal. But the Appellant has control over bulk buying. If the retail area use is greater in the summer there is a higher recovery for that area. Mr Thomas added that none of the Appellant's criticism of this review letter was included in the Appellant's case before the hearing.

In paragraph 31 of the Appellant's skeleton argument, it is contended that that it is self-evident that the Commissioners reached their decision based upon the misunderstanding of the proposed method, and summarised the Commissioners reasons in the following terms:

"(i) that floorspace apparently relating to taxable supplies actually related to mixed supplies;

(ii) that the different valuations of different parts of the premises do not justify the suggested method because the premises are rented as a whole; and

(iii) that some property related costs (e.g. carpeting) will not vary according to the value of the property concerned."

None of those reasons, Mr Thomas said, appeared in Mr Laney's decision letter. The basis for the misunderstanding was absent from that letter. The decision was not vitiated because Mr Laney had not retracted the previous correspondence. The letter of 8 September could be read and understood without reference to the previous correspondence. It was treated as a new method, so that there was no reason to refer to that previous correspondence. The Commissioners were not relying upon the previous correspondence. In paragraph 33 of the amended statement of case, the words:

"The Commissioners have considered the arguments put forward in correspondence by the Appellant's representatives and have rejected them for the reasons given."

did not mean that all the proposals were rejected for all the reasons, though the proposals were considered.

Referring to Mr Ward's letter of 29 February 2000, in paragraph numbered (i), described the zoning principle (see paragraph 36 above). Mr Thomas said that the first three sentences of that paragraph were accepted: zoning was the method by which properties were valued for rent, and was a useful technique, referable also to rating and rent reviews. But it is not relevant, and is artificial, he contended, in the context of input tax recovery. The commercial reality is that the tenant does not pay more for the front than for the back of the premises, but pays a cohesive whole for the premises as a whole. The next sentence, that the Appellant pays different amounts for different areas was a non sequitur; the Appellant pays different amounts for different shops, not for different areas. The matter of advertising (paragraph (iii) of the letter) was not separately pursued, but it was a fact that the space occupied by advertisements was used for mixed supplies, including advertisements on mirrors and sale notices on the shelves, all of which were in the retail area. The second bullet point in Mr Ward's letter of 18 April 2000 (see paragraph 38 above) is also a non sequitur and is not the fact. It is not true to say that more rent was paid for areas where more taxable activities took place than for areas where fully exempt activities took place. The whole rent, Mr Thomas contended, was paid for the whole of the premises. Mr Thomas said that the third bullet point shewed a complete ignorance of dispensing. No attempt had been made to meet the question of identifying a dispensing area or shew that dispensing does not take place in the retail area. The second paragraph under the third bullet point gives a good reason why a floor-space method is inappropriate, namely the practical difficulty of identifying and measuring the floor space attributable to taxable and to exempt supplies. The letter says that a purchase based method was preferable. Mr Thomas agreed with that, adding that those difficulties were real.

Referring to Mr Ward's letter of 22 May 2001, Mr Thomas pointed out that it had not been previously agreed that the standard method was inappropriate for input tax referable to the property, as the second paragraph of that letter said. The third paragraph stated, in its first sentence, that it had been established that it is possible to apportion the floor accurately. Mr Thomas contended that if a claim such as that was made one would expect to find the result of the research making it possible, but no such result was to be found, as Mr Laney said in his reply on 25 May 2001. There was no evidence upon which to found that claim. The mere assertion was not accepted. Referring to that letter, Mr Ward said again, in his of 5 June 2001, that the Appellant was able to analyse accurately specific use of floor-space, and promised the detailed analysis as soon as the measurements had been completed. The analysis was not set out in that letter. Mr Ward also asserted that the analysis would demonstrate that the mixed area was significantly smaller than the taxable and exempt areas, and that the whole of the retail area was not mixed, although both taxable and exempt supplies were made from it. Mr Thomas said that it would be difficult to reconcile those claims unless such an analysis was really capable of being made.

The further information relating to the floor-space method, offered in the letter from Mr Ward of 19 June 2001 (see paragraph 45 above), provided no analysis of the use to which the property was put. There was nothing, Mr Thomas said, to substantiate what the previous letter had said, nor did that of 19 June set out to do so. All that the letter said was that measurements had been taken. That was not enough to say what the use of the area was. In those paragraphs (set out in paragraph 45) what the Appellant has identified are not areas with no connexion with supplies of services. Trying on spectacles or looking at and choosing frames are particularly a part of dispensing. The analysis is seriously flawed. The allocation of space for viewing the frames in the displays was not a concession at all, but part of the method itself. Mr Thomas contended that it could not be properly or accurately done. The method actually posits artificiality, in the imaginary area for customers to stand at the counter and the imaginary walkways. In a shop like that at Bluewater, with tables and display cases in the middle, there was no evidence that customers took one path to the counter rather than another. Furthermore, no allowance was made for the space actually taken up by the tables and chairs used for dispensing. Mr Thomas contended that the method took no account of the dispensing which was admitted to take place in the retail area. The Appellant had asked, "What else could a customer do in the two-foot area but look at the frames?" The answer to that question was, try frames on, consider whether they were suitable, ask the assistant's advice about the frames. Dispensing, Mr Thomas argued, was not uncomplicated, involving physical acts and the giving of advice; the idea that that can be attributed to a defined space was inherently implausible.

In the table of measurements relating to Bluewater (Appendix I), the retail area is seen to be used almost entirely for taxable supplies, Mr Thomas observed. There was no mention of the dispensing tables, nor of the general nature of dispensing services supplied in zone A. The only thing taken into account was the walkway. Two things would be fundamental to such an analysis: to measure floor-space and give the value, and to attribute space to taxable, exempt, or mixed use. But there was no such analysis. The letter addressed only zoning, and also made some assertions. Mr Laney and Mr Mattacks rejected these assumptions and unrealistic assertions on the basis of experience, Mr Laney on the basis of his dealings with the Appellant and his two visits. Mr Mattacks had had all the material available and all the correspondence. Mr Allen and Mr Kane would not have had the experience to be able to say whether dispensing took place in the retail area. All the evidence suggested that dispensing does take place in the retail area: Messrs Laney and Mattacks said so, and Professor Jalie's definition confirmed what Messrs Laney and Mattacks had understood, and suggests that that was the case. Dispensing started at the point of hand-over from the optometrist to the sales staff, and that would continue in the retail area. The result of the analyses produced a result which varied between 85.37 per cent and 97.98 per cent taxable use.

