United Kingdom VAT & Duties Tribunals Decisions

---

You are here: BAILII >> Databases >> United Kingdom VAT & Duties Tribunals Decisions >> Wadham College Oxford & Anor v Revenue & Customs [2007] UKVAT V20233 (02 July 2007)
URL: http://www.bailii.org/uk/cases/UKVAT/2007/V20233.html
Cite as: [2007] UKVAT V20233

---

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

---

Wadham College Oxford & Anor v Revenue & Customs [2007] UKVAT V20233 (02 July 2007)

20233

Partial Exemption - subsequent claim for input tax not claimed under operation of special method based on the "CVCP Guidelines" for universities and colleges – special method allowed or approved, not directed – determination of nature of special method – if the method did not provide recovery in respect of all relevant input tax how to fill gaps - with standard method or method approved for later periods - jurisdiction of tribunal in relation to a special method which had been allowed or approved – could the taxpayer appeal?

LONDON TRIBUNAL CENTRE

WADHAM COLLEGE OXFORD First Appellant

and –

MERTON COLLEGE OXFORD Second Appellant

and –

THE COMMISSIONERS FOR HER MAJESTY'S REVENUE & CUSTOMS Respondents

Tribunal: CHARLES HELLIER (Chairman)

DIANA WILSON

Sitting in public in London on 8, 9 and 10 May 2006 and 11, 12 and 13 April 2007

Penny Hamilton and Sadiya Choudhury, Counsel, instructed by Hutchinson Mainprice for both Appellants

Neil Sheldon, Counsel, instructed by the Acting Solicitor for HMRC for the Respondents

© CROWN COPYRIGHT 2007


 


DECISION

Introduction

1. This is a decision in relation to two appeals heard at the same time. The appeals concern the recovery of input tax by Wadham College Oxford and Merton College Oxford. Between 1973 and 1994 these colleges gave consideration for goods and services which included VAT charged by their suppliers (input VAT), and made both taxable and exempt supplies. Each college had an entitlement under the relevant Directive to the recovery of part of that input tax. The arguments in each appeal were the same. In this decision we address the arguments in the appeals together but indicate where any relevant facts were different.
2. In that period the Colleges claimed, and were allowed, deductions in respect of some of their input VAT under arrangements which were based upon guidelines promulgated with the agreement of the Respondents by the Committee of Vice Chancellors and Principals (the CVCP Guidelines) from 1973 to 1994.
3. The college now claims a deduction for, and repayment of, further amounts representing a further portion of its input tax.
4. Originally there had been an issue between the parties as to whether regulation 29(1A) of the VAT Regulations 1995 prevented the college from making claims for periods ending on or prior to 31 July 1994; but in the light of the judgment of the Court of Appeal in Fleming v HMRC 2006 EWCA Civ 70, the Respondents now accept that the college is not barred from claiming in respect of these periods.
5. Where a business makes taxable supplies only there is no difficulty in determining the allowable input tax – broadly it is all recoverable; where it makes exempt supplies only there is likewise no difficulty - none is allowable. But when a business makes both taxable and exempt supplies and some of the input tax relates to inputs not wholly attributable to taxable or exempt outputs some method is required to determine what input tax is deductible. At all material times the relevant Acts have provided for the Commissioners to make regulations for securing a fair and reasonable attribution of input tax to taxable supplies. Regulations have been made from time to time under those powers. We shall set out the detail later in this decision, but broadly they provided for a "standard" method for determining the input tax which was to be taken as attributable to taxable supplies (and was therefore deductible), but also provided that the Commissioners could direct or permit an alternative method (which we shall as customary call a "special method"). Up until 1992 the precise wording was that the Commissioners might "allow" or direct the use of an alternative method; after 1992 the provision was: the Commissioners might "approve" or direct the use of an alternative method.
6. Under the CVCP Guidelines the activities of a university or college were classified under various headings or "tunnels", and a calculation made in relation to each tunnel of the input VAT which could be recovered in respect of the related activity. They may well have been called tunnels because they permitted some sort of path to be cut through the maze of the university's activity from the relevant input to the related output.
7. It is common ground that in the relevant period each college used a special method based on the CVCP guidelines. Under that method it reclaimed input tax in relation to three particular tunnelled areas of activity: outside conferences, separate catering, and bar sales (we call these the "three formulaic tunnels" for reasons which will become apparent). For these activities the CVCP guidelines provided a formula for input tax recovery which was used by the college.
8. In University of Sussex (VAT decision 16221) the university had for some period made input tax claims on the basis of the CVCP Guidelines. The tribunal found that the input tax recoveries made by the appellant in that case did not properly give effect to that appellant's rights under the applicable EU Directive to the recovery of input tax. Broadly speaking it was on the basis of that decision that the Appellants in this appeal claimed additional repayments of input tax. The Respondents' position at the first hearing was that they acknowledged that in principle additional amounts were repayable. What was in dispute was the way in which that additional amount should be calculated. Thus when the appeal first came before this tribunal it was asked to determine in principle the method by which the under-recovery should be computed. Two bases were put before the tribunal and we were asked to choose one of them.
9. However, on consideration, this tribunal was not clear that there had been an under-recovery at all. That was for two reasons. First there was no evidence before the tribunal that the colleges' taxable activities extended beyond those encompassed in the three formulaic tunnels it used; and second we were not clear that in the colleges' circumstances the method applied in relation to those areas of activities failed to give effect to the college's EU rights. Our tentative view was that it did not seem to us that all of the reasoning in University of Sussex was applicable, and that we had no jurisdiction to adjudicate on a hypothetical claim.
10. Accordingly we directed a further hearing at which these issues could be aired. At that hearing the Respondents' position was different from that they had adopted at the first hearing: whilst they accepted that the colleges had a right to claim further input tax in relation to any taxable economic activities which were not embraced in the three formulaic tunnels and in respect of which input tax had not otherwise been claimed, they supported the tentative view advanced in our direction that, in relation to the three formulaic tunnels, the method used provided full recovery of all input tax to which the college was entitled.
11. At the second hearing we agreed that we would produce a decision in principle in which we would set out our views on each of the issues which had been before us. Those issues are:

Issue (1) whether the colleges' taxable economic activities extended beyond the three formulaic tunnels (thus addressing the first point in the October 2006 direction);

Issue (2) first, whether in relation to the three formulaic tunnels the CVCP guideline method adopted by the colleges gave proper effect to the colleges' rights under the relevant Directive to the recovery of input tax. In this regard we agreed that would make findings of fact and inference therefrom as to the nature of the special method operated by the colleges, and in particular as to whether in relation to the three formulaic tunnels, overhead residual VAT was intended to be taken care of by the 20 per cent. 20 per cent, and 5 per cent formulae. Second, we would also address a fallback argument advanced by Mrs Hamilton that the method encompassed the ability retrospectively to elect for apportionment based on actual percentages. We have called this, Issue 2A below; and

Issue (3) if the method used in relation to the three formulaic tunnels or otherwise did not take care of residual overhead VAT then what, properly construed did the special method which was operated from 1973 to 1997 provide for the recovery of residual VAT: which of the two methods proposed by the parties should be adopted (those methods we call the Conference Method and the Modified Standard method as we shall explain later).

12. The nature of the tribunal's jurisdiction in relation to some of these issues is not wholly clear. We shall address this issue in our conclusions, but at this stage we should say that we have approached this appeal as requiring us to determine the amount of input tax which is creditable under the special method which was in operation rather than to decide which of two different special methods should be applied.
13. In the remainder of this decision, after discussing the evidence before us, we shall first discuss the relevant statutory provisions and the more general relevant case law. We shall then turn to the history and content of the CVCP guidelines and the tribunal decisions relating to them – in particular the decision in University of Sussex. We shall then address the three issues set out before reaching our conclusions.

The Evidence

14. At the first hearing we heard oral evidence from Derek Button the clerk of Council at Wadham College, a position he had held for the last 19 years, from Peter Thomas Nash of Mainprice Nash Miller, a firm which advised many Oxford colleges on VAT issues, and from Richard Harvey and Mrs Stephanie Jean Farr both officers of the Respondents. We had bundles of documents which included copies of the CVCP guidelines in force from time to time. At the second hearing we heard further oral evidence from Mr Nash, Mr Button and Mrs Farr, and also oral evidence from Clifford Ronald Webb, the domestic bursar of Merton College. All of whom provided witness statements. We also had a witness statement from Ian Honeyman who was senior Bursar of St Hugh's college from 1980 to 2003. He did not give oral evidence. In addition the Respondents provided some further documentary evidence relating to the circumstances surrounding the settlement of the CVCP guidelines in 1973. We shall set out our findings of fact in the relevant sections below.

1. The relevant statutory provisions

15. Article 11.1 of the Second Directive, whose provisions were relevant from 1973 to 1977, provided that where inputs were "used for the purposes of his undertaking, the taxable person shall be entitled to deduct" the input tax. Article 11.2 provided that no deduction was permitted for inputs used in non-taxable or exempt transactions, and that where inputs were "used" both in taxable and non taxable transactions deduction was allowed only for that part of that input tax which was "proportional to the amount relating to the transactions giving entitlement to deduction (pro rata rule)."
16. Paragraph 22 of Annex A to the Directive (which formed an integral part of the directive) provided that:

"The pro rata figure shall, in general, be determined in respect of all transactions carried out by the taxable person (general pro rata figure). However a taxable person may, exceptionally, obtain administrative permission to determine special pro rata figures for certain sectors of his activities."

17. We note the following features of these provisions: first they confer an entitlement to deduction of input tax on the taxpayer, second the deductible amount depends upon the "use" of the input (although the drafting of paragraph 22 is not crystal clear on that issue), third input tax directly on inputs used only for taxable outputs is wholly allowable in all circumstances, fourth the pro rata rule applies only to residual input tax, fifth a limited discretion is given to member states to approve special residual input tax pro rata figures for particular sectors and sixth that member states are not given a right to calculate the reclaimable VAT as a proportion of all input VAT; instead the directly attributable VAT had to be dealt with first.
18. The CVCP guidelines were first promulgated when this Directive was applicable. As we note in more detail below the "tunnelling" provisions of the Guidelines permit the recovery of directly attributable input tax and a sector by sector approach to the recovery of residual tax broadly as suggested by paragraph 22 of this Directive.
19. The Sixth Directive of 17 May 1977 provides that the taxpayer has a right to deduct input tax in so far as the inputs are used for taxable transactions. The key concept is the use of the input Article 17.2 provides:

"In so far as the goods and services are used for his taxable transactions the taxable person shall be entitled to deduct from the tax which he is liable to pay: (a)value added tax …in respect of [supplies] to him…"

Article 17(5) provides that as regards inputs used for both taxable and exempt transactions "only such proportion of the VAT shall be deductible as is attributable to" taxable transactions. It continues:

"This proportion shall be determined in accordance with Article 19 …

"However member States may:

(a) authorise the taxable person to determine a proportion for each sector …

(b) compel … a proportion for each sector …

(c) authorise or compel a taxable person to make the deduction on the basis of the use of all of the part of the goods and services …"

20. Article 19 provides for the standard method: the deductible proportion of the input tax is broadly:

taxable outputs
total outputs

21. We note, as did the tribunal in Merchant Navy Officers Pension Fund Trustees Ltd (VAT Decision 14262), that "Article 17.5 and 19 are only concerned with inputs used for mixed taxable and exempt outputs. Inputs used exclusively for taxable or exempt outputs are not covered by Article 17.5 or 19: if inputs are used wholly for taxable supplies they are deductible under Article 17.2; if wholly used for exempt supplies they are not deductible.
22. We gratefully adopt paragraphs 23 and 24 of the tribunal's decision in Merchant Navy:

"23. Once attribution on the basis of actual use is impossible or impractical any other method of attribution can only be designed to approximate to actual use being only estimated or assumed. The standard method is very rough and ready and may result in attribution which is demonstrably very different from probable actual use however it has the important merit of simplicity.

