United Kingdom VAT & Duties Tribunals Decisions

You are here: BAILII >> Databases >> United Kingdom VAT & Duties Tribunals Decisions >> R Twining and Company Ltd v Revenue & Customs [2007] UKVAT V20230 (05 July 2007)
URL: http://www.bailii.org/uk/cases/UKVAT/2007/V20230.html
Cite as: [2007] UKVAT V20230

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

R Twining and Company Ltd v Revenue & Customs [2007] UKVAT V20230 (05 July 2007)

20230

Value added tax – zero-rating – whether product a powder for the preparation of beverages – whether product a preparation or extract of milk – meaning of milk considered – Value Added Tax Act 1994, Schedule 8, Part II, Group 1

LONDON TRIBUNAL CENTRE

R TWINING AND COMPANY LTD Appellant

and –

THE COMMISSIONERS FOR HER MAJESTY'S REVENUE AND CUSTOMS

Respondents

Tribunal: Dr. David Williams (Chairman)

Mr. J. Robinson (Member)

Sitting in public in London on 16 and 17 April 2007

James Henderson of counsel, instructed by Ernst and Young for the Appellant

Owain Thomas of counsel, instructed by the Solicitor to Her Majesty's Revenue and Customs for the Respondents

© CROWN COPYRIGHT 2007

DECISION

The Appellant company ("Twinings") is appealing against a decision of the Respondents ("the Commissioners") refusing to accept that a product ("the Product") in a product range marketed as "Max for Milk" ("the Product Range") is a product entitled to be zero-rated for value added tax ("VAT") purposes on sale. The Product Range is a range of powders designed for dissolving into water or milk to produce a tasty but also healthy and nutritious drink for younger consumers. The Product Range is linked both in its presentation and in its manufacture to the well-known range of Ovaltine drink products.

The issue

The immediate context of the disagreement is that the Product in question is one of three products in the Product Range. The three products are the versions of Max for Milk with banana, chocolate and strawberry flavours. The banana flavoured product is accepted by Twinings to be subject to the standard rate of VAT. The chocolate flavoured product is accepted by the Commissioners to be subject to the zero rate. The tribunal is asked to decide on the status of the strawberry flavoured Product. It is obvious from this context that the case is a borderline case. The way in which part of the Product Range is subject to VAT, and part zero-rated, is yet another illustration of the complexity of the United Kingdom's value added tax laws. But this is a case that does not involve any direct application of European Union value-added tax laws beyond issues of general principle. The detailed decision is to be made by reference to United Kingdom law alone.

The evidence

The tribunal heard sworn evidence for Twinings from Michael Friedrich Baumann, head of the central research and development team of the group of companies of which Twinings is a member. He was a most impressive witness. It also heard sworn evidence from Elizabeth Sarah Edwards, a marketing manager at Twinings with responsibilities, among other matters, for the Product Range. It received agreed bundles of documents including full details of the manufacturing and marketing of the Product. It was invited to, and did, examine the websites about the Product Range. It was given samples of the Product and invited to, and did, conduct a formal testing and tasting of the Product. And it watched a short marketing video film about the Product Range. The Commissioners offered no evidence.

The Product

The Product is marketed as "Max for Milk STRAWBERRY". It is one of three similar products marketed together in the Product Range. As noted above, the VAT status of the other two products in the product range was agreed and was not consistent.

In some of the marketing materials aimed at schools and caterers the Product Range is described as "Ovaltine Max for Milk". It can also be read that way on the labelling of the Product Range. Mr Henderson drew our attention to the fact that the Commissioners publicly stated that Ovaltine is zero-rated as a milk-based drink. Paragraph 9.9.2 of the relevant Customs and Excise Manual at Section 9, Beverages, states:

"Drinks which are substantially based on milk, such as milk shakes, Ovaltine, Horlicks, or Complan are zero-rated as preparations of milk.

