[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	United Kingdom VAT & Duties Tribunals (Excise) Decisions
You are here: BAILII >> Databases >> United Kingdom VAT & Duties Tribunals Decisions >> United Kingdom VAT & Duties Tribunals (Excise) Decisions >> Continental Wine and Food Ltd v Revenue & Customs [2006] UKVAT(Excise) E00965 (23 June 2006) URL: http://www.bailii.org/uk/cases/UKVAT/Excise/2006/E00965.html Cite as: [2006] UKVAT(Excise) E965, [2006] UKVAT(Excise) E00965

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

E00965

EXCISE DUTY – Excise Duty Directive (92/83/EEC) having direct effect – whether article 16 (intermediate product) or article 20 (ethyl alcohol) applies – Directive incorporates test by reference to tariff classification – Combined Nomenclature (CN) – rum and coconut tasting alcoholic beverage produced from a fermented made-wine base and fortified with distilled alcohol – BTI classifying the beverage under CN heading 2208 – whether beverage properly classified under CN heading 2206 (fermented beverage) or 2208 (spirituous beverage) – objective characteristics to be considered to determine which component gives the beverage its essential character – General Rule of Interpretation (GRI) 3(b) – held the distilled alcohol gives the beverage its essential character – therefore properly classified under CN heading 2208 – therefore article 20 of the Excise Directive applies – appeal dismissed

MANCHESTER TRIBUNAL CENTRE

CONTINENTAL WINE AND FOOD LIMITED Appellant

- and -

THE COMMISSIONERS FOR HER MAJESTY'S

REVENUE AND CUSTOMS Respondents

Tribunal: JOHN WALTERS QC (Chairman)

MISS CAROLE ROBERTS

Sitting in public in Manchester on 9 and 10 March 2006

Alan Powell, of Triple Two Consulting, for the Appellant

Owain Thomas, Counsel, instructed by the Acting Solicitor for HM Revenue and Customs, for the Respondents

© CROWN COPYRIGHT 2006


 


DECISION

The appeal and the Tribunal's jurisdiction

1. Continental Wine and Food Limited ("the Appellant") appeals in terms (according to the Notice of Appeal sent to the Tribunal) against a decision ("the Original Decision") made by Mr. R. Lamb, an Officer of HM Customs and Excise (predecessors of the Commissioners of HM Revenue and Customs and generally under both styles referred to in this Decision as "Customs") to reject a claim for repayment of £21,002.73 excise duty and to assess additional excise duty of £31,528.73, communicated to the Appellant by a letter dated 23 January 2003. The Original Decision was in fact confirmed by a decision on review, communicated to the Appellant by Mrs. J. A Birnie, a Reviewing Officers of Customs, in a letter dated 1 April 2003. The assessed duty was paid by the Appellant.
2. The rejection of the repayment claim, and the assessment, followed from the ruling of Customs Tariff Classification Section (a binding tariff information ("BTI")) that the Appellant's product "Bali Tropical Alcoholic Drink with Caribbean Rum & Coconut flavour" (to which we refer in this Decision as "Bali") should be classified under tariff heading 2208, and not under tariff heading 2206, as contended for by the Appellant.
3. These are tariff headings of the Combined Nomenclature ("CN") and, put shortly, heading 2208 covers spirituous beverages, while heading 2206 covers fermented beverages. In detail, these headings are as follows:

"2206 – Other fermented beverages (for example, cider, perry, mead); mixtures of fermented beverages and mixtures of fermented beverages and non-alcoholic beverages, not elsewhere specified or included.

2208 – Undenatured ethyl alcohol of an alcoholic strength by volume of less than 80% vol.: spirits, liqueurs and other spirituous beverages."

