The Commissioners for HM Revenue and Customs v General Motors (UK) Limited: [2015] UKUT 0605 (TCC) [2015] UKUT 0605 (TCC)

UPPER TRIBUNAL

TAX AND CHANCERY CHAMBER

[2015] UKUT 0605 (TCC)

Case No FTC/18/2014

Between

The Commissioners for HM Revenue and CustomsAppellant

General Motors (UK) LimitedRespondent

Before

MR JUSTICE HENDERSON

JUDGE SINFIELD

Mr James Puzey and Mr Jospeh Millington (instructed by the General Counsel and Solicitor for HMRC) for Appellants

Mr Roderick Cordara QC (instructed by KPMG LLP) for Respondent

Date 17 November 2015

Category: Tax

[1]This is an appeal to the Upper Tribunal by the Commissioners for Her Majesty’s Revenue & Customs (“HMRC”) from the decision (“the Decision”) of the First-tier Tribunal (Tax Chamber) (Judge Charles Hellier and Michael James) (“the FTT”) released on 14 August 2013 after a hearing in London spread over 10 days in June, July and October 2012: see [2013] UK FTT 443 (TC). The appeal is brought with permission granted by Judge Hellier on 16 January 2014.

[2]We heard the appeal over three days on 16 to 18 June 2015. The parties were represented by the same legal teams as they had been before the FTT, with Mr James Puzey and Mr Joseph Millington appearing for HMRC and Mr Roderick Cordara QC, instructed by KPMG LLP, appearing for the respondent General Motors (UK) Ltd (“GMUK”).

[3]An appeal to the Upper Tribunal from a decision of the FTT lies only on questions of law: see section 11 of the Tribunals, Courts and Enforcement Act 2007. We emphasise this point at the outset, because Mr Cordara submits that, apart from a self-contained question of law which forms the first of the five grounds of appeal, the remaining grounds represent an impermissible attempt by HMRC to find an error of law in the approach of the FTT to, and its evaluation of, a very complex question of fact involving much narrative and technical evidence. HMRC maintain, for their part, that the grounds of appeal are focused on narrow and discrete issues, and are not an invitation to reopen all of the many complex factual issues covered by the evidence and submissions at the hearing and in the Decision. Background

[4]During the years relevant to this case (1987 to 1996), a car manufacturer in the UK which took a car which it had manufactured into use in its own business was treated for VAT purposes as making a “self supply” of the vehicle. The central issue in the case is for what consideration that deemed supply should be treated as having been made.

[5]GMUK is part of the General Motors Group, which has its headquarters in the USA. GMUK manufactures cars in the UK, including many well-known models such as Vauxhall. It has sister companies in Germany, France, Spain and other European countries. During the relevant period, as the FTT record in the first paragraph of the Decision:

“[GMUK] sold and manufactured cars and car parts in the UK, and imported them from, and exported them to, sister companies in Europe. In the UK it sold its cars to individual 3 retail purchasers and to corporate entities through a network of dealers. It also used some of the cars it produced or imported to provide cars for its own staff and business.”

[6]The cars which GMUK took into its own use in its business were used as demonstrator cars, press cars, pool cars and cars for GMUK’s staff. After they had been so used for a relatively short period (sometimes less than 6 months, and usually for not much longer than a year) they were sold as second-hand cars. The number of cars that GMUK took into its own use varied between 13,000 and 20,000 in each year. These cars were of a higher than average specification, because many of them were supplied to senior GMUK employees, whose preference was for higher specification vehicles. The FTT found that, for different reasons, cars used as demonstration and pool vehicles may well also have been of a higher specification.

[7]The cars which GMUK imported from its sister companies included models which it manufactured itself in the UK. Although there is no express finding to this effect, it is implicit in the Decision, and we do not understand it to be disputed, that the imported models were in all material respects identical to their UK-manufactured counterparts. The only difference between them lay in their place of manufacture.

[8]By virtue of Article 4 of the Value Added Tax (Cars) Order 1980 (SI 1980 No. 482) and its successor, Article 7(1) of the Value Added Tax (Input Tax) Order 1992 (SI 1992 No. 3222), cars were subject to a “blocking order” which prevented the recovery of any input tax by a business purchaser of the vehicle on its supply in the UK or its importation into the UK. There were certain exceptions to this treatment, none of which is material. The underlying reason for the block, which was authorised by EU law, was the practical difficulty of ensuring that cars purchased for business use were in fact exclusively so used: see Chalke v Revenue and Customs Commissioners

[2009] EWHC 952 (Ch)

[2009] STC 2027

, at [12] to [16].

[9]If nothing further were done, this treatment would have discriminated in favour of car manufacturers which took their own cars into business use, because in such circumstances there would have been nothing to prevent recovery by the manufacturer of the input tax referable to the production of the car. In order to maintain the principle of fiscal neutrality, Article 5 of the 1980 Order, and its successor Article 5 of the Value Added Tax (Cars) Order 1992 (SI 1992 No. 3122), provided for a deemed self-supply when a car manufacturer took a car it had manufactured into use in its business. The car was then to be treated “as both supplied to him for the purposes of that business and supplied by him in the course or furtherance of that business”. The result of the self-supply was thus to generate input tax on the deemed acquisition of the car by the manufacturer, which (by reason of the blocking order) the manufacturer could not then set against the VAT due on the deemed supply of the car by him in the course of his business. In this way the tax treatment of the transaction was made equivalent to that of any trader purchasing a car for use in his business. 4

[10]As a matter of EU law, the amount of the consideration for the deemed selfsupply was specified by Articles 5(7)(a) and 11A(1) of the Sixth VAT Directive (Directive 77/388/EEC). Article 5(7) provided that:

“Member States may treat as supplies made for consideration: (a) the application by a taxable person for the purposes of his business of goods produced, constructed, extracted, processed, purchased or imported in the course of such business, where the Value Added Tax on such goods, had they been acquired from another taxable person, would not be wholly deductible; …”

[11]Article 11A(1) then provided that:

“The taxable amount shall be: … (b) in respect of supplies referred to in Article 5(6) and (7), the purchase price of the goods or of similar goods or, in the absence of a purchase price, the cost price, determined as [at] the time of supply; …”

As the FTT noted in the Decision at [15], the words “determined as [at] the time of supply” apply to both purchase price and cost price; and cost price is to be resorted to only if there is no purchase price, even for similar goods. That is the force of the words “in the absence of a purchase price”. The omission of the word “at” in the phrase “determined as [at] the time of supply” appears to be an error, and nobody has suggested that anything turns on it.

[12]In the first part of the period of GMUK’s claim (“the Claim Period”), from 1987 until 31 July 1992, the amount of the notional consideration on a selfsupply of cars under domestic law was fixed by paragraph 7 of Schedule 4 to the Value Added Tax Act 1983 (“VATA 1983”) as “the cost of the goods to the person making the supply”, subject to an immaterial exception. It is clear that this provision did not correctly transpose Article 11A(1)(b) of the Sixth Directive, under which cost was the default measure in the absence of a purchase price for the goods or similar goods. This deficiency was not rectified, however, until 1992, when with effect from 1 August of that year paragraph 7 of Schedule 4 to VATA 1983 was amended so as to provide that:

“(2) The value of the supply shall be taken to be – (a) such consideration in money as would be payable by the person making the supply if he were, at the time of the supply, to purchase 5 goods identical in every respect (including age and condition) to the goods concerned; or (b) where the value cannot be ascertained in accordance with paragraph (a) above, such consideration in money as would be payable by that person if he were, at that time, to purchase goods similar to, and of the same age and condition as, the goods concerned; or (c) where the value can be ascertained in accordance with neither paragraph (a) nor (b) above, the cost of producing the goods concerned if they were produced at that time.”

[13]These provisions were then consolidated as paragraph 6(2) of Schedule 6 to the Value Added Tax Act 1994 (“VATA 1994”), and remained in force until the end of the Claim Period.

[14]To avoid confusion, we should point out that in paragraph [33] of the Decision the FTT appear to have overlooked the fact that the provisions in paragraph 6 of Schedule 6 to VATA 1994 were first enacted in 1992, and that the first part of the Claim Period therefore came to an end on 31 July 1992, not on 1 September 1994 when VATA 1994 came into force.

[15]In paragraph [35], the FTT expressed the view that the relevant provisions of Schedule 6 correctly implemented Article 11A(1)(b) of the Sixth Directive. In his oral submissions to us, Mr Cordara made a number of criticisms of the wording of Schedule 6, including in particular the fact that it introduces a hierarchy between the purchase price of identical goods and the purchase price of similar goods, with recourse to be had to the latter only if the former cannot be ascertained. He submits that there is no warrant for this distinction in Article 11A(1)(b), where the words “the purchase price of the goods or of similar goods” are true alternatives, in contrast to the cost price which only applies “in the absence of a purchase price”. We will return to this point when considering the first ground of appeal, but we note at this stage that there is no response to HMRC’s notice of appeal from GMUK challenging the correctness of the FTT’s conclusion in paragraph [35] of the Decision.

[16]The FTT also held, and the parties are in agreement, that during the first part of the Claim Period the provisions of Article 11A(1)(b) had direct effect, with the result that GMUK had the choice of relying either on them or on the cost measure of the deemed consideration contained in paragraph 7 of Schedule 4 to VATA 1983: see the Decision at [46].

[17]The cost measure was itself first introduced in 1978. Before then, the domestic provisions required (or at least were interpreted by HMRC as requiring) the use of open market value: see the Decision at [32], but note that here (as elsewhere) paragraph 7 of Schedule 4 to VATA 1983 is mis-described as paragraph 2(7). The change in 1978 to cost as the sole criterion for 6 calculating the notional consideration on self-supplies was discussed in advance between HMRC and the Society of Motor Manufacturers and Traders (“the SMMT”). As the FTT record, at [75]:

“Correspondence ensued about the practicalities of determining “cost”

. Reports were received by HMRC from manufacturers including GMUK of the percentage which their calculation of costs represented of the retail price of their cars. GMUK provided a weighted average cost as a percentage of list price of 66.23%. HMRC accepted the practical and theoretical difficulties involved in the determination of cost and wrote to SMMT on 1 February 1978 saying that the use of the retail list price less 33⅓% would be acceptable as the basis of tax on self supplies of cars by volume manufacturers.”

[18]This proposal was then accepted by GMUK, in common with other volume manufacturers, and throughout the period from 1978 to 1996 GMUK used two thirds of retail list price as a proxy for the cost of all cars, whether imported or manufactured in the UK, which were taken into its own use.

[19]The rest of the relevant procedural history is conveniently summarised by the FTT in paragraphs [78] to [85] of the Decision, which we reproduce: “78. Until the decision of the ECJ in 1997 in the case of Commission v Italy C-45/95

[1997] STC 1062

(the Italian Republic case), it had been assumed by the UK legislature (and accepted by GMUK) that GMUK’s sales of the “second-hand” cars were subject to VAT. Article 6 of the 1980 Cars Order and Article 7(4) of the 1992 Input Tax Order provided that the VAT should be charged on the excess of the consideration received on the sale of the second-hand car over the value of the self supply.

[79]Thus until Italian Republic if a car had a list price of £100, GMUK says it accounted for VAT on £66⅔ when it took [it] into its own use, and when it then sold it “second-hand” for, say, £80, it would account for tax on a further £13⅔. GMUK say that it was rare for a car to be sold after its own use of it at less than the equivalent of £66⅔.