In Mr Ward's file note of the meeting of 17 July 2001 (see paragraph 48 above), it was recorded that the Appellant maintained that the entire front of the shop was intended to attract customers by displaying standard-rated goods. Mr Thomas said that the question was not why any part of the property was used, but how. He said also that the Commissioners did not accept that the actual display areas with a floor area round them were wholly taxable.

In his letter of 12 September 2001 (paragraph 49 above), in the paragraph numbered 3, Mr Laney had reached a perfectly good conclusion, Mr Thomas contended. He had set out clearly what he had seen at the shop, from first-hand observation. It included having seen that dispensing activities were going on in all parts of the retail area. His visit report also described activities in the retail area which were dispensing. Paragraph numbered 2 of the letter of 12 September shewed how many assumptions had to be made, and shewed also how inaccurate the Appellant's measurements of the area were. Also in that letter, Mr Laney gave as his view that the proposal would allow a distorted recovery for refurbishment costs. He accepted, however, that the Appellant was not achieving a sufficient recovery in respect of the spectacle display racks.

Turning to the section headed "Retail area" in Mr Hartley's letter of 16 October 2001 (see paragraphs 54 et seq. above), Mr Thomas observed that the first of the above quoted paragraphs contained nothing that was very different from the original proposal, and that that paragraph was not in the nature of a concession. The second of those paragraphs revealed a basic misunderstanding: the last sentence of that paragraph was mistaken, what was described, the fitting of a frame on a customer was actually dispensing. The Appellant was seeking to define "dispensing" out of existence. In saying that "consulting rooms have been treated as exempt notwithstanding that opticians are also actively involved in selling standard rated goods," shewed a fundamental misunderstanding: it was not a question of who was carrying on the activities, but where they were carried on. As soon as the Appellant started supplying prescription spectacles it was making exempt supplies when the customer moved into the retail area to choose a frame. As to the use of the front area, the point, Mr Thomas said, was that it is dispensing that takes place there, not eye tests. The Appellant said that the floor-space method was not predicated on splitting up the rent, but that was exactly what that method did. That section of the letter made an important concession, that the Appellant was prepared to apportion property costs, other than refurbishment costs, on the basis of the standard method.

The critical factor in the reasons given by Mr Dowling for rejecting the proposed method, in his letter of 26 November 2001 was that the proposed method, including weighting of rent, was not just that it would allow a higher recovery of tax but that it would allow an unjustifiably higher recovery. What the trader was paying was the rent for the property, and the artificial breakdown into zones was not acceptable. The Commissioners were entitled to look at what a trader recovered under the standard method. Mr Dowling also looked at what the result would be without the weighting the of the rent into zones, which was also perfectly legitimate. Where the standard method had not been shewn to give an unfair result it was legitimate to compare the recovery produced by it with the proposed special method. If there is a large discrepancy it is necessary to examine why, and to see whether that difference is produced by a fair method of analysing the use to which the input tax is put. If the proposed method is significantly different from the standard, it cannot be said that that is something which you may not take into account.

In his letter of 30 April 2002, Mr Behan criticised the use of the standard method as a control method, but said that a request for a special method was predicated on the standard method not giving a fair result. But, Mr Thomas contended, he had not used it as a control mechanism but as a factor for consideration. Further, Mr Thomas contended, that any special method was predicated on that special method itself, and not on the result produced by the standard method.

Mr Behan's reply (see paragraph 65 above) to indent 1 of the letter of 26 November 2001 expressed the Appellant's view that it could not sensibly be stated that any dispensing activity took place in the retail area. Mr Thomas said that it was not the Appellant's view of the use of the retail area which mattered, but what it was actually used for. It was wrong to state that an area round a display was used for a wholly taxable purpose, forming "part of the curtilage" so as to form a unity with the display unit in terms of its use. There was nothing to suggest that such an area was used for wholly taxable supplies, and nothing in Mr Behan's letter countered what Mr Laney had seen on his visit. The display units were used to display spectacles; the floor area was not exclusively so used.

That Mr Mattacks came to his own decision on a fresh look at the case was evident, Mr Thomas contended, from his letter of 10 May 2002. This was based upon his experience. That he did not visit Bluewater did not vitiate the decision because his conclusions were fair. Mr Mattacks referred to both Leightons in the High Court and Southport Visionplus in the Tribunal in both of which it was found as fact that dispensing was a major requirement of customers who went to an optician to buy spectacles, and he had given his view that an attractive shop front must be seen as bringing in customers to enjoy the whole range of products offered, which included dispensing. It was also the case that Mr Allen, in his evidence, had referred to the full package which draws the customers in. It was conceded that a fleeting use of an area for exempt purposes would not break the exclusivity of taxable use if taxable use were otherwise the exclusive use. Mr Laney's visit to Bluewater shewed him that the claim that the measurements were accurate was readily contestable, Mr Thomas argued. That arose in large part from the suggested walkways, which were wholly artificial, and the ignoring of the tables. The shop was naturally divided into two walkways which were divided by the tables.

Summarising the facts, Mr Thomas submitted, first, that the Commissioners' conclusion, based upon the evidence of Mr Laney's visits to Ealing and Bluewater, that the use of the retail area was mixed was correct. Professor Jalie's conclusion, that dispensing starts when a customer is handed over to sales staff must also be a correct and reasonable view. There was no basis for finding that the retail area was used wholly for taxable supplies. Secondly, what constitutes dispensing is a question of fact. It was the Commissioners' case that all dispensing was carried out either by qualified staff or under direct supervision, and they were prepared to accept that, in the absence of evidence to the contrary, as a fact. The output apportionment of 58.8 per cent included in it as component attributable to time spent by dispensing opticians and unqualified sales assistants. The selection of frames could only be done in the retail area. It was the Appellant's evidence that the profit of the business was not in the eye tests but in the sales of spectacles, and that included a dispensing element.

The Appellant's reply

Mr Ghosh made five points relating to the appropriateness of the Commissioners' response to the special methods that were put forward by the Appellant. First, that the Commissioners should, in making a considered response, have had regard to whether the standard method was fair. Article 17.5 gives a right to deduct input tax, and provides that a standard method is given; but to safeguard the rights of taxable persons special methods can be used. Regulation 102 imposed a condition: the special method has to be fair or unfair in the context of the standard method. What else can reasonably be done but to start by looking to see if the standard method was fair.