24. The methods authorised under Article 17.5(c) are clearly different from that under Article 19, since otherwise Article 17.5(c) would be otiose. However it cannot mean actual direct use since, if the actual use of the input could be determined, it would not be a mixed use and Article 17.5 would not apply at all. It seems to me that Article 17.5(c) must authorise attribution on the basis of estimated or assumed use."

23. Mrs Hamilton took us to the judgement of Auld LJ in University of Sussex v CCE 2003 EWCA Cir 1448 at paragraph 173 where he said:

"In my view, … [the University] had accrued rights under arts 17-20 of the Sixth Directive … I consider those provisions are unconditional and sufficiently precise to give rise to a directly effective Community Law right …"

He then discussed the procedural mechanics related to the exercise of this right in Article 18(2) and (3) and the discretion there given to member states, and continued:

"[175]. The existence of an element of discretion in a member state as to how such right is to be exercised cannot, in my view, sensibly deprive it of direct effect …"

That last statement does not in our view indicate that the quantity (rather than the existence or quality) of a taxpayer's right to input tax recovery cannot be affected by a determination made by a member State under the provisions of Article 17(5).

24. The Finance Act 1972 section 3(3) provided that:

"the input tax that may be deducted by a taxable person shall be :

(a) the whole of that tax if all his supplies of goods or services are taxable supplies; and

(b) such part of that tax as, in accordance with regulations made under this section, is attributable to taxable supplies, if some but not all of his supplies are taxable supplies…

(4) The Commissioners shall make regulations for securing a fair and reasonable attribution of input tax to taxable supplies…"

25. These provisions did not expressly follow those of the Second Directive, since they provided in the case of a person making taxable and exempt supplies, not for the direct attribution of inputs to taxable supplies, and then for a fraction to be applied to the residual input VAT, but for these matters to be left to regulations.
26. The Finance Act 1977 substituted new section 3 and 4 into the Finance Act 1972 but the new section 4 contained provisions identical to those in section 3 above. Thus the provision of that Act did not mirror the Sixth Directive which required direct attribution first and then a proportion of residual input tax, but left such matters to the regulations.
27. Section 15 Finance Act 1983 made provision for the entitlement to credit in broadly, the same terms as Finance Act 1977, providing for regulations to determine the allowable proportion where a person made taxable and non taxable supplies. The amendments made by Finance Act 1987 to section 15 left the position the same in that respect.
28. Thus, as far as concerned taxpayers who made taxable and exempt supplies, the Finance Acts in force at the relevant times threw onto the Commissioners the burden of making regulations which would comply with the requirements of the relevant Directive and secure a fair and reasonable attribution of input tax to taxable supplies.
29. From 1973 until 1984 it appears that the regulations provided, so far as is relevant to the appeal:

"(1) Subject to paragraph (2) of this Regulation, the proportion of input tax to be attributed to taxable supplies by any taxable person who makes exempt supplies shall be determined in any prescribed accounting period by either of the following methods:

Method 1

Subject to Regulation 25 [which dealt with certain exclusions] he may deduct such part of his input tax as bears the same ratio to his total input tax as the value of taxable supplies by him bears to the value of all supplies made by him.

Method 2

[was broadly relevant to persons who brought and sold goods in the same state such as retailers].

(2) The Commissioners may allow or direct the use of a method other than the one specified in paragraph (1) of this Regulation."

30. Thus in the absence of a special method "allowed or directed" by the Commissioners these regulations did not expressly carry in UK law the initial deduction of the input tax directly attributable to taxable supplies and the proportionate treatment of residual input tax specified by either the Second or the Sixth Directive. Neither, apart from such allowance or direction, did the regulations provide for the emphasis on "use" as in that Directive.
31. The 1984 amendments to the regulations removed Method 2 but otherwise made no material changes. The 1985 regulations made similar provision.
32. In 1987 the 1985 regulations were amended. After amendment they provided, (with our emphasis):

"30(1) Subject to paragraphs (2), (3), (4) and (5) of this regulation, the amount of input tax to be provisionally attributed to taxable supplies by a taxable person shall be determined … by the following method:-

(a) …

(b) the input tax on such importations and supplies as are wholly used or to be used by him in making taxable supplies may be deducted;

(c) the input tax on such importations and supplies as are wholly used or to be used in making exempt supplies or if any activity other than the making of taxable supplies may not be deducted;

(d) the deductible proportion of any remaining input tax shall be provisionally calculated as follows:

(i) importations by and supplies to the taxable person in the period which are partly used or to be used by him in making taxable supplies shall be identified;

(ii) the extent to which the above importations and supplies are used or to be used by him in making taxable supplies shall be ascertained, and expressed as a proportion of the whole use made or to be made by him of such importations and supplies;

(iii) there may be deducted such proportion of the remaining input tax as corresponds with the proportion ascertained above.

(2) The Commissioners may in the case of a taxable person who incurs exempt input tax [i.e. input tax wholly or partly attributable to an exempt supply] allow that paragraph (1)(d) of this regulation shall not apply, in which case the deductible proportion of any remaining input tax may be provisionally calculated as follows:

Subject to paragraph (3) of this regulation, there may be deducted such proportion of any remaining input tax as bears the same ratio to the total remaining input tax of the taxable person as the value of taxable supplies by him bears to the value of all supplies by him.

(3) [provided for exclusions not relevant in the instant case.]

(4) Where the Commissioners consider it necessary in order to secure a fair and reasonable attribution of input tax to taxable supplies, they may in the case of any taxable person or class of such persons, direct the manner in which the extent of use of importations and supplies is to be ascertained under paragraph (1)(d) of this regulation.

(5) The Commissioners may allow the use of a method other than that specified in this regulation."

33. It will be seen that these regulations contained the emphasis on use inherent in the Sixth Directive, followed the initial identification of directly attributable input tax and the later apportionment of residual input tax of the Directive, but, rather than specifying the standard method of Article 19 of the Directive as the starting point, took advantage of the permission in Article 17.5(c) to compel deduction of the residual tax on the basis of use, with Regulation 30(2) providing for the possibility of allowing a standard method.
34. In 1992 these provisions of the 1985 regulation were amended and from 1 January 1993 provided for the replacement of the use based criteria for the attribution of residual tax in the former Regulation 30(1)(d) with the standard method (by reference to the proportion of taxable outputs). This approach was re-enacted in the 1995 regulations.
35. Until 1992 the regulations provided that the Commissioners might "allow" or direct a special method; in 1992 the regulations instead provided that the Commissioners might "approve" or direct such a method.
36. Thus, in summary, and so far as material to this appeal:

(i) from 1973 to 1987 the regulations provided for the use of Method 1 (See paragraph 29 above);

(ii) from 1987 to 1992 direct attribution applied, and residual input tax was deductible on the basis of use;

(iii) from 1992 onwards direct attribution applied, and residual input tax was deductible on the standard method,

in each case unless the Commissioners directed or permitted otherwise.

37. Stephanie Farr's witness statement included a useful summary showing how the regulations had been reflected in the Commissioners' published practice over the period since 1973. Two points are worthy of particular note in that summary:

(i) first for the period April 1983 to March 1984 it is noted in that published practice that if "the bulk of inputs could be directly related to taxable and exempt supplies taxpayers could apply to use the direct attribution method." This involved direct attribution to taxable and exempt supplies and the use of "Method 1" to apportion remaining output;

(ii) second, in the period April 1987 to March 1992 (when it will be recalled from paragraph 32 above that residual input tax was required to be recovered on the basis of use) she notes that traders were advised to discuss with their local offices how tax should be apportioned on the basis of use, and that an output-based calculation (such as that in Article 19) was a special method.

Overhead and attributable residual tax

38. In HMRC v Mayflower Theatre Trust [2006]EWCA Civ 116, Carnwath LJ quoted with approval Hart J's identification of two types of residual inputs. He had said:

"Inputs may fall into the residual category for one of two reasons. First they may not be specifically attributable to either exempt or to taxable supplies. In that situation they are described in the European jurisprudence as being overheads, i.e. cost components of the business as a whole…Secondly they may be specifically attributable both to particular exempt and taxable supplies…"

In relation to overheads which cannot be attributed to particular supplies, the eligibility of the input tax attributable to them is assured by establishing the appropriate link with the "whole economic activity" of the taxpayer.

39. This distinction between the different elements of the residual VAT of the colleges was helpful in the argument before us, and we adopt the terms "overhead residual tax" and "attributable residual tax" for those elements in the rest of this decision.

2. The CVCP Guidelines and the Grid

40. In 1973 an agreement was reached between the Commissioners and the Committee of Vice Chancellors and Principals concerning the recovery of input tax by universities. Following that agreement, the CVCP Guidelines were issued by that committee. The Guidelines were periodically updated and were withdrawn in 1997.
41. As noted above, at the time the guidelines were originally issued the default method for recovery of input tax by partially exempt persons did not require the identification of the inputs relating solely to taxable outputs and of those relating solely to exempt outputs before dealing with residual input tax, but gave input tax recovery on the basis of a portion of the entire input tax which an amount equal to



42. Paragraph 11 of the March 1973 Guidelines (the earlier paragraphs dealt with the tax liability of various outputs) explained that universities would be entitled to recover input tax or a portion of it included in the cost of supplies. It noted that the normal arrangements for partially exempt organisations were explained in C&E notice 706 which set out Method 1: the use of the single fraction applied to all input VAT as described in paragraph 29 above; and explained that since for a university the value of its exempt outputs (principally its teaching and research efforts) was much the greater part of its outputs, the denominator of the fraction would have been substantially larger than the numerator and so very little input tax would be recovered.
43. The Guidelines indicated that "Customs and Excise have accordingly approved a special arrangement for universities whereby each taxable activity… can be dealt with separately or "tunnelled".". It seems to us that such tunnelling would follow more closely the requirements of the Second Directive than the default method then in operation (and took advantage of the sectorisation permitted by paragraph 22 of the Annex to the Directive).
44. Paragraphs 12 and 13 of the March 1973 CVCP Guidelines dealt with three specific areas of activity. These were the three formulaic tunnels. It is worth setting out parts of these paragraphs in detail:

"In relation to the making of some taxable supplies – such as the letting of accommodation to an outside, non educational body for a conference…it will be apparent that goods and services will be used which are also required for the making of exempt supplies..."

[We pause to note that this specifically contemplated inputs used for both taxable and exempt supplies]:

"the two areas in which this factor is of special relevance are…the supply of conference/holiday facilities and the supply of non exempt catering. Here a formula approach has been agreed under which the universities will not be required to keep any records of the amounts of tax paid in the cost of related supplies. On the basis of evidence collected from a sample of universities Customs and Excise have agreed that each university shall be entitled to reclaim 20% of the output tax payable in respect of such supplies: this proportion will be regarded as representing related, deductible input tax…"

[We note here that these percentages were based on data gathered from a number of universities. As we record later, the data and the calculations from it are no longer available, but some correspondence did emerge in relation to Voluntary Colleges of Education.]

"the group has welcomed this as a major contribution towards eliminating time-consuming detailed accounting work…its expectation is that the saving in administrative costs which will result will more than compensate any university which might believe that it could substantiate a case for recovering greater amounts of input tax…"

[So the formula was not precise, could lead to a lesser recovery, but was probably worth it.]

"…the alternative …which any university is entitled to explore and opt for is full rigorous accounting…and the presentation to Customs and Excise of a detailed breakdown of…input tax between taxable and exempt outputs"

45. Paragraph 13 of those Guidelines dealt with bars. There separate accounts of the direct inputs (drink and tobacco) were to be kept, but it was "recognised …that there [would] be other [supplies] in relation to bars – for example on glasses, cleaning materials and items of equipment where input tax [would] be payable – and ought to be recoverable. Customs and excise [had] agreed that each university would be entitled to reclaim 5% of the total output tax payable.. as representing related deductible input tax on these items". The example of glasses etc disappeared in later editions of the Guidelines.
46. Paragraph 14 dealt with all other tunnelled areas of activity:

"In relation to all other areas where an output tax liability will arise, schemes similar to those devised for commercial retailers for the recovery of input tax – of which examples are given on pages 6 and 7 of notice 706 – will have to be agreed locally in respect of each distinct taxable activity. In this connection it should be remembered that input tax can only be recovered in the proportion to which the value of taxable outputs bears to the value of total inputs" Those pages describe Method 1, the standard partial exemption method recorded in paragraph 29 above.