The tribunal considers that Mr Henderson is entitled to draw the tribunal's attention to the inclusion of Ovaltine in that list as that is also a product produced by the group of companies of which the Appellant is a company. No third party is therefore involved. Mr Thomas indicated that he could not deny that but had no instructions as to its accuracy. The tribunal does not take that issue further, as the use of a name or description clearly does not carry any binding significance in VAT law. Nor does any guidance issued by the Commissioners do more than indicate the Commissioners' thinking. Nevertheless, the tribunal bears this in mind as part of the context within which the Product Range was developed. This decision must be guided, and guided only, by the legislation and clear authority.

The law

As in all borderline cases, it is important to note precisely under what provisions the parties argue the product should fall. The general principle, under European Union law, is that all products are subject to value added tax at a positive rate.

It is common ground that European Union law accepts, and section 30(2) of the Value Added Tax Act 1994 ("VATA") enacts, that specified supplies of goods are zero-rated in the United Kingdom.

The extent of the zero-rating for food is specified in Group 1 of Schedule 8 to VATA. The Schedule is in entirety an exception to the general rule that value added tax should be charged on the supply of products. Group 1 sets out the general exception for food followed by a heirachy of exceptions to the exception and of exceptions to those exceptions. The specific rules in question here are:

"GROUP 1 – FOOD

"The supply of anything comprised in the general items set out below, except- …

(b) a supply of anything comprised in any of the excepted items set out below, unless it is also comprised in any of the items overriding the exceptions set out below which related to that excepted item.

General Items

Item No

1 Food of a kind used for human consumption

…

Excepted items

Item No

…

4 Other beverages (including fruit juices and bottled waters) and syrups, concentrates, essences, powders, crystals or other products for the preparation of beverages.

…

Items overriding the excepted items

…

6 Milk and preparations and extracts thereof

NOTES:

(1) "Food" includes drink

…

(6) Items 4 to 6 of the items overriding the exceptions relate to item 4 of the excepted items.

Mr Henderson argued for Twinings that the Product is a preparation or extract of milk within override 6 of excepted item 4 to general item 1 and so is zero-rated. Alternatively he argued that the Product is not a beverage within excepted item 4 and so is within general item 1 and is zero-rated.

Mr Thomas argued for the Commissioners that the Product was a powder for the preparation of beverages but not milk or a preparation or extract thereof. It was therefore within excepted item 4 to general item 1, but not any of the overrides, and so is standard-rated.

European Union value-added tax law is, Mr Thomas, submitted, relevant to this analysis because the European Court of Justice has laid down the approach to be taken to any exception to the general rule that all goods are subject to a positive rate of value added tax. In his submission it is clear law that exceptions must be strictly construed as exceptions to the general regime of VAT. They must also be construed by reference to the purpose of any exception and the scope permitted in European Union law for such exceptions. In European terms, zero-rating is exemption with credit. Exemptions are allowed only within the terms of the EC Second VAT Directive, Directive EEC/67/228, article 17:

"for clearly defined social reasons and for the benefit of the final consumer".

The tribunal accepts that approach. Bearing it in mind, the tribunal considered that it should look first at the arguments of the parties about whether the Product is a powder for the preparation of a beverage, and, if it is, whether it is a preparation or extract of milk.

Beverages

It is common ground that the Product is "food of a kind used for human consumption" when that descriptor is read with the note that ""food" includes drink".

It is also common ground that the Product as sold is not as such a beverage.

Mr Henderson accepted that that was not the critical point.

The parties disagreed about whether the Product is properly described as a "powder … or other product for the preparation of beverages".

Mr Henderson invited the tribunal to read the decision of the tribunal in Alpro Ltd, Decision 19911. In that case the tribunal found that flavoured soya milk products were not beverages. The tribunal had the benefit of expert evidence from Professor Sean Strain, Professor of Human Nutrition of the University of Ulster. That evidence included the following (at paragraph 22):

"Professor Strain agreed that there was no scientific definition of a beverage but he gave it as his opinion that a beverage would contain water and not much else except sugars and that it would therefore slake the thirst as well as hydrating the body without making a consumer feel uncomfortably full."