Secondly, the Commissioners made no visits. It was a question of fact whether there was direct supervision and also what activities were supervised. The proper question was not what is meant by dispensing, but whether an unqualified person was doing it under supervision. The question, what activities were being supervised, should have been addressed. They did not include greeting a customer, asking him to wait, taking payment, or administration. It is impossible to find out except by asking. To say that dispensing starts when a customer is handed over to sales staff did not sit well with the notion that greeting a customer was part of dispensing. Mr Laney did not ask whether the activities numbered (i) to (xi) were supervised, nor could he have known whether the retail area was used for supervised activities without asking. The Commissioners Manual, at paragraph 7.10, pointed out that one cannot rely upon assumption because traders vary. It is a question of fact, not of experience. The decision was vitiated because improper reasons were given because the Commissioners asked the wrong questions or did not ask at all, in the face of what the Manual said was good practice.

Thirdly, whether Mr Laney's decision was infected by his misunderstanding was a matter of fact for the Tribunal to find. If he was not so affected, the response to the letter of 25 October 2000 was a rejection without any consultation, negotiation, or comment, simply a bare refusal, which cannot be said to be a proper exercise of discretion. The second rejection, in the letter of 12 September 2001 was also without discussion after any of the letters. Again, it is not a proper response to say simply "No".

Fourthly, in the first appeal, the consideration of the possible bulk purchase was misunderstood. If there were a large upturn in sales of sun-glasses that would give a higher recovery. If on the Appellant's purchase method a very large purchase of sun-glasses was made, what would be wrong with that? The retail area would be used to a greater extent for taxable sales. Mr Ghosh adopted the Commissioners' argument on that point. Again, no inquiry was made by the Commissioners on this relevant point.

Fifthly, it was a non sequitur that zoning was not relevant to input tax attribution. There was a distinction between the valuation method and consideration for a supply, but no difference. The test was, what was the Appellant paying for, and what was it using the property for? The point of the site visit was to let the Tribunal see how the premises were used, segmented into different areas. But it was at least possible that there could be heavy use of the retail area for unsupervised activities, and the Commissioners should have asked about that. The question, what is dispensing, was not mentioned in Group 7 of Schedule 9. The Commissioners should consider the special method properly, look at the premises and ask the right questions. If they did that their decision would be unimpeachable. They should have confined finding facts to whether the retail area could have been used for a type of activity which was not directly supervised.

The Tribunal's jurisdiction

Mr Ghosh referred the Tribunal to the decisions in Cliff College Outreach v Customs and Excise Commissioners (2001) (Decision 17301), University of Exeter v Customs and Excise Commissioners (2003) (Decision 18117), BMW (supra), and Merchant Navy (supra) on the question whether the Tribunal has a full appellate jurisdiction or only a supervisory jurisdiction in these appeals. Mr Ghosh also said that he was content to base the matter on the reasonableness of the Commissioners' decision in the Wednesbury sense, because the remedy for a successful appeal was the same, namely a direction to the Commissioners to make a proper decision on a further review of each appeal. We considered the relevant paragraphs of each of those decisions.

In Cliff College, at paragraph 35, the Tribunal preferred the view that the Tribunal had an appellate jurisdiction as being the only conclusion consistent with section 83(e) of the 1994 Act,

"which confers jurisdiction on the Tribunal in respect of 'the proportion of input tax allowable under section 26'; it does not confine that jurisdiction to the validity or otherwise of any direction made by the Commissioners relating to such matters."

In University of Exeter, the Tribunal came to a similar conclusion, which adopted the observation cited above of the Tribunal in Cliff College Outreach. In BMW, the Tribunal took the contrary view, for these reasons (in paragraph 64 of that decision):

"Section 26 provides that the amount of input tax for which a taxable person is entitled to credit is so much of his input tax as is allowable under regulations. Section 26(3) provides that the regulations are 'for securing a fair and reasonable attribution of input tax'. Regulation 102(1) provides that Customs and Excise 'may direct' the use of a method of attribution of input tax. As regulation 102(1) confers a statutory discretionary power on Customs and Excise it follows, from the principles which we have identified above, that the jurisdiction of the Tribunal in the present appeal is limited to determining whether the discretionary power was properly exercised."

And in paragraph 67, the Tribunal held that its task was not "to decide whether the directed method 'secures a fair and reasonable attribution of input tax' but only to decide if the decision to direct the method was properly taken, bearing in mind the statutory condition."

In Merchant Navy, both parties had submitted that the Tribunal's jurisdiction was appellate, and that the Tribunal should decide for itself on the evidence whether, at the time of the directions, the 1991 standard method continued to achieve a fair and reasonable result, rather than being limited to a consideration of whether, in making the direction to use a special method, the Commissioners had acted unreasonably in the sense in which that expression was used in John Dee Ltd v Customs and Excise Commissioners [1995] STC 941, 952, applying the judgment of Lord Lane in Customs and Excise Commissioners v J H Corbitt (Numismatists) Ltd [1980] STC 231. However, the Tribunal in Merchant Navy declined to decide the issue, though inclining to the view that the jurisdiction was appellate and proceeding with the appeal on that footing. For that reason evidence as to the matters to which the officers had regard in making the direction was not called. The appeal was successful, and the Tribunal said also that had the jurisdiction been limited to the John Dee approach it also would have succeeded.

It appears to us that since the Tribunal can exercise an appellate jurisdiction on an appeal as to the proportion of input tax allowable under section 26, it must follow that the Tribunal's jurisdiction is also appellate in considering whether that proportion has been rightly decided, that is, over the fair and reasonable attribution of input tax to taxable supplies, in accordance with the regulations made for that purpose. The Tribunal must therefore be allowed to examine the method by which the Commissioners came to their fair and reasonable attribution, and look to see whether it is both fair and reasonable. In this context, that use of "reasonable" appears to us not to be the same as the Wednesbury or John Dee concept, because it is the attribution that has to be reasonable. However, it might also be in issue whether the method whereby the Commissioners reached their decision (which they would maintain was fair and reasonable) was reasonable also in the Wednesbury sense. For that reason we agree with the Tribunal in Cliff College Outreach, that an appellate jurisdiction is the only jurisdiction that is consistent with section 83(e). The appeals have been argued before us on the basis that the special methods themselves were more fair and reasonable than the standard method, and also on the basis that the Commissioners acted unreasonably in the Wednesbury sense in rejecting them. It appears, therefore, that a case such as this falls into both categories. The Tribunal may consider whether the actual proportion of input tax allowable is a fair and reasonable apportionment, and may also consider whether the refusal to exercise their discretion to approve a special method was Wednesbury reasonable.