4. The BTI was issued on 1 November 2002 in response to the Appellant's request dated 14 October 2002. The Appellant made this request (notwithstanding the fact that the Appellant did not intend to export the product Bali) because Mr. Lamb indicated that a BTI was necessary for the purpose of his consideration of how, if at all, an extra-statutory concession ("ESC" – paragraph 4.8 of Notice 48) which took effect on 27 April 2002 could apply to the product Bali.
5. Mr. Powell, for the Appellant, suggested in argument that the BTI was invalid because it does not relate to an envisaged import or export operation. This is not right, however, because it is clear from the provision for requesting and supplying a BTI (article 11 of Council Regulation 2913/92/EEC establishing the Community Customs Code) that a BTI may be requested in any circumstances, but that the customs authorities are only bound to supply one where an import or export operation is actually envisaged. In any other case, the customs authorities have a discretion to supply a BTI, and, in this case, the BTI has been supplied by Customs in the lawful exercise of its discretion. It is therefore valid.
6. The ESC, in short, allowed relief from excise duty to certain mixed beverages, which would otherwise be dutiable under the domestic excise duty legislation as spirits, by waiving the application of a condition laid down by sections 57 and 58 Alcoholic Liquor Duties Act 1979 ("ALDA") – that a permitted mixture of made-wine, or wine, and spirits in an excise warehouse must be in a proportion not exceeding 12 litres of alcohol to 1 hectolitre (100 litres) of made-wine, or wine – "so long as the resultant product does not assume the essential characteristics of a spirit". Mr. Lamb needed to know whether the product Bali "assumed the essential characteristics of a spirit" in order to decide whether the ESC applied to it, and he suggested to the Appellant that it should request a BTI to assist him in deciding this issue.
7. By a letter dated 25 September 2003 sent to Customs by KPMG, who were then representing the Appellant, a further departmental review was requested. This request was treated by Customs as a requirement for a further review within section 14(5)(a) Finance Act 1994 ("FA 1994"). The decisions to be reviewed fell within section 14(1)(b) FA 1994 with regard to the duty assessed and within section 14(1)(bb) with regard to the refusal of the claim for repayment of duty. A different Reviewing Officer, Mr. S. Palmer, was appointed to conduct the review and requested samples for independent organoleptic testing by CCFRA Technology Ltd. ("CCFRA"). CCFRA is, the Tribunal understands, a laboratory accredited by the United Kingdom Accreditation Service to carry out tests such as the organoleptic tests envisaged in this case. Apparently both CCFRA and the Appellant are members of the Wines and Spirits Association. (Organoleptic testing, the Tribunal understands, is a scientific sensory process to identify taste and smell.)
8. The Appellant objected to Customs' proposals for testing on a number of grounds and refused to supply the samples unless the specific terms of reference for the testing were agreed between the Appellant and Customs. Agreement for the provision of samples was not reached. Customs applied to this Tribunal for a direction that the Appellant provide the samples requested, without any condition, and that application was dismissed (Tribunal Chairman: Mr. Colin Bishopp) after a hearing on 14 February 2006. In these circumstances the further review was not completed and, pursuant to section 15(2) FA 1994, Customs are assumed to have confirmed Mrs. J. A Birnie's decision of 1 April 2003.
9. It is of course not within this Tribunal's jurisdiction to decide whether or not the ESC should apply to the product Bali. However an appeal lies to this Tribunal under section 16(1)(a) FA 1994 with respect to a deemed confirmation under section 15(2) FA 1994. The Tribunal heard oral evidence from Mr. Palmer. It seems to the Tribunal that, for the purposes of determining the issues in dispute in the appeal, the Notice of Appeal ought to be amended so that it specifies as the decisions appealed against, the deemed confirmation under section 15(2) FA 1994 of Mrs. Birnie's decision of 1 April 2003, which resulted from Mr. Palmer's abortive review. We therefore direct an amendment to this effect pursuant to our power under rule 14(1) of the VAT Tribunals Rules 1986 (SI 1986 No. 590). The correctness or otherwise of the BTI (that the product Bali is properly classified under heading 2208) is therefore in issue. The powers of the Tribunal include power to quash or vary the deemed confirmation of Mrs. Birnie's decision and power to substitute our own decision for the deemed confirmation or Mrs. Birnie's decision (section 16(5) FA 1994).

The facts

10. The facts relevant to the dispute are in large part agreed between the parties. An agreed bundle of documents was put in evidence and (as we have said) we heard oral evidence from Mr. Palmer. From the evidence we find the following facts.
11. The Appellant manufactures the product Bali. It is manufactured by fortification of a fermented wine base made from grape juice concentrate. The drink is flavoured by the addition of rum and coconut flavouring essence.
12. The product Bali was launched in 1998 at a strength of 17% alcohol by volume ("abv") and was classified for excise duty purposes as a made-wine falling with CN heading 2206. In this Decision we refer to the product Bali in this formulation as "the Original Formulation".
13. In 2002 the product Bali was re-formulated to a higher strength of 21% abv (the Reformulation I formulation) in order to compete with similar products on the market, predominantly imported from other EU Member States. Mr. Powell, for the Appellant, submits, and the Tribunal accepts, that fortification with spirits is carried out to stabilise the product, increase alcoholic strength and reduce raw material costs. Distilled alcohol is cheaper by volume than base wine.
14. The product Bali in the Reformulation 1 formulation was produced by the Appellant with the aim of taking advantage of the ESC mentioned above.
15. As we have already said, the Appellant was invited by Customs to apply for a BTI of the product Bali in the Reformulation I formulation in order to assist in the classification of the product Bali for excise duty purposes.
16. The Appellant's application for a BTI was made on 16 October 2002. In the application (on form C103) under the heading "Description of the Goods", the Appellant described the manufacturing process as follows:

"First stage is the fermentation of re-constituted concentrated grape juice (under temperature controlled conditions). This yields an alcoholic strength of approx. 14%.

Having completed fermentation, the base wine is centrifuged and carbon treated to remove any "off flavours" developed at fermentation stage.

The base wine is sweetened with liquid sugar / glucose and fortified with alcohol and rum to ME 22% in accordance with [Customs] regulations in the manufacture of British wine.

White rum flavour and coconut flavour are added to the product at 21.5% immediately prior to bottling.

All stages of the manufacturing process are monitored and tested in house by our technical department. Alcoholic strengths are determined by the Anton Paal density meter."

17. The formulation of the product Bali in the Reformulation I formulation, which the Tribunal is considering and to which the BTI relates, is as follows

"Base wine (14% abv) 583 litres

Liquid sugar 263 litres

Alcohol (96.3% abv) 85 litres

Rum (95% abv) 61 litres

Flavours 6.1 litres"

18. The elements of alcohol derived from distillation in the Reformulation I formulation are represented by the alcohol and the rum (96.3% of 85 litres, plus 95% of 61 litres i.e. 139.805 litres), while the elements of alcohol derived from fermentation are represented by the base wine (14% of 583 litres, i.e. 81.62 litres). On this basis the ratio by volume of distilled alcohol to fermented alcohol is 63:37.
19. The description of the goods given in the BTI is as follows:

"A fortified alcoholic beverage consisting of a base wine (derived from the fermentation of concentrated grape juice) to achieve an alcoholic strength of 14% – 583 litres; liquid sugar / glucose mix 263 litres; alcohol (neutral spirit at 96.3%) – 85 litres; Caribbean rum (alcoholic strength 95%) – 61 litres; white rum flavour – 0.1 litres and coconut flavour – 6 litres. Although the bulk volume of the product is derived from the wine base, the greater alcohol by volume is produced from the spirits. In addition taste and smell tests confirm that due to the addition of neutral spirit and rum, the resulting beverage no longer retains the character of a wine, but has the characteristic of a spirituous beverage of heading 2208."