In discussing the law, the tribunal commented (at paragraph 46):

"had we been asked to decide if milk is a beverage we would have found that it is not for the same or at least similar reasons to those which we will give below for finding that soya milk is not a beverage."

Those reasons are set out at paragraphs 55 to 58. The tribunal first indicated that it accepted that "beverage" is an ordinary English word and that it followed the guidance in the Bioconcepts case. It then found that the products were not marketed in the same way as beverages are marketed, and that they could be - but were not much - used in cooking. Finally, it concluded that as a matter of impression and fact the products were not beverages.

Mr Henderson also drew attention to the recent decision of the tribunal in Unilever Best Foods UK Ltd, [2007] UKVAT V20016. In that case the tribunal, chaired by Sir Stephen Oliver QC, also found that a product was not a beverage. The products were "Knorr Vie shots", a range of fruit and vegetable combinations with a substantial vegetable content and therefore not fruit juices. The tribunal noted in particular "the Bioconcepts test" as it was referred to in a decision of the Irish Revenue Appeals Commissioner, namely:

"that notwithstanding the Oxford English dictionary definition of "beverage" meaning drink, it is not used in the sense of meaning all drinkable liquids. Its meaning … covers drink or liquors that are commonly consumed. This is the primary meaning in the Oxford English Dictionary. Liquids that are commonly consumed are those that are characteristically taken to increase bodily liquid levels, to slake the thirst, to fortify or to give pleasure."

The tribunal reviewed other tribunal decisions and concluded that "the Bioconcepts test is workable and produces an intelligible set of results." It then analysed the products in that case with the test in mind. This tribunal finds the discussion of the test and its background in Unilever Best Foods UK Ltd are most useful in this case, but the application of the test to the specific product is of limited assistance.

Finally, Mr Henderson relied on the decision of Warren J in Kalron Foods Ltd v Commissioners for Her Majesty's Revenue and Customs, [2007] EWHC 695 (Ch), a decision released shortly after the decision in Unilever Best Foods UK Ltd. The issue was again whether products, in this case "Zumo Fresh Blend" products sold at Zumo Fresh Smoothie Bars, were beverages. The tribunal had agreed with the Commissioners that the products were beverages. Warren J found, applying the well known test in Edwards v Bairstow, that the tribunal was entitled to conclude the case by saying that the appellants had not met the burden of proof on it to show that the products were not beverages. That is the test to be met by Twinings in this case.

The argument by which the appellants failed in Kalron was that the products were a soft form of food but not a beverage. This was argued on a different basis before Warren J. In his thorough and helpful judgment, he accepted that the ordinary meaning of "beverage" is a question of fact and therefore a matter on which the appellants were required to advance evidence to establish their contentions. He subjected the extensive argument put to him for the appellants that "beverages" has a special meaning to close forensic examination in the light of the various court and tribunal decisions. He concluded:

"It is at that point that I wish to say no more about whether "beverage" is to be given a special meaning in the VAT legislation. None of the tribunals which have looked at the meaning of the word in the past have done so: they all regarded themselves as applying the ordinary meaning of the word… I reject [counsel's] submission that beverage has a special meaning in Group 1 different from its meaning as a matter of ordinary language."

Mr Thomas, in replying on the question of the meaning of "beverage", adopted a much simpler approach than Mr Henderson. He took the basic approach that the argument for Twinings simply cannot be right. This case was about a product that was clearly a powder. Further, on the evidence, the powder was used to by adding it to milk. He also relied on the evidence given as showing that the resulting milk-based drink is a beverage by reference to the Bioconcepts test. In this case the resulting liquid from mixing the Product as instructed was as a matter of fact both a drink and a beverage.

Milk and preparations and extracts of milk

The wording of the overriding item "milk and preparations and extracts thereof" is important. The express inclusion of "milk" as number 6 in the list of overriding items to the excepted item "other beverages", along with tea and also cocoa and coffee confirms, in the view of this tribunal, that milk is now for these purposes a beverage. The tribunal is conscious that in taking this view it is disagreeing with the tribunal in Alpro. At paragraph 49 that tribunal commented:

"Whilst having difficulty in ascertaining the full reasoning behind the zero rating provisions in social policy terms we are quite satisfied that milk is not referred to in the overriding provisions because it is a beverage but rather at most to avoid any suggestion that it might be."