47. The question discussed under Issue 2 below is whether or not these 20 per cent, 20 per cent and 5 per cent allowances represented all the relevant deductible input tax i.e. the input tax attributable solely to the particular activity, the relevant part of the input tax attributable to that and other taxable activities, the relevant part of the input tax which was "attributable residual tax", and an appropriate part of the residual overhead tax.
48. The guidelines were accompanied by a Grid or worksheet for calculating the recoverable input tax. This Grid in our view helpfully illustrates how the Guidelines were intended to work. We therefore reproduce it below:

Supporting Worksheet for VAT Quarterly Summary, Quarter ended



Note: the boxes marked 'X' were those customarily filled in by Wadham College

49. The later revision to the guidelines made no material amendments to the effect of the provisions set out above. Save that the phrase "for example on glasses, cleaning materials and items of equipment" was at a later time omitted from the section relating to bar sales.
50. Column E of the Grid, "Input tax Wholly Attributable" is to be contrasted with Columns F and G "Input tax partially Attributable". Where an activity was partially exempt and no direct attribution of an input wholly relating to that activity was possible it seems to us that the Grid invited that input tax to be dealt with in Columns F and G (we discuss later whether overhead input tax was to be dealt with similarly). Where the activity was wholly taxable – as for example was "(vi) Bars" then the Grid invited the tax wholly attributable to that activity to be put in Column E. The shading on the grid indicated areas not to be used. Later versions shaded (iv) E to G, (v) E to G and (vi) F and G to indicate they were not to be used.
51. There is one last aspect of the guidelines to which we should advert. It is the statement noted in paragraph 14 of the 1973 revision, reproduced above, that in relation to areas of activity other than the three formulaic tunnels "schemes… will have to be devised and agreed locally" and that "it should be remembered that input tax can only be recovered in the proportion to which the value of taxable outputs bears to the value of total outputs." So far as the first part of this quotation goes it expresses what must be the case in relation to other activity: the filling in of boxes F and G in the Grid for that activity must be in an allowed or approved way; so far as the second part goes, it seems to us that it did not represent the law and is inconsistent with the first part. The second part can be no more than an indication that the standard method might be a starting point for discussions with the local office.

3. Relevant case law

52. Ampleforth Abbey Trust v CCE (VAT decision 15763) concerned the position under the 1972 Act and the Second Directive. The tribunal noted that Directive gave a right to deduct directly attributable input tax and a right to pro rata deduction of residual tax.
53. The taxpayer had entered into an arrangement with the Commissioners under which they would recover directly attributable tax. Residual tax was not addressed in the arrangements.
54. The tribunal held that the arrangement entered into between the Commissioners and Ampleforth Abbey Trust was not a "special method" within regulation 24(2). It held so because:

(i) the arrangement was simply an arrangement which dealt with directly attributable input tax;

(ii) section 3(3)(b) of the 1972 Act was to be construed if possible so as to give effect the Second Directive;

(iii) section 3(3)(b) of that Act should therefore be construed so as to ascertain how much residual input tax should be recovered;

(iv) that question was simply not addressed by the arrangement; and

(v) it was therefore not a special arrangement - to place it any higher would be to deprive the taxpayer of the residual tax pro rata right.

55. The essential feature of this reasoning was effectively to treat the right to credit for directly attributable input tax as permeating the 1972 Act so that the effect of section 3(3)(b) and Regulation 24(2) was limited to residual input tax recovery. Since such recovery had not been addressed by the arrangement it could not be a special method within Regulation 24(2).
56. The tribunal noted that the Sixth Directive went further than the Second Directive; whereas the Second Directive permitted member states to require proportions to be determined sector by sector, the Sixth Directive also permitted a use based special method to be applied to all inputs.
57. In University of Edinburgh v CCE (VAT decision 6569) the university claimed input tax in relation to the period July 1984 to July 1990 in respect of the taxable supplies of its computer services department. It was conceded that the university was entitled to invoke the "tunnelling" provisions of the CVCP guidelines to recover directly attributable input tax; the question related to residual input tax recovery. The Commissioners argued that a special method had been agreed under which the university had renounced its right to residual input tax as part of the acceptance of the tunnels. The tribunal found that there was no such agreement to preclude the entitlement to residual input tax.
58. In other words, residual input tax was not dealt with by the method actually employed and was recoverable.
59. In University of Sussex v CCE (VAT decision 16221) the consequences of the CVCP guidelines arose again in relation to the period April 1973 to July 1996. The university claimed residual input tax (calculated on the standard method) after taking account of the residual input tax it had recovered in making tunnelling claims in accordance with the CVCP guidelines (see paragraph 25). The Commissioners asserted that the university's special method did not provide for the recovery of additional residual input tax (see paragraph 27 at the end): they did not maintain that the university agreed not to claim input tax that it was entitled to claim, but that under the special method there was never an entitlement to claim: the content of the special method was not to be found only in what had been allowed or approved. The tribunal:

(i) followed Ampleforth in that it held that there was a right to a pro rata deduction under the Directive;

(ii) was influenced by the fact that if the special method was confined to the limited recovery actually made under the tunnels operated by the university it did not address all the economic activity of the university and it produced anomalous results;

(iii) found that the CVCP guidelines formed a setting for the special method, and potentially some part of it; and

(iv) held that the content of the special method allowed or approved was not to be confined to what had actually been claimed, but should be construed as including the ability to recover residual input tax.

60. The tribunal considered Kwik-Fit [1998] STC 159 - but held that the method at issue was not void for uncertainty or lack of fairness and reasonableness. The tribunal did not give detailed consideration to the proportion of residual input tax which was recoverable.

4. The Three Issues

Issue 1: Did the colleges' activities extend beyond the three formulaic tunnels and was input VAT claimed in respect of any of those activities?

61. From the evidence of Mr Button and Mr Nash we find the following facts: (i) Wadham College made taxable supplies which were not encompassed by the three formulaic tunnels. These included, the sale of stationery and postcards, the sale of cycles and second hand furniture, gown hire, and more recently rent and other income from land which was taxable or in respect of which the option to tax had been exercised; (ii) the income from those activities was between £1K and £29K per annum between 1973/74 and 1990/92, and was considerably greater in later years; and (iii) in general no VAT input tax recovery was claimed in respect of input VAT relating to those activities.
62. From the evidence of Mr Webb and Mr Nash we find in relation to Merton College that: (i) Merton College made taxable supplies which were not encompassed by the three formulaic tunnels. The included laundries, telephones, gaming and vending machines, guided tours, milk quota sales, car parking, and again more recently dispositions of land on which VAT was payable; (ii) some input VAT was recovered in relation to some of these activities outside the three formulaic tunnels: when land had been disposed of there had been a recovery of VAT on professional fees, and when a charge had been made to students for electricity on which VAT had been charged, part of the input VAT on the supply of electricity to the college had been allowed as deductible under arrangements made with the Respondents' officer responsible for the college; but these instances did not encompass the whole of the taxable activities of the college.
63. Therefore in relation to both of the colleges we find that there was input VAT which did not relate to the activities of the three formulaic tunnels in respect of which the college did not claim nor receive credit. HMRC did not dispute before us that credit was due, and we were asked by the parties not to address the quantification of the college's claim. There seemed to be every prospect that it would be agreed between the parties.
64. Thus unless there is later disagreement our job in relation to this issue is done. But before we proceed to Issue 2 we should mention the basis for the Colleges' claims. It is in our view this. We find that: (i) the colleges used a special method in the relevant years (see also paragraph 171 below); (ii) the use of that method was allowed or approved by the respondents (see paragraphs 172 to 174 below); and (iii) that method was based on but not prescribed by the CVCP Guidelines. In relation to matters outside the three formulaic tunnels, those guidelines indicated that a scheme would have to be devised and agreed locally. The ability so to do was part of the special method. Failure timeously to seek to agree a scheme does not deprive the taxpayer of the right to recovery of input tax or constitute an agreement not to claim. Once agreed the relevant amount falls to be treated as recoverable under that method.

Issue 2: Did the three formulaic tunnels permit recovery of all relevant input VAT?

65. This was the issue raised by the tribunal. In our direction we set out our concern and tentative conclusions thus:

"15. These guidelines permitted what was called tunnelling. This consisted of the identification of certain lines of business (or tunnels) of the universities and the calculation or estimation of the recoverable input VAT relating to each tunnel. The method of calculation would vary with the line of business

"16. For three particular lines of business (which we will call the three tunnelled outputs or the three formulaic tunnels) there was a formula provided for calculating the relevant recoverable VAT. For other lines of business no formula was provided.

"17. In the case of the provision of outside conferences the formula provided that no tally would be kept of the inputs directly, or partially related to that output but instead a blanket figure of 20% of the output tax would be treated as recoverable input tax. In our view this figure appears to have been set to take account both of any input tax directly related to a particular taxable output and also any residual input tax only partially related to that output and partially to other exempt or taxable outputs.

"18. In the case of a supply of catering the formula was to treat as recoverable both any input VAT actually related to that output and also 20% of that output. Again in our view this formula appears to have taken into account residual VAT.

"19. In the case of bar sales the formula was directly related input tax plus 5% of outputs. Again in our view this formula appears to have taken into account residual VAT.

"20. Thus for these three tunnelled outputs there were formulae which estimated the properly recoverable attributable input VAT by taking account of the directly attributable inputs and also a proportion of those inputs which were attributable both to the tunnelled output and also to other taxable or exempt outputs.

"21. In relation to other taxable activities the guidlelines provided no estimation method for recoverable input tax merely noting that invoices (received and rendered) relating to other taxable activities would have to be retained and records kept.

"22. The CVCP guidelines do not on their own constitute a method for the recovery of input tax which automatically applied to any college or university. Before any method (a "special method") for such recovery can be used in place of a standard method prescribed by the relevant legislation or regulations that method has to be allowed or approved (whether "allowed" or "approved" depends on the relevant time).

"23. But the Appellants in the instant appeals clearly submitted their VAT returns on the basis of calculations prepared using the three formulaic tunnels in the CVCP guidelines, and for 20 years or more those calculations were accepted by the Respondents. The way they submitted their returns involved a "method" of calculating their recoverable VAT which was allowed or approved by the Respondents at all relevant times. It was a method other than the standard method and therefore on our view (as is accepted by the Respondents) was a special method for these purposes.

"…44.It seems to this tribunal that unless it is shown that the Appellants' taxable economic activity extended beyond those three formulaic tunnels, or that the estimation of input tax recovery entailed in the formulae applicable to those tunnels took no account of residual VAT. The Special Method in this case was confined to the operation of those formulaic tunnels. This is the issue on which the tribunal seeks the parties' assistance."

66. From the evidence of Mr Button we find the following facts in relation to the period from 1973 to 1987:

(i) Wadham college used the grid to calculate its claims for deductible input tax;

(ii) the only entries made in the grid were those marked with an "X" at paragraph 48 above together with some entries relating to miscellaneous outputs in relation to which columns A to D only were used – and thus no input tax claimed;

(iii) Row (iv) Column A represented the cash takings through the lodge for casual accommodation; Row (iv) Column C represented invoiced conference income;

(iv) Row (iv) Column I, and Row (v) Column I were filled in with 20% of Column D Row (vi) Column I was filled in with 5 per cent of Column D;

(v) the total of Column I was the amount of the input tax claim for the college. No other input tax deduction was claimed.

And from Mr Webb's evidence we find the following facts:-

(i) Since the 1970s the standards expected by conference delegates have become more exacting and significantly greater expenditure has been incurred in more recent years in responding to those expectations. Proportionately the costs are substantially higher than they were in the 1970s.

(ii) Generally colleges were unwilling to act independently or to rock the boat in relation to the application of the CVCP Guideline methods. There was a feeling that the Guidelines were set in stone.

(iii) Over the period from 1973 Conference income has risen as a proportion of total income. Conference income can be 20% of residential and tuition fees income.