20. While the Appellant denies that the product Bali in the Reformulation I formulation has the characteristic of a spirituous beverage by reason of its taste or smell, or for any other reason, Mr. Powell, for the Appellant, accepts, and the Tribunal finds, that "although the bulk volume of the product is derived from the wine base, the greater alcohol by volume is produced from the spirits".
21. The date of start of validity of the BTI is 1 November 2002. On receipt of the BTI the Appellant stopped production of the product Bali in the Reformulation I formulation considered above, and further reformulated the product using a higher strength base wine, thereby reducing the proportion of added spirits, so that the greater alcohol by volume was produced from the fermented wine base, and not from the spirits. This further reformulation of the product Bali was submitted to Customs' Classification Unit and was classified as falling within CN Heading 2206 (the made-wine category), with a date of start of validity of 14 November 2002. The Appellant re-commenced production of the product Bali using this further reformulation on 15 November 2002.
22. The further reformulation of the product Bali which the Appellant started to produce on 15 November 2002 – and which we will refer to as "the Reformulation II formulation" – is as follows

"Base wine (17% abv) 715 litres

Liquid sugar 180 litres

Alcohol (96.3% abv) 37 litres

Rum (95% abv) 61 litres

Flavours 6.1 litres"

23. The elements of alcohol derived from distillation in the Reformulation II formulation are represented by the alcohol and the rum (96.3% of 37 litres, plus 95% of 61 litres i.e. 93.581 litres), while the elements of alcohol derived from fermentation are represented by the base wine (17% of 715 litres, i.e. 121.55 litres). On this basis the ratio by volume of distilled alcohol to fermented alcohol is 43:57.
24. The Appellant submits that the product Bali in its Original Formulation and in the Reformulation I formulation has the same organoleptic characteristics (this refers to the taste and smell of the product). The Appellant also submits that the product Bali in the Reformulation II formulation has the same organoleptic characteristics of the product in the earlier two formulations. The Appellant submits that the source of the alcoholic content of the product (whether from distillation or fermentation) does not affect the organoleptic characteristics of the product, and indeed that the source of the alcoholic content, whether from distillation or fermentation, will vary from batch to batch, because fermentation is a natural process. The Appellant points out that the rum and coconut must remain constant – as they do in the formulations being considered – as these are, in the Appellant's submission, the major contributors to the essential organoleptic characteristics of the product Bali, and, if they did not remain constant, consumer and customer complaints would arise, due to variability of the product. We accept these submissions and find facts accordingly.
25. The Tribunal did not see, smell or taste a sample of the product Bali. Mr. Palmer tasted samples from one formulation accepted by Customs to be properly classified as 2206 (a fermented beverage) and from the Reformulation I formulation (the formulation in dispute). The samples were unidentified when he tasted them. His evidence was that he could taste coconut and rum in both samples. He accepted that he was not an expert in tasting (i.e. in assessing organoleptic characteristics of a product) and confirmed that he could detect no difference in taste between the samples.
26. The reason why the Appellant did not submit samples of the product Bali to be organoleptically tested by CCFRA was that it insisted that such testing should take the form of a comparison of the organoleptic characteristics of the product Bali in the Reformulation I formulation and in the Original Formulation or the Reformulation II formulation, the Appellant maintaining that there was no, or no significant difference between the organoleptic characteristics. Customs on the other hand, and this was Mr. Palmer's evidence, insisted that this was the wrong basis on which to invite CCFRA to carry out organoleptic testing. They insisted that as the review was a review of the BTI, the testing should be of the product Bali in its Reformulation I reformulation in isolation.
27. At the hearing, the Tribunal indicated that it would reconsider Mr. Bishopp's decision not to direct that a sample of the disputed product, i.e. the product Bali in its Reformulation I reformulation, be provided for testing by CCFRA.
28. We have decided not to disturb Mr. Bishopp's decision. We have concluded that testing by CCFRA is not necessary for us to reach a decision in this appeal. We note that Mr. Thomas, for Customs, stated that he was content to rest on the evidence as it presently stands.
29. We accept the evidence of Mr. Palmer that there is no or no significant difference in the taste and smell of the product Bali in its Original Formulation and in the Reformulation I reformulation. We also accept his evidence that the product in both formulations smells and tastes of rum and coconut.
30. The Tribunal was shown an empty bottle of the type in which the product Bali is put up for sale. It is a large (70 cl) glass bottle, produced for normal sale to the off-licence trade, with an art-work design featuring palm trees with the word "Bali" prominently displayed, and with the description "Tropical Alcoholic Drink with CARIBBEAN RUM & COCONUT flavour". The word "Caribbean" is repeated in stylised lettering.
31. We find from this evidence that the product Bali is marketed as a drink with characteristics designed to evoke tropical places (Bali and the Caribbean) and a tropical lifestyle of relaxation, and a tropical taste of rum and coconut. We take note of the fact that rum is and is commonly known to be an alcoholic beverage and, in fact, a spirituous beverage, specifically mentioned under CN heading 2208.
32. The product Bali is sold in the UK market with a retail price of between £5 and £7.
33. Mr. Palmer applied a general test in considering whether to uphold the BTI. This was on a provisional basis pending testing of the product Bali in the Reformulation I reformulation by CCFRA, and subject to any evidence that such testing might produce. The test was set out in his letter dated 27 February 2004 to KPMG, then acting for the Appellant, and was, relevantly, as follows:

"Three tests are applied to products that are comprised of a mixture of fermented beverage and added alcohol to establish whether it should be characterised as a spirituous or a fermented beverage … :

Whether the spirituous or fermented liquor contributes more to the overall volume of the liquor.

