Invicta Foods Ltd v The Commissioners for HMRC

The appellant, Invicta Foods Limited (“Invicta”), imports various meat products from outside the European Union. These include uncooked chicken breasts (“the Product”) which it imports from Brazil. In response to an application by Invicta, HMRC issued what is called a Binding Tariff Information (GB 120015154) (“the BTI”) on 20 October 2010 pursuant to article 12(1) of Regulation 2913/92 which established the Community Customs Code (“CCC”). A BTI remains binding on the customs authorities as against the importer who requests it unless and until revoked or annulled under article 9 or article 12(4). The BTI classified the Product under code 16023211 in Chapter 16 of the Combined Nomenclature (“CN”). This is the name for the system used to classify imported products for customs duty purposes and to impose a common customs tariff on imports from outside the EU. It is based in part on the Convention on the Harmonised Commodity Description and Coding System (“HS”) to which the EU is a signatory and is composed of divisions and sub-divisions of several thousand groups of goods each of which is allotted an eight-digit code. The CN and the HS are very similar (although the HS uses six-digit codes).

The common customs tariff is provided for by Council Regulation (EEC) 2658/87 of 23 July 1987 and each year the European Commission produces by regulation a revised version of the CN incorporating any amendments which have taken place since the last version took effect. The CN includes a description of the goods in each category which is supplemented by Explanatory Notes to the Combined Nomenclature (“CNENs”) that are published by the Commission. These are not legally binding on the courts of the Member States but have been treated by the CJEU as highly persuasive in relation to any issue about classification: see Intermodal Transports BV v Staatssecretaris van Financiën (C-495/03) [2005] ECR 1-8151.

On 12 May 2011 the BTI issued in 2010 was revoked by HMRC and the Product was re-classified as falling under Chapter 2 of the CN. This produced a higher rate of customs duty. Invicta appealed challenging both the legality of the revocation and the correctness of the new classification and on 10 January 2014 the First-tier Tribunal (Tax Chamber) (“FtT”) allowed the appeal holding that the revocation was unlawful and that the correct classification of the Product was under Chapter 16. On 19 January 2016 the Upper Tribunal (Tax and Chancery Chamber) (David Richards LJ) allowed HMRC’s appeal (see [2016] UKUT 1 (TCC)) in relation to both the revocation and the classification issues. Invicta now appeals with the leave of this Court limited to the classification issue.

In terms of classification the contest is between Code 02071410 in Chapter 2 and Code 16023211 in Chapter 16. Code 02071410 includes:

“Meat and edible offal, of the poultry of heading 01.05, fresh chilled or frozen

- of fowls of the species Gallus domesticus

[…]

- cuts and offal, frozen:

- cuts:

- boneless.”

Code 16023211 covers:

“Other prepared or preserved meat, meat offal or blood:

[…] Of fowls of the species Gallus domesticus:

- containing 57 % or more by weight of poultry meat or offal:

- uncooked.”

The Product is uncooked chicken which has been treated by being vacuum tumbled in a non-cure brine solution made up of water (73.12%), salt (10.00%), sugar (8.13%), dextrose (5.00%) and phosphates (E 450/E 451) (3.75%). The rate of application was 8% solution to 92% of raw chicken breast and the solution was fully absorbed by the Product. Invicta’s case is that the application of this process means that the Product is “prepared meat” within the meaning of Code 16023211. It is common ground that the Product is not “preserved meat”. The interpretation of the classification is assisted by Explanatory Note 1 to Chapter 16 which states:

“This Chapter [i.e. Chapter 16] does not cover meat, meat offal, fish, crustaceans, molluscs or other aquatic invertebrates, prepared or preserved by the processes specified in Chapter 2 or 3 or heading 05.04.”

In the FtT HMRC contended that the Product had been prepared or preserved by one of the specified processes but the FtT rejected that and there was no challenge to that finding in the Upper Tribunal. But it is also necessary to consider Additional Note 6(a) (“AN6(a)”) to Chapter 2 which states:

“Uncooked seasoned meats fall in Chapter 16. ‘Seasoned meat’ shall be uncooked meat that has been seasoned, either in depth or over the whole surface of the product, with seasoning either visible to the naked eye or clearly distinguishable by taste.”

The CNENs to AN6(a) state:

“Salt is not considered to be a seasoning within the meaning of this additional note.”