It also follows, in our judgment, that we must look to see whether the standard method, which has in effect been imposed upon the Appellant, is fair and reasonable or not, and whether either of the special methods proposed is fair and reasonable. We do not consider that we should decide whether, if we were to conclude that more than one of the three methods was fair and reasonable, one method (standard or special) is more fair and reasonable than another, but simply whether one or another of the three methods under consideration is or is not fair and reasonable in its result. We should also consider whether the refusal of either of the special methods was Wednesbury reasonable. The question, as it seems to us, is not whether one method is fairer than another, in particular whether a special method is fairer than the standard method. That is to postulate that the standard method is fair and another method is more so. It is difficult to see how, if one method is fair another fair method can be more fair, without suggesting that the first of those methods is not entirely fair. Therefore, as it appears to us, we should consider whether one method is unfair and another is fair. Even if our jurisdiction were only supervisory, we would still have had to hear the evidence in order to ensure that the Commissioners had arrived at their decision having taken into account all relevant matters and ignored all irrelevant matters.

The issues

The principal issues which the parties have raised, and which we have to consider, are the following:

(1) Whether either of the proposed special methods would have produced a fair and reasonable recovery of input tax; whether the standard method produced, in this case, a fair and reasonable result?

(2) Whether, in rejecting the two special methods, the Commissioners considered all relevant factors and left all irrelevant matters out of consideration in coming to their several conclusions that neither special method would have produced a fair and reasonable result?

A number of secondary issues fall to be examined on the way to determining the principal issues. These are largely factual, and are as follows:

(A) Whether the retail area was used for both taxable and exempt supplies. That involves consideration of what, so far as the evidence shews, the retail area was actually used for, and what is meant by "dispensing". Also within this issue falls the question, whether the front of the shop was used to attract customers to buy standard-rated goods or to buy the "whole package" of standard-rated and exempt goods and services? The relevance and effect of zoning also comes within this issue.

(B) Whether unqualified employees under direct supervision were concerned in dispensing activities?

(C) Whether Mr Laney's misunderstanding of the First Special Method coloured his view of that method after 28 February 2000, and if so with what consequences?

(D) Whether the Commissioners' conclusions were based upon evidence or upon the officers' assumptions and experience?

(E) Whether it was necessary to look at the standard method in order to see if a special method is fair? Also, whether it was legitimate to use the standard method as a comparator against which to set the special method?

Reasons for conclusions

While the site visit to Bluewater was useful and informative, with hindsight it would have been much more so had it taken place on day three or four of the hearing. By then the factual issues, and, in particular, as to what the retail area is actually used for, would have become clear, and our attention would have been more keenly directed to what was actually taking place when customers came in. It would have been useful if our attention had been specifically directed to those matters, so that we could have spent considerably more time observing those activities. As it was, we did not know enough about what we were supposed to observe, nor were we told at the time. We therefore do the best that we can on the evidence that was put before us at the actual hearing.

We begin with the secondary issues which, as determined, should give the conclusions on the principal issues.

(A) The actual use of the retail area

We were assisted by a number of pieces of evidence. Mr Kane said that the Appellant was less interested in selling eye-tests than in selling spectacles; it was critical that the customer would be attracted into the shop to buy frames. He said that the retail area was designed to attract the maximum number of customers in to buy spectacles. We were told that the bulk of the Appellant's profits came from the sale of spectacles, by which was meant the complete package, a complete pair of dispensed spectacles, or sun-glasses. Sight tests were offered as part of the package. Mr Kane also told us what the normal procedure was when a customer came to buy spectacles, which included being shewn and advised on frames in the retail area. Mr Allen told us that when the optometrist had concluded his tests and report, the customer is handed over to the sales staff in order to select a frame. The staff were capable of dispensing, and many of the shop managers were qualified dispensing opticians.

In his letter of 23 October 1996, Mr Ward said that it was not possible to identify space used exclusively for making taxable supplies, which was why the floor-space method was calculated in the was in which it had been. He also said that most of the residual floor space was used for taxable purposes and the remainder for mixed purposes, resulting, he said, in the calculation of exempt floor space being biased towards the exempt side. Mr Laney said that on his two visits, to Ealing and Bluewater, he had observed dispensing activities going on in the retail area. That consisted largely of customers seeking advice on frames from sales staff. We were also told during the visit that the tables in the retail area might be used for such things as measuring a customer's eyes and face for the purpose of fitting frames. We did not understand that it was disputed that those activities were carried out in the retail area. It is also the case that sun-glasses, of which only a small proportion are dispensed, and spectacle cases and other accessories are displayed and sold in the retail area. With the exception of the dispensed sun-glasses, these are all taxable goods.

Professor Jalie gave evidence as an expert witness. He said that when a customer has had an eye test and is then handed over to the sales staff was the moment when dispensing began. It included the taking of measurements by the dispensing optician, ensuring that the positions of the optical centre were correct, and that the lens type was suitable for the prescription, and correct positioning of the segments of multifocal lenses, choosing the lens shape and the dimensions of the frame. He said that all the activities listed in paragraph 7(i) to (xi) of the statement of case were all included in dispensing, though later he gave the view that items (vii) to (xi) were only ancillary to dispensing, and did not come within the definition of dispensing given in paragraph 21 of Southport Visionplus (supra). In that paragraph reference is made to the ABDO's Advice and Guidelines on Professional Conduct for Dispensing Opticians, which, in paragraph 14, begins:

"The purpose of the dispensing function is to translate an optical prescription into an order for a pair of spectacles, or other optical appliance, appropriate to the individual patient's needs. A registered dispensing optician should normally ensure:...."

and there then follow ten subparagraphs detailing what the dispensing optician should do. These include: ensuring that the spectacles supplied should be appropriate to the individual requirement, that adequate measurements are taken, that frames are suitable, giving reasons where it is considered that a frame is not suitable; checking the finished spectacles for quality and accuracy, and also for comfort and function on the individual. Patients should also be told the cost of the professional services and of the spectacles before they commit themselves to payment. In our judgment, those two statements of the meaning of "dispensing" appear to agree with each other, to be realistic, and to be correct. Accordingly, for example, we cannot accept Mr Ward's contention in his letter of 16 October 2001 (paragraph 54 above), that "If a member of staff fits frames on a customer in [the retail] area that activity is more readily characterised as retail than dispensing."