Whether the spirituous or fermented liquor contributes more to the alcohol content of the liquor.

Whether the external characteristics of the product (for example, how it is labelled and presented) are those of a spirituous or fermented beverage.

If both of the tests looking at the relative contributions to the product show that it is made up of more spirituous rather than fermented liquor, the product would be classified as a spirit and will be duties as such. If similarly, both of these tests show that a product is comprised primarily of the fermented element, then the fortified product will retain its duty status as a made-wine or wine.

However, if the two tests on the composition of the liquor show different results, then the third test on the external characteristics of the product will determine how it is to be treated. For instance, if the majority of the volume of a beverage was sourced from its fermented component, but the majority of the alcohol within the beverage was sourced from distilled liquor, then the status of the drink would be decided by whether its is presented/marketed as a spirituous beverage."

34. Mr. Palmer, correctly, looking at the product Bali in its Reformulation I reformulation, concluded that while the fermented liquor contributes more to the overall volume of the product, it is the spirituous liquor that provides the bulk of the alcohol content.
35. Therefore he applied the third test, and concluded that as the product was put up as a "Tropical Drink with Caribbean Rum and Coconut flavour", the emphasis on rum suggested that the product Bali was being marketed as a spirituous beverage. The taste and smell of the product Bali (in both formulations) as a drink with a rum and coconut flavour confirmed this conclusion. He acknowledged that he had tasted the product in both the Original Formulation or the Reformulation II formulation and in the Reformulation I reformulation and both tasted the same, but concluded that since he was only concerned with the Reformulation I formulation in his review, this factor was irrelevant.
36. Mr. Palmer said in evidence and the Tribunal accepts that the test described above represented Customs' policy in this area.

The law

37. The base wine element in the product Bali is "made-wine" within section 1 of ALDA. Section 57 of ALDA permits the mixing of made-wine and duty-free spirits in an excise warehouse (as in this case) provided that the mixture is not raised to a greater strength than 22 per cent, and provided also that the mixing is in a proportion not exceeding 12 litres of alcohol to 1 hectolitre of made-wine. The mixture in this case, it is common ground, does not meet the second proviso. Therefore, as a matter of domestic law (and subject to any application of the ESC), the Appellant may not treat the product Bali in its Reformulation I reformulation as liable to excise duty at the rates applicable to wine or made-wine of a strength not exceeding 22 per cent (see: Schedule 1 to ALDA). The product is, instead, liable to duty as spirits. Mr. Powell, for the Appellant, agrees that this is the position.
38. Customs however accept that the Appellant is in principle entitled to rely on the Excise Directive (92/83/EEC) which has direct effect for these purposes. The Excise Directive provides for fixed rates of taxation of spirits only where products fall within CN headings 2207 or 2208, and for a lower rate of taxation for products which fall within CN heading 2206. The references to the CN headings are to those headings as they were in force at the date of the adoption of the Excise Directive 19 October 1992 (see: the recitals to the Excise Directive), but we were told that the relevant CN headings had not changed since that date.
39. In this way the Excise Directive makes the question of liability of products to excise duty dependent on their correct classification for CN purposes. Mr. Thomas, for Customs, told the Tribunal that it was for this reason (to put the CN classification of the product Bali in issue, if necessary) that the Appellant was invited to apply for a BTI even though it had no stated intention to import, or export, the product Bali.
40. Customs accept that if the product Bali in the Reformulation I formulation is properly classifiable in CN heading 2206, then the Appellant is entitled to the benefit of the lower rate of duty.
41. Under the Excise Directive, for relevant purposes, a product properly classifiable under CN heading 2208 (spirits) would fall within article 20 ("ethyl alcohol"), and a product properly classifiable under CN heading 2206, but not exceeding 22 per cent. abv, (higher strength fermented products) would fall within article 16 ("intermediate product"). The classification relates to the CN in force at the date of adoption of the Directive, 19 October 1992 (see: the recitals to the Directive).
42. The CN is published as Annex 1 of the Council Regulation 2658/87/EEC and is updated annually by the Commission. Part 1 of the CN contains, at section 1A the General Rules for the Interpretation of the CN ("GRIs"). Part 2 of the CN contains the headings themselves, which amount to a comprehensive goods nomenclature designed to classify all goods imported into the EU in the appropriate heading.
43. The CN is based on the Harmonized Commodity Description and Coding System ("HS") of the Customs Cooperation Council or World Customs Organisation ("WCO"). The WCO produces explanatory notes to the HS and these are known as "HSENs".
44. The Tribunal was referred by Mr. Thomas to the case of Holz Geenen GmbH v Oberfinanzdirektion München (Case C-309/98) from which it is clear that the Court of Justice has regard to the HSENs in interpreting the scope of the headings of the CNs, although they do not have legally binding force (see: ibid. at paragraph 14). This is supported by other Community law authority such as Develop Dr. Eisbein GmbH v Hauptzollamt Stuttgart-West (Case C-35/93).
45. Chapter 22 of the CN covers "Beverages, Spirits and Vinegar". Within Chapter 22, the scheme of the headings is as follows:

2201 – Waters not containing added sugar or other sweetening material

2202 – Waters containing added sugar, etc. and other non-alcoholic beverages

2203 – Beer made from Malt

2204 – Wine of fresh grapes, including fortified wines

2205 – Vermouth and other wine of fresh grapes flavoured with plants or aromatic substances

2206 – Other fermented beverages – see [3] above

2207 – Undenatured ethyl alcohol of an alcoholic strength by volume of 80% vol or higher; ethyl alcohol and other spirits, denatured, of any strength

2208 – Undenatured ethyl alcohol of an alcoholic strength by volume of less than 80% vol – see [3] above

2209 – Vinegar and substitutes for vinegar obtained from acetic acid

46. Of the GRIs in Part 1 of the CN, the third GRI is particularly relevant. It is introduced by the final sentence of GRI 2(b) as follows:

"The classification of goods consisting of more than one material or substance [including mixtures and combinations] shall be according to the principles of rule 3".