(iv) For the later years in the relevant period the 20% recovery under the conference tunnel when compared to a more rigorous use based analysis did not represent a good deal.

The Parties' arguments.

67. Mrs Hamilton did not accept that the three formulaic methods took account of both directly related input tax and all other input tax only partially related to the relevant output and also partially related to other exempt or taxable supplies. She said that tunnelling, and in particular the nature of the three formulaic tunnels was not intended to, and was not capable of dealing with overhead residual tax which did not form a direct cost component of the supplies made in a particular tunnel. She says that this proposition is consistent with the terms of the CVCP Guidelines, the context of the partial exemption rules, and the way the Guidelines were in fact operated.
68. In relation to the terms of the CVCP Guidelines, Mrs Hamilton makes four points. First, she notes that the paragraphs of the guidelines which deal with tunnelling refer to: "related deductible input tax", the cost of "related inputs, and "input tax included in the costs of goods and services used to make taxable supplies", and "input tax …in relation to each …activity". The 1986 Guidelines referred to the formulaic tunnels thus: "Customs have agreed a formula for recovery of input tax in three areas of taxable supplies where goods and services are used which are also used for exempt use". These words she says suggest close relationship and refer to the situation where there is a direct link between input and output rather than to overheads. That impression is strengthened by the paragraph which deals with bar sales: there, apart from the directly used drink and tobacco inputs, the Guidelines recognised that there would be other expenditure, and the earlier versions of the Guidelines included the phrase: "-for example on glasses cleaning materials and items of equipment" which expenditure was to be covered by the 5% allowance of that tunnel. There was no mention of overheads and the language was redolent of closely related consumables only.
69. Mr Sheldon says that the language is at best neutral. It does not refer to overhead inputs, but neither does it refer to "closely linked consumables"; inputs which "relate" to an activity naturally include overheads even if they cannot be directly connected to it.
70. On the language generally we agree with Mr Sheldon. The language does not seem to us to be that of direct connection and attribution only. But the examples given in relation to the 5% allowance did suggest that overheads were at the very least not uppermost in the parties minds.
71. Second Mrs Hamilton looks at the columns in the grid (reproduced at para [46] above). Columns F and G deal with "Input tax partly attributable". The "Total relevant tax" is put in column F and the attributable amount in column G. That she says is a clear indication that the input tax "partially attributable" was not the total residual input tax incurred by the College in respect of all its activities.
72. We find this aspect of the Grid and the Guidelines difficult. The Guidelines do not provide any detailed explanation of how these columns F and G should be filled in in the case of activities other than the three formulaic tunnels. Indeed paragraph 14 of the guidelines (reproduced at paragraph 46 above) simply says that a scheme has to be devised for these tunnels. Is the whole of the university's partly attributable tax meant to go into column F and then some fraction of it into column G? How is the proportion to be determined? Does it depend on whether the input is shared with another activity? The relevant paragraph of the guidelines says "in respect of activities where both taxable and exempt outputs arise, apportionment of input tax will be necessary". But neither is that exercise reflected in the headings of columns E, F or G or is there any other help in the text other than the reference to a method like that in notice 706. This issue worried the tribunal in University of Sussex. At paragraph 34 they said:

"A most telling point in our opinion is that, whilst the guidelines permit some recovery of input tax used for both taxable and exempt supplies where these have been gathered together in a single tunnel, they do not describe any way of dealing, for example, with shared input common to two taxable activities that are separately tunnelled. Neither in the guidelines is there a method specified for dealing with input tax that is common to a wholly taxable activity that has been tunnelled and an activity also separately tunnelled giving rise to both taxable and exempt supplies – a so called mixed tunnel. Similarly the guidelines do not spell out how input tax shared by two or more mixed tunnels or attributable both to wholly exempt activities and to any number of exempt activities is to be dealt with…"

73. It seems to us however, that a comparison between the heading of column E: "Input tax wholly attributable", and that of columns F and G: "Input tax partially attributable" suggests that F and G are there to deal with some form of apportionment of input tax. Where the activities of the tunnel are wholly taxable, the question of apportionment between exempt and taxable does not arise, and if F were the total of the inputs (including overhead inputs) attributable to more than one activity, G could be an appropriate fraction of that amount for the activity in that tunnel. And where the activity in a tunnel does potentially comprise exempt and taxable activity (such as car parking), then the grid could perfectly well be used to achieve a result consistent with the Directive if two entries were put in each of the relevant columns F and G: the first entry for the inputs directly related to the activity only – which would be apportioned on the basis of the ratio of taxable to exempt outputs in that activity, and the second entry for VAT on other inputs which would be apportioned by a fraction obtained by multiplying that ratio by the proportion which the total (taxable and exempt) activity in that tunnel represent of the total outputs of the university (assuming that such an attribution was for that output a fair proxy for use, and if not by some other agreed portion). (Indeed the lines under each type of taxable activity "Totals Group (i)" etc suggest that more than one entry in each box was contemplated.) Alternatively some combined agreed weighted average fraction might be used. The words of the guidance are vague, but the method we outline seems to us to flow naturally from the format of the Grid. It would be a scheme for completing the Grid and making claims which would be consistent with the reference in paragraph 14 of the Guidance to Notice 706, and which could have been agreed locally.
74. On that basis it does not seem to us that the Grid supports the contention that overhead input tax is not intended to be taken into account in the computation of other tunnels' recoverable VAT under the method which the Guidelines would give rise to. The most that can be said is that the Guidelines are vague, uncertain or in need of interpretation or the devising of detailed methods or fractions, but we incline strongly to the view that the Grid can, and was intended to, be used to produce an apportionment which encompasses all the categories of input tax which should be recoverable in respect of each particular tunnelled activity.
75. Third, Mrs Hamilton points to the special arrangements for the self supply of stationery in Annex B to the 1973 guidelines. In that Annex there is a worked example which distinguishes between "Supplies - input" and "wages and overheads"; the deductible amount is shown as being the input tax included in the cost of supplies "to the print room" – the former only. Even though "overheads" in this context could mean NI and pension costs, or other non VATable supplies received, the context suggested that overhead residual input VAT was being ignored. We thought that this point had some weight.
76. Fourth, paragraph 44B of the 1990 version of the Guidelines deals with the advent in 1990 of the Capital Goods Scheme. It explains the terms of the scheme they will apply to universities thus:

"When you acquire a capital item covered by the scheme you follow the normal rules for claiming the input tax you incur on it…if you use it for both taxable and exempt supplies you claim the proportion of the input tax to reflect the extent of taxable use"

She notes that there is no reference to tunnelling or in particular to the formulaic tunnels.

77. Mr Sheldon says that a 1990 amendment to the Guidelines is a poor aid to the construction of the 1973 thinking applicable to them. He says that perhaps the author of the amendment simply overlooked the formulaic tunnels, or did not consider the Capital Goods Scheme relevant to conferences, catering or bar sales, or had in mind the ability of the university to opt out of the tunnelling system
78. It seems to us that the formulation of this paragraph makes perfect sense in the context of the non formulaic tunnels. There, as we have suggested above, overhead residual input VAT could, under a scheme to be agreed locally, be apportioned between the various activities and within those activities as between the portion relating to taxable and the portion relating to exempt supplies and a different fraction applied to each or alternatively a composite fraction to the whole. There is nothing inconsistent with that in the paragraph. What is less clear is the applicability to the three formulaic tunnels. The percentages used for the recovery in these formulaic tunnels were (see para 44 above) estimates based on figures collected in 1973. It is odd, but not wholly unsurprising if no consideration was given to revisiting those figures in the light of this change to the VAT system; of course it is possible that it was considered and a decision made not to revisit the percentages or that it was simply overlooked because the system was by then regarded as set in stone.
79. In relation to the background to the Guidelines, Mrs Hamilton makes two points. First she says that they were drawn up on the premise that universities would not suffer from the imposition of VAT because their grants would be increased correspondingly. Paragraph 18 of the 1973 Guidelines said: "universities can expect to receive full compensation for cost increases attributable to VAT from the time it becomes operative". She also points to paragraph 20:

"…cost increases may also arise in relation to activities which are…within the scope of the exemption and where there will be no opportunity to recover related input tax. Such activities include the provision of food in exempt dining outlets and the supply of accommodation in halls of residence (where such factors as higher charges for maintenance and repairs work undertaken by outside contractors could have effect on the total operating costs)"

She says that this is not consistent with such input tax being recovered by means of tunnelling.

80. Mr Sheldon says that the government's intention that universities should not lose out is not inconsistent with the inclusion of overhead residual VAT within the formulaic tunnels, and that in any event the greater part of the VAT costs referred to in this paragraph would have not have been reclaimable because the inputs were used for exempt purposes.
81. It seems to us that the paragraph Mrs Hamilton refers us to does not indicate any understanding that there would be no recovery in respect of the overhead costs. It is dealing with the effect of exempt supplies and saying that in relation to such supplies there would be no recovery. That meant that the universities' grants would increase correspondingly. There is nothing inconsistent in that paragraph with overhead VAT being encompassed by the formulaic percentages.
82. Second Mrs Hamilton points to the changes in the VAT regime. Until 1990 there was no capital goods scheme; fuel and power became subject to VAT in 1989, and more building supplies bore VAT after that date. Two of these changes had been noted in the revised versions of the Guidelines, but there were no changes in the formulaic tunnels to compensate for the additional irrecoverable VAT. Surely if it were intended that overhead VAT had been included in the three formulaic tunnels, Customs and Excise, if they were acting fairly or reasonably, should have considered a change to the percentages.
83. Mr Sheldon says that this observation cuts both ways. Surely the colleges and universities should have recognised their additional costs and made a claim either for a separate credit for that tax (if they believed that it was not covered by the tunnels) or for an adjustment to the tunnel percentages (if they thought that it was).
84. It seems to us that if the formulaic tunnels were originally intended to cover overhead residual VAT, then a failure to revisit the computation after the percentages had been used for 15 or more years would not be surprising. It does not seem to us to point clearly to any institutional memory that that the formulae were not intended to cover such VAT.
85. In relation to the operation of the Guidelines in the context of the partial exemption regime, Mrs Hamilton says that including overhead VAT in each tunnel would have been inconsistent with the principles underlying partial exemption in three ways. First, if each tunnel were intended to deal with all the residual input tax there would have been a risk of overclaiming input tax: in a fully taxable tunnel if all the input tax were included, including overhead input tax, there would be full recovery of all that input tax, notwithstanding the fact that the university made exempt supplies. But it was a fundamental principle of the partial exemption regime that a partially exempt taxpayer should not get all his input tax back.
86. This concern echoes that of the tribunal in University of Sussex which is set out at para 72 above. It seems to us, however, that the existence of the concern is dependent on the assumption that the "total relevant tax" in column F is not divided between directly attributable and overhead input tax, and that only one percentage is applied to the combined total to arrive at the figure for column G. Since the Guidelines are silent about these matters, it seems to us that one should interpret the Grid in such a way that makes sense in the context of the VAT regime rather than a way which is clearly nonsensical. Do you start by assuming that where the Guidelines are silent you adopt an approach which makes sense in the context of the VAT regime, or do you adopt one which gives a silly result? If you adopt the latter then you can conclude that you also get a silly result for the formulaic tunnels. But why not adopt a sensible approach to the Grid?
87. Second, Mrs Hamilton says that a taxable person is entitled to recover all his input tax attributable to his taxable activities. The formulaic method is only a sensible proxy for taxable use where costs increase proportionately with turnover. That is the case where in relation to inputs which are directly related to supplies made (such as beer glasses) but is not the case in relation to overheads, such as refurbishment or capital expenditure. Thus if the method had been intended to be fair – and that must be the assumption because only a fair method could have been directed - it cannot have been intended to be a proxy for the recovery of overhead tax on the basis of use.
88. It seems to us that it is inherent in a partial exemption method that there is some element of approximation, of rough and ready calculation. By permitting recovery of VAT calculated as a percentage of taxable turnover there is a possibility that the recovery which the percentage suggests may not accurately reflect the use of the overhead. Any method which attributes overhead tax on the basis of the turnover of a particular sector will suffer from the criticism that the use of the overhead may not increase with turnover. But that is the method suggested by the relevant Directive. Making the link to turnover absolute rather than by reference to the ratio of taxable to exempt turnover does not significantly affect the fundamental criticism of the method. The real criticism arises where the effect of what is done significantly distorts the picture. But given that in 1973 most of a university's or a college's turnover would have been exempt, the effect of this distortion seems to us to fall within the rough and readiness of a fair approximation.
89. Then Mrs Hamilton turns to the conduct of the parties in relation to the Guidelines. We accept that, particularly when first promulgated, the way the Guidelines were operated can properly be taken to shed light on what the method actually was. It is also possible, since this was a method which was not "directed" but was "allowed" or "approved" that the method may have changed over time, and the method as originally conceived was not the method later used. Later conduct may also evidence an institutional memory of what was originally intended.
90. The Appellants note that the Respondents paid earlier claims for input tax on the basis that they reflected "an apportionment of residual input tax… according to the use to which the supplies were put", and that similar claims were paid for Queen's College, and St Catherine's College Cambridge, and Balliol, St Hilda's and Exeter College Oxford.
91. In 1994 Balliol College made a series of retrospective claims dating back to 1980. Customs accepted that such a claim could be made in respect of general overheads on the basis that they were not embraced within the 20 per cent already allowed. The college argued in addition that the input VAT on major repairs and equipment purchases was not included in the 20 per cent figure. The Commissioners accepted and paid a number of claims in relation to those years on that basis.
92. In 1991 St Catherine's College Cambridge applied for the retrospective recovery of input VAT in respect of expenditure on refurbishment of accommodation. The period for which the claim was made started in 1989. It appears that the claims were paid.
93. In 2003 St Hilda's College Oxford made a claim in respect of the years 1973-1995. Mr Sutton, the Commissioners' officer wrote to the college indicating that although the formulae were accepted for many years "it became increasingly clear that the colleges were not recovering any of the residual input tax they could have had on residual costs beyond those I have mentioned above e.g. on capital works to buildings used for conferences…on repairs and maintenance to the fabric of the colleges, on office overheads, utilities, computer facilities, professional fees and so on." The claim in an agreed amount was paid.
94. Exeter College Oxford made a similar claim, received similar correspondence, and its claim too was paid. Wadham Colleges early claims were also paid. (See also the accepted evidence of Mr Harvey at paragraph 117 below).
95. Mr Sutton, onetime the Respondents officer controlling the Oxford Colleges, gave an address on VAT changes on the withdrawal of the CVCP Guidelines. In it he referred to the effect of the Edinburgh University decision that "just because the method you were using was silent on [the right to reclaim residual input tax] did not mean that you had signed away your entitlement to claim it… Many of you have already found it worthwhile to make retrospective claims of the input tax". He then stresses that this is not a retrospective change of method but retrospective claims of amounts found to have been claimable under an existing method:

"The existing method is considered to require that this input tax should be recovered according to the use for which the purchases are put, but does not specify how that use is to be calculated. Whichever method is used should not cover any supplies where part of the supply is considered to have been claimed under the tunnelling percentages."

"The attitude of Customs to this is that if we had known that was what we were agreeing to we wouldn't have agreed to it…"

96. It seems to us that these actions reflect a more recent understanding of the Guidelines and shed little light on what the method was at the time it was used. The evidence of Mrs Farr was that these claims were paid on the basis of the Respondents' understanding of the effect of the Edinburgh and Sussex decisions rather than on the basis of their understanding of the method at the times it was used.
97. Mrs Farr told us, and we accept, that despite a search of the Respondents records nothing had been found which shed direct light on the computation of the percentages for the three formulaic tunnels: records were routinely destroyed after a period. She had however found some documents in a file relating to Voluntary Colleges of Education which were referred to on her witness statement. Late in the course of the proceedings, on the morning of the third day of the second hearing, Mr Sheldon produced a document which Mrs Farr had found in the Respondents' files on the previous evening and which we decided to admit in evidence. It was a letter dated 13 April 1973 from the Respondents to the Department of Education and Science and was about VAT and Further Education Colleges. It appears that when the discussions with the Committee of Vice Chancellors and Principals were taking place, other Higher Education Colleges made representations that they too should be embraced by whatever VAT rules had been deemed appropriate to the universities. There was a meeting and this letter resulted from that meeting.
98. An early paragraph of that letter says:

"(a) the figures of 20% in paragraph 12 and 5% in paragraph 13 of the [March 1973 CVCP Guidelines] are provisional pending the production of evidence justifying their adoption. Any Voluntary College of Education which used those percentages to calculate deductible input tax must be prepared to submit, from its normal accounts, figures to show whether the percentages are appropriate."

99. The letter encloses two pro forma statements showing the information required for the calculation of these percentages, and asks for the figures to be supplied by two representative colleges. The first proforma is for halls of residence accounts and the second for catering accounts. In each proforma expenditure is to be listed under various headings together with the associated input VAT (if any); and income is to be categorised in classes which would permit the ratio of taxable to exempt supplies to be calculated. The following categories of expenditure are listed for halls of residence:

(i) Salaries and Wages

(ii) Food

(iii) Laundry

(iv) Fuel light and Water

(v) Cleaning materials

(vi) Printing Postage and Stationery

(vii) Telephones

(viii) Miscellaneous

(ix) Repairs renewals and replacement of equipment.

And for Catering:

(i) Food

(ii) Salaries and Wages

(iii) Fuel Light and Water

(iv) Office expenses

(v) Maintenance of equipment

(vi) Cleaning and Laundry

(vii) Vending machines

(viii) Transport

(ix) Provision for renewals

(x) Building Alterations

(xi) Sundries.

100. Mr Sheldon says that the inference must be drawn that this information was needed to compare with the exercise which had been done for universities and that the computation of the 20 per cent, 20 percent and 5 percent for the universities must therefore have been done by reference to the same or very similar expenses and it is clear that these expenses included overheads. Thus he says the input tax recovery under the three formulaic tunnels was intended to include overhead residual VAT. Mrs Hamilton says that the letter shows that the percentage formulae were not intended to take in to account all the relevant overhead VAT. She notes at the time of the computation fuel light and power would not have borne VAT and so the formulae cannot have taken into account the input VAT which later arose on these items. Neither is construction mentioned, nor in the case of halls of residence Building alterations. The method adopted to give an amount of recoverable VAT was so deficient that it cannot have been a fair proxy for the recoverable input VAT.

Issue 2: Discussion

101. First we recall that the Second directive said that "a taxable person may, exceptionally obtain administrative permission to determine special pro rata figures for certain sectors of his activities". In our opinion the Guidelines and their computational expression in the Grid are capable of being or forming a framework for such permission given in a manner consistent with that Directive. Likewise the Sixth Directive permitted member states to authorise a taxable person to determine a proportion for each sector of his business or to compel a proportion for each sector, and the Guidelines are capable of being consistent with that authority.
102. Second we recall that the standard method is a rough and ready proxy for use, and that any method other than an impractical one which analyses each input's use is going to have rough edges. As the tribunal said in Merchant Navy, Article 17.5(c) must authorise attribution on the basis of estimated or assumed use. The power given by the regulations to the Commissioners is to make regulations to secure a fair and reasonable attribution. They are not required to secure an exact attribution
103. Third, it seems clear that the effects of changes in VAT legislation, of the increase in conference activity, and of the increase in the overhead expenditure (including capital expenditure) occasioned by the greater demands of conference attendees, all meant that over time the 20 per cent figure, even if it fairly reflected related overhead and other residual properly recoverable VAT in 1973, ceased so to do at some time in the 24 years in which the Guidelines operated. But with any special method circumstances will change and the method become more or less fair or reasonable. But such a change in circumstances does not mean that the original method – or the continuing method until it is changed- does not give the taxpayer his right to the recovery of input tax given by the Directive. Changes normally give rise to a review of the calculation. The fact that, because of the reluctance of the Colleges to rock the boat, or because of the inertia of the Respondents, there was no review, does not in our view affect the nature of the method or that, until replaced it defined and gave effect to the taxpayers' Directive rights. (To the extent that colleges made claims for additional input tax recovery substantially contemporaneously with the submission of their returns, the acceptance of those claims can be seen as an acceptance by the Respondents of a variation of the special method being used. So seen they indicate that absent such contemporary or prospective variation and approval the special method did not encompass the right to additional recovery reflected in the variation. To the extent that the claims were truly retrospective we do not see how their acceptance is evidence of what was used and allowed or approved.)
104. Fifth, we do not find any particular help in Sussex or Edinburgh in relation to this issue. In our view the CVPC guidelines and the Grid provided the basis for a university or college to adopt a special method. Properly applied and with local agreement in relation those classes of taxable activity (including, it must be noted, the Miscellaneous Class) other than the formulaic tunnels, the Guidelines and the Grid were in our view capable of delivering a method which permitted the recovery of all directly attributable input tax (whether relating to one taxable output activity or many) and the recovery of properly attributable residual input tax.
105. On the evidence before us we conclude that the percentage formulae were calculated so as to deliver to universities and colleges what at the time was thought to be a fair approximation of all the residual input VAT to which a taxpayer was entitled in respect of those particular activities. The contents of the letter regarding the Voluntary Colleges and the general approach in paragraphs 11 to 14 of the 1973 Guidelines weighed heavily in reaching this conclusion. We conclude that overhead residual VAT was included in the 20, 20, and 5 percent formulae.

Issue 2A: Mrs Hamilton's fallback argument: the option.

106. If we found that overhead residual tax was included in the formulaic tunnels, Mrs Hamilton had a fallback argument. In this case she says that it was inherent in the method permitted that a college or university had the option as part of the method to elect at any time for an apportionment based on actual figures rather than the percentage formulae. Paragraph 12 of the 1973 Guidelines said:

"The alternative to acceptance of the formula, which any university is entitled to explore and opt for, is full and rigorous accounting for all input tax included in the cost of goods and services used to make these particular supplies [i.e. the formulaic tunnel supplies] and the presentation to Customs and Excise of a detailed breakdown of the apportionment of that input tax between taxable and exempt outputs.".

107. Similar language appears in the 1977 version of the Guidelines and, in the 1987 and 1990 version of the Guidelines the following:

"An alternative approach to the methods outlined…must be agreed with the local VAT office and will normally be required to apply for at least two years."

108. Mrs Hamilton says that once Edinburgh had made it clear that failure to make a claim was not an agreement not to make a claim, the Respondents were prepared to accept claims for residual input VAT. These were claims not to a new retrospectively permitted special method, but within the terms of the special method which had been used and permitted and was framed by the Guidelines.
109. Mrs Hamilton took us to claims which other colleges had made referred to above. She says that the approach taken by the Commissioners in relation to those claims must be indicative of what the special method was. If the Commissioners' conduct did not evidence that the method did not embrace overhead VAT, at the very least it evidenced that the Commissioners considered that the colleges had an option under the method to seek full recovery.
110. It seems to us that the Guidelines indicated to universities and colleges what methods would be acceptable – i.e. what methods would be allowed or approved by the Respondents. The colleges could make their returns and claim their input VAT using a method outlined in the Guidelines in the knowledge that it would be accepted by the Commissioners. The Appellants used the formulaic tunnels. That was in our view the selection of one type of method from those available. But we are not able to read the quoted paragraphs in the Guidelines as indicating that having used the formula for the three tunnelled outputs there remained an option for another method. The nature of the whole method might be wider than simply the formulaic tunnels but it did not embrace a further option once those formulae had been used to opt for a totally different method as respects the input recovery relating to those outputs.
111. Neither does the correspondence with the other colleges affect this conclusion. The correspondence indicates to us that the Commissioners' officers accepted that the method which had been operated had become unfair, and that they considered that the taxpayers had an entitlement to further input tax credit. That is not the same as an understanding that the agreement permitted an option for a further calculation.
112. We therefore reject the argument that the method contained within it an option to elect for further recovery on the basis of more rigorous accounting.

Issue 3: If we are wrong in our conclusion on Issue 2, what should be used to "fill in the gaps"?