[80]After Italian Republic it was recognised the sale of such a second-hand car was in fact exempt from VAT as a result of Article 13B(c) of the Directive.

[81]This case sparked a claim by GMUK for repayment of the VAT it had paid on the sale of the second-hand cars. The claim was made in 2005/6 and, after verification, was agreed by HMRC and paid in 2007. The making of the claim involved the presentation of schedules estimating the numbers of cars which had been sold second-hand in each year from 1973 until 7 the formal recognition of the exemption for the second-hand car sales.

[82]GMUK say that before the settling of the Italian Republic claim it did not matter much what amount was used for the value of the self supply. If the amount under the relevant legislation properly construed was VAT on £50 rather than on £66⅔, then the difference would have been picked up when the car was sold. The only effect was that some of the VAT would have been payable earlier: there would be a continuing cash flow disadvantage. But when the Italian Republic claim was agreed it became clear to GMUK that the VAT on any excess of the 2/3 proxy figure over the proper figure had been a real cost.

[83]On 30 March 2009, the day before the deadline imposed by Finance Act 2008 for the making of claims more than three years old, GMUK wrote to HMRC making a claim for the repayment of overpaid VAT in relation to self supplies of cars in the period from 1 January 1978 to 31 October 1996. The claim was made on the basis that the self supply charge should have been calculated by reference to the cost of the cars used and that that cost was less than the 2/3 proxy. The letter explained that there were no longer records of actual costs and set out a method for estimating actual cost on the basis of the information available to GMUK.

[84]HMRC rejected this claim and GMUK appealed. Following the making of the appeal there were further exchanges of information and discussions between HMRC and GMUK. GMUK amended the basis of its claim. By the time of the hearing GMUK had limited its claim to the years 1987 to 1996. …

[85]HMRC accept that the amendments to the claim did not constitute fresh claims which would be out of time. This decision therefore relates to GMUK’s claim that, between 1987 and 1996, VAT was overpaid in relation to the self supply of cars.”

[20]We would add that GMUK’s claim was made under section 80 of VATA 1994, which provides (so far as material):

“(1) Where a person has (whether before or after the commencement of this Act) paid an amount to the Commissioners by way of VAT which was not VAT due to them, they shall be liable to repay the amount to him. (2) The Commissioners shall only be liable to repay an amount under this section on a claim being made for the purpose. … 8 (6) A claim under this section shall be made in such form and manner and shall be supported by such documentary evidence as the Commissioners prescribe by regulations …”

[21]Regulation 37 of the Value Added Tax Regulations 1995 (SI 1995 No. 2518) provides that:

“Any claim under section 80 of the Act shall be made in writing to the Commissioners and shall, by reference to such documentary evidence as is in the possession of the claimant, state the amount of the claim and the method by which that amount was calculated.”

The Decision: an overview

[22]The Decision is long and detailed. It runs to some 70 pages and 384 paragraphs. It is clearly the product of much labour and careful analysis by the FTT. What follows is no more than a bare summary intended to place the grounds of appeal in context.

[23]The final version of GMUK’s claim was set out in calculations sent by GMUK’s representatives to HMRC on 11 May 2012 and amended grounds of appeal. The claim was not only split chronologically, by reference to the two parts of the Claim Period which we have mentioned, but a distinction was also drawn throughout between cars manufactured in the UK and those which were imported. Different cost price ratios were relied upon for the vehicles in each category.

[24]In respect of the later part of the Claim Period (which, as we have explained, in fact began on 1 August 1992, not on 1 September 1994), the FTT held that the VAT due on the self-supplies was to be calculated according to the purchase price of the vehicles. In the case of cars manufactured in the UK during this period, the FTT rejected HMRC’s submissions that cost was the only available criterion because GMUK did not in fact purchase such cars, or similar cars, from anybody: it was a manufacturer, not a purchaser. The FTT said that this was wrong, and what the legislation contemplated was a notional purchase by GMUK of the cars from a third party such as a dealer.

[25]Accordingly, the FTT concluded at paragraphs [41] to [42]: “41. Thus we consider that merely because the taxpayer is a manufacturer of the relevant goods, it cannot be said that there is not a purchase price. We have accepted that for some objects, perhaps those made in the course of the manufacture of others, it will not be possible to purchase them or something like them because they are never sold: in that case cost will rule. But where the goods in question are produced to be sold and 9 routinely purchased and sold, the legislation intends that there will be a price at which they may notionally be bought.

[42]Thus in our view in this period …, the question for us is [for] what price GMUK could have purchased the relevant cars from a third party. If that price is less than that on which GMUK accounted for VAT, it will have overpaid VAT. We shall address the circumstances (the number of units and the conditions attached) of that notional purchase later.”

[26]In the case of imported cars during the later period, there was an actual purchase price because a transfer price was always payable by GMUK to the sister company which had manufactured the cars. The FTT described the relevant transfer pricing arrangements, and after a detailed review of the evidence concluded at [227] that for the years 1993 to 1996 the ratio of the list price of imported cars to their cost price (i.e. the transfer price paid to the sister company) was no greater than 61%, although they were unable to say that it was less than that amount.

[27]Pausing at this point, there is no appeal by HMRC against any of the conclusions of law or fact which we have so far mentioned. The FTT’s conclusion in relation to imported cars during the later part of the Claim Period meant that this part of GMUK’s claim succeeded, because GMUK had accounted for VAT on the imported cars on the basis of the proxy consideration of two thirds of the UK list price. A substantial sum has therefore been repaid by HMRC to GMUK in respect of this part of the claim.

[28]We revert to the question which the FTT posed at [42], quoted above: namely, for what price could GMUK have purchased the UK-manufactured cars from a third party? The FTT considered this question, by reference to the whole of the Claim Period, at [109] to [119], concluding that overall it had not been proved that GMUK would have obtained a total aggregate rebate and discount of more than 33.33%. In other words, it had not been proved that the notional price would have been less than the proxy price (two thirds of list price) on the basis of which VAT had been paid. The FTT added, at [119]:

“In relation to cars which were or could be imported by GMUK from sister companies different considerations may apply.”

[29]The FTT addressed those “different considerations” at [131] to [133], under the heading “Purchase Price: Imported or Importable Cars”. They concluded that the purchase price for cars which were imported, or could be imported, subject to the relevant transfer pricing arrangements did not exceed the price agreed under those arrangements. Accordingly, for such cars the transfer price “must be the maximum purchase price”.

[30]There is no appeal by GMUK from the FTT’s conclusion that the notional purchase price of UK-manufactured cars throughout the Claim Period had not 10 been shown to be less than two thirds of list price. If matters stopped there, it would seem to follow that GMUK’s claim in respect of such cars during the later part of the Claim Period must fail. On the basis of the FTT’s findings, there was a purchase price for such cars, but it was no lower than the proxy price. Cost would not be an available alternative from August 1992 onwards, although it would be for the earlier part of the Claim Period when paragraph 7 of Schedule 4 to VATA 1983 was in force.

[31]The FTT clearly considered, however, that their conclusion on the notional purchase price of UK-manufactured cars was not the end of the matter in relation to the later part of the Claim Period. When they came to summarise their conclusions in relation to the later part of the Claim Period at [374] and following, they said at [375], under the heading “Purchase Price”:

“375. This means … the price at (sic) which someone in the Appellant’s position would have paid for the cars had it bought them at the time of their appropriation: (1) …; (2) in the case of UK assembled cars that means the list price less the discount and rebates the appellant would have got as a bulk purchaser in its bargaining position. We found that it was not shown that that would be less than 2/3 of the list price ([118]); (3) where a car could be purchased from a sister company for a price less than that determined under (2), that import price would be the purchase price.”

[32]It can be seen, therefore, that the FTT regarded the import price of UKmanufactured cars, where they could be purchased from a sister company, as an available alternative measure of the purchase price, with the consequence that GMUK’s claim would succeed in respect of such cars if it could show that the import price was less than two thirds of the UK list price. In view of the FTT’s unappealed findings about the import price of cars during the later part of the Claim Period, the result of applying this alternative measure of price would be that GMUK’s claim succeeded, during the later period, in respect of all its UK-manufactured models which could also be purchased from a sister company.

[33]It is this part of the FTT’s reasoning and conclusions which HMRC challenge by their first ground of appeal.

[34]We now turn to the earlier part of the Claim Period, from 1987 until 1992. The issues were essentially the same as for the later period, but with the important difference that cost was agreed to be an available alternative measure of the deemed consideration for the self-supplies. 11

[35]As to the primary measure of price, we have already recorded the FTT’s unchallenged finding that for UK-manufactured cars it had not been shown to be lower than the proxy of two thirds of list price.

[36]In relation to imported cars, the FTT were also not satisfied that the transfer prices paid during the earlier period were less than two thirds of the UK list price, although they had been so satisfied in relation to the later period: see the decision at [227] to [228]. There is no appeal against this conclusion.

[37]It follows that GMUK’s claim in respect of the earlier period could only succeed on the alternative basis of cost. Moreover, it is now clear that it could only succeed in relation to UK-manufactured cars, because the FTT’s conclusion in relation to the price of imported cars extended to the cost of imported cars, as they made clear at [380(3)] where they said:

“For the period 1987-1992 we are not persuaded that purchase price or cost of an imported car was less than 2/3 of list price …”

Accordingly, the relevant question was whether GMUK could show that the cost of its UK-manufactured cars between the years 1987 and 1992 was less than two thirds of their list price. It is this part of the claim to which the remaining four grounds of appeal relate. Before we come on to those grounds, however, we propose to deal first with the self-contained first ground. Ground 1

[38]As we have explained, the issue is whether it was open to GMUK to rely on the price of imported vehicles as an alternative measure of the price of UKmanufactured cars in the later part of the Claim Period.

[39]The FTT did not address this issue directly in the Decision, and the conclusion which they stated in paragraph [375(3)] may perhaps have been something of an afterthought. In granting permission to appeal on ground 1, Judge Hellier said (in paragraph 8 of the permission decision) that the relevant reasoning of the FTT was to be found in [28] to [30] of the Decision, “in which the tribunal found that the purchase price is the price at which similar goods could be purchased from a third party”, and in [131] to [133] under the heading “Imported and Importable Cars”. This may suggest that Judge Hellier regarded the imported versions of cars manufactured in the UK as “similar goods” rather than the same goods, but we doubt whether this was his intention because in paragraph 10 of the permission decision he said he was satisfied:

“that HMRC had an arguable case that the tribunal had erred in law in its conclusion that the purchase price for the purposes of the Directive was the lower of the price at which the cars could 12 have been obtained on a notional purchase in the UK and the price for which they could have been obtained on import.”

This formulation indicates that he regarded the imported cars as being the same as their UK-manufactured counterparts, so we suspect that the reference to “similar goods” in paragraph 8 was a slip for “the same or similar goods”.

[40]Be that as it may, we see no reason to doubt that the imported cars should be regarded as the same goods as the corresponding models manufactured in the UK. They were intrinsically identical, and the only difference lay in their place of manufacture. Nor is it a material distinction, in our view, that GMUK distinguished between UK-manufactured and imported vehicles for the purposes of its claim. This was a choice made for convenience in the presentation of a complex claim. It cannot preclude GMUK from arguing, as a matter of law, that the price of identical imported cars may be taken as a relevant measure of “the purchase price of the goods” within the meaning of Article 11A(1)(b).