We accept Mr Kane's evidence that zoning of retail premises is, now, a normal commercial method of arriving at rent, which is invariable used. We accept also that the greater part of the expenditure of the Appellant upon its premises is devoted to the retail area, which is in the front of the shop: zone A. It was the repeated contention of the Appellant that the important consideration in these appeals is the purpose for which the premises are used, including the purpose for which the particular zones are used, the purpose for which the Appellant spent its money on the premises. As a principle, that appears to us to be correct, and indeed we did not understand that that was in dispute. If we have correctly understood the argument of the Appellant, it was that since the rental value of zone A was very much higher than zones B and below, and that therefore taxable activities took place in zone A rather than the other zones, that fact should be taken into consideration by apportioning the input tax on the rent of zone A at the rental value put upon zone A. Not so, say the Commissioners: rent is a single charge for a single supply, and the values for rent and rating purposes have nothing to do with expenditure for input tax purposes. Examples were suggested, which we did not find very helpful. One, suggested by Mr Laney, related to carpeting the premises, and he observed that carpet would cost a particular sum and would be the same in all zones. We suspect that if the Appellant were to carpet the whole premises, the carpet used in the retail area would be of significantly higher quality than that in the back premises. We did not regard the argument that zoning is artificial as carrying very much weight. It clearly is artificial, but then so is the whole system of attributing and apportioning residual input tax where both taxable and exempt supplies are made.

The logic of the Appellant's approach is not readily apparent, especially in the context of the actual use to which zone A (and, indeed, the other zones) is put. It may well be that the expenditure upon zone A would be higher than on other zones, since it consists entirely in retail area, and is the part of the shop which the public see from the outside and when they first enter the shop. But one cannot have zone A without the other zones forming the particular premises, and the rent is expressed as a single rent for the whole premises. The zoning is, it appears to us, simply a method of calculation, for the purposes of ascertaining the rent and the rates, and not a designation of a separate rent for a separate zone. We conclude, therefore, that the rent valuation of the zones has no place in the calculation of input tax relating to the premises.

Dispensing services are exempt, and the sales of frames and lenses is taxable. It follows from the above, that in our judgment both taxable and exempt supplies were made in the retail area.

(B) Were unqualified employees engaged in dispensing under supervision?

The evidence was that 13 out of the 29 shops had qualified dispensing opticians as shop managers. Evidently the remaining 16 had not. Note (2) to Group 7 of Schedule 9 provides that Item 1(b) applies to services supplied by a person who is not registered or enrolled in any of the specified registers or rolls if directly supervised by a person who is so registered or enrolled. But there was no evidence before us that any such services were supplied in any of the Appellant's shops by any person who was not either so registered or enrolled or was directly supervised by someone who was. We feel sure that there would have been such evidence had it been the case. In the circumstances, the facts and decision in Land (supra) are of no assistance in this case. We therefore assume that there was no unqualified employee engaged in dispensing otherwise than under direct supervision. On the basis of that assumption, it follows that all dispensing services supplied by the Appellant were exempt under Note (2).

(C) The extent and effect of Mr Laney's misunderstanding of the First Special Method

It is common ground that Mr Laney's misunderstanding of the nature of the First Special Method persisted until at least 28 February 2000, when Mr Ward, as his note (paragraph 35 above) records, explained to Mr Laney that the method was not floor-space based and that for the purpose of input tax relating to property a purchase based method had been used. Also according to the note, Mr Laney conceded that he had misunderstood. The explanation was reinforced in a letter from Mr Ward the following day. It was the Appellant's case that in reality Mr Laney continued after that to misunderstand the First Special Method. This contention the Appellant based upon Mr Laney's reply on 20 March 2000 (paragraph 37 above), in which he described the method as "being one based on a purchase cost of the property by reference to valuations of areas within the property". Further, one of Mr Laney's reasons, in that letter, for rejecting once again the special method was because a single price was paid for the property, and it was not rented in separate areas for separate considerations. Requesting a review of Mr Laney's decision in his letter to Mr Rishi of 18 April 2000, Mr Ward again pointed out that Mr Laney had misunderstood, and repeated that the proposed method was not based upon floor space but upon the use of the property.

It was at that point that the matter was referred to Miss Nagarajah, who considered that the formula ignored the input tax that was not directly attributable to either taxable or to exempt supplies and omitted it from the denominator, and that such an apportionment would give a very high recovery, possibly 100 per cent. Such an apportionment would not give a fair and reasonable recovery rate, and for that reason she could not accept the method. In his reply of 8 September 2000 Mr Ward made the concession that the standard method should be used for all input tax except that relating to the properties, and that for tax relating to properties the ratio of taxable input tax to all input tax, expressed as a percentage should be used. Miss Nagarajah considered that to be the proposal of a new method, and had passed the matter back to Mr Laney, who then wrote the letter, dated 25 October 2000, which was the decision now under appeal. It is that letter (paragraph 41 above) that we examine, to see whether Mr Laney's earlier misunderstanding persisted and influenced his consideration and rejection of the method considered by Miss Nagarajah to be a new method.

The first paragraph quoted in paragraph 41 above makes no reference at all to floor space, and, indeed, treats the proposed formula, as set out in the letter of 8 September, as a new method. Its contents appear to us accurately, if briefly, to reflect the formula itself, amplified in the telephone conversation of 23 October. The following three paragraphs appear to reject the proposed method on the basis that it had no direct link to the supplies involved and did not reflect the ultimate use to which the purchases were put. Two examples were given in order to illustrate this. In our view the first example does not do so. If a "bulk purchase" of new shelfing were made (and it is difficult to see how such a purchase could be other than a "bulk" purchase) being for taxable purposes there is no reason why the input tax should not be deducted. Nor is it certain that there would be no increase of taxable usage of non-attributable items; there might be an increase or a decrease: the purchase and the amount of use of the shelves may well be unrelated. It also seems to follow that, if this is an apt example, any replacement of equipment which is used exclusively either for taxable or for exempt supplies is going to give rise to a distortion. The second example appears to be correct, since, as Mr Laney said, the VAT on the rent would appear only in the denominator. But the important thing is that there is no reflexion of Mr Laney's misunderstanding in that letter. We conclude from the above that Mr Laney's misunderstanding continued until his letter of 20 March 2000. However, in our view, the letter of 25 October 2001 is not so infected.

(D) Whether the Commissioners' conclusions were based upon evidence or assumptions and experience?