47. GRI 3 is relevantly as follows:

"When … goods are prima facie classifiable under two or more headings, classification shall be effected as follows:

(a) the heading which provides the most specific description shall be preferred to headings providing a more general description. However, when two or more headings each refer to part only of the materials or substances contained in mixed or composite goods or to part only of the items in a set put up for retail sale, those headings are to be regarded as equally specific in relation to those goods, even if one of them gives a more complete or precise description of the goods.

(b) mixtures, composite goods consisting of different materials or made up of different components, and goods put up in sets for retail sale, which cannot be classified by reference to 3(a), shall be classified as if they consisted of the material or component which gives them their essential character, so far as this criterion is applicable.

(c) when goods cannot be classified by reference to 3(a) or (b), they shall be classified under the heading which occurs last in numerical order among those which equally merit consideration." (emphasis added)

48. The emphasised passage in GRI 3(b) also appears in the general interpretative rules relative to the HSENs, but there – in paragraph (VIII) – a further provision is made to the effect that the factor which determines essential character will vary as between different kinds of goods. It may, for example, be determined by the nature of the material or component, its bulk, quantity, weight or value, or by the role of a constituent material in relation to the use of the goods.
49. Finally, we mention the HSEN to classification 2206 (other fermented beverages, etc.) which makes clear that that classification includes inter alia cider, perry, mead, raisin wine, wines obtained by the fermentation of fruit juices other than juice of fresh grapes, malton (a fermented beverage prepared from malt extract and wine lees), spruce (a beverage made from leaves or small branches of the spruce fir, or from spruce essence), saké or rice wine, palm wine, ginger beer and herb beer. The HSEN adds that:

"all these beverages may be either naturally sparkling or artificially charged with carbon dioxide. They remain classified in the heading when fortified with added alcohol or when the alcohol content has been increased by further fermentation, provided they retain the character of products within the heading." (emphasis added).

50. The proviso in the emphasised passage was apparently first introduced by the WCO on an amendment to the HSEN in 2001. Before that amendment, the HSEN to code 2206 (other fermented beverages) stated that products remained classified in the heading when fortified with added alcohol.
51. The European Court in Holz Geenen GmbH emphasised that in the interests of legal certainty, and for ease of verification, the decisive criteria for the classification of goods for Customs purposes is in general to be sought from their objective characteristics and properties as defined in the wording of the relevant heading of the CN (ibid. at [14]).

The parties' contentions

52. The Appellant's case is that the Bali product in the Reformulation I formulation is an intermediate product within article 16 of the Excise Directive. Mr. Powell submits that intermediate products within that article taste similar whether produced as made-wine, fortified made-wine or spirits. This is in effect a submission, which the Tribunal has accepted, that there is no discernible organoleptic difference between the product Bali in the Reformulation I formulation, and in the Reformulation II formulation.
53. He relies on Commission Regulation (EC) No. 2802/95 in the case of Apfelroyale, which is to the effect that a sweet, apple-tasting, alcoholic beverage made from cider, sugar syrup, apple flavouring and alcohol, in containers holding less than 2 litres with an alcoholic strength of 19.5 per cent. abv was properly classified under CN heading 2206. The reasons given in the Regulation for the classification are that it is determined by the provisions of general rules 1 and 6 for the interpretation of the CN (which are not controversial) as well as the wording of CN codes 2206 00 (see [3] above) and 2206 00 51 ("Still, in containers holding: 2 litres or less: cider and perry), and reference is also made to the second paragraph of the HSEN to heading 2006. The second paragraph referred to is presumably in effect the provision that "other fermented beverages" within code 2206 remain classified in that heading when fortified with added alcohol … but before amendment to include the proviso that they must retain the character of products falling in the heading.
54. Mr. Powell made reference to a European Commission document (emanating from the Directorate General XXI, Customs and Indirect Taxation and marked "Working Document for Official Use Only") containing the view of the Commission Services, given apparently as advice to the Excise Committee, on a question raised by an Italian delegation on the classification of alcoholic beverages in CN headings 2206 or 2208. This document appeared to confirm submissions made by Mr. Powell to the effect that the product considered in Apfelroyale was a drink of 19.5 per cent. abv of which almost all the alcohol was of distilled origin.
55. Mr. Powell objects to the Customs' test as applied by Mr. Palmer, and in particular to the importance attached to the fact that the greater part of the alcohol by volume in the product Bali in the Reformulation I formulation is derived from spirits rather than fermentation. He submits that this approach disregards the wording of GIR 3(b) and the objective approach in that GIR, and enjoined generally by the European Court. He referred the Tribunal to the earlier "Guidance Note to trade associations" issued by Customs in May 2002 (after the Budget of 2002) detailing the ESC and explaining that the "set criteria" to be invoked in tariff classification on a case by case basis are:

•    Relative alcohol content by volume (i.e. the proportion of alcoholic strength within the product which is sourced from a distilled as opposed to a fermented origin)

•    Relative alcohol content by bulk and weight (i.e. the proportions within the product by bulk and weight which are derived from fermented as opposed to distilled alcohols)

•    The manufacturing process involved (i.e. principally looking at whether the manufacturer is carrying out a genuine process of fortification)