[41]HMRC argue that it was not open to GMUK to rely on the price of imported vehicles, because the test posited by the UK legislation after 1 August 1992 is a notional purchase of “goods identical in every respect (including age and condition) to the goods concerned”. They submit that a car imported from a sister company cannot be regarded as identical in every respect to a car of the same model manufactured in the UK. More generally, HMRC submit that imported vehicles “were not the vehicles in respect of which that element of the claim is made”.

[42]We do not agree. As to the first objection, we do not consider the place of manufacture to be a relevant attribute of the goods in considering whether they are identical. In our judgment the focus is on the physical attributes of the goods, including their age and condition. We also note that Article 11A(1)(b) refers only to “the goods”, and it may be questioned whether the requirement of identity in every respect goes further than is strictly necessary to transpose the relevant test into UK law.

[43]As to the second objection, it is true that imported vehicles were not the vehicles in respect of which this part of the claim was made, but that does not answer the question whether the price of imported vehicles may be taken as an appropriate measure of the purchase price of UK-manufactured vehicles.

[44]For these short reasons, we consider that the FTT’s conclusion on this issue was correct. We can find nothing in the relevant post-1992 legislation which prevents the purchase price of identical imported goods from being used as a measure of the consideration for the self-supply of the UK-manufactured vehicles. We also think the FTT were right, by implication if not expressly, to regard the imported cars as identical in all material respects to their UKmanufactured counterparts. 13

[45]This conclusion makes it unnecessary for us to consider GMUK’s fallback argument that the imported cars were in any event “similar goods”, and Article 11A(1)(b) on its true construction allows the purchase price of similar goods to be taken as an alternative measure to the purchase price of the original goods. This argument involves the contention that Article 11A(1)(b) has been incorrectly transposed in the post-1992 UK legislation. We doubt whether this contention is properly open to GMUK, given the absence of any challenge by GMUK in its response to HMRC’s notice of appeal to the FTT’s conclusion that the Article was correctly implemented in the post-1992 legislation. More importantly, we prefer not to express a view on a question of EU law which may not be acte clair in a case where it is unnecessary to do so.

[46]It follows that HMRC’s appeal on ground 1 will be dismissed. Grounds 2 to 5: Introduction

[47]Before we consider the remaining grounds of appeal individually, it is convenient to say some more by way of background to the task which the FTT faced in trying to ascertain the cost of cars manufactured by GMUK. The FTT considered this question throughout the duration of the Claim Period. Although the criterion of the cost of the goods was directly relevant only to the years 1987 to 1992, there were at least two reasons why they looked at the question throughout the Claim Period. First, the methodology for ascertaining cost advanced by GMUK necessarily involved the whole of the period. Secondly, it was material to consider the position in the years from 1992 onwards in case the FTT were wrong in their conclusion that price could be taken as the measure of the consideration for the self-supplies in those years.

[48]An important initial issue is what the concepts of “cost price” in Article 11A(1)(b), and of “cost” in the domestic legislation, mean and include. The FTT discussed this question at [50] to [59]. In [52], they said they would use the expressions:(a) “variable costs” to mean the specific extra cost of producing one extra car;

(b) “manufacturing cost” to mean variable cost plus an attributable portion of the fixed costs directly connected with manufacture;

(c) “operating cost” to mean manufacturing cost plus the costs of design and development together with their own fixed costs, plus marketing and selling costs; and

(d) “total cost” to mean operating cost plus costs such as those of long term finance costs and superior (strategic) management overheads. 14

[49]After referring to the decision of the ECJ in

Case C-72/05

, Wollny v Finanzamt Landshut,

[2008] STC 1618

, the FTT held at [54] that “cost price” must be an autonomous EU concept (as the ECJ had held “full cost” to be in the context of Article 11A(1)(c) in Wollny), although there was no case in which the Court had given guidance on its meaning. The FTT then concluded in [56] that “cost price” should be given “a meaning which includes all the expense of the business attributable to bringing the product to its condition and location at the time of appropriation”. Accordingly, it should include:

“(1) The costs of the purchase and transport of materials (2) The direct expenses which are attributable to the production including direct labour costs and subcontract costs, (3) Overheads, labour and services for the production of the goods including the depreciation of assets used in production, (4) Other overhead costs attributable to the production including those relating to the design of the product and the means of production. ”

[50]On the other hand, the FTT considered it to be inherent in the idea of cost that expenses incurred in the marketing of the product should not be taken into account: see [58]. Finally, they held that the same principles should apply in construing “cost” in the domestic legislation as “cost price” in the Sixth Directive, given the presumed intention of the domestic legislation at least in this respect to implement the directive.

[51]There is no challenge to the FTT’s conclusions on any of these matters.

[52]In seeking to establish the cost of the cars which it manufactured during the Claim Period, GMUK faced the obvious difficulty that it no longer held any accounting records for that period. It was company policy to retain records for ten years and then to destroy them. As the FTT said, at [88]:

“Although we had published accounts for the period, figures for cars sold by model, and there were one or two other bits of paper before us which had slipped through this net, we had no contemporaneous records of costs.”

[53]What GMUK did have was its so-called “FIN 51” accounting material for the years from 1998 onwards. The FTT recorded at [198] that the FIN 51 material was compiled from inputs from the whole of GM’s European operations, and included figures for all the costs incurred by that consolidated operation under various headings such as the costs of materials, labour costs and variable costs. The FIN 51 data were prepared for each variant of each model, and represented a snapshot at the time of their preparation. A fuller description of the FIN 51 reports, which were produced every six months and were an important management tool in the group, may be found at [240] to [242]. 15

[54]In a lengthy section of the Decision running from [230] to [372], the FTT dealt with the cost of UK-manufactured vehicles as a percentage of list price. They called this ratio “C %”. The witnesses whose evidence was of particular importance in relation to this issue were (for GMUK) Mr John Fulcher and Mr Bill Robinson, and (for HMRC) Dr Matthias Holweg.

[55]Mr Fulcher was the chief financial officer of GMUK. He had worked for General Motors since 1974, and in 1993 he had been responsible for collating the group’s FIN 51 material. Between 1994 and 1996 he had worked on transfer pricing in Europe, followed by a year in Ireland as chief financial officer before returning to Luton as financial controller. Mr Fulcher prepared calculations of the costs of the self-supplied cars which formed the basis for GMUK’s claim.

[56]Mr Robinson is head of economics at KPMG. He has extensive experience in economic forecasting. He provided a model based on macro-economic indicators which estimated how the cost/price ratio of GMUK’s cars had moved between 1978 and 2003, linking his model to the results for cost and list price obtained by Mr Fulcher from the FIN 51 data for the period 1998 to 2003.

[57]It should be noted that Mr Robinson’s evidence was not tendered as expert evidence, but as evidence of fact, albeit of a highly technical nature.

[58]Dr Holweg is a Reader in operations management and director of research at the Judge Business School in Cambridge. He has considerable expertise in the economics of the car manufacturing industry. Although the FTT record that he gave his opinions on the method used by Mr Fulcher and the economic model used by Mr Robinson, he too was not tendered as an expert witness.

[59]In [231], the FTT outlined the methodology which Mr Robinson had used to estimate “C %” throughout the Claim Period:

“In outline, the appellant, through Mr Robinson, estimated C for each year in the Claim Period in the following way: (1) for 1978 the figure of 66.66% had been provided by GMUK as part of the discussion between HMRC and SMMT …; (2) for each of the years 1998 to 2003 the appellant estimated the cost/price percentage for UK assembled cars by taking data from its FIN 51 accounting system; (3) using indices of material and labour costs produced by ONS [the Office for National Statistics] and adjustments to reflect the change in car specifications during the period Mr Robinson estimated C for each year starting with the 1978 figure and finishing with 2003. 16 (4) He then compared the results of the extrapolation with the FIN 51 results for the period 1998 to 2003. (5) He concluded that the “fit” of the extrapolation from 1978 to the period of the FIN 51 results for 1998 to 2003 validated the extrapolation method so that the percentages produced by the extrapolation from 1978 could be used as C% in the calculation of any overpaid VAT.”

[60]The FTT then said that this outline “hides the iterations” of GMUK’s claim. Initially, the claim was made on the assumption that the cost/price ratio for the whole Claim Period was, for domestically produced cars, the average of the FIN 51 results for 1998 to 2003. The claim was then refined, with Mr Robinson’s advice, by using general cost indices to extrapolate the 1978 fraction forwards to 2003. Dr Holweg questioned the indices used, so Mr Robinson recalculated the extrapolation using where possible indices specifically linked to car production. This showed little variation from the earlier figures, and produced what the FTT called “the “green line” extrapolation”.

[61]The FTT continued: “233. That extrapolation however resulted in lower percentage cost/price ratios for 1998 to 2003 than that obtained from the FIN 51 results. Mr Robinson considered that the shortfall was likely to be the result of increases in the specification of vehicles over the period and by changes in the efficiency of production. He estimated [the] effect of these changes as an annual percentage (the “Q Factor”) which he applied to his extrapolation for C. The result was an extrapolation (which we shall call the “red line”), which for 1998 to 2003 was a close fit to the FIN 51 results (in particular to their variation) in that period. (Because the Q Factor was calculated as the amount necessary to bring the extrapolation up to the FIN 51 figures it is not significant that the red line is on average close to the FIN 51 results for that period: for the purposes of assessing the accuracy of the method what matters is the fit of the red line to the variation in those figures).

[234]In his first witness statement Dr Holweg provided a source of data from which the increased cost in real terms of increased specification of cars over the period 1967 to 2006 could be estimated. Mr Robinson used this data to adjust to the Green line extrapolation to produce what we shall call the “blue line” extrapolation. In this extrapolation the Q factor was replaced by Dr Holweg’s “Ward” data source. This extrapolation resulted in a fairly close fit of the blue line to the FIN 51 data for 1998 to 2003 both in absolute values and in variation in that period. 17

[235]In the remainder of this decision we proceed on the basis that the appellant’s case is that the values of C in the Claim Period are likely to have been the blue or the red line figures for those years. Whatever else may be the case, it seems to us that the variations in cost/list price produced by these means make it unlikely that a fixed cost/list price ratio obtained in the Claim Period.

[236]These estimates and procedures give rise to a number of issues:(1) To what extent [do] the FIN 51 figures give rise to proper estimates of the value of C between 1998 and 2003?

(2) How accurate was the 1978 starting point of a 2/3 ratio?

(3) How reliable – what uncertainties are there and how large are they – is the extrapolation in the blue or red line?

[237]We deal with each of these issues in turn in the following sections …”

[62]The FTT dealt with the first of the questions they had identified at [238] to [290]. We do not understand there to be any challenge to this part of the Decision.

[63]The FTT then considered the second question at [291] to [303]. Again, there is no challenge to this part of the Decision. The FTT concluded, broadly speaking, that the 2/3 starting point in 1978 was accurate, and included substantially the same expenses as GMUK’s calculations. The FTT also pointed out, at [292], that it would in principle be possible to “consider the extrapolation as working backwards from known data of 1998 to 2003 rather than forwards from 1978”.