113. This was the issue which formed the meat of the submissions put to us at the first hearing. We set out below the arguments advanced by the parties at that hearing. In the case of the Respondents' arguments these must now be seen as a fall back from the arguments on Issue 2
114. The Respondents put the issue thus in their original skeleton argument:

"The principal issue before the Tribunal is: which is the applicable methodology for calculating the value of the Appellant's claim for under-recovered residual input tax between 1973 and 1994? There are two options:

(i) The apportionment known as the Standard Method as set out in Regulation 101 of the 1995 Regulations (and its predecessor provisions).

(ii) The Conference Method operated by the Appellant since September 1997 and formally approved by the Commissioners on 1^st August 1998."

115. The Appellant says that the most appropriate method to determine the creditable tax is the "Conference Method". Under this method a fraction of the residual input tax is treated as recoverable. The fraction is the product of two other fractions. The first represents the portion of the year for which the college was available for conference activity, the second represents the portion of that activity which was taxable. Thus it assumes that inputs accrue evenly over the year or are evenly used over the year in the activities of the college, that in term-time those inputs relate to exempt activities and out-of-term-time to conference activities, and that the inputs out of term-time are used for taxable and exempt conference activity in the proportion that the income from those activities arises.
116. The Respondents say that where a special method does not address a portion of the residual tax that unaddressed input tax must be apportioned using the standard method i.e. in the proportion of taxable to total supplies.
117. From the evidence of Mr Harvey we find the following facts in relation to Wadham College. The position in relation to Merton College was not materially different:

(i) following the withdrawal of the CVCP guidelines in 1997 the college adopted a special method based on the Conference Method from 1 August 1998.

(ii) On 22 September 1997 the college made a claim in respect of the years ending 31 July 1995 and 1996 for residual input tax under-recovered under the input tax recovery procedures it had adopted. The claim was based on the Conference Method.

(iii) Mr Sutton, the officer of the Commissioners who controlled the VAT affairs of the Oxford colleges at that time, accepted that claim and it was paid in full on 15 October 1997.

(iv) A further claim for the year to 31 July 1997 was made on the same basis on 5 May 1998 and was paid in full on 10 May 1998.

(v) Whilst litigation over the applicability of the 3 year cap in regulation 29(1A) was being pursued, Mr Harvey entered into correspondence with Mainprice Nash Miller (MNM) the college's VAT advisers in relation to the quantum of any claim the college might have for the years 1973 to 1995 if the Courts upheld the challenge to regulation 29 (1A).

(vi) In that correspondence Mr Harvey and MNM came to an agreement of the amount of the claim that would be accepted if (a) the challenge to regulation 29(1A) succeeded, and (b) the Conference Method was used to calculate the recoverable input tax.

(vii) Later, following an exchange of emails with other officers of the Commissioners, Mr Harvey informed MNM that the Commissioners' decision was that the quantum of the claim should not be determined by the Conference Method: "as no special method had been agreed prior to 1998 all claims for input tax prior to that date should have been made on the basis of the standard method."

(viii) Mr Harvey and MNM then corresponded on the quantum of the claim on the basis of the Commissioners' argument that the standard method (as provided by the legislation at any time) should be used to plug any "gaps" in the college's recovery.

(ix) When the output based standard method (i.e. one in which the standard method fraction had outputs in the nominator and denominator) was used for the whole period Mr Harvey calculated that the claim would be for £88,168; if for the years 1998 to 1992 a use based standard method (i.e. using the faction: input tax recovered under tunnels ÷ residual input tax) was applied, this figure fell to £55,383.

(x) The calculations prepared by MNM and discussed with and adjusted by Mr Harvey were not based on records of actual figures for the relevant inputs, outputs and VAT for the relevant years because such records were not maintained or have not been retained. Instead they are on the basis of extrapolations from available figures. Mr Harvey's assent to the calculations was an assent to their being calculations based on reasonable assumptions and being correctly calculated to determine the input VAT which would be payable on the basis of the applicable method.

118. From the evidence of Stephanie Farr we find that there was no improper motive in the Commissioners' policy decision not to apply the Conference Method for the years 1973 to 1995. Mrs Farr put the argument thus: whereas the Conference Method is accepted as fair and reasonable after 1998, and whereas some officers may have accepted it as fair and reasonable for 1995 to 1997, it is only one of many possible methods each of which is a proxy for use of the relevant input. What is a fair and reasonable proxy depends upon the circumstances of the business; the circumstances were or could have been very different 30 years ago: who knows whether that method would have been fair and reasonable then? The standard method, on the other hand, (in its various enacted manifestations) is a statute based reasonable proxy for use which for the majority of traders is a fair and reasonable method.
119. The calculations in relation to the conference method discussed by Mr Harvey and MNM were concerned with the estimation of the figures in the Conference Method Formula:

Input Tax x Taxable Conference Income Fraction x Availability Fraction

120. Proper records were not available for the input tax in any year, or the percentage of conference incurred which was taxable. So estimates were needed. The estimates were made on the basis of the VAT returns (and the grids) and the accounts of the colleges. We were shown computations which we accept were the basis for a fair estimate under this method. But we have not however been asked to settle upon an amount of input tax but to leave this to discussion between the parties following a decision in principle as to the method to be used.

Issue 3: The contentions of the parties.

121. It is common ground that in the relevant period each College used a method other than the standard method. It is also common ground for the purposes of Issue 3 that the effect of using that method was that the college failed to claim deduction of input tax which it was entitled to deduct: the college claimed its deductions using a method which was based on the CVCP guidelines but its claims did not relate to all the college's taxable activities and did not (for the purposes of this Issue) deal adequately with the residual input tax which related to the three formulaically tunnelled activities: its claims left "gaps" or "holes".

The Commissioners' contentions

122. For these purposes the Commissioners do not dispute that the college is entitled to reclaim additional input tax. Mr Sheldon invites us to fill in the gaps with the standard method. That method he says was the only method for which the regulations provided in the absence of a directed or permitted special method, and is therefore the only permissible method.
123. The Commissioners contend that the gaps are to be filled in by applying the standard method as it was constituted from time to time to the activities of the college to determine what the allowable input tax would have been under that method. The under recovered tax is then to be computed as the excess of that amount over the tax actually claimed.
124. Mr Sheldon says that the CVCP guidelines are not themselves a special method but a setting for the construction and adoption of a special method. He says that the issue in this appeal is the input tax which fell into the gaps in the CVCP guidelines. By definition he says the CVCP guidelines can say nothing about the gaps - because they are gaps in those very guidelines.
125. He accepts that the college had a de facto special method - one which was allowed or approved by the Commissioners - but its terms he says were particularly sparse. Even sparser than the sieve-like guidelines.
126. Where should we look for the methodology to fill the gaps in the de facto special method? There is nothing in what was done which helps. The gaps in what was done replicate gaps in the CVCP guidelines: we cannot look there. All that remains, he says, is the Regulations.
127. What do the Regulations say? They say that where there is no special method, the standard method applies. The words are clear.
128. Mr Sheldon says that if one searches for pointers that other tribunals have filled in the gaps with a standard method, one finds them in University of Sussex. There the tribunal at paragraph 25 accepts calculations on the standard method, and at paragraph 36 refers to "ordinary principles of VAT law": that means, he says, the standard method.
129. Mr Sheldon's secondary submission is to ask what is the nature of the Appellants' alternative. He says that in reality the Appellant is inviting the retrospective application of a special method, but attempting to disguise this by saying that the special methodology should be applied retrospectively. There is no magic in an "ology" he says: if you are filling a gap or perfecting a method, it is the same thing: creating retrospectively something which was not there at the relevant time.
130. Regulation 101 of the current VAT Regulations and its predecessors used mandatory words: "shall be the amount". A retrospective approval or allowance is required to displace that requirement and there has been none.

The Appellant's Contentions

131. Mrs Hamilton invites us to fill in the gaps using the Conference Method. She says that each college had a right to claim input tax deduction and only a fair and reasonable method will give effect to that right. The college says that the standard method is not possible to apply in its full rigour and, in any event, is not fair and reasonable and does not give effect to the college's right under the relevant EU Directives to recover relevant input tax: it says that the proper method to evaluate its entitlement is the Conference Method. She says that the Conference Method is a method which has been accepted as fair and reasonable by the Commissioners in relation to later years and is the method which should be used. She says that the standard method does not achieve a fair and reasonable attribution and therefore cannot be used.
132. Mrs Hamilton says that the college had a directly enforceable Community Law right to deduct input tax. That right is given by Article 17.1 and 17.2 by reference to use of the input. Article 17.5 provides for the calculation and exercise of that right: it provides for the Article 19 standard method to apply as a proxy for use but subject to special methods. She says that a fair and reasonable method must give effect to that right and that the Article 19 method could not in all circumstances be fair.
133. We asked Mrs Hamilton why the Community Law right which had accrued to the college was not simply the right to an Article 19 proportion; after all the Directive did not compel member states to adopt any of the methods in Article 17.5(a) to (e): the adoption of sectorisation and use based methods were optional. Why was the right not limited to the Article 19 proportion? She said that in the context of domestic legislation which took advantage of Article 17.5 (a) to (e) and required fair and reasonable attribution, it could be (and in this case, was) the case that only a non-Article 19 method could give fair and reasonable effect to that Community law right.
134. She makes the following points:

(i) the basis of the CVCP guidelines was that the standard method was not fair and reasonable. That was accepted by the Commissioners between 1973 and 1997;

(ii) the Conference Method has been accepted as fair and reasonable by the Commissioners for other periods. She pointed to the actions of the Commissioners officers: Mr Sutton who agreed to pay the first claims in 1997, Mr Madelen who paid Wadham College's second claim and Mr Curtiss who agreed to that method going forward in 1998.

(iii) even for 1973 to 1997 certain officers of the Commissioners accepted the Conference Method as fair and reasonable in relation to that period. Mr Harvey had clearly been of that view until he had had his discussions with his colleagues.

(iv) the Conference Method had been accepted for other colleges.

(v) if there was a direct EU right, it should be capable of exercise without discrimination between taxpayers: similar colleges in similar circumstances had received repayments based on the Conference Method;

(vi) the difference between the claim under the Conference Method (£400k) and that under the standard method (£55k) was vast. The difference showed that the standard method was not fair and reasonable.

135. It was common ground, she said, that the college had operated a special method, that the method had been allowed or de facto approved, and that it was not perfect. There were gaps. But she says that the standard method in what is now Regulation 101 is irrelevant to those gaps. Once it is agreed that there is a special method Regulation 101 does not apply. Regulation 102 reads "the Commissioners may approve or direct the use… of a method other than that specified in regulation 102". The earlier regulation said "subject to" allowance or direction of a special method. Once you get into a special method Regulation 101 is otiose.
136. It might be that a standard method calculation could form part of a special method but that would be specific borrowing of the methodology as part of that method. When Regulation 102 applied, it applied to the whole business not part. It was all or nothing.
137. Mrs Hamilton said that there was an accepted special method; what was required was a means of claiming residual input tax on the basis of use. The college contended that this was not a new method but just the old one with that right properly recognised. Mr Sutton had said the same in 1997. In doing this the college's argument was simply that adopted by the tribunal in University of Sussex: there the tribunal had found that the method was not confined to what was actually done but must incorporate the right to residual input tax reclaim. The tribunal there had not given express consideration to the computation of the amount due under that right but that computation was just as much within the purview of the method as was the existence of the right.
138. That computation, which was part of the method, had to be fair and reasonable. The standard method filler offered by the Commissioners was not fair and reasonable. As the tribunal accepted in Merchant Navy what was fair and reasonable depended upon the alternatives. The Conference Method was the only alternative accepted as fulfilling this criterion.
139. The Commissioners use of the reversion to the standard method was flawed. The standard method applied to the inputs and outputs of the whole business; it had to be adapted when some of those inputs and outputs had already been dealt with under the special method for example:



140. It was impossible to apply the relevant standard method in its full glory without ignoring the special method. The Commissioners' case could not be simply that the standard method should be used to fill the gaps, it had to be that there should be the use of an adapted standard method. That was not reverting to Regulation 101. It opened the question: what in these circumstances is fair and reasonable. Once asked that question did not admit the mechanical answer: the standard method.
141. The Commissioners were arguing for a special method which incorporated elements of the standard method suitably adapted to fill the gaps. The question for us was whether that was fair and reasonable, or more fair and reasonable than the Conference Method.