This tribunal respectfully begs to differ. That tribunal appears to have reached that conclusion partly because it did not consider milk a beverage as a matter of fact. This tribunal, as a matter of fact - as we explain below - disagrees. If the wording of the law is approached from our assumption rather than that of the tribunal in Alpro, then our view is that one cannot find the same satisfaction in reaching that tribunal's conclusion. Indeed, in our view, it argues to the opposite conclusion. The famous campaigns in the past encouraging everyone to "drinka pinta milka day", the provision of free milk for all schoolchildren at school, and also of free milk for younger children through local welfare programmes, clearly gave milk the status that that tribunal referred to as "near iconic". But we remain entirely unpersuaded that milk was nonetheless not a beverage in fact.

The tribunal also notes from the Commissioners' publication to which it was referred that milk was originally accepted as a food and not a beverage. (paragraph 9.9.1 of the Customs and Excise Manual). That also suggests that milk powder used for babies is also food and not a beverage (paragraph 9.9.3). The approach of the tribunal in Alpro Ltd appears to follow that line of reasoning. But it was one reached without being fully tested. It must now be tested by reference to the accepted Bioconcepts tests. That is discussed further below.

Mr Thomas' reply also accepted Mr Henderson's main argument, under which he accepted that the Product, when mixed, produced a beverage. Having accepted that, Mr Henderson argued that it was entitled to exception from the general treatment of beverages because the Powder, as sold, is a preparation or extract of milk.

In presenting this argument, Mr Henderson based his argument on the use of the terms "preparation", "extract" and "milk" in the override as being used in their ordinary senses. He submitted that the dictionaries did not help with either "preparation" or "extract" and that neither terms was capable of precise definition.

In his submission, the Product is a "preparation" or "extract" for these purposes because it has a significant milk content and is intimately connected with the preparation of a milky drink. The significance of the extent of the link may be measured by the amount of milk in the Product, its importance to the Product, and by the design of the Product to produce a nutritious milky drink.

At the end of the hearing the tribunal asked the parties to consider whether the meaning of the term "milk" should take into account the European and United Kingdom laws under which milk was sold for human consumption. The question posed by the tribunal was whether "milk" should be said to bear only its ordinary meaning, whatever that might be, when both European and British laws prevented the description "milk" being given to products that did not meet strict legal criteria. The tribunal invited the parties to comment on whether the tribunal should adopt an approach of considering the meaning of "milk" in its context here as something that can be supplied to consumers legally only within those laws. The tribunal also raised the issue that, on an initial view, "milk" properly covered a wide range of forms of milk from full cream Channel Islands milk to skimmed milk.

The parties responded to the tribunal's direction with thorough surveys of the laws presented as written arguments after the close of the hearing. The tribunal is grateful to both parties for the thoroughness of their responses. Both parties concluded, after detailed surveys of the relevant European and United Kingdom legislation, that the various precise definitions of "milk" in that legislation did not inform the use of the term in the United Kingdom VAT legislation. Although the submissions were made separately, they came to a clear common conclusion that necessitated no further argument or submissions.

The tribunal accepts that common approach. The parties' analyses are therefore not repeated in this decision. But the tribunal does draw from the surveys one point that it considers important in reaching its findings of fact. "Milk", even when that term is limited to cows' milk, is not a single product with a precise content or specification. Nor is it simply the milk that is taken from a cow. As European Council Regulation 2597/97 of 18 December 1997 provides, "milk" as sold in the United Kingdom can have its fat content added, reduced or standardised. It can be enriched with milk proteins, mineral salts and vitamins. And it can have its lactose (a natural sugar present in milk) reduced. Correspondingly, "preparations or extracts thereof" must include a similarly wide range of milk-based products, including products based on only some of the ingredients of the milk as taken from a cow and products to which other nutrients have been added, or both.