[64]The FTT then considered the third question, under the heading “The extrapolation of the blue line and the red line”, at [304] to [372]. This is the part of the discussion on which the remaining grounds of appeal are focused. We will refer to the reasoning of the FTT, to the extent we consider it necessary to do so, when considering those grounds. At this stage, it is enough to say that the FTT were clearly satisfied that Mr Robinson’s methodology, in its final iteration, had sufficient probative force to ground the conclusion, on the balance of probabilities, that throughout the Claim Period C % was likely to have been less than two thirds. On the other hand, they did not agree with all of Mr Robinson’s calculations, and accepted some of the criticisms made of them by Dr Holweg. With the benefit of all the evidence they had heard, the FTT then devised a solution which differed in material respects from that advanced by either side, but they did not attempt a precise calculation of the amount of VAT which had been overpaid. Instead, they adjourned the appeal in the expectation that the parties could agree the figures on the basis of the principles which they had laid down. 18

[65]The FTT stated the result of the appeal in [382], as follows:

“Thus the Appellant succeeds in the appeal to the extent that the VAT which it would have paid had it accounted for VAT on cost or purchase price as determined above [is less than] the VAT on 2/3 of the list price.”

By an unfortunate slip, the FTT said “exceeds” rather than “is less than”, but the intended sense is clear, and both sides agreed with us that this is what the FTT meant to say. Ground 2

[66]Ground 2 is expansively set out in paragraphs 18 to 32 of the Grounds of Appeal. In our judgment it would have benefited from a conciser formulation. In general terms, however, it alleges an error of law in the FTT’s approach to determining cost price. We find it helpful to read these paragraphs of the Grounds of Appeal with the conclusion in paragraph 46, which encapsulates many of HMRC’s submissions on this part of the case:

“This was a complex and difficult matter for the Tribunal to determine; however, its approach in certain important aspects gave rise to errors of law. Fundamentally, these trace back to the misconception by the Tribunal that its role was to provide a solution rather than determine whether the claim, as advanced by the Appellant, had been proved on the balance of probabilities.”

[67]We begin with the alleged “misconception” by the FTT that its role was to provide a solution rather than to determine whether GMUK’s claim had been proved on the balance of probabilities. In our view there is a false dichotomy in this formulation of the FTT’s role. Of course the FTT had to be satisfied, on the basis of the evidence adduced before it, and to the civil standard of proof on the balance of probabilities, that GMUK had overpaid VAT on the deemed self-supplies of its UK-manufactured cars between 1987 and 1992. As a step in the reasoning which might lead it to that conclusion, the FTT had to consider and evaluate the methodology for ascertaining the ratio of the cost of production of such cars to their list price, throughout the whole of the Claim Period, as advanced and explained by GMUK’s witnesses, including in particular Mr Robinson.

[68]The FTT were not, however, confined to choosing whether to accept or reject Mr Robinson’s model in its entirety. So far as they could properly do so, it was their duty (applying their own expertise as a specialist tribunal) to ascertain the true amount of VAT (if any) which GMUK had overpaid. This result could be achieved either by the FTT performing the appropriate calculations itself, or by stating the principles by reference to which they 19 considered the calculation should be made. In performing this task, the FTT had to act with procedural fairness, and there had to be a proper evidential foundation both for their findings of fact and for their conclusions. But their preferred solution did not have to be one for which either side had specifically contended, either before or in the course of the hearing.

[69]We were referred to a number of authorities which touch on the role of the tribunal when it hears a VAT appeal. In Customs and Excise Commissioners v Pegasus Birds Ltd

[2004] EWCA Civ 1015

[2004] STC 1509

, Carnwath LJ said at [38], giving guidance to the tribunal when facing a challenge to “best of judgment” assessments to VAT:

“The tribunal should remember that its primary task is to find the correct amount of tax, so far as possible on the material properly available to it, the burden resting on the taxpayer.”

To similar effect, Chadwick LJ referred at [92] to:

“the underlying purpose of the legislative provisions – to ensure that the taxable person accounts for the correct amount of tax.”

We agree with Mr Cordara that Carnwath LJ’s formulation of the primary task of the tribunal is not confined to “best of judgment” cases, but is of general application.

[70]We would add that this approach to the tribunal’s task is arguably reflected in rule 2 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009, SI 2009/273, which states that:

“(1) The overriding objective of these Rules is to enable the Tribunal to deal with cases fairly and justly. (2) Dealing with a case fairly and justly includes – … (b) avoiding unnecessary formality and seeking flexibility in the proceedings; … (d) using any special expertise of the Tribunal effectively; … (4) Parties must – (a) help the Tribunal to further the overriding objective; and 20 (b) co-operate with the Tribunal generally.”

[71]This is not to say, however, that the tribunal is bound to find a way, if it can, of upholding a claim, where the quality of the evidence before it is insufficient to justify reaching any conclusion in the claimant’s favour. As Lord Tyre has recently said, in the context of a claim to recover under-claimed input tax between 1974 and 1997, in Lothian NHS Health Board v HMRC

[2015] UKUT 264 (TCC)

at [19]:

“The onus of proving that “an amount” of input tax has been paid and not recovered rests upon the claimant. The standard of proof is the balance of probabilities. At the conclusion of a hearing, it is open to a Tribunal to hold that the claim fails for either of two reasons: (a) because the Tribunal is not satisfied, on the balance of probabilities, that there is any unrecovered input tax; or (b) because the Tribunal, although satisfied that there is unrecovered input tax, is unable to find, on balance of probabilities, that any particular – even a minimum – amount of input tax can be ascertained as having been paid and not recovered. In the latter alternative the Tribunal does not function as a detective with a duty to fix a figure – even a minimum figure – for input tax paid but not recovered, regardless of the quality of the evidence placed before it by the claimant.”

[72]This decision is also of value for the rejection by Lord Tyre of a submission that there is “a different, more relaxed standard of proof for historic VAT claims”. He said at [23]:

“23. I do not consider that such a distinction can or ought to be drawn. In all cases the standard of proof remains the balance of probabilities: that applies equally to historic claims for unrecovered input tax [or, we would add, for overpaid output tax]. There is no rule of law or procedure restricting the exercise of the right of recovery in such cases; proof by means of estimates, assumptions and extrapolations was open to [the Tribunal] as it is in all cases. The problem for the appellant was that the Tribunal was not satisfied that the material placed before it was of sufficient value to enable any reliable conclusions to be drawn, whether by way of estimation, assumption, extrapolation or otherwise.”

[73]We now turn to the passages in the Decision in which the FTT described their approach to their task.

[74]First, the FTT said at [64] that “the only question we have to address is whether the appellant can prove that it is more likely than not that it overpaid VAT”. Rightly, HMRC do not disagree with that formulation, while adding 21 that, in order to reach a conclusion that VAT had been overpaid, the FTT would have to determine that VAT had been overpaid in an identifiable sum.

[75]Secondly, the FTT said in [69]:

“It seems to us that there might in theory be an absolute answer to the question “if GMUK paid too much VAT, how much did it overpay?”, but we are not called upon to answer it; instead we have to answer the question “in the VAT periods in the Claim Period what, by reference to the nature of the claim made under reg 37, is the maximum amount of tax which GMUK is likely to have overpaid”.”

HMRC criticise the words “the nature of” in this passage, arguing that the FTT’s task was to adjudicate upon the claim as presented, and not, as it turned out, to substitute a different method of calculation. As we have already said, we do not think this is a fair criticism. The task of the FTT was to ascertain the correct amount of VAT overpaid, as far as it could properly do so on the material before it.

[76]HMRC also criticise the reference to “the maximum amount of tax which GMUK is likely to have overpaid”, saying that this would simply lead to continued dispute as to the correct amount of tax, as opposed to the maximum. Again, we do not think there is anything in this point. The FTT’s language may be slightly clumsy, but in our view they were clearly referring to the maximum overpayment that GMUK could satisfy them on the balance of probabilities had been made. In other words, they were referring to the upper limit of the claim which GMUK could make good on the evidence.

[77]Thirdly, the FTT went on to say in [72]:

“It seems to us that in determining such an appeal the jurisdiction of the tribunal is not limited to allowing or dismissing the appeal. It has a duty to determine the amount of the claim. Otherwise for example small inaccuracies in the claim could make an otherwise good claim ineffective and the right of appeal would be all but illusory; indeed a taxpayer would not have an effective route to obtaining those rights afforded to him by EU laws in relation to the repayment of overpaid output VAT.”

We cannot see anything objectionable in this paragraph.

[78]Finally, the FTT concluded the discussion of their task in [74], saying:

“It seems to us that we cannot conclude that tax was overpaid without concluding that it was likely that at least a particular 22 amount was overpaid. A conclusion on the evidence that it is likely that at least £X (where X is greater than 0) was overpaid means that the appellant’s claim would succeed as to £X. But it is a conclusion which shirks the Tribunal’s duty to settle the appeal, because it leaves open the question of whether it is likely that more than £X was overpaid. Our duty must be to determine, or at least, having found all relevant facts on the evidence before us, to set out the principles for determining, how much tax (if any) was likely to have been overpaid.”

[79]Again, we cannot find any misdirection of law in this paragraph. HMRC complain that the FTT’s reference to “setting out the principles” was wrongly construed by them “to permit a new method of calculation to be adopted … without reference to the parties and one which was in contradiction of agreed expert evidence and the case for both parties”. This submission elides a number of different points. We have explained why we do not consider that the FTT was confined to the methods of calculation advanced by the parties. Whether the adoption of a different method was open to the FTT on the evidence, and whether HMRC had been given a fair opportunity to deal with it, are different questions, which are in substance raised by the other grounds of appeal.

[80]It follows that we can find no fault with the FTT’s general approach to their task.

[81]A separate point, which appeared to be implicit in HMRC’s grounds of appeal, was that regulation 37 of the VAT Regulations 1995 (quoted in [21] above) somehow confined GMUK to the methodology of quantification of the claim which had been set out in the original claim under section 80 of VATA 1994. In his oral submissions, Mr Puzey made it clear that this was not HMRC’s contention, and that no objection was made to any of the evolutions in the claim which had taken place before the start of the hearing. We also understood him to agree that the FTT were not necessarily confined to adopting solutions for which the parties had argued, provided always that, if they decided to adopt a different solution, the parties had been given a proper opportunity to comment on it. In these circumstances, we do not need to rehearse the detailed arguments presented to us by Mr Cordara about the purpose and limited scope of regulation 37. It is enough to say that we agree with him that it is a provision of an essentially administrative nature, designed to ensure that, when a section 80 claim is first made, it has sufficient particularity for HMRC to engage with it and decide whether or not to accept it.

[82]Apart from the questions of general approach which we have so far discussed, HMRC also make some specific criticisms under this ground of appeal of the way in which the FTT treated Mr Robinson’s model. In essence, the criticisms were these: 23(1) First, in his written and oral evidence Mr Robinson had repeatedly stressed the importance of his fixed start and end points, in 1978 and 1998 to 2003 respectively. The correlation of his projection forward from 1978 with the FIN 51 data for 1998 to 2003 showed the success of his model, and (in his view) justified its adoption. As a corollary of this, Mr Robinson conceded that, if the FIN 51 data were incorrect, then his calculations would likewise be invalidated.