•    Product labelling

56. An objective approach would, Mr. Powell submits, be to recognise that the product Bali, both in the Reformulation I formulation in issue, and also in the Reformulation II formulation, which Customs accept falls within CN heading 2206, in terms of bulk, weight of alcoholic product and value, derives more from made-wine than from spirits. He relies on the General Rules for the Interpretation of the Harmonized System, particularly the reference under rule 3(b) in paragraph (VIII), for the guidance as to what factor determines the essential character of a mixture product, to which we have referred at paragraph 48 above.
57. He contends that it is immaterial to the objective classification rules that the majority of the volume of alcohol in the product Bali in the Reformulation I formulation is provided by spirits. The factors suggesting the predominance of the made-wine are, in his submission, more powerful in determining the essential characteristics of the product. He submits that the filtration process in the manufacture of the product Bali is designed to remove taints and flavour compounds. It does not remove all the wine flavour and he submits that the product remains a made-wine. He accepts that the Appellant uses a made-wine base because of the excise duty implications and that there are very little marketing advantages to using made-wine.
58. He adds that the "vague wording" on the bottle is not determinative, when the factors associated with bulk, weight and value are properly considered.
59. He made anecdotal reference to a case called Smirnoff Mule considered by the WCO in or about 2001 where a product resulting from a mixture of fermented alcohol and added spirits was classified as falling within 2208. This was the case, he told the Tribunal, which had given rise to the amendment to the HSEN to classification 2206, introducing the proviso that the product must retain the character of a fermented beverage.
60. He stated that the composition of the product Bali in the Reformulation I formulation is weighted far more in its essential character towards a classification within 2206 than either the Apfelroyale or the Smirnoff Mule product. He pointed out that labelling was not a factor addressed in either of those cases.
61. Mr. Thomas, for Customs, invited the Tribunal to uphold the BTI for the reasons given by Mr. Palmer in correspondence. He submitted that it is appropriate and correct to take account of the wording of the HSEN to heading 2206, including the proviso to which we have referred in paragraphs 49, 50 and 59 above. In any case, he says, the HSEN does no more than clarify what is already apparent. Where a product is a mixture of a fermented beverage and a spirit it falls (by GRI 2(b)) prima facie to be classified within both CN headings 2206 and 2208. Neither of these CN headings describes a mixture product such as the product Bali (in any formulation). Therefore, as is the case with all mixture products, the GIRs are engaged to reach a decision whether a mixture product is essentially a fermented beverage strengthened by a spirit, or a spirituous beverage diluted with a fermented beverage. If the mixture product has does not have (or has lost) the character of a fermented beverage, it has (or has assumed) the essential character of a spirit.
62. He cites the factors relied on by Customs in making the BTI, namely that although the bulk of the product Bali in the Reformulation I formulation is derived from the wine base, most of the alcohol is obtained from the spirit content. In addition, he relies on the decision of Customs in making the BTI that the product tastes and smells more like a spirituous beverage than a fermented beverage. He submits that taste is a relevant factor (and not to be dismissed on grounds of subjectivity), citing Gijs van de Kolk – Douane Expéditeur BV v Inspecteur der Invoerrechten en Accijnzen (Case C-233/88). He submits that, besides taste and smell, the presentation and marketing of the product Bali points to its classification as a spirituous beverage.
63. He emphasised that it was the Appellant's own evidence that as part of the process of manufacture of the product Bali, there is a centrifugal filtration process applied to the base wine to remove any "off flavours" developed at the fermentation stage. This achieved a neutralised taste for the fermented alcohol, which, in effect, as the Appellant had submitted, tastes no different from the spirits added to strengthen the product.
64. Mr. Thomas submitted that the reliance placed by the Appellant on the classification of the product Bali in the Reformulation II formulation as within CN heading 2206 is wholly misconceived, because the exercise engaged on by Customs in issuing the BTI, and the exercise in which the Tribunal is presently engaged, are both, he submits, limited to assessing the objective characteristics of the product to be classified, i.e. the product Bali in the Reformulation I formulation.
65. He further stated that the correctness (as a matter of strict classification law) of the classification of the product Bali in the Reformulation II formulation as within CN heading 2206 is not conceded. Nevertheless, he assured the Tribunal that there was no question of revisiting the classification of the product Bali in the Reformulation II formulation.
66. Mr. Thomas submitted (inconsistently, we note, with the arguments advanced by Mr. Palmer in correspondence on which he also placed reliance – see: paragraph 61 above) that GIR 3 is not engaged, because the product Bali in the Reformulation I formulation falls squarely within heading 2208. This is the position, he said, having regard (as is required) to the relevant HSEN, which shows that a fermented beverage which has been fortified with spirits only remains classified in heading 2206, where it retains the character of a fermented beverage. This is not the case with the product Bali in the Reformulation I formulation and, therefore, the heading 2206 is not applicable, and it falls to be classified in heading 2208 without reference to GIR 3. He referred the Tribunal to a document emanating from the European Commission entitled "FISCALIS project group on the classification of alcoholic beverages – Discussion document – April 2005)" which was supportive of his submissions.
67. He pointed out, as a matter of fact, that when the product Bali was reformulated to the Reformulation II formulation, the amount of rum in the mixture did not change. The spirit which was removed in the reformulation was a neutral spirit, and he submitted that that was a significant fact in suggesting that the essential character of the product Bali in all formulations was that it tasted of rum (and coconut).
68. He submitted that the Apfelroyale case could be distinguished from the present case, because there the product remained, as a matter of essential character, a cider – albeit, as he said, a "souped up" (alcoholically fortified) cider. The cider had not lost its apple-tasting characteristic. He submitted that the product Bali is not a British fortified wine.
69. He submitted that the change in the HSEN by the addition of the proviso was not significant, because the HSEN does no more than clarify and explain the scope of the classification headings, and does not alter them.
70. He relied on the evidence of taste and smell, that the product Bali tasted and smelt of rum and coconut and submitted that that was consistent with the way in which the product was marketed, mentioning the format of the label.
71. As to the law, Mr. Thomas pointed out that the Excise Directive puts much more emphasis on the strength of the alcoholic content of a beverage, whereas the CN is more concerned with the type of beverage. These are two parallel concepts. He submitted that a correct application of the classification procedure leads to the conclusion that the product Bali in its Reformulation I formulation is classifiable as a spirit within CN heading 2208.