The evidence

Mr Henderson adopted the approach of counsel in Rivella (UK) Ltd, VATD Decision 16328 of asking us to adopt the "four factors" test from Snapple Beverage Corporation,VATD Decision 13690 as a fraework within which to examine the evidence. Rivella is a case about the override that concerns the tribunal here, while Snapple concerned the override about tea. The tribunal in Rivella accepted that the approach taken in Snapple was a sound approach also to the override relating to milk, and this tribunal is content also to accept it as a sound approach. The four factors are:

a. ingredients

b. manufacturing process

c. appearance and taste

d. marketing and packaging

This tribunal therefore turns to the evidence and arguments of fact based on that evidence presented to it within that framework. In doing so, it follows the approaches set out above but draws no prior specific conclusion from the factual analyses or findings of the courts or tribunals in any of the cases cited to it.

Ingredients

The tribunal accepts Mr Baumann's expert evidence about the content and production of the Product. He was a most impressive witness with a detailed knowledge of the products of the group of companies of which Twinings is a member. He is a graduate of both German and British universities in food technology and science. He clearly had not only the necessary scientific and technical knowledge, and an impressive knowledge of the technical aspects of the Product Range, but also a detailed knowledge of how the Product was treated under the relevant tax and other laws of all the European Union states into which the Product Range is marketed.

Mr Baumann explained the history of the Product as an Ovaltine related product aimed at the children's drink market. Like Ovaltine, it was designed as a healthy product with a good nutritional profile but it also had to meet the marketing requirements of colour, taste and mouth feel. Ovaltine, we were told and accept, is made of malt, eggs, milk, barley and cocoa. The challenge with the Product Range was to remove the flavour of malt and also, for some of the Product Range, including the Product, to remove the cocoa. At the same time, the nutritional value of the Product Range was to be a key point. In particular, the Product Range avoided using sugar as a major element in the ingredients, and relied instead for the carbohydrate element on maltodextrin. The Product Range also avoids the use of artificial colours and sweeteners.

The tribunal accepts that the Product is manufactured to a precise standard recipe. The main ingredients of that recipe are:

maltodextrin 42%

"crunchy base" 36%

fructose 12%

strawberry flavour 3%

xanthum gum 3%

No other individual ingredient, including vitamin mix and natural colouring, constituted more than 2% of the recipe or is directly relevant to the arguments of the parties. The "crunchy base" had the following main ingredients:

Skim milk concentrate 41%

Malt extract 25%

Granulated sugar 18%

Skimmed milk powder 8%

Glucose syrup 4%

Full milk powder 1%

No other individual ingredient formed more than 2% of the "crunchy base" or is directly relevant to the arguments of the parties.

This information is given on the jar labels of the Product as follows:

"INGREDIENTS

Maltodextrin, malted powder 36% (skimmed milk concentrate, barley and malt extract, sugar, skimmed milk powder glucose syrup, vegetable oil, whole milk powder, yeast, salt, magnesium carbonate, vitamins, glucose, flavouring), fructose, flavouring, thickening agent (xantum gum), calcium carbonate, magnesium carbonate, calcium phosphate, colour (beetroot red), vitamins (C, E, niacin, A, pantothenic acid, B1, B6, B2, folic acid, biotin, B12), zinc carbonate.

Contains: milk, barley, gluten

Suitable for vegetarians."

The tribunal notes that this list also features yeast. Yeast is one of the items (together with meat and egg) which, together with any preparation or extract, form override 7, to the same effect as override 6 dealing with milk. Twinings put no weight on that point and the tribunal, noting that yeast is a very minor element of the ingredients, accepts that approach.

Mr Baumann gave the following evidence in his statement on the question of the milk content of the Product, and of the use of the Product with milk and water:

"5.3 In view of the instructions from the marketing team that the product was to be healthy, we used only milk products. That was the more expensive choice…

5.4 I understand that the issue in this case is whether the Strawberry Max for Milk is a preparation or extract of milk. In my opinion, it is difficult to define precisely what is meant by a preparation or extract of milk. However, Strawberry Max for Milk does contain milk ingredients for the specific purposes that I have outlined and I would consider it to be a preparation or extract of milk.