(2) Secondly, however, the FTT implicitly rejected Mr Robinson’s model and provided their own method of calculation. Instead of starting from 1978, the FTT took the FIN 51 figures for 2003 as their starting point, and then worked backwards to the Claim Period. Not only did this procedure reverse the Robinson model, but it started from a point when the actual cost/price ratio was uncertain, instead of from the agreed position in 1978.

(3) Thirdly, the FTT held that the FIN 51 figures for 1998 to 2002 required significant revision, the details of which are summarised in the Decision at [380(5)]. HMRC do not seek to appeal the findings of fact which led to these adjustments, but point out that they would lead to alterations of the cost price ratio shown on the Robinson model.

(4) The FTT concluded, on the basis of a schedule and graph at pages 954G and 954I of the bundle (to which we will have to return), that the cost price ratio for 2003 produced by Mr Robinson’s model almost coincided with the FIN 51 data point for that year. Accordingly, the FTT approached the model “as if it predicted from the 2003 figures the figures for earlier years”: see the Decision at [358]. However, GMUK had previously excluded the 2003 data from its model calculations as a “rogue observation”. Thus, the FTT adopted as the starting point of its model a year which GMUK had discounted as being a rogue observation; and the FTT did this without reference to either party.

[83]We see little force in these objections, whether they are viewed singly or collectively. As to the FIN 51 data, there is no appeal against any of the adjustments which the FTT made. They were all discussed at the hearing, and were in fact adverse to GMUK. The original FIN 51 data for 2003 were indeed unreliable, but the error had been corrected by the time when the schedule and graph at pages 954G and 954I were submitted on 19 July 2012, some three months before the conclusion of the hearing in late October 2012. The decision of the FTT to work backwards from the 2003 figures, rather than forwards from the 1978 figures, was ultimately a matter of factual evaluation for them, and anyway should in principle have made no difference to the outcome as they acknowledge in [358]. We will return to this objection in the context of ground 5: see [131] to [133] below.

[84]More generally, we consider it an exaggeration to say that these differences in the FTT’s approach from that adopted by Mr Robinson amounted to a 24 substitution by the FTT of its own method of computation. Mr Cordara was able to satisfy us, by taking us in detail through the history of the litigation, that the overall contours of GMUK’s claim had remained essentially unaltered since the methodology was first outlined in a letter from GMUK to HMRC dated 30 March 2009. The basic steps in that methodology were all respected by the FTT in their Decision, and the points on which they differed from Mr Robinson were ones of detail rather than underlying substance.

[85]We find confirmation for this conclusion in Judge Hellier’s own comments on ground 2 when granting permission to appeal. He said, at paragraph 14 of the permission decision:

“I confess that I was surprised by the contention that the method of computation decided upon by the tribunal was regarded as materially different from that discussed in the hearing. The mechanism of the model was to estimate annual percentage change. If to some extent the model was accepted, it worked in the same way whether the changes were applied forwards from a 1978 starting point or backwards from a later one. The tribunal preferred the more recent starting point for the reasons it gave.”

[86]All in all, we remain unpersuaded that any of the matters canvassed by HMRC under ground 2 reveal an error of law on the part of the FTT. Ground 3

[87]We turn to ground 3, at the heart of which lies a procedural objection that the FTT acted unfairly in first admitting, and then relying upon, the schedule and graph at pages 954G and 954I of the bundle. We have already made brief reference to these documents in our discussion of ground 2. As they are central to grounds 3 and 4, we must now describe them, and the circumstances of their production during the hearing, in more detail. We will refer to them as “the Schedule” and “the Graph” respectively, and together as “the Disputed Documents”.

[88]In form, the Schedule was a revised and expanded version of similar schedules which Mr Robinson had produced earlier in the proceedings, and upon which he had been cross-examined. It was first sent to the FTT and HMRC on Friday, 13 July 2012, that is to say between the first and second sittings in the appeal, after Mr Robinson had given his initial evidence, and after he had been recalled to give further evidence by GMUK. The Schedule was accompanied by the Graph, which represented certain data derived from the Schedule in graphical form.

[89]There were at least four major differences between the data contained in the Disputed Documents and the earlier schedules and graphs which Mr Robinson 25 had produced. First, the “rogue” FIN 51 pricing figures for 2003 were corrected. Secondly, entirely new figures were produced for each year on the basis of the number of vehicles produced in each year, alongside the previous figures based on the number of vehicles sold. Mr Robinson had accepted in cross-examination that his calculations ought to have been based on the number of vehicles produced. On the face of it, this deficiency in his calculations was now remedied, but the consequential amendments set out in the Schedule were very extensive: they included two new columns in the calculation section of the Schedule, and two new “Index” sections relating to the Q factor, based on sales and production respectively. Thirdly, the Schedule also included two new “Error” sections, again based on sales and production respectively, for the years from 1997 onwards. Finally, entirely new figures were also produced for the years from 2004 until 2011.

[90]In this last respect, the Disputed Documents responded to questions raised by the FTT during the earlier part of the hearing, which had concluded on 29 June 2012. As KPMG said in their covering letter of 13 July 2012 to the FTT:

“During the earlier hearing between 22 and 29 June 2012 the Tribunal raised a number of questions concerning the possibility of extending Dr Robinson’s model for periods post 2003 in order to assess the correlation between cost price ratio predicted by reference to the movement in the underlying economic indices and FIN 51 data for the same period.”

The letter then pointed out a number of difficulties with this approach, including changes from 2004 in the basis of the FIN 51 data, the fact that the period after 2004 was “highly volatile”, the closure of the Luton manufacturing centre in 2002, cessation of production of the Vectra model in the UK in 2004, and the economic crisis from 2008 onwards.

[91]The letter continued:

“Despite these difficulties and in order to provide an illustration of the correlation which exists, over the last 2 weeks the Appellant has taken the available FIN 51 data (which from 2004 no longer measured operating profit but contribution margin and accordingly manufacturing costs, sales, administration and general costs and tooling were excluded) and added manufacturing costs data in a consistent way with the FIN 51 data from the prior period. We have taken the 6 monthly data in order to provide as many data points as possible. This data has then been plotted against the output of the following exercise undertaken by Dr Robinson’s team:  The ONS series data through to 2011 were obtained.  Average list prices for the Astra were obtained for the years 2003-2011 (note: the pricing 26 information was updated from 2003 rather than 2004 as it has become clear that the pricing data previously used for the year 2003 was erroneous as indicated in Dr Robinson’s witness statement).  The Astra prices were re-based in 2004 …  Sales and production figures for 2005-2011 were obtained from GM’s annual accounts.  The “Q factor” was recalculated taking the average difference between all the available data point[s] (1998-2011). Separate versions of the calculations and graphs have been produced showing the calculation based on either sales or production figures.”

[92]In their covering letter of the same date to HMRC Solicitor’s Office, KPMG said:

“As you will note we have provided the tribunal with the Appellant’s efforts to extend Dr Robinson’s model and the actual information reconstructed from the FIN 51 data for illustrative purposes. What if any use the tribunal wants to make of this exercise can also be debated.”

[93]At the resumption of the hearing on 19 July 2012, GMUK was represented (in Mr Cordara’s absence) by Mrs Amanda Brown, a non-practising solicitor with KPMG. She was the author of the two covering letters of 13 July 2012 from which we have quoted. After making an application to admit the evidence of three further witnesses, Mrs Brown turned to the Disputed Documents. She explained that GMUK’s claim had not changed: it remained exactly as it was when Mr Robinson and the other witnesses were in the witness box. The claim was “pegged to the period 1998 to 2003 and the correlation of the lines there”. She said of the Disputed Documents:

“It is not something on which we rely; it is something we have done.”

[94]After some further exchanges, in which Mrs Brown stressed that the FTT had yet to hear Dr Holweg’s further evidence, she concluded as follows:

“What this material is is in some ways neither here nor there: the appellant does not rely on it, it produced it because the Tribunal was talking about it. So the appellant is entirely in your hands as to what, if anything, you wish to do with that material. 27 So I don’t make an application to admit it. I probably make an application for the Tribunal to tell us what, if anything, it wants to do with it at all … I don’t really know what to say other than that. In some regards, I guess we put out heads in our hands and think: wish we’d never done that now, but we were just trying to be helpful.”

[95]In response, Mr Puzey for HMRC made it clear that Dr Holweg had looked at the new calculations, and could offer a view on them. Nevertheless, Mr Puzey submitted that the FTT should disregard the Disputed Documents for three reasons. First, GMUK effectively accepted that the new figures were unreliable. Secondly, if the FTT were invited to place credence in them, HMRC “have had no opportunity to properly test them, for example, by discovery of the FIN 51 figures”. Thirdly, there was no witness HMRC could ask about the calculations, because they were not accompanied by a witness statement. HMRC’s position was that Mr Robinson had already had two opportunities to give evidence, and he should not have another.

[96]In the discussion with counsel which ensued, Judge Hellier said he would prefer to put off a decision on what to do with the Disputed Documents until after Dr Holweg had given his evidence. In response to a submission by Mr Puzey that it was “not a particularly satisfactory approach to invite the FTT to make of the documents what they would”, and that GMUK could not “throw out material and hope that something catches”, Judge Hellier pointed out, correctly in our view, that:

“The Tribunal is also charged with getting to the right amount of tax, I think, rather than just simply deciding between two parties.”

Mr Puzey accepted this, and he also agreed that a decision on the Disputed Documents should be deferred until Dr Holweg had given evidence, while reserving the right to make further submissions at that stage on whether it would be fair to place any reliance on the new figures at all, given the circumstances in which they had arisen.

[97]On the following day, Dr Holweg returned to give his further evidence and he was cross-examined by Mr Cordara. Mr Cordara questioned Dr Holweg on a number of matters, including the Graph (and its counterpart at page 954H, which was based on figures for sales rather than production). After Mr Cordara had asked a number of questions on the document at 954H, Mr Puzey intervened to object that GMUK were not relying on this material. He said:

“I am very cautious about questions being asked of this witness on the basis that matters are established when there is no evidence of them, and there is to be none, we were told.” 28 Mr Cordara then said he wished to give Dr Holweg an opportunity to comment on the new documents, and was willing to do so “de bene esse”

. Mr Puzey was unhappy with this proposal, on the footing that the documents were either in evidence, or they were not: there was no half way house. Mr Cordara replied that the documents could be put to Dr Holweg on a hypothetical basis, asking for his comments on the assumption that they were correct. Mr Puzey said he would still prefer the documents not to be put at all to the witness, but

Judge Hellier adopted Mr Cordara’s suggestion, saying:

“I think it depends on what benefit we would get out of hearing Dr Holweg’s view upon the difference between the two lines. Lets take it as a theoretical matter.”

The cross-examination then proceeded on that basis, without further objection from Mr Puzey.

[98]Against this background, HMRC now argue that the FTT should never have admitted the Disputed Documents, and they were therefore wrong to place any reliance upon them. Apart from the fact that GMUK itself disclaimed any reliance on the documents, HMRC complain of the fact that the FTT never ruled on their admissibility, nor did they give any indication of their intention to rely on the documents before the Decision was produced. Furthermore, although the Disputed Documents were clearly based on work carried out by Mr Robinson’s team, they were not accompanied by any witness statement explaining what had been done and verifying the new figures. HMRC and Dr Holweg were therefore placed at an unfair disadvantage in having to respond to this new material at short notice, without any certainty whether it was going to influence the thinking of the FTT or not.