Our Decision

72. The issue for the Tribunal's decision is whether the product Bali in the Reformulation I formulation falls within article 16 of the Excise Directive (92/83/EEC) as an "intermediate product" or whether, instead, it falls within article 20 of the Excise Directive as "ethyl alcohol". If the product is properly classified within CN heading 2206 ("other fermented beverages"), it falls within article 16 of the Excise Directive. If, on the other hand, the product is properly classifies within CN heading 2208 ("spirituous beverages"), it falls within article 20 of the Excise Directive.
73. Therefore we must decide whether the product Bali in the Reformulation I formulation is a "fermented beverage" (of the same type as cider, perry or mead) or a "spirituous beverage". This test arises from the wording of the relevant CN headings – cf. Holz Geenen GmbH at [14]. In making this decision we must have regard to the objective characteristics and properties of the product Bali in the Reformulation I formulation (ibid.).
74. Clearly, the reformulation of the product, set out at paragraph 17 above, showing the bulk, weight and value of the base wine as being preponderant in the mixture of the product, is an important factor, which is an objective characteristic and property of the product, and which, by itself, would suggest that its essential characteristic is that of a fermented beverage.
75. We consider, however, that other factors which have been identified are also objective characteristics and properties of the product, and they also must be considered. These other factors are: the relevant alcohol content by volume; the manufacturing process involved, and, in particular, whether that impacts on the essential characteristic of the product as a fermented, or alternatively a spirituous, beverage; and the product labelling. We add also to these factors the taste and smell of the product to the extent that we are satisfied that these can be objectively assessed – rather than being the subjective opinion of the tester. The relevance of taste is borne out by the reference in the Apfelroyale Regulation to that product being "apple-tasting".
76. We note that the relevant alcohol by volume is accepted to be preponderantly derived from distilled alcohol in the product Bali in its Reformulation I formulation. This is in the proportion:– distilled alcohol 63: fermented alcohol 37 (see: paragraph 18 above).
77. We accept the submissions of Mr. Thomas regarding the manufacturing process involved. We regard it as significant that, according to the Appellant's own evidence, as part of that process there is centrifugal filtration applied to the base wine to remove any "off flavours" developed at the fermentation stage and that this achieves a neutralised taste for the fermented alcohol. The Appellant agrees that the fermented alcohol, after the filtration process, tastes no different from the spirits added to strengthen the product (see: paragraph 63 above).
78. The product labelling, as we have said at paragraphs 30 and 31 above, is evidence that the product Bali is marketed as a drink with characteristics designed to evoke a tropical relaxed lifestyle and a tropical taste of rum and coconut. We do not regard the wording as "vague" in context as Mr. Powell contended (see: paragraph 58 above), but indicative that the product is a beverage of the character of rum, which is a spirit. We consider that the fact that product labelling apparently played no part in the Apfelroyale classification (or in the Smirnoff Mule case) is irrelevant. We agree with Mr. Powell, however, that this factor is not determinative, only adding that no single factor is determinative of the issue we have to decide.
79. As to the taste and smell of the product to the extent that these can be objectively assessed, we note the Appellant's submission that rum and coconut are the major contributors to the essential organoleptic characteristics of the product Bali, in all its formulations, and that if they did not remain constant, consumer and customer complaints would arise, due to variability of the product (see: paragraph 24 above). We have accepted Mr. Palmer's evidence that the product Bali in the Reformulation I formulation, and also in the Reformulation II formulation, tastes and smells of rum and coconut (see: paragraph 29 above). We accept Mr. Thomas's submission that the fact that the amount of rum in the mixture did not change when the product Bali was reformulated in the Reformulation II formulation supports the conclusion that objectively the product is required, for commercial reasons, to taste of rum and coconut, and that we can safely conclude that it does so (paragraphs 67 and 70 above). The above matters, taken together, suggest to us that an objective assessment of the taste and smell of the product would be that it tastes and smells of rum and coconut, and we so find This conclusion enabled us to proceed without the results of an organoleptic test by CCFRA.
80. The factors of: relevant alcohol content by volume; the manufacturing process involved; the product labelling; and the taste and smell of the product to the extent that we are satisfied that these can be objectively assessed; all point towards a conclusion that the essential characteristic of the product is that it is a spirituous beverage.
81. We go on to consider what use to make, in the interpretation of the CN headings, of the HSEN to classification 2206 (see: paragraph 49 above). The HSEN in its original (pre-2001) form, without the final proviso, stated in effect that fermented beverages remained classified in heading 2206 when fortified with added alcohol (or when the alcohol content has been increased by further fermentation). The proviso was added to ensure that this applied only where the fermented beverages so treated retained their character as fermented beverages. In order to gain assistance from this HSEN, Mr. Powell must persuade the Tribunal that the product Bali in its Reformulation I formulation is indeed a fermented beverage (which is, of course, the Appellant's case). However we understood Mr. Powell to be making a further point in reliance on the fact that the proviso was not present in the HSEN when the Excise Directive was adopted (on 19 October 1992). This point, as we understood it, was that the absence of the proviso at that time is an indication that CN heading 2206 is to be interpreted for Excise Directive purposes as including a fermented beverage fortified with added alcohol even if the fortification caused the beverage to lose its character as a fermented beverage. We cannot accept this argument. We consider it much more likely that the proviso was added to the HSEN to clarify its meaning rather than to change it. On this basis, even if the proviso is ignored, the HSEN will help the Appellant only if we agree with its submission that the product Bali in its Reformulation I formulation is a fermented beverage. This is where Mr. Thomas takes issue. He says that, having regard to the preponderant weight of the factors we have considered, which point to the product being a spirituous beverage, the only conclusion to be drawn is that it is not a fermented beverage. In the result, we agree with Mr. Thomas.