Mr Laney's experience consisted, evidently, of a visit to the Ealing and Bluewater branches of the Appellant and, as he said, some knowledge of opticians' business. (Our own knowledge is based upon the evidence we have heard and seen and a visit of about an hour and a half to Bluewater.) Mr Laney said that his observation shewed him that dispensing activities were being carried on in the retail area of the Bluewater shop. The Appellant's case was, as we understand it, that there should have been more visits, and that the review officers should have visited one or more shops, and that they should have asked a good many questions about what the procedures were in the Appellant's shops. They relied upon advice given in the Commissioners' Manual: "Every case needs to be considered on the basis of all the available information; do not assume that a method that has been approved for an apparently similar trader will automatically be acceptable for your trader. There may be small but significant differences between the two traders or the person who approved the method for the other trader could have overlooks something relevant...." That appears to us to be sound, if not obvious, advice, and in cross-examination Mr Mattacks said that he did not consider that that his not having visited any of the premises and not having asked questions was entirely consistent with that advice. We are led to the conclusion that Mr Mattacks may well not have considered all the available material, because he did not avail himself of it. Mr Laney, we consider, clearly considered more material. But the Appellant complained that he could not have known enough about the manner in which the retail area was actually used without longer observation and asking relevant questions. It is pertinent to observe here, that during the long course of correspondence and the several meetings between representatives of the parties, it would have been very easy for the Appellant to ensure that the Commissioners understood exactly what they ought to know, and to direct their minds towards those things, suggesting visits so that they could be shewn what they ought to have been considering.

What effect did the failure to visit or to inquire have upon either of the review decisions? we ask ourselves. The important part of Mr Laney's letter was contained in the third of the paragraphs quoted in paragraph 41 above, where he says that the formula set out in Mr Ward's letter (paragraph 40) "does not have a direct link to the supplies made and does not reflect the ultimate use to which the purchases are put. In other words ... it does not provide a reasonable apportionment of input tax to the ultimate use of the supplies." That declaration is followed by the two examples referred to above. That is all that is said by way of reasons for rejection of the method regarded as being a new method after the First Special Method. (The actual First Special Method had already been rejected, and is apparently not the subject of this appeal.) The reasoning, which is not expressed by Mr Laney, must presumably be that, unlike the standard method, the proposed formula does not multiply the ratio of taxable input tax to all input tax by the ratio of taxable supplies made to all supplies made. It is difficult to see that there is any significant difference, if any difference at all, between the formula proposed in the appendix to the letter of 26 October 1999 (see paragraph 33 above) and the formula regarded by Miss Nagarajah as a new method, unless the denominator of the former, taxable inputs plus exempt inputs, is intended to omit input tax relating to mixed supplies, so that the denominator is smaller than that under appeal. Neither of them, however, includes any function of the supplies made, taxable and exempt, by the Appellant. Given that the inclusion of the ratio between taxable supplies and all supplies is a necessary ingredient, it is not easy to see how any special method can differ significantly from the standard method. If that is the only reason for rejection (and at no time were the reasons made specifically clear), then it does not appear that any observation or inquiry was necessary beyond what was contained in the figures which are included in the calculation for the VAT return. We therefore conclude that the rejection of the First Special Method, with which we include the New Method, did not depend upon observation or visits but only upon the concept as shewn in the two formulae.

As to the Second Special Method, it had been contended that there was no visit by Mr Mattacks, that he relied wholly upon his experience of other shops, nor had he asked any questions relating to dispensing or any other aspect of the business; all he did, the Appellant said, was compare the standard method with the special method. Mr Mattacks's reasons (set out in paragraph 68 above) refer to the "assumption" that the main sales of the Appellant's business are taxable sales of spectacles and exempt sales of eye tests, giving little recognition to dispensing. It was also, he said, an "assumption" that an attractive shop front would attract customers to buy spectacles (taxable) only, and he stressed the statement in correspondence that a member of staff fitting frames on customers was "more readily characterised as retail than dispensing". He stressed also that a sale of prescription spectacles was a mixed supply of taxable goods and exempt dispensing. We find his use of the word "assumption" in this context strange, since there was no assumption of those matters, they were averred by the Appellant. However, disregarding that word, the evidence that we have heard has shewn that what he said was factually correct. It was the Appellant's evidence that the principal sales were those of spectacles, and that the shop front was intended to attract customers in to buy spectacles, including "designer" frames. The sale of spectacles included, to a substantial degree, the sale of dispensed spectacles, which on any view were sales of taxable goods and exempt services. The customers were attracted in to purchase the whole package, as Mr Allen said. Mr Mattacks did not set out to define dispensing nor to specify what amount of dispensing went on in the retail area, but based his decision upon the evident fact that what was sold was a combination of taxable goods and exempt services. It was not apparently necessary to his decision that he should have specified what proportion of dispensing went on in the retail area. We find that he did, indeed, fail to make any visits or inquiries such as the Appellant contends that he ought. But it also appears to us that the facts which underlay his decision were correct. He had apparently based it upon the facts set out in the correspondence, and not upon assumption or his experience alone.

Mr Mattacks's second point was that high levels of assumption were needed to quantify the areas. By that we understand him to refer to assumptions that there were notional walkways and a notional space of two feet in front of and round display cabinets. This objection relates to the ease of calculation and of review of the method when necessary. Since the walkways and the two-foot area were indeed notional, it did not appear to us that any visit or inquiry would elucidate the matter usefully. But from the diagram that was produced it did appear to us that, if anything was artificial, these notional areas were, and that they would be very difficult to calculate if that were necessary.

The expression "artificial and arbitrary" as applied to the practice of zoning we regarded, in the context of partial exemption, as being unrealistic. The very concept of the standard method is itself artificial, necessarily so, since it is an attempt to put some kind of rule on to a form of attribution that would be very difficult to calculate with accuracy. It is, as it seems to us, a method of compromise between a complicated calculation and guess-work. Mr Mattacks's statement that "some surveyors may employ a zoned calculation in setting rents" appeared to us to be mistaken; we accept the evidence that it is now common practice and more or less invariably used in calculating rents and rates. But we repeat our view that it has no application in the attribution of input tax relating to the use of premises.