[99]In support of these submissions, Mr Puzey referred us to the pertinent observations of Lightman J in Mobile Export 365 Ltd v Revenue and Customs Commissioners

[2007] EWHC 1737 (Ch)

[2007] STC 1794

, when giving guidance to the VAT and Duties Tribunal on the future conduct of a “missing trader” fraud case. He said, at [20]:

“(3) The tribunal cannot (as it has proposed in the decision) decide to admit evidence on the basis that it can later reverse this decision if it considers it just. The tribunal must (at least in any ordinary case such as the present) make a final decision either way. Pending such a final decision, the tribunal may find it necessary to allow evidence to be read and referred to “de bene esse” before finally deciding on its admissibility. The availability of this course does not afford a green light to postponing a final decision on admissibility longer than is necessary.”

[100]According to Mr Puzey, the procedure against which Lightman J warned is essentially what has happened in the present case. The FTT were prepared to admit the Disputed Documents on a provisional basis, and permitted Dr 29 Holweg to be cross-examined upon them on a hypothetical basis, but never gave a ruling on their admissibility.

[101]In answer to these submissions, GMUK denies that the Disputed Documents were never admitted in evidence. On the contrary, says GMUK, they were produced at the request of the FTT, they were put before HMRC, they were debated in the courtroom, and Dr Holweg answered questions on them from both Mr Cordara and the Tribunal. There are no formal rules of evidence in the FTT: see rule 15(2)(a) of the FTT Procedure Rules, which empowers the Tribunal to “admit evidence whether or not the evidence would be admissible in a civil trial in the United Kingdom”. Accordingly, it is said, anything that is put before the FTT is in evidence.

[102]GMUK goes on to submit that the question of reliance on the material is a separate issue. Even if evidence is not relied upon by the party who produces it, another party may rely on it against the producing party. Thus, HMRC now seek to rely on the Disputed Documents in an attempt to discredit GMUK’s case based on Mr Robinson’s model. HMRC are fully entitled to do so. What is not open to them is, at the same time, to say that the documents are not in evidence. If necessary, GMUK submits that there has been a waiver by HMRC of their objection to the admissibility of the Disputed Documents.

[103]GMUK further submits that it is important to identify what was new, and what was not, in the Disputed Documents. They certainly contained new material, but they were also well rooted in what had gone before. To a large extent, the changes introduced were responsive to criticisms made by HMRC, and had the result of reducing GMUK’s claim. This is therefore far from being a case of a new piece of “killer” evidence coming in late. It is, rather, one of adjustments being made to earlier evidence to account for points made by HMRC. True, the Disputed Documents also contained new data for the years 2004 to 2011, but this was done in response to questions raised by the FTT, and the FTT were entitled to make of it what they would. GMUK cannot be criticised for providing material which the FTT had said they might find helpful.

[104]On this last point, we think it is material to bear in mind rule 15(1) of the FTT Procedure Rules, which provides that without prejudice to the general case management powers of the FTT, the Tribunal may give directions as to:

“(a) issues on which it requires evidence or submissions; (b) the nature of the evidence or submissions it requires; … (e) the manner in which any evidence or submissions are to be provided, which may include a direction for them to be given – (i) orally at the hearing; or (ii) by written submissions or witness statement; 30 …”

The FTT did not give any formal directions of this nature, but the existence of these powers serves to make the point that the FTT may take the initiative in requiring evidence on particular matters, and is not necessarily confined to the evidence which the parties have chosen to place before it.

[105]Finally, GMUK submits that Dr Holweg was given his opportunity to deal with the new material. He said whatever he wanted to say. Mr Robinson was not questioned about it, but that is because HMRC’s stance was to object to its admissibility. HMRC never applied for Mr Robinson to be recalled to answer questions about it.

[106]Before we state our conclusions on this ground of appeal, we should note that in the permission decision Judge Hellier described the Schedule as “a schedule of calculation results” and the Graph as “a graph which displayed the results of the figures in column 26 of [the Schedule] against other figures”. He said the genesis of the Schedule was described in the Decision at [320], [324] and [355]. He acknowledged that the Tribunal had relied on the Schedule, as paragraph [355] of the Decision makes clear. He added:

“We did not formally admit it into evidence or formally reject it. We regarded it as an arithmetical submission rather than primary evidence.”

[107]What, then, are we to make of HMRC’s objections to the admission of this material? We have not found this an altogether easy question, which is one reason why we have set out the relevant procedural history at some length. We can understand why HMRC may have been left with a lingering feeling of unfairness, when GMUK had expressly disclaimed any reliance on the new material, HMRC had raised objections to its admission, and the FTT never gave a final ruling on those objections, but nevertheless made substantial use of the new material in reaching its conclusions. If we were satisfied that there was a real procedural injustice, we would see no alternative to remitting the case to the FTT so that it could receive further evidence and submissions directed to the Disputed Documents.

[108]After careful consideration, however, we have come to the conclusion that any unfairness is more apparent than real. With the benefit of hindsight, we think that the procedural history which we have recounted is a little unfortunate in a number of respects. It would have been better if the new material had been accompanied by an explanatory witness statement from Mr Robinson, and if Mrs Brown had made it clear that reliance was being placed upon it at least to the extent that it reflected justified criticisms of Mr Robinson’s model which HMRC had elicited in cross-examination. We also think it would have been better if the FTT had ruled on the admissibility of the material before permitting Dr Holweg to be cross-examined upon it on a hypothetical basis. And having done so, the FTT should then have taken stock of the position 31 with Mr Puzey, clarified whether he still wished to maintain any objections to the reception of the material in evidence, and (if so) ruled upon the objections without more ado. In general terms, we respectfully endorse the guidance given by Lightman J in the Mobile Export case, although it must now be read in the light of the FTT Procedure Rules.

[109]We are also anxious to dispel any suggestion that the FTT Procedure Rules envisage a kind of evidential free for all where anything goes. Particularly in heavy and complex cases of the present type, it is important that directions for evidence should be given and adhered to on both sides, and that there should be no dispute about the evidential status of documents placed before the Tribunal. Otherwise, as the present dispute shows, there is much scope for misunderstanding and potential unfairness.

[110]Nevertheless, the fact remains that KPMG’s letter of 13 July 2012 to the FTT gave a basic description of the nature and content of the Disputed Documents. To the extent that the new material corrected acknowledged errors in Mr Robinson’s model, which had been exposed when he gave evidence, it is hard to see how there can have been any prejudice to HMRC. Likewise, the correction of the “rogue” data for 2003 would seem to have removed any objection to using 2003 as one of the fixed FIN 51 data points on the Graph. Whether those fixed data points were sufficient to ground the inferences which the FTT drew from them is another matter, to which we will have to return. To the extent that the new material extended the previous iterations of the model from 2004 to 2011, this was done in response to indications from the Tribunal that it might be helpful, and KPMG clearly explained the limitations on the usefulness of the exercise. In those circumstances, it was in our view legitimate for GMUK to place the material before the FTT, while maintaining their existing case. It was also legitimate, in principle, for the FTT to decide to make use of the new material, even though GMUK placed no reliance upon it. As we have explained, the role of the FTT is in some respects an inquisitorial one, and it is not merely a passive umpire in a dispute between the taxpayer and HMRC. This more active role reflects the interest of the general body of taxpayers in ensuring that the correct amount of tax is paid.

[111]In assessing HMRC’s complaints of unfairness, we also bear in mind that Dr Holweg was able to, and did, comment on the new material; that Mr Puzey’s objections to the material being put to Dr Holweg were rather half-hearted, and not vigorously pursued; and that, for whatever reason, Mr Puzey chose not to apply for Mr Robinson to be recalled for questioning on the material. Had such an application been made, we cannot think of any grounds upon which GMUK could reasonably have objected to it. Furthermore, if HMRC had any residual concerns after Dr Holweg had given his evidence, they could have raised them with GMUK and/or the FTT in the period of over three months which elapsed between 20 July and the resumption of the hearing on 29 October 2012.

[112]Taking all of these matters into account, we are ultimately not persuaded that there was any procedural error of law in the way in which the FTT dealt with, 32 and placed reliance upon, the Disputed Documents. The third ground of appeal therefore fails. Ground 4

[113]We will deal with this ground briefly, because it seems to us to lie uneasily between grounds 3 and 5, to neither of which does it add anything of substance.

[114]As formulated in the original Grounds of Appeal, the nub of ground 4 is that, even if the Disputed Documents were rightly admitted, it was wrong for the FTT to place any reliance upon their contents. It is said that this goes beyond a criticism of the weight to be attached to the calculations, but is rather a challenge on Edwards v Bairstow grounds to the effect that no Tribunal, properly directing itself, could have had regard to those calculations.

[115]This contention is then explained by reference to the range of FIN 51 data points shown on the Disputed Documents, as compared with the five data points (from 1999 to 2003, excluding 1998 because it was the starting point) in earlier versions of the model. This argument is in our view better considered with ground 5, which is a more broadly based Edwards v Bairstow challenge to the FTT’s acceptance of a model-based approach to the problem.

[116]In their skeleton arguments, Mr Puzey and Mr Millington seek to broaden the scope of ground 4 in some respects, but they all relate to the use made by the FTT of the Disputed Documents, and mount a challenge to such use on Edwards v Bairstow grounds. We therefore consider that these further points are again better considered under ground 5, to which we now turn. Ground 5

[117]The fifth ground of appeal is that the FTT erred in law in accepting a modelbased approach to the question of the cost of manufacturing cars in the UK during the Claim Period. The alleged error of law is put squarely on Edwards v Bairstow grounds, that is to say HMRC accept that they need to bring their challenge within the well-known principles stated by Lord Radcliffe in Edwards v Bairstow

[1956] AC 14

at 36:

“When the case comes before the court it is its duty to examine the determination having regard to its knowledge of the relevant law. If the case contains anything ex facie which is bad law and which bears upon the determination it is, obviously, erroneous in point of law. But, without any such misconception appearing ex facie, it may be that the facts found are such that no person acting judicially and properly instructed 33 as to the relevant law could have come to the determination under appeal. In those circumstances, too, the court must intervene. It has no option but to assume that there has been some misconception of the law and that this has been responsible for the determination. So there, too, there has been error in point of law. I do not think that it matters whether this state of affairs is described as one in which there is no evidence to support the determination or as one in which the evidence is inconsistent with and contradictory of the determination, or as one in which the true and only reasonable conclusion contradicts the determination. Rightly understood, each phrase propounds the same test. For my part, I prefer the last of the three, since I think that it is rather misleading to speak of there being no evidence to support a conclusion when in cases such as these many of the facts are likely to be neutral in themselves, and only to take their colour from the combination of circumstances in which they are found to occur.”