5.5 Analysing fresh milk, its content in terms of total dry matter (ie dairy solids) is between 10-12%. The amount varies depending on the type of milk: skimmed milk is about 10% dry matter; and whole milk about 12%. Other factors have influence eg, genetic variant of cow, time of yar, geographical location. Strawberry Max for Milk actually contains 11.2% dairy solids just like fresh milk. If one were to prepare Strawberry Max for Milk with water, the milk content of the resulting mix would be the same as fresh milk. Indeed, 20g of Strawberry max for Milk mixed with 200ml water would result in the Max for milk contributing 2.2g of dairy solids, which compares to 200ml fresh milk containing between 2.0g and 2.4g dairy solids. Strawberry Max for Milk mixed with water does actually produce a form of reconstituted milk. However the preference would be to use milk instead of water when preparing Strawberry Max for Milk since it is a smoother taste."

In reply to questions on this evidence, Mr Baumann indicated that he considered the product to be a preparation rather than extract of milk. As a drink, he considered it not to be a thirst quencher because – unlike drinks such as fruit juices – a litre of milk would make you feel full. That was much the same effect as a glass of milk. The milk was present in the Product, however, not only for taste and mouth feel but also for its protein content.

Mr Baumann also gave a full explanation of the nutritional, cost, and market-led reasons for the other main ingredients of the Product. The presence of maltodextrin was explained as a carbohydrate that was a preferred energy source for many sports products. But it had the advantages for the Product of good overall mouth feel and being rather bland in taste so able to support fragile tastes such as strawberry. The tribunal also saw the detailed recipe of the Banana Max for Milk product in the Product Range but not of any other related product. Some comparisons were made with the banana product but the tribunal did not find that these assisted its fact finding.

Manufacturing process

The tribunal was told by Mr Baumann of the manufacturing process for Max for Milk. One element of that process is directly relevant to the tribunal's analysis, namely the extent to which the powders and essences were deliquescent or hygroscopic and were used or marketed at above or below their natural water contents. The tribunal drew on its own expert knowledge in asking these questions. A deliquescent product is one that naturally takes liquid from the air and then dissolves in the liquid naturally absorbed in this way. A hygroscopic substance is one that tends to absorb moisture from the air, but not to the extent of a deliquescent substance. Both kinds of substance will naturally change their compositions by increasing their water content when exposed to the air without any intervention.

Mr Baumann confirmed that the powder used in the Product is highly hygroscopic. This is because of several of the ingredients in the Product, including the milk powders and the fructose. Making the crunchy base involved vacuum evaporation to preserve flavour. The milk powder used comprised the solid elements of the milk after drying. There was some confusion in the figures used in some of the product analysis because of this, as Mr Baumann accepted. The dry powder weight of the relevant milk powders was 11.2% of the total, while the same powders when retaining or gathering water to a natural balance, without outside intervention, constitute 17.69% of the product.

The tribunal takes the view as a matter of fact that it should take into account the naturally hygroscopic nature of the Product and the level of liquid naturally acquired. While it is manufactured and sold subject to vacuum conditions, those conditions are lost the moment the Product is opened for use. The Product, including in particular the milk element within the Product, will then gain water from the atmosphere without outside intervention. The tribunal considers that as a matter of fact it should take this into account when considering the ingredients and manufacture of the Product.

Appearance and taste

On the reverse side of the container below the statutory information and key selling points and above the ingredients are directions as follows:

DIRECTIONS

Mix 3-4 teaspoons of Ovaltine Max for Milk Strawberry with a little cold milk to create a smooth paste. Pour on 200 ml of cold milk, stir well and enjoy!

The tribunal was asked to test and taste the Products and we did so during the lunch break. The Product was tasted after a light lunch on a warm day. It is a light coloured powder with a scent reminiscent of strawberries and cream. It was tested both with water as directed on the jar and also with semi-skimmed as a comparison. We were also invited to try it with full cream milk. We did not want to do so after those tastings because we both felt that the product when mixed with semi-skimmed milk tasted like flavoured full cream milk, and we felt that we would learn nothing by that further test. We therefore did not test it with full cream milk. We had no skimmed milk available, none having been made available, with which to test it.