(E) Comparison of special method with standard method

Regulation 101(2) sets out the standard method of attribution and, while mandatory, is expressed to be subject to regulation 102, which gives the Commissioners a discretion to allow the use of a special method. Nowhere is the concept of fairness mentioned in the regulation, nor in Articles 17.5 or 19 of the Sixth Directive. However, it must be axiomatic that a special method must be fair; it would be absurd to contemplate either the request of a method which was unfair to the taxpayer or the approval or imposition by the Commissioners of one which was unfair to the revenue. The way in which the 1985 Regulations are expressed suggests that the concept of comparative fairness as between the standard method and a special method does not come into the matter when a special method is requested. The requirement of fairness arises in the enabling section, section 26(3), which empowers the Commissioners to make regulations for "securing the fair and reasonable attribution of input tax" to supplies for which the taxpayer is entitled to credit. Article 17.5 decrees that where goods and services are to be used for the making of both taxable and exempt transactions only such proportion of the VAT shall be deductible as is attributable to taxable transactions. The next part of Article 17.5 provides that that proportion shall be calculated according to Article 19, and sets out a formula for that calculation, which is in fact the standard method. Where Article 17.5 permits other methods of calculation, these refer to the determination of a proportion for each sector of the business, deductions on the basis of the use of all or of part of the goods and services, or the authorisation or compulsion of the taxpayer to make the calculation in respect of all goods and services used for all transactions. The scope for special methods is therefore narrow, but it is not suggested that any comparison of the methods is contemplated. The 1985 Regulations appear to be a little broader in the scope for special methods. However, it appears to us that there is no reason why a special method should not be approved even if, in a different way, the standard method is fair. On the other hand, if the standard method is fair, there appears to be no reason why a special method should be approved, provided that the discretion to refuse approval is properly and reasonable exercised. We do not think that it would be proper for the Commissioners to say, that the standard method is fair, so that there is no need to look at any special method. In our view there is no reason why, if a special method is requested, the standard method should be looked at at all, unless the proposed special method is considered by the Commissioners to be unfair; in that case the special method would not be approved and the standard method would prevail, as in the present case. We conclude, therefore, that it is not necessary to look at the standard method to see if a special method is fair, nor is there any relevant purpose in comparing a proposed special method with the standard method: if a special method is not fair the standard method is imposed in any event. But if the standard method is not fair, then clearly a special method should be adopted.

We return now to the principal issues in these appeals.

(1)(a) Whether the standard method produced a fair and reasonable result

One of the surprising things about these appeals was that although it was many times asserted that the standard method produced a distorted result, at least so far as rent was concerned, and that either of the two proposed special methods would produce a fair and reasonable recovery, none of the methods was worked out using the actual figures, or at all, to illustrate just how the fairness or distortion appeared in reality. It was stated many times that the essential was to reflect the actual use to which the properties were put. By the Commissioners it was stated that it was necessary for the special method to reflect, or be linked to, the actual use to which the input tax was put. Both those statements appear to us to be obviously correct. In these appeals we are concerned only with the treatment for VAT of the rent of the various premises, since in the correspondence the main issue was fined down to that. For brevity, the standard method can be expressed thus: (IT_tx S_t) ÷ (IT x S), where IT_t = input tax referable to taxable supplies, IT = all input tax, S_t = taxable supplies, and S = all supplies. If one applies that method to the Appellant's situation, IT will include all input tax including that in respect of rent (where an election to tax has been made) and S will include all supplies made by the Appellant. The link or connexion with the actual use to which the properties, and therefore the input tax, have been put is therefore apparent. It is not apparent, on the face of it, where any distortion lies, unless and until the concept of zoning is introduced so as to attribute a higher level of input tax to the retail area. It is, of course, the Appellant's case that the use to which it put the inputs was concentrated in the retail area because it was in zone A. However, we have already decided that the rent of the premises must be taken as a single sum for the whole premises in each case. In the light of that decision, it does not appear to us that there is any distortion in the standard method. That does not preclude an application for a special method, nor does it prohibit the Commissioners from approving a special method.

(1)(b) Whether the First Special Method would have produced a fair and reasonable result

In considering the First Special Method, we include the method considered to be a new method by Miss Nagarajah (to which we will refer as "the New Method"). The formula for residual input tax in the First Special Method can be expressed in terms similar to those above: (IT_t ÷ (IT_e + IT_t)) x IT_m= R, where IT_e = input tax referable to exempt supplies, IT_m = input tax referable to mixed supplies, R = recoverable element of IT_m.This is derived from the formula in Mr Ward's letter of 22 September 1997 (paragraph 31 above). In this method there is no reference to the proportion of taxable supplies to all or other supplies, and the fraction is in effect taxable input tax divided by taxable plus exempt input tax, neglecting that referable to mixed supplies, rather than taxable input tax divided by all input tax. The connexion between the input tax and supplies is implied, being present only in so far as input tax is necessarily related to purchases for the purpose of making supplies. But the formula does not appear to us to have a direct connexion with the supplies actually made.

The formula proposed in Mr Ward's letter of 26 October 1999 (paragraph 33 above) was: (IT_t÷ (IT_t + IT_e)) x 100, being the recoverable percentage of residual input tax. That was not the formula for the New Method, however, which was different again (see paragraph 40): (IT_t÷ (IT_t+ IT_e + IT_m)) x 100, which was intended to be the recoverable proportion of residual input tax. As in the original First Special Method, the connexion of the input tax with the supplies is present only by the implication of purchases being related to supplies, and there is no direct connexion.

We are led, therefore, to the conclusion that an essential feature of a special method, that the input tax should be directly linked to the supplies made, in other words that the method should be based upon the actual use made of the inputs, is absent in the First Special Method and its ramifications.

(1)(c) Whether the Second Special Method would have produced a fair and reasonable result

This method, dependent upon accurate apportionment of the floor areas of the Appellant's shops to reflect taxable, exempt, and mixed use, and the consequent analysis of the floor space into taxable or exempt usage (we paraphrase here from Mr Ward's letter of 22 may 2001), was said by the Appellant to reflect accurately the actual usage of the input tax incurred in relation to the floor space. However, no specific evidence was adduced which indicated how this result was or might be achieved, nor how that approach would be applied to the ultimate management and control of the business. The method was further explained in the letter of 19 June 2001 (paragraph 45 above) in respect of the defined areas, and the letter concluded that it could be seen that the mixed element amounted to less than one third of the floor-space by value; the apportionment of mixed areas was calculated by reference to the ratio of taxable to exempt areas, which, as a percentage, would be applied to all unattributable input tax incurred in respect of property, but no accounts were produced to shew the application or effect of this. The Commissioners' objection to that method were, first, that there was no need to apply the method to shopfitting since that could be broken down into taxable and exempt without difficulty. Secondly, rent should not be broken down into areas of the premises; thirdly, the manner in which the Appellant proposed to do this was not acceptable. It was also not accepted by the Commissioners that the six shops of which measurements had been taken were representative of all the Appellant's shops. Also, an attempt to verify the measurements at Bluewater suggested to the Commissioners that not all the areas designated taxable by the Appellant were in fact so. Lastly, even if the six shops were representative, all of them would require a visit by the Commissioners to verify the details, which, in Mr Laney's view, was not an easy method of verification.