82. We do not, however, agree with Mr. Thomas's approach in oral submissions to the impact of the GRIs. He says that GIR 3 is not engaged because the product Bali in the Reformulation I formulation is not a mixture or composite goods made up of different components, because it is a spirituous beverage.
83. We have noted above (at paragraph 66 above) that this is a different approach to the GRIs from that taken by Mr. Palmer in correspondence and also relied on by Mr. Thomas (see: paragraph 61 above). Mr. Palmer's approach was to accept that the product Bali in the Reformulation I formulation was a mixture of a fermented beverage and a spirit and therefore GRI 3(b) is engaged so that the eventual classification of the product is "as if [it] consisted of the material or component which gives [it its] essential character", which he thought was the character of a spirituous beverage.
84. We prefer the approach taken by Mr. Palmer in correspondence to the alternative approach advanced in oral submissions by Mr. Thomas. We consider that the product Bali is indeed a mixture made up of different components and that therefore GRI 3(b) is engaged. The centrifugal filtration process applied to the fermented base-wine does not, in our judgment, deprive the base-wine of its character as a fermented liquor. Therefore inasmuch as the whole issue in this case is which of the two types of alcohol gives the product its essential character, it would in our judgment be perverse to hold that the product was not a mixture of different components. Thus we return to the question of whether the fermented alcohol base or the distilled alcohol used in fortification gives the product Bali in its Reformulation I formulation its essential character.
85. We have rejected the Appellant's argument that we should exclude consideration of (or give little weight to) the product labelling and taste and smell. We turn now to the two other main arguments advanced by the Appellant in seeking to persuade us that the essential character of the product Bali in the Reformulation I formulation is that it is a fermented beverage. They are: that it is a stronger case than the Apfelroyale case, where the product was ruled to fall within CN heading 2206, and that it is not materially different from the product Bali in the Reformulation II formulation, which Customs accept falls within CN heading 2206.
86. We accept that there is the distinction between this case and the Apfelroyale case which Mr. Thomas has suggested. The beverage in that case was apple-tasting and made from cider (with added apple flavouring and alcohol). The conclusion that that product was, as a matter of essential character, a fortified fermented apple beverage does not, in our judgment, compel, or even predicate, a conclusion in this case that the product Bali in the Reformulation I formulation is a fortified fermented British wine beverage. We have found that, objectively assessed, the product tastes of coconut and rum, and that it is put up as a tropical alcoholic drink with rum and coconut flavour. It would, in our judgment, be inconsistent with this conclusion to hold that the product has the essential character of fortified fermented British wine, notwithstanding the fact that the preponderant bulk of product is derived from the fermented base-wine base. In saying this, we take into account that the centrifugal filtration process applied to the fermented base-wine certainly goes a long way towards negating the fermented base-wine's character as a fermented base-wine (even if it does not negate that character altogether).
87. We also accept Mr. Thomas's submission that it would be misconceived to approach the decision of the issue before us by way of a comparison of the product Bali in the Reformulation I formulation, with the product in the Reformulation II formulation. Our function is to determine the essential character (in terms of GRI 3(b)) of the product in the Reformulation I formulation. The agreed classification of the product in the Reformulation II formulation is irrelevant to this determination. That agreed classification may (or may not) be wrong – and generous to the Appellant – as Mr. Thomas hinted in oral submissions. As to that, we make no comment. It is sufficient to say that the agreed classification of the product Bali in its Reformulation II formulation, when that classification does not have the authority of a Commission Regulation (of the status of that in the Apfelroyale case), is neither determinative, nor even persuasive, in the context of the decision we must make in relation to the product in the Reformulation I formulation.
88. We do not endorse the test applied by Mr. Palmer in considering whether to uphold the BTI (see: paragraph 33 above). We recognise that it is reasonable for Customs to have a policy to determine on a practical basis, and as an administrative matter, where to draw the line between beverages classified under CN headings 2206 and 2208 respectively, and we do not criticise the test applied by Mr. Palmer as representing an unfair policy. Nonetheless we take note that Mr. Thomas did not support the policy decision taken in relation to the product Bali in the Reformulation II formulation as being correct in law, and it must follow from that that Customs do not assert that the test applied by Mr. Palmer achieves results in all cases that are consistent with the law in this area. The approach we have adopted has not been to formulate a test which could be applicable in all doubtful cases, but (following the approach taken in the General Rules for the Interpretation of the Harmonized System under rule 3(b) – paragraph (VIII)) we recognise that "the factor which determines essential character will vary as between different kinds of goods" and that the correct CN classification as a matter of law will be determined by an examination of all relevant objective characteristics of the goods on a case by case basis.
89. The examination which we have carried out in this case, taking into account all the objective factors to which we have made reference above, has led us to the conclusion that the distilled alcohol used in fortification of product Bali in the Reformulation I formulation gives it its essential character, and that therefore the product in that formulation is properly classified under CN heading 2208 as a spirituous beverage. Duty is accordingly attracted under article 20 of the Excise Directive. Therefore we uphold the BTI, the assessment and the refusal of the claim for repayment of duty, and dismiss the appeal.

JOHN WALTERS QC

CHAIRMAN

RELEASE DATE: 23 June 2006

MAN/2003/7012