It is relevant to note that both members of the tribunal came from families where ordinary milk was drunk regularly, and we both drank milk regularly as a matter of choice as children and still do. We were both also familiar with Ovaltine and similar drinks.

Both of us separately formed the view that the water-based mix was at the same time both milky and creamy in flavour. Nonetheless the effect, for instance the mouth feel, was of a water-based drink not a milk-based drink. We noted that there was a distinct difference in that sense between the Product as prepared with water and the Product as prepared with semi-skimmed milk. Both of us also agreed that the drink was not a thirst-quenching drink for us on a warm day such as that on which the test took place. But we considered it could well be a thirst-quenching drink for a child and when it was not so warm. The overall general impression was, as we had anticipated before the test, that of a sort of strawberry-jam flavoured Ovaltine.

Marketing and packaging

Both Mr Baumann and Mrs Edwards gave evidence relating to the marketing and packaging of the Product and the Product Range. Mrs Edwards' evidence confirmed and to some extent expanded that of Mr Baumann. It was valuable corroborative evidence, but overall the tribunal felt it added only a little to the tribunal's understanding and evaluation of the Product.

The evidence of the witnesses is accepted as establishing that the Product Range was aimed at a younger market than the traditional British market for Ovaltine, now an older market. It was also an attempt to break away from the "bedtime drink" perception of Ovaltine in the British market, a perception not shared in other Ovaltine markets elsewhere in Europe or more globally. At the same time it sought to carry forward the approach that Ovaltine was a nutritious product and not sugar based as were some of its rivals. This was a focus for marketing the Product Range direct to school canteens and similar outlets. The Product Range met the various nutritional requirements of such outlets while market rivals, being largely sugar based, did not. But it had presented them with a marketing problem at the same time because sugar based products are cheaper to produce than products with the Ovaltine base materials.

The packaging of the Product has similarities to the packaging of the main Ovaltine brand. The Product is sold in 300g plastic containers designed like jars with screw tops. This is said to be 15 servings, so each serving uses 20 grams of powder. The labels are, as Elizabeth Edwards described them, busy. The trade marked product name "Max for Milk" is the most prominent part of the labelling. It is designed to evoke ideas of milk, for example using a milk bottle instead of the letter "i" in "Milk". The registered mark "Ovaltine", with an orange background, is also prominent. The Product is then described as a strawberry flavour powder mix. This is positioned over a picture of the top of a glass containing a thick milk shake drink. In smaller print below is added "just add cold milk".

Our decision

The tribunal found a full answer to this appeal difficult. The first stage in our answer was, by contrast with the second stage, relatively easy despite the apparent contrary authority of the tribunal in Alpro Ltd. In this tribunal's view the Product is clearly within the category of foods, including drinks, described as "powders" … for the preparation of beverages". It is therefore within excepted item 4 and so outside general item 1 in Group 1 of Schedule 8. As such, it is subject to VAT at the standard rate unless one of the overriding items applies.

The tribunal follows the authorities set out above in regarding "beverage" as an ordinary English word. The Product is unquestionably a powder. And the tribunal finds that the powder is used to produce a beverage. Or, putting it the other way round, Twinings failed to satisfy the tribunal under the civil burden of proof that it was not a powder for the preparation of a beverage. The tribunal bases that on the Bioconcepts tests:

"drinks or liquors that are commonly consumed … those that are characteristically taken to increase bodily liquid levels, to slake the thirst, to fortify and to give pleasure."

The tribunal in the Alpro decision indicated that, if asked, it would not have found milk to be a beverage (paragraph 46). Formally, as that reference suggests, it was not asked, and its decision is not technically authority for its answer. It was concerned on the facts only with soya milk. But we also indicated above that this tribunal did consider that milk is a beverage in the common meaning of that term in the VAT context. In this tribunal's view the Bioconcepts test approach leads to this answer.