The Appellant took issue also with Mr Dowling's reasons for rejection (paragraph 60 above), contending that it begged the question and made unverified assertions of fact. His principal reason, or the summary of the four reasons, was that the method would allow an unjustifiably higher recovery of tax than would either the standard method or the Second Special Method without the weighting in respect of zones. The schedule to Mr Dowling's letter which, he said, contains a comparison of the standard method with the special method with and without weighting, shewed that the standard method gave a recovery proportion of 54 per cent, the unweighted special method 51.78 per cent, and the weighted special method 85.86 per cent, each of these referring to the Bluewater shop. The standard method was calculated by taking the standard-rated outputs as a percentage of total outputs, while the two forms of the special method were the total taxable floor space as a percentage of the overall floor space. He pointed out that the unweighted special method produced a result which was very close to the standard method. The view was expressed by the Commissioners in correspondence that a "fairer" method did not simply mean one which produced a higher recovery of tax for the appellant. We consider that that is a correct statement, since any special method must be fair and reasonable, that is, fair and reasonable alike to the taxpayer and to the Commissioners.

We have already concluded that zoning of the premises has no part to play in the calculation of input VAT relating to premises. It therefore follows, in our judgment, that there is no reason for applying the proposed weighting in the Second Special Method. It therefore appears that the Appellant is slightly better off with the standard method than with the unweighted Second Special Method. For this purpose we assume that Mr Dowling's calculations are correct: no alternative was offered.

We therefore conclude that the weighted Second Special Method has not been shewn to produce a fair and reasonable result.

Our conclusion on the first principal issue is, therefore, that neither of the proposed special methods has been shewn to produce a fair and reasonable recovery of input tax, not having been shewn to relate recovery of input tax to the use of the inputs concerned; and that the standard method produces a result in which the recovery is directly connected to the use to which the inputs are put. We have been unable to check the figures, not having been supplied with them, but in our view this is of secondary importance, since what governs the recovery of input tax is the formula for calculation.

(2) Whether the Commissioners came to their several decisions reasonably

The final question is whether the Commissioners came, in the case of each proposed special method, to their decision reasonably in the Wednesbury sense. In the case of the First Special Method, we are in no doubt that if it had been left at the stage of the first method proposed, Mr Laney's mistaken view of the nature of the proposed method would have rendered the decision to reject not just unreasonable but erroneous. The mistaken matters which he took into consideration could not have been relevant, and were central to the decision. But also he was, in effect, deciding upon a method which had not been proposed, which was erroneous in itself. We agree with the Appellant that, as we have already said, his misunderstanding survived the explanations which took place on 28 February 2000, and had the proposed First Special Method been the method upon which the Commissioners finally made their decision we would have held that that decision was thereby vitiated.

However, there was interposed the New Method. That was treated by Miss Nagarajah as a new method, and passed back to Mr Laney as a new method. Mr Laney therefore considered it unaffected by his earlier misunderstanding. The decision was reached upon consideration of the principles upon which the method was based, and it was upon those principles that Mr Laney made his decision. Should he, we ask ourselves, nonetheless have considered all the facts relating to what dispensing was carried out in what part of any property, and whether given areas were devoted to taxable or mixed supplies? It appears to us that if it is concluded that the principles upon which a special method is based lack an essential ingredient, then that method is flawed and cannot be such as to provide a fair and reasonable recovery. The missing principle in the case of the two, or perhaps three, special methods under appeal was the direct link between the recovered input tax and the purpose for which the inputs were to be used. In the case of the Second Special Method there were the added ingredients of weighting according to the zones, the measurement of the shops, and the verification of the measurements. We have already given our decision upon the weighting. We also take the view that the measurement of the shops, especially if such nebulous concepts as walkways and the two-foot area are to be taken into account, would be no simple matter, even if it were established that six of the shops were actually representative (which would itself require to be established by evidence) would be far from straightforward and would be likely to be time-consuming and prodigal of the time of the officers who would be concerned. We therefore conclude that there was nothing else, in the circumstances, which Mr Laney and Mr Mattacks were obliged to consider in respect of the New method and the Second Special Method.

Did the officers take into consideration anything which was irrelevant? The Appellant contended that Mr Laney had fatally misunderstood the First Special Method and Mr Mattacks had addressed the wrong questions, and that both had made assertions of fact without any basis in evidence, none of which could be relevant. We have found that both rejections were made, and properly made, on the basis of the principles behind the special methods, and not the surrounding facts. It does not appear to us that any other things were taken into consideration.

We therefore conclude that the decisions to reject the New Method and the Second Special Method were reached reasonably in the Wednesbury sense.

Summary of conclusions

(1) That the retail area was used for both taxable and exempt supplies.

(2) That dispensing began when the optometrist handed over the customer to the sales staff, and continued until the customer was supplied with a finished pair of spectacles.

(3) That the front of the shop was used to attract customers in for the purpose of buying "the whole package" of dispensed spectacles, as well as frames, sun-glasses and accessories.

(4) That the zones which are now commonly used as a means of arriving at a proper rent of retail premises, and for the purposes of rating, have no place in the attribution of input tax in partial exemption.

(5) That there was no evidence that any of the Appellant's employees were carrying on dispensing activities otherwise than under direct supervision by those who were qualified for the purposes of Note (2) to Group 7 of Schedule 9 to the 1994 Act.

(6) That Mr Laney's misunderstanding of the nature of the First Special Method coloured his view of that method until and including his letter of 20 March 2000, but not his view of the New Method.

(7) The Commissioners' conclusions were based, in the case of each special method, upon the principles upon which each of the special methods was based.

(8) It was not necessary to look at the standard method, or compare the standard method with a special method, to see if a special method would give a fair and reasonable recovery. There was no purpose in using the standard method as a comparator between the standard method and a special method.

(9) Since an essential ingredient was missing from each of the New Method and the Second Special Method, that of direct connexion between recovery and the purpose of the inputs, neither would have produced a fair and reasonable recovery. The First Special method was similarly flawed.

(10) The decisions to reject were reached reasonably in the sense in which that word was used in Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223.

For the above reasons this appeal is dismissed. The Commissioners did not apply for their costs in the event of success, and accordingly we give no direction as to costs.

ANGUS NICOL

CHAIRMAN

RELEASED:

LON/00/1281