[118]HMRC therefore do not shrink from submitting that, on the basis of all the evidence which the FTT heard and considered, the “true and only reasonable conclusion” open to them was that neither Mr Robinson’s model, nor the modified version of it which the FTT adopted, was rationally capable of answering the question of cost of manufacture to the civil standard of proof on a balance of probabilities. In a case of this nature, it is essential to bear in mind the emphatic warning given by Mummery LJ in Procter & Gamble UK v Revenue and Customs Commissioners

[2009] EWCA Civ 407

[2009] STC 1990

, at [74]:

“I cannot emphasise too strongly that the issue on an appeal from the tribunal is not whether the appellate body agrees with its conclusions. It is this: as a matter of law, was the tribunal entitled to reach its conclusions? It is a misconception of the very nature of an appeal on a point of law to treat it, as too many appellants tend to do, as just another hearing of the selfsame issue that was decided by the tribunal.”

See too the observations of Jacob LJ at [9] to [11] and [19] and Toulson LJ at [48] and [60] to [62].

[119]It is equally important for an appellant on Edwards v Bairstow grounds to observe the principles stated by the Court of Appeal in Georgiou (trading as Mario’s Chippery) v Customs and Excise Commissioners

[1996] STC 463

at 476, where Evans LJ (with whom Saville and Morritt LJJ agreed) said:

“It follows, in my judgment, that for a question of law to arise in the circumstances, the appellant must first identify the finding which is challenged; secondly, show that it is significant in relation to the conclusion; thirdly, identify the evidence, if any, which was relevant to that finding, and, fourthly, show that that finding, on the basis of that evidence, 34 was one which the tribunal was not entitled to make. What is not permitted, in my view, is a roving selection of evidence coupled with a general assertion that the tribunal’s conclusion was against the weight of the evidence and was therefore wrong.”

[120]Mr Puzey and Mr Millington did not explicitly follow these principles in drafting either the grounds of appeal or their skeleton argument in support of the appeal, but in their reply to GMUK’s skeleton they attempted to remedy the omission by explaining how the rather discursive discussion in their first skeleton argument was meant to fit in with the Georgiou principles. Thus, they said: “30. It must be remembered that the Tribunal’s findings of fact have not been challenged. There is no challenge for instance to the following conclusions:  The acceptance, in principle, of FIN 51 figures as being an acceptable means of establishing cost;  The Tribunal’s proposed alterations to the cost price ratio derived from those figures;  The Tribunal’s conclusion that a model could in theory be used for the task;  The Tribunal’s acceptance that the most appropriate macroeconomic cost indices available had been used to attempt to model cost.

[31]What is challenged is whether the model as proposed by Dr Robinson and adapted by the Tribunal could produce an outcome that was more likely than not for the claim period. The Tribunal’s conclusions on this key issue were evaluative judgments and the challenge to these follows the approach set out by the Court of Appeal in Georgiou.”

[121]HMRC’s key criticism is then identified as being that the FTT “adopted and then adapted Dr Robinson’s model in circumstances which did not accord with the prevailing standards of those with expertise in modelling”. This criticism does not get off to a promising start, however, because HMRC expressly accept the correctness of the statements of principle by the FTT at [306] and [309] of the Decision, as follows: “306. A model’s usefulness depends upon its likelihood of predicting an outcome sufficiently close to reality to enable the user to have confidence relying on it. In the context of this appeal that means that it is useful only if it delivers a result which is more likely than not. 35 …

[309]Dr Holweg says that a good model is one whose assumptions are justified and whose results can be verified against actual outcomes. Mr Robinson did not take issue with this. We agree. For our purposes the accuracy of the predictions of a model and the validity of its assumptions are the factors which point to the likelihood that it delivers a result which represents reality. These issues formed much of the debate before us, and we discuss them below.”

[122]The FTT went on to say, at [312], that they would, first, consider the detail of the model proposed (and variations on it); secondly, address the justification of the assumptions inherent in the model; and, thirdly, consider its validation against actuality.

[123]At this point, we need to refer to the FTT’s discussion of the third of these topics, namely validation, because it is only in the light of this discussion that HMRC’s criticisms can be understood.

[124]The FTT began by contrasting some of the main points made by Dr Holweg for HMRC and Mr Robinson for GMUK: “351. Dr Holweg agreed that the object of modelling was to reduce reality by creating a number of assumptions and thereby ignoring certain variables which occur in reality; thus the failure to take into account a particular feature of reality was not something wrong with the model but an inherent feature of any model. The question was to what extent does the model represent reality and that is tested by the congruency or otherwise of the results of the model with reality. The question becomes whether the degree of error when the results of the models are compared with reality suggest that too much has been ignored in making the relevant assumptions. That, he said, was where subjectivity came in: the level of error which would be acceptable varied by discipline. In the social sciences 90% or more was treated as the minimum.

[352]Dr Holweg says that statistics is a game of large numbers. He says that four or five data points provide very little comfort for the accuracy of a model. He says that he would expect about 30 data points before a reliable estimate could be made of the degree of error in the model. The figure of 30 was a convention and the reliability even then was dependent on the lack of significant error in the prediction. It depended to some extent on the confidence levels sought. In his profession levels of 90% or more were required. The convergence of the model and the FIN 51 data was not of sufficient quality to present to an academic audience as proof of the model. There was in his 36 view insufficient data to prove that there had been a significant deviation from the 66% cost/list price ratio.

[353]Mr Robinson compares the Blue and Red line predictions produced by his model as extrapolations of the 1978 starting point to the actual FIN 51 figures for 1998 to 2003. He says that they differ by only a few per cent. He says that given the periods between the starting point in 1978 and the first FIN 51 data, the accuracy of the prediction is impressive and indicates that the model is likely to be reliable.”

[125]The FTT then turned to the document at page 954G, i.e. the Schedule. They recorded that Dr Holweg had commented upon it in his evidence. They gave a brief description of the data which it recorded. They continued: “356. It seems to us that the closeness of the predicted figures from Dr Robinson’s revised blue line to the movements of the FIN 51 ratios is persuasive that the model does produce something which is close to the FIN 51 data. It does not convince us that it produces it accurately: as the absolute mean error of 4.15% on 954G for the comparison of FIN 51 with the production based model shows.

[357]The schedule compares the results of the figures produced for the cost/list price ratio by the model on the basis:(1) of sales car numbers and a Q of 1.303, and

(2) of production car numbers and a Q of 0.934 with the FIN 51 figures for 1998 to 2011. (Mr Robinson’s comparison in his earlier evidence had been against Mr Fulcher’s numbers for 1999 to 2003. We asked during the hearing whether that data could be extended to a longer period. The appellant produced figures for 2004-2011 but told us that they did not rely on them.) The use of production rather than sales figures to moderate fixed costs appears to us to be a justified input to the model. We thus concentrate only on the results of (2).

[358]Under basis (2) the ratio produced by the model for 2003 almost coincides with the FIN 51 figures for that year. We therefore approach the model as if it predicted from the 2003 figure the figures for earlier years (there is no difference in the computational methodology between running forwards and backwards). The comparison showed that the FIN 51 ratios moved broadly in harmony with, but exceeded, the model’s predictions. [The FTT then set out two tables, showing the differences between the predictions and the actual FIN 51 figures, first for 1998 to 2002, and then for 2003 to 2010] 37 The first set of figures, and the general correspondence of the figures, suggested to us that it was likely that the model delivered a result for periods before 1998 which was within a margin of the result which would have been produced by the FIN 51 numbers had they been available. The question was, what was that margin?”

[126]We pause here to deal with a point of detail. In our view it is clear from the context that in [356] the FTT were referring to the Disputed Documents, even though there is no “blue line” on the Graph. The revised line shown on the Graph was in fact red. We are satisfied that this was a minor slip, of the kind which we have noticed elsewhere from time to time in the Decision. Although HMRC sought to make something of this point at the hearing, and we permitted the parties to file further written observations on it after the hearing had ended, we do not consider it at all plausible that the FTT were confused about the document to which they were referring.

[127]The FTT went on to discuss a model produced by Dr Holweg, which showed the effects of uncertainty in the variables on the eventual prediction, and the effects of using a slightly different model. They accepted the point Dr Holweg was making. The FTT then recorded at [360] that they were dealing with the version of the model “in which Q rather than the Ward index is used and in which the figures for vehicles produced, rather than sold, are used to moderate the fixed costs”. They said, correctly, that this was one of the results displayed on the Schedule.

[128]The FTT then made some important observations on their approach to the statistical issues: “361. It seems to us that the exercise we are required to conduct is different from that a medical statistician might attempt in this respect. A medic would not want to offer a new treatment unless really very confident that it would do no harm. The phrase 95% confidence describes that quantitatively as well as mathematically. We are in a different position. Our concern is whether something is more likely than not. Not quite the toss of a coin but not beyond reasonable doubt. We are not in the land of academic papers.

[362]The experts could not give a figure for the likelihood that the model predicted the right results. They could not say “it is more likely than not that the model overstates or understates the cost/list price ratio in any year” or “there is a 26% chance that the model is right to within 7%”. We are left with subjective impressions; but comforted to some extent by the fact that Dr Holweg tells us that the level of confidence required or accepted in more rigorous academic circles is in the end a subjective decision.” 38 The FTT then noted that Dr Holweg did not think it possible to provide a “probability density” for results in earlier years, and GMUK had not volunteered one. Accordingly, they again said they were “stuck with a more subjective approach”.

[129]The FTT then concluded their substantive discussion, as follows: “364. We are persuaded that the assumptions inherent in Mr Robinson’s model are unlikely to produce errors in the Claim Period which are not reflected in (by which [we] do not mean equal to) the degree of error produced in the period of comparison with the FIN 51 results. The assumptions are not unjustifiable or incapable of being reflected in the evaluation against actuality, and our consideration of them did not indicate that the simplifications which they introduced were likely to give rise to significant changes in the Claim Period which were not encompassed by the variables used … As a result it seems to us that if comparison of the outcomes of the model with reality support a conclusion that the actual Claim Period ratios lie within a range of the model prediction, the nature of the assumptions underlying the model is not such as to upset a conclusion to that effect.

[365]If one considers the model as predicting from the 2003 data the FIN 51 data for 1998 to 2002 – i.e. one applies its methods, assumptions and data to see what it predicts for those periods, one may assess its validity in that period by reference to the fit to those known points and the possible variations of the real Q in that period against the presumed fixed Q become part of the evaluation of the fit to actuality. If one can conclude that there is a likelihood that within particular parameters the model does predict reality, then one may be able to assign a likelihood to its doing so within a margin for years before 1998.

[366]It seems to us that the use of a starting point of 1978 has shown that the model does have some predictive value. But it is better to use it to regress back to the Claim Period from the FIN 51 ratios for 1999-2003 to obtain for the Claim Period estimates of the FIN 51 data than to try to estimate from the more distant past (1978) the ratios in the Claim Period. This has the practical advantage that any changes that arise as a result of our conclusions on the compilation of the FIN 51 ratios are taken into account in any conclusion as to the ratios which applied in the Claim Period, rather than requiring a subsequent adjustment.

[367]The figures in [the Schedule] use Mr Robinson’s original Q factor rather than the interpolated Ward index. The difference between the use of the two factors results in the period of comparison with the FIN 51 ratios is that the Ward index ratios are lower in each year by an average of 3%. That indicates that if the Ward index is used in the model regressing 39 the FIN 51 figures back to the Claim Period it would deliver a greater cost/list price ratio in that period. But the Ward index accounts for only some changes. It seems to us that the conclusion that the Q predictions lie within 6% to 7% of the actual is enough to encompass a similar conclusion that they lie within a lesser range of the Ward figures.