Issue 3: Discussion

142. Mrs Farr referred us to the Commissioners published policy in relation to gaps in special methods as it stood at the time the college's claims were submitted. In VI-15 Partial Exemption at paragraph 7.8 that policy indicates that a modified form of the standard method should be used to fill the gaps. The existence of that policy now may make clear that in relation to a special method approved after its publication the approval is on the basis set out. But that policy does not help with the college's position. There was no evidence that that policy was published or drawn to the college's attention in 1973 and the terms of the CVCP guidelines suggest it was not in existence at that time. Thus that policy cannot in our judgement be used to determine what was allowed between 1973 and 1994.
143. Our task under this heading is to determine what the special method adopted by the college was rather than to decide between one method and another. We remind ourselves of the tribunal's comment in University of Sussex (paragraph 36): "In our judgement we should seek to interpret the material before us as containing a special method to achieve [the result that a special method should address all the economic activity of the University]." We do this against the background of the relevant EU Directive, the requirements of the relevant Acts that the method be just and reasonable, the Regulations in force at the relevant times and the CVCP Guidelines which formed part of the method.
144. The EU Directive(s) give a right to recovery of attributable residual input tax. The Directive(s) based the right to recovery on use, and as a proxy, give a right to the standard fraction. Absent the exercise by a Member State of a right to impose a different method, the taxpayer's right under the Directive is to the standard fraction. However, where a Member State authorises a method other than the standard method that exercise must, if it is to be within the powers granted by the Directive to the Member State, be attribution on the basis of estimated or assumed use. As the tribunal said in Merchant Navy:

"22. There is normally no problem with inputs directly used for taxable or exempt supplies. The problem arises with overheads and other indirect costs. It also arises however with inputs which could in part or in whole be attributed on the basis of actual use but where such attribution would be burdensome or difficult to check: an example might be telephone calls.

"23. Once attribution on the basis of actual use is impossible or impractical any other method of attribution can only be designed to approximate to actual use being only estimated or assumed. The standard method is very rough and ready and may result in attribution which is demonstrably very different from probable actual use however it has the important merit of simplicity.

"24. The methods authorised under Article 17.5(c) are clearly different from that under Article 19, since otherwise Article 17.5(c) would be otiose. However it cannot mean actual direct use since, if the actual use of the input could be determined, it would not be a mixed use and Article 17.5 would not apply at all. It seems to me that Article 17.5(c) must authorise attribution on the basis of estimated or assumed use."

145. It is a principle of the application of the Directives that any discretion must be exercised with regard to the principle of fiscal neutrality inherent in the common system of VAT (see e.g. Solleveld v Staatssecretaris van Financiën C - 43/04 paragraph 39): if the college undertakes an activity the method used to determine its VAT recovery should so far as practicable leave it no better and no worse off in relation to the VAT treatment of the economic results of that activity than a person whose business is confined to that activity. This is one reason for the importance of use as the criterion for recovery.
146. The UK legislation requires a special method to be just and reasonable: that is to say it must both have due regard to the principle of fiscal neutrality and the right to recovery based on use, and also, as the tribunal said in Merchant Navy (at paragraph 25):

"[t]he method must be reasonable for the trader to operate, in that it does not involve disproportionate or unreasonable resources, and it should be capable of being checked by the Commissioners again without unreasonable effort."

147. Last we note that the CVCP Guidelines envisage the recovery of input tax by reference to each sector of the university's activities and not by reference to an overall formula.
148. We therefore decide this issue on the following basis:

(x) a special method was used by the Appellants and was allowed or approved in all relevant periods;

(xi) that method must be construed as giving rise to an entitlement to residual input tax recovery;

(xii) that entitlement arises under the relevant Directive but is to an amount computed by a method consistent with the directive;

(xiii) such implementation requires an approximation to use;

(xiv) that approximation must as a result of domestic legislation be fair and reasonable; and

(xv) that method involves the recovery of input tax by reference to sectors of activity or tunnels.

149. It is difficult and somewhat unsatisfactory to attempt to determine after a period of 30 years what the method was which was properly to be operated in those years. There may have been other methods not considered by any of the parties to this appeal which could have fulfilled the relevant criteria. But we can only do our best on the basis of the material and the contentions put to us.
150. It seemed to us that because the method we were construing already involved sectorisation, the simple application of the relevant standard method to the whole of the college's activities was not the right answer. The method would scarcely be "special" if it just permitted the standard amount to be reclaimed but by a two stage route – first part of the tax recoverable under the Grid, and then the balance. The Grid and the Guidelines would have no effect. It would just be the standard method.
151. Mr Sheldon's approach thus can only be treated as requiring some adaptation of the standard method to each tunnel or sector of activity. That meant in effect reading some additional words into what is now regulation 102: so that it reads thus (with the additional words underlined):

"102(1) Subject to paragraph (2)… the Commissioners may approve or direct the use by a taxable person, in relation to all or to a part only of his business, of method other than that specified in regulation 101…"

152. So the issue in relation to each tunnel was whether the Conference Method should apply if necessary with suitable adaptations or the modified standard method.
153. The Appellants' claim was conducted on the basis that the conference fraction should be applied to the whole of the input tax of the college so that the amount due would be that amount so ascertained less the amount recovered over the last 24 years.
154. Unless the conference and catering activity (taking education as conference activity for these purposes) was insignificant as a proportion of all activity this method could not possibly be regarded as part of the special method as properly construed. It does not involve sectorisation and would not properly attribute input tax to other activities and take into account their taxable or exempt nature. In our view it must be adapted to each relevant sector and so adapted its application must be confined to each relevant sector.
155. We consider first the three formulaic tunnels. If the 20 per cent, 20 per cent, 5 per cent recovery in relation to the three formulaic tunnels does not provide for recovery of overhead residual input VAT, what should the method provide for in relation to those three activities to determine the proper entitlement and thus the under recovered amount? We start with the first two of those - the 20 per cent conference and catering tunnels.
156. In our view the method should be construed as incorporating a method in respect of the these two formulaic outputs which most fairly and reasonably reflects the "use" object in the relevant Directive. Given that income is a poor proxy for use, the Conference Method in our view better reflects that object than the modified standard method, and in our judgment would form part of the special method for those outputs.
157. But where the college has activities other than conference and catering activities, the mere application of the conference fraction to the residual VAT will produce a distortion because it is based on the assumption that the colleges' activities are either term-time (exempt) activities or out-of-term conference activities (taxable or exempt) only . Thus the fraction it produces can overstate the proportion of residual input tax attributable to the Conference (and catering) income.
158. Some adjustment is therefore needed. In the absence of any other way of attributing input VAT to conference rather than other activity, an initial division on the basis of outputs is fair and reasonable and not contrary to the approach of the Directive.
159. On that basis in our judgment the method requires that the recoverable amount in respect of the first two of the three tunnelled outputs would be (using 'conference income' to represent income from conferences and catering) :

(conference income/total income) x conference availability fraction x taxable conference income percentage.

160. In relation to the 5 per cent 'bars' tunnel, an attribution of overheads on the basis of conference activity does not make sense: the bar is as likely to be open and making taxable supplies in term-time as it is in the vacation. On the basis that the 5 per cent represents input tax directly attributable to the bar only, some attribution of the remainder of the overhead tax has to be made. The Conference method is not appropriate for that exercise so we would in this case find that the method required an attribution of residual overhead input tax on the basis of the modified standard method i.e. in the ratio of the taxable supplies made by the bar to the total supplies of the college.
161. We found the diagram below helpful to show the effective division of the input tax.

Activity Type of input	Taxable Conference and Catering	Bars	Other Tunnels
Tax directly attributable only to this taxable activity	[unlikely to be any]	Covered in Grid	To be agreed
Tax directly attributable to more than one taxable activity	1	2	To be agreed
Tax directly attributable to taxable and exempt activities	3	4	To be agreed
Overhead tax	5	6	To be agreed

The shaded box represents the total recoverable input VAT under the method. The recovery under the claims already made by the college under the 20 per cent tunnels will form part of the tax in boxes 1, 3and 5. The total of those three boxes should be calculated using the modified conference method described above. The recovery claim of 5 percent made in relation to the bar outputs will form part of the recoverable tax in boxes 2, 4 and 6. The total of the tax in those latter three boxes being determined as set out in paragraph 160 above.

5. Conclusions

1. The Tribunal's jurisdiction.

162. There are three issues. First the nature of the tribunal's jurisdiction in relation to an appeal against a special method; second whether this is an appeal against a method under section 83(e), or in relation to an amount of a credit under section 83(c), and lastly what effect the changes in stance of the Respondents has on the jurisdiction of the tribunal.

(a) the jurisdiction in relation to appeals against a method.

163. The Court of Session in Edinburgh Telford College v HMRC [2006] CSIH 13, said at paragraph 28:

"It was, furthermore, a matter of agreement that the Tribunal had exceeded its jurisdiction by stating, at page 21:

"We also determine that the appellant is entitled to a new ascertainment of a fair and reasonable special method of dealing with the attribution of tax. That in our judgment must involve sectorisation"

"While it was within the jurisdiction of the tribunal to decide whether or not a determination by the Commissioners was fair and reasonable it did not fall within the Tribunal's jurisdiction to specify a method that the Commissioners should adopt." [our emphasis]..

And in Kwik-Fit GP v C & E Commissioners 1998 STC 159 Lord McCluskey said at page 164a:

"In my opinion, this Court has no responsibility in this appeal to determine what would be a "fair and reasonable" method of regulating the attribution of input tax to supplies for this group. The responsibility of the Court in this appeal, as I see it, is simply to examine the Direction and to determine if it is properly to be described as not securing a fair and reasonable attribution of input tax to supplies in the known circumstances. In the present case where both parties are agreed what the objectives of the Direction are, it should be possible to attempt to make a judgment as to whether or not the scheme does achieve the agreed objective. If it does not then the Direction which expresses it is necessarily one which fails the "fair and reasonable test". I consider that it is both possible and entirely appropriate to make that judgment."

164. Banbury Visionplus Ltd v HMRC [2006] EWHC 1024 Ch, Etherton J held that the jurisdiction of the tribunal on an appeal under section 83(e) was not a limited jurisdiction. He said, starting at the end of paragraph 48:

"I can discern nothing in the Regulations or in section 26 or in any other part of the relevant legislative framework that confers on the Commissioners alone the right to decide whether a particular method would in fact achieve the section 26 objective, to the exclusion of the tribunal on appeal.

"49. That conclusion on the jurisdiction of the tribunal is not undermined by the fact that there may be a number of different ways in which the section 26(3) objective may be achieved, and… the task of the Commissioners is to decide, when exercising their discretion…, whether the new method better secures a fair and reasonable attribution…

"52. I can see no practical or jurisprudential difficulty in conferring on the tribunal a full appellate jurisdiction to determine, on the basis of the facts found by it at the time of its decision, whether a decision of the Commissioners under Regulation 102 substituted a method which secures, [or better secures] a fair and reasonable attribution of input tax to taxable supplies for the purposes of section 26(3) of the 1994 Act. That would be consistent with the unqualified wording of the appeal provisions in section 83(e) of the 1994 Act. It imposes an objective test which… is consistent with the provisions of Article 17 and 19 of the Sixth Directive, and which the tribunal… is well qualified to conduct."

165. That conclusion was followed in MBNA and also by Warren J in St Helen's School Northwood Limited v Commissioners for Revenue and Customs where he said that he followed the approach of Etherton J in Banbury, but continued:

j. "This is not to say that the tribunal is able to put forward its own version of a more reasonable special method (if there is one). It cannot do so…If on an appeal by a taxable person against a refusal of Customs to allow a proposed special method the tribunal decides that the method is fair and reasonable and also that it is more fair and reasonable than the method in operation… the appeal should be allowed. But if the tribunal thinks that both the existing method and the proposed method are unfair or unreasonable, it could not allow the appeal even if it considers that the proposed special method is less unfair and unreasonable than the existing method."