There appears to be an unstated assumption in the Alpro decision that the consumers against whom the Bioconcepts tests are to be applied are adult consumers. That may be so of soya milk. But is that a valid assumption for ordinary cow's milk? The policies for distributing milk mentioned above suggest otherwise. At its height the policy provided, or aimed to provide, free cow's milk daily to every child in the land. And it should not be forgotten that milk is the natural product that has evolved to provide both fortification and the relevant bodily liquid levels for baby and young mammals of all kinds. Nor can it be denied that it provides human babies with pleasure. That point is particularly relevant for the Product Range as it was designed, marketed and packaged expressly with children in mind. The personal experience of the tribunal's members is also that for some youngsters milk is used to slake thirst. Again, perhaps, that may not be so true of older adults who will more readily feel full than an active child or teenager after drinking milk.

The Product is, of course, not milk. It is a drink, or beverage, that may be created by mixing the powder with milk (as recommended on the packaging and marketing material) or water. The tribunal discounts the use of the product as powder without adding liquid to it (or the reverse). Even when taken with water the powder creates a milk-style drink. It is essentially the same kind of beverage as many others made by adding liquids to powders: instant coffee, Ovaltine, drinking chocolate, some forms of herbal drinks.

The tribunal finds it harder to decide whether the powder is properly described as a "preparation or extract" of milk. It has expressed its view above that "milk" while being a common English word with no specific technical meaning as used in this override nonetheless covers a range of products rather than a single substance. It notes and puts weight on Mr Baumann's view, as the person responsible for the research and development that led to the product, that the Product "contains milk ingredients for the specific purposes that I have outlined and I would consider it to be a preparation or extract of milk (paragraph 5.4 of his evidence)." It notes that the published view of the Commissioners is that a beverage falls within this category if

it is "substantially based on milk" (paragraph 9.9.3 of the Customs and Excise Manual) or it is a mixed drink that has the "texture and nature of a milky drink" (paragraph 9.9.5).

Taking the constituent elements of the analysis used above, the tribunal finds as follows. The Product is one of a Product Range packaged and marketed in a uniform manner. But one of that Range is zero-rated while another is standard-rated. That test is inconclusive. The taste and feel of the Product when mixed with both milk (as instructed on the packaging) and water (as tested by the tribunal) is that of a milky drink. That element of the analysis clearly points to the powder being a preparation or extract of milk. In the tribunal's view it is correct to look at the ingredients and the manufacturing process together. This is because the Product as packaged contains less liquid than it would contain if left on its own by reason of its hygroscopic nature. That nature should be taken into account, in the view of the tribunal, in evaluating the extent to which the Product consists of milk or milk-derived substances. Accepting the evidence of the Appellants, the natural element of the Product that is milk-based (that is, with naturally acquired water) is about 18 per cent. Some of that is skimmed milk in which the natural solid levels are less than those of full milk. Of itself, that does not sound substantial. However, its effect on the Product, as noted from the tasting tests, is substantial. That is deliberate. Other elements in the ingredients were chosen as bland elements specifically selected not to affect that appearance, taste and mouth-effect.

Does the evidence satisfy the tribunal that the Product is within override 6 as being a preparation or extract of milk that, when mixed with milk or water, forms a beverage? The tribunal did not at first reach a clear answer, with one member taking the view that the Product fell on one side of the line and the other tending to take the opposite view. That is perhaps not surprising as the decision is one of impression and of its nature at the margins of the relevant categories. But the tribunal does not consider that it should decide the appeal by casting vote or reference to the burden of proof alone. There was considerable evidence about the Product and the Product Range. The tribunal took the view that it should reach a single decision on the facts. On balance, and after careful consideration - and reconsideration - of all the evidence, it reached the conclusion that the Product was shown sufficiently to be a preparation or extract of milk such as to fall within override 6.

The tribunal therefore reaches the conclusion that the Appellant's appeal succeeds on its primary argument, but not its secondary argument. The Product should be zero-rated.

CHAIRMAN

RELEASED: 5 July 2007

LON 2006/305