[368]At paragraph [358] above we ask within what margin it could be said that it was likely that the model’s results coincided with those which would have been derived from FIN 51 had they existed. Given the variation shown in the table in that paragraph it seemed to us that that margin should in periods adjacent to 1998 to 2003 be 5% - so that one could say that it was likely that in that period the FIN 51 result fell within 5% of the result forecast by the model.

[369]But we were unwilling to conclude that we could draw the same conclusion in relation to the same margin to periods further away from 1998-2003. That was for three reasons:(1) the more iterations of an algorithm, the greater the scope for the significance of a compounding error;

(2) the comparison of the model’s results with the new (unattested) FIN 51 figures for 2005-2011 gives a feeling of some unease;

(3) the possibility that one or more of the constituents of Q did not change such that Q was uniform over the period became more significant.

[370]On this basis we have concluded that it was likely that the model delivered results which were no less than:(1) 6% less than the FIN 51 figures would have been in 1991 to 1996,

(2) 7% less than what the FIN 51 figures would have been for 1986-1990 and that it was not possible to say that it was likely that the difference would have been less than that. [The FTT then dealt with the “rich mix” factor in [371] and [372]. Nothing turns on this.]”

[130]As we read these paragraphs, we think that the main steps in the FTT’s reasoning may be summarised as follows:(1) The assumptions inherent in Mr Robinson’s model are unlikely to produce errors in his figures for the Claim Period which are any greater than the range of error observable in the period for which FIN 51 data 40 are available, when those data are compared with the projections of the model.

(2) Mr Robinson’s use of 1978 as a starting point shows that his model has some predictive value, but it is better to use the model to work backwards from the FIN 51 figures for 1991-2003 to the Claim Period. There are two reasons for this: (a) the Claim Period is nearer in time to 1999/2003 than it is to 1978; and (b) the FTT’s changes to the FIN 51 data can be taken into account at the beginning of the exercise, rather than by way of subsequent adjustment.

(3) The adjustments to the model based on Mr Robinson’s Q factor are substantial enough to subsume the adjustments which application of Dr Holweg’s Ward index would require, bearing in mind that the Ward index accounts for only some changes in vehicle specification.

(4) Between 1998 and 2003, the likelihood is that the actual FIN 51 ratios fell within 5% of those predicted by the model (using the adjusted 2003 FIN 51 figures as the starting point).

(5) A larger margin of error should be allowed, however, for years before 1998, for the three reasons given in [369].

(6) The appropriate margins of error to adopt for the years within the Claim Period are: (a) for 1991 to 1996, 6% (but no less than that); and (b) for 1986 to 1990, 7% (but no less than that). This means that the cost price shown by the adjusted model for those years is more likely than not to have fallen within 6% (or, for the earlier years, 7%) of the actual cost price which the FIN 51 data for those years (were they still available) would have established.

[131]An aspect of the FTT’s reasoning which may at first sight seem puzzling is their decision to take 2003 as their starting point, apparently for no better reason than the near coincidence in that year of the red line on the Graph with the FIN 51 data point. The FTT can safely be credited with realising that there is no particular significance in this coincidence, given the margins of error 41 shown on the Schedule and recorded by them in the Decision at [358]. Why, then, did they prefer to work backwards from 2003, rather than forwards from Mr Robinson’s solid (and validated) starting point in 1978? And why did they allow a larger margin of error for the years in the Claim Period which were nearer in time to 1978, than they did for the more distant period of 1991 to 1996?

[132]This question has caused us some anxiety, but we think the answer lies in the fact that the only attested FIN 51 data still available are those for 1998 to 2003. Those figures are “hard” data, just as much as the two thirds cost price ratio in 1978. The gap between them had to be filled, if at all, by extrapolation, either forwards or backwards. Mr Robinson chose to extrapolate forwards, and (crucially) the FTT were satisfied that his model had real predictive value. Thus the coincidence between his projection and the 2003 FIN 51 data point itself forms part of the pattern which (for the FTT) validated his model, even though the coincidence in that year was statistically no more or less significant than the results for the other FIN 51 years when the red line and the FIN 51 data points diverge by a few per cent. The reasons given by the FTT for preferring to run the model backwards rather than forwards do not appear to us particularly strong, but that is not the point. It was a matter for the FTT to decide, and we certainly cannot say that their decision was so unreasonable as to amount to an error of law.

[133]Once the decision to run the model backwards had been taken, it then made good sense to pick 2003 as the starting point. It was clearly necessary to start from one of the “hard” FIN 51 data points, and taking the latest of them permitted the maximum comparison between the red line and the FIN 51 figures back to 1998. Again, it seems to us fanciful to argue that the FTT’s choice was so unreasonable as to involve an error of law. Furthermore, the effect of running the model backwards was that the FTT allowed a larger margin of error in the years in the Claim Period which were closer to 1978. Even on the assumption that this was erroneous, it was an error which favoured HMRC, because its effect was to reduce the gap between the red line adjusted for error and a cost price ratio of two thirds.

[134]We must now deal with a more fundamental objection raised by HMRC, which lies at the heart of ground 5. They argue that no reasonable tribunal properly directed as to the law could have used Mr Robinson’s model, or any variant of it, as the basis for making findings on a balance of probabilities about the cost of cars during the Claim Period. It is said that the modelling methodology employed by Mr Robinson was too flimsy to justify the conclusions which the FTT purportedly drew from it, because it did not conform to the minimum requirements for justification and validation accepted by experts in the field. The methodology therefore had no probative value, or at least had insufficient probative value to found findings to the civil standard of proof. The work carried out by GMUK failed to satisfy the basic requirements of modelling, and the FTT were wrong to ignore Dr Holweg’s evidence to this effect. 42

[135]HMRC further submit that Mr Robinson himself recognised that his work would not withstand scrutiny by experts in the field. HMRC rely, for example, on the following evidence given by Mr Robinson in re-examination on day 6:

“I mean, there is a huge uncertainty. All I’m really asking the tribunal to accept is that all this information I’ve brought to bear from the macroeconomy, from the accounts and from FIN 51 compared with the starting point, tells you more – tells you usefully more than you would know in its absence, because in its absence I think you just would be used to drawing a straight line between there and there. I think there’s important evidence here that should be brought to bear, and it’s not perfect, and Dr Holweg is quite right to say this wouldn’t – you know, I couldn’t write a paper at an econometric conference based on this work, I’m just trying to help the court as best I can with what we have.”

[136]On the previous day, Mr Robinson had said in his evidence in chief:

“If I can just very briefly be the business school lecturer I once nearly was, what I’ve done here is describe first of all – all my model does – I think even calling it a model is to exaggerate its complexity – is to say that the cost-price ratio which starts at something like 66 over 100, will have fallen over this period, because costs grew less fast than prices.”

[137]HMRC also point to the explicit acceptance by the FTT, at [362], that the experts “could not give a figure for the likelihood that the model predicted the right result”, with the result that they were “left with subjective impressions”. These observations must, however, be read in their context. The FTT had just directed themselves, correctly, in [361] that their concern was “whether something is more likely than not”, and that they were “not in the land of academic papers”. The point they were making in [362] was that neither expert could assign a probability of correctness to the projections of the model. They were not saying that the model was incapable of grounding a finding of fact on the balance of probabilities.

[138]One aspect of the model which HMRC particularly criticise is the perceived congruence of the model’s outcome with the FIN 51 data for 1998-2003. Dr Holweg’s evidence was that any perceived congruence with so small a number of data points had no statistical significance at all, and Mr Robinson himself conceded that forecasting on the basis of such a small number of data points “seems very slender”.

[139]In fairness to Mr Robinson, we should point out that he immediately went on to say (day 6, page 90 of the transcript): 43 “But I think that is to make the problem seem worse than it is, for the reason that I keep banging on about, which is that, actually, I think we have subjected this model to quite a severe test by starting it in 66 [i.e. the two thirds ratio in 1978], running it through with just macro data plus the Holweg Q, and then lo and behold, we are 3% away. Again there can be noise in that, but it looks to me as if the trend is pretty accurate, if it has come so close after a 20-year blind forecast.”

[140]To summarise, HMRC say the FTT should have concluded that the absence of sufficient data prevented the successful completion of the exercise proposed by GMUK. As Mr Puzey and Mr Millington put it in their skeleton argument in reply:

“To continue to “muddle through” on the basis of [the FTT’s] more “subjective approach” wrongly ignored the weight of the evidence that it was not possible to produce a model that generated results that were in any sense reliable.”

[141]In considering these submissions, we begin with the argument that the FTT were obliged to disregard Mr Robinson’s model because it failed to meet the minimum standards for credible expert evidence in the field of econometrics. One obvious difficulty with this argument is that Mr Robinson was not, strictly speaking, giving evidence as an expert at all. As we have already pointed out, neither Mr Robinson nor Dr Holweg was tendered as an expert witness, and they gave their evidence as witnesses of fact. In those circumstances, it seems to us unfair to criticise Mr Robinson as if he had presented his evidence in an expert’s report, and the question for the FTT was simply whether, on all the evidence before them, they felt able to draw any conclusions, on a balance of probabilities, about the cost of cars during the Claim Period.

[142]In the second place, we agree with Mr Cordara that the exercise which the FTT had to perform was essentially a unitary one. They did not first have to decide whether the econometric evidence before them met a particular qualitative threshold, but rather had to assess it together with all the other evidence before delivering their answer on a balance of probabilities. They were right to observe that they were not in the land of academic papers. Their overriding duty was to ascertain, if they properly could, the true amount of VAT due during the Claim Period, drawing for that purpose on their own experience and expertise as well as the evidence presented to them. If the performance of this duty involved an element of subjectivity, that means no more in the present context than that they had to bring their own skill and judgment to bear on the problem, and they were not bound by the standards of evaluation which would apply if the same question arose in an academic context. In our view the FTT discharged their duty carefully and conscientiously, and with a full awareness of the potential weakness of the data which underlay Mr Robinson’s model. 44

[143]For similar reasons, we consider that the FTT were plainly entitled to conclude that, despite the paucity of the FIN 51 data points, there was a sufficient correlation between them and Mr Robinson’s “blind” projections from 1978 to establish that the model had probative value. We find it impossible to say that, in reaching this conclusion, they must have erred in law.

[144]In our view it would be disproportionate to deal in detail with each and every argument advanced by HMRC in support of ground 5. Mr Puzey rightly recognised in his oral submissions that, in a case of the present nature, an Edwards v Bairstow challenge faces very considerable difficulties. Once his arguments based on the quality of the “expert” evidence, and the use of the FIN 51 data points, have been disposed of, it seems to us most implausible that any other grounds could justify the conclusion that the FTT must have erred in law in adopting a model-based approach to this part of the case, and then drawing the conclusions which they did about the cost of vehicles during the Claim Period. We will merely say that we have done our best to consider all of the points raised by HMRC in their written and oral submissions, and remain wholly unpersuaded that any such error of law can be discerned.

[145]For these reasons, HMRC’s appeal on ground 5 will also be dismissed. Overall conclusion

[146]Since we have upheld none of the grounds of appeal, it follows that this appeal must be dismissed. MR JUSTICE HENDERSON JUDGE SINFIELD RELEASE DATE: 17 November 2015

Cited in 13 later judgments

and 4 more of the most senior