166. On that basis this tribunal would not have jurisdiction to substitute its own method for one advanced by the Respondents. But since what we are deciding is the proper construction of the special method used it does not seem to us that we are affected by this restriction.
167. We note that all the cases referred to above dealt with a "direction" of the Commissioners. No direction was made to the colleges in respect of the relevant years. None can now be made because the Commissioners do not have power to make retrospective directions (see regulation 102 (4)). And although Etherton J's reasoning in Banbury is in theory as applicable to an allowance or approval as it is to a direction (because the initial question is the same: is what has been allowed or approved fair and reasonable), it is, in relation to a special method which has been allowed or approved, more difficult to see how an appellant can appeal against what it has done and had allowed or approved by the Commissioners for he would effectively be appealing against his own conduct. For the following reasons we consider that an appeal does not lie against a method which has been used and the use of which has been allowed or approved.
168. If a taxpayer uses a method and the respondents do not allow or approve its use, then the taxpayer can appeal in relation to the proportion which is at issue under section 83(e); likewise if the Commissioners direct a method and the taxpayer does not like it, he can appeal under the same section. Also where a taxpayer has used method X for a number of years with the permission or approval of the respondents and wishes to change to method Y but the Commissioners object, an appeal lies. But if a method has been used and allowed or approved for a period, then in our view there can be no appeal against the refusal of a request to use another method for that period. The taxpayer cannot choose method A and get its returns approved on that basis, and then change his mind and ask for method B and then change again for method C or D and so on, and at each turn appeal against the refusal of the Commissioners to allow or approve the new basis on which he wants to account for his VAT. Even if each subsequent method is fairer or more fairly reflects the EU or statutory rights of the taxpayer, the process of settling the method by which those rights are given quantification must take place with a degree of finality when the taxpayer makes his bed and the Commissioners approve his occupancy of it.
169. For this reason if it is the case that the colleges were using a special method in the relevant periods, then whilst we have jurisdiction to determine what the creditable VAT should be on the basis of a proper construction of that method, we cannot in our judgement treat this appeal as an appeal against the method which has been allowed or approved. The point can be illustrated by the effect plethora of methods which might otherwise be regarded as in play:

(i) The Conference Method as proposed by the Appellants

(ii) The CVCP type method actually used over the period by the colleges;

(iii) The method which the tribunal in Sussex held was being used – namely one which was not confined to what was claimed but which included the right to residual tax

(iv) The special method as the Commissioners now understand it to be, namely that the three formulaic tunnels gave a right of full recovery in respect of those three activities, and that for other activities a fair and reasonable method had to be agreed;

(v) The special method as used with the gaps filled in by a modified standard method (the method the Commissioners originally contended for at the first hearing);

(vi) The CVCP method with the benefit of "Mrs Hamilton's option".

170. For the tribunal to attempt to adjudicate between all these seems perilously close to what Warren J said that it did not have jurisdiction to do in an appeal against a method.

(b) An appeal under section 83(c)

171. Despite the fact that the argument has related to the principles of the quantification of the claim, the effect of the decision will formally (and eventually) be the determination of an amount of repayable VAT.We have approached this appeal as an appeal under section 83(c) and not on the basis that this should be treated as an appeal against the special method allowed directed or approved by the Commissioners if such an appeal were permissible. And that is because, for the reasons which follow it seems to us that the issues raised relate to the proper construction of the special method rather than to competing special methods.
172. First, the colleges used a method other than the standard one. If what had been used had been simply a method under which directly attributable tax was recovered, then since under both the Second and the Sixth Directive the taxpayer had a right to recover residual input tax as well, a method which simply set out the calculations effecting the recovery of directly attributable tax only would not have been a special method. But the method used also embraced (at least some) residual input tax (because in relation to the conferences tunnel the 20 per cent must have had regard to the fact that some of this activity would have been exempt, and in relation to the tunnels which comprised mixed taxable and exempt activity the intention of columns F and G must have had that effect) and did so in a manner which differed from the standard method in the regulations. Unlike that in Ampleforth Abbey Trust, the method used to calculate input tax did embrace elements which relate to residual input tax. It was therefore a special method.
173. Second, that method was allowed, approved or directed.
174. In the period 1973 to 1992 a special method had affect if it was "allowed or directed". It seems clear that the method used by the colleges was allowed: computations were prepared on the basis of the Grid and the VAT was paid, credited or repaid accordingly.
175. In the period from 1992 onwards a special method had affect only if it was approved or directed. Acceptance of calculations over a period of time coupled with the existence of the CVCP guidelines in our view was approval for these purposes.
176. Third, the method was not in our view so uncertain that it could not amount to a method at all. In Kwik-Fit GP v C & E Commissioners 1998 STC 159 Lord Johnston (at page 1669j) said:

"The requirement of fairness and reasonableness involves clarity and a lack of ambiguity and if ambiguity is present I do not consider that the method so construed meets the fair and reasonable test."

177. Other members of the Court also indicated that ambiguity would lead to the conclusion that a direction of a special method failed the fair and reasonable test. That case, however, was in the context of a formal written direction. The Court was asked to construe the words of a direction; it was not asked to determine what was "allowed" or "approved".
178. But there was no suggestion from either side that the method was void on Kwik-Fit principles.
179. If we found the method void then the taxpayer would be thrown back upon the standard method. In such circumstances, the parties would need to come to an arrangement under which a new just and reasonable method was allowed or approved by the Commissioners (the Commissioners having no power (as a result of Regulation 102(4) to direct a method retrospectively). If the Commissioners failed to agree a reasonable method then the taxpayer's only remedy would be via judicial review. This would be a drastic conclusion to reach in the context of a method which had been operated over a long period and one which would give rise to unnecessary litigation. As the tribunal noted in University of Sussex, it would be a "most unattractive result". It is therefore one which we would seek to avoid if at all possible.
180. But we do not find that the method is void. In the case of a method allowed, or approved, a greater degree of uncertainty is capable of being present by the nature of the statutory words. An inconsistency or uncertainty in a formal written direction may render it void but greater latitude must exist in the divination of what is allowed or approved. The question of uncertainty must be addressed in relation to what has been permitted. If the specifics of what as been permitted are ambiguous or uncertain, then the method permitted could be void. For example, if the method which was allowed used one percentage of outputs in one period and a different percentage of inputs in another period without rhyme or reason, it would seem to us to be too uncertain to be a method at all. But where the specific elements which have been allowed are clear and doubt exists only about gaps which have not specifically been addressed in what has been allowed, the proper approach is not to treat the method as void but to seek to fill those gaps.
181. Thus in our view the issues before us related to the proper construction of the method used and permitted.

(c) the changes in the Respondents' position.

182. In its original notice of appeal the appellant appealed against the decision of the Commissioners given in a letter dated September 2004 refusing a claim for under recovered input tax of £402,087.33. That refusal was on the basis that regulation 29A imposed a time limit on the making of claims. By the time the appeal came before us the dispute between the parties was as to the quantum of the claim (or the method to be used to quantify the claim), and as the case has unfolded there has been a change in the method proposed by the Respondents to quantify the claim.
183. The Respondents did not write fresh letters setting out new grounds for refusal of the claim as made nor amend their Statement of Case to reflect the different approach to quantum (although they did amend it for the first hearing to reflect the move from the capping argument to that about quantum). Neither did we direct them so to do. Their position at each hearing was clear from Mr Sheldon's skeleton argument.
184. Section 83 VATA 1994 provides that:

"Subject to section 84, an appeal shall lie to a tribunal with respect to any of the following:…

(c) the amount of any input tax which may be credited to any person,…

(a) the proportion of input tax allowable under section 26".

185. We note that, whereas section 84(5) provides that, in relation to an appeal under section 83(p) against an assessment, the tribunal can of its own motion increase the assessment, there is no corresponding provision in relation to an appeal under section 83(c). The tribunal is not expressly given a power to say "what HMRC have agreed is creditable is too much". But in effect, by directing further argument on the issues raised in our direction, and as a result of the conclusions reached above, that is what we have done. It seems to us however that given that the tribunal procedure is intended to be flexible and relatively informal, and that there has been no prejudice to the Appellant (otherwise than perhaps in relation to costs), that our decision falls within the jurisdiction granted by section 83(c).

2. Quantifying the taxpayers' right – construing the method

186. During the period to which the college's claim relates, one of two Directives, one of two different Acts each as amended, and one of at least eight different forms of Regulations applied to determine at any time the application of a special method. However at each time:

(i) the relevant Directive provided for a directly enforceable right to recover input tax and permitted special methods;

(ii) the relevant Acts provided for the making of regulations to secure fair and reasonable attribution of input tax; and

(iii) the regulations provided that the Commissioners could permit (using this word for "allow" or "approve") or direct a method (a special method) other than the standard method.

187. The EU right is a right to such recovery as may be permitted by the directive. The proper exercise by a member state of a power given by the directive to use a method different from the standard method may affect the quantification of that right but does not affect its quality.
188. Under domestic legislation the permitting or directing by the Commissioners of a method other than the standard method gives the taxpayer such right to repayment as the method produces. Only if the method fails to give the taxpayer his EU rights – i.e. only if the power given by the Directive is not properly used – can the EU right give an additional right to recovery.
189. Thus in Ampleforth the tribunal rejected what purported to be a special method since it did not address residual VAT, and held that the state could not impose a special method otherwise than in relation to such input VAT. In Sussex, the tribunal held that a special method had been operated but that method had to be construed consistently with the directive. Once so construed the right to additional input VAT recovery arose under that method rather than separately.
190. Although different considerations may arise in relation to a method which has been directed, it seems to us that the proper approach in relation to a method which has been permitted is to seek if at all possible to construe it with the grain of the directive
191. Under the method as operated various areas of activity were tunnelled or sectorised. For those areas other then the three formulaic tunnels the method for recovery of directly attributable input VAT was clearly expressed; the method of recovery of input tax partly attributable was less clearly expressed. At paragraph 73 above we express a view as to how it could have been operated. The tribunal in Sussex took a different view. But under our approach or that of the tribunal in Sussex the method can be construed to give in relation to these areas of activity a right to recovery consistent with the relevant Directive, and so to give a domestic right to a claim for the relevant input tax attributable to the non formulaic tunnels.
192. We have found that the colleges did engage in activities outside the formulaic tunnels. No input tax was recovered in respect of all or many of those activities. The colleges are entitled under the special method to recover that tax. The parties have not asked us to determine the amount repayable. In relation to this aspect therefore the appeal will formally be adjourned pending agreement of the amounts or a fresh application to us to determine them.
193. In relation to areas of activity within the three formulaic tunnels, in our view the method which was used and permitted: (i) was created consistently with the power given in the relevant directive to determine input tax entitlements, (ii) was in no need of further amplification, and, (iii) by virtue of a fairly rough and ready formula created for the taxpayer the EU right which was determined and limited by that formula. Accordingly no further input tax credit is due in respect of these activities.
194. Save in respect of the calculation of the recovery due in respect of the non formulaic tunnelled outputs, these conclusions dispose of the appeal. We have set out our views in relation to Issue 3 above in case the issue re-emerges as the result of a later appeal.
195. Our decision was unanimous.
196. The appeals are now adjourned for the parties to agree the amounts involved or, failing agreement, to bring the calculation back before us.
197. Our conclusions apply both to the appeal of Wadham College and that of Merton College.

7. Costs

198. We were not addressed in detail on the issue of costs. But taking into account the change in the position of the Respondents at the second hearing and the rather late production by them of a relevant document, we award the costs of the second hearing to the Appellants, such costs to be agreed, and failing agreement to be determined by either the Chairman sitting alone, or together with the other member of the tribunal as the tribunal shall direct. We make no award of costs in relation to the first hearing.

CHARLES HELLIER

CHAIRMAN

RELEASED: 2 July 2007

LON 2004/1456