It was common ground before the Upper Tribunal that the Product can only fall within Chapter 16 if it is “uncooked seasoned meat” within the meaning of AN6(a). It is also agreed that the Product had not been flavoured with seasoning visible to the naked eye. The only issue therefore (which is determinative of the classification) is whether the seasoning that was used is “clearly distinguishable by taste”. If it is not then the Product falls within Code 02071410 under Chapter 2.

In order to understand how what one might consider should be a simple question of fact has led to a second appeal to this Court it is necessary to say a little more about the background to the revocation of the BTI and the re-classification of the Product under Chapter 2. The relevant facts are set out in some detail in the decision of the FtT at [26]-[62] but the following summary is sufficient to describe how the issue on this appeal has arisen.

Invicta applied for the BTI on 30 September 2010 and provided a description of the Product and the details of the brine solution which I referred to earlier. The BTI was issued on 20 October 2010 which described the Product as:

“A frozen boneless, skinless chicken breast preparation. It is 92% of raw chicken with an 8% seasoning. The seasoning consists of water, salt, sugar, dextrose and phosphates.”

The Product had until then been classified under Chapter 2 in the UK but Invicta considered that this was at odds with its classification in other Member States. HMRC raised various requests for further information in response to the application for a BTI including whether the meat had been impregnated throughout with salt or was changed in appearance. Photographs were asked for and supplied. Invicta provided HMRC with a power point presentation which notified the viewer that the seasoning had been fully absorbed and that, as a consequence, the appearance of the Product had changed.

On 23 March 2010 HMRC notified Invicta that the information so far supplied was insufficient for the purpose of classifying the Product. The letter said:

“From the information provided we do not consider the product to be classifiable within Chapter 16 as the ingredients of the brine solution do not necessarily exclude it from Chapter 02. If you wish to continue with this classification request please supply an analytical taste test results of the chicken before and after the solution has been added.”

The request for a taste test was clearly linked to the requirement in AN6(a) that the seasoning should be “clearly distinguishable by taste”. But it is also significant that the test involved a comparison between treated and untreated samples of the Product. Invicta therefore instructed Leatherhead Foods Research (“Leatherhead”) to carry out an appropriate test or tests and to produce a report. For this purpose, Leatherhead was supplied by airfreight from Invicta’s supplier in Brazil with two samples of the Product, one with the seasoning and the other in its original raw state.

Between April and July 2010 discussions took place between Invicta, HMRC and Leatherhead about the testing format. The tests were carried out and on 12 August 2010 Leatherhead produced its report. It is not necessary for the purposes of this appeal to consider the method of testing in any detail. The samples were cooked and then tasted by a group of trained assessors. All samples were tested in triplicate using the unseasoned samples as a control. The conclusions set out in the Leatherhead Report were:

•

“Statistically significant differences between the treated and untreated samples were found in terms of mouthfeel (texture), flavour, aftertaste and after feel.”

•

“The changes in mouthfeel show clearly that the mouthfeel/textural characteristics of the treated sample are very difference [sic] from that of the control. The nature of the changes is consistent with the treated sample as being perceived as a more tender product.”

•

“The changes in flavour and aftertaste again show clearly that the flavour of the treated sample is very different from that of the control. The changes show increased flavour levels of the overall flavour and individual components (salty, sweet and savoury) in the treated sample.”

•

“One afterfeel attribute (bits in teeth) was significantly lower in the treated sample, which shows different breakdown characteristics in the mouth.”

•

“Our overall conclusion is that the treated chicken sample differs substantially from the control sample in terms of the wide range of textural and flavour characteristics.”

Following the submission of the Leatherhead Report the BTI was issued and a refund claim by Invicta for the period up to 31 August 2010 was approved. But in May 2011 Invicta was informed by HMRC that the BTI had been revoked. It appears that HMRC took the view that the use of sugar and salt did not amount to seasoning and did not therefore take the Product outside Chapter 2. HMRC also considered that the testing process was defective because no sample of the prepared Product had been sent to HMRC for testing or had been preserved from the samples tested by Leatherhead.

In June 2011 Invicta requested HMRC to review its decision on revocation of the BTI and on 4 August 2011 HMRC wrote confirming the original decision. In this letter HMRC wrote:

"When the classification for the chicken breasts was considered, although you provided an informative taste test report, no sample was taken for analysis by HMRC which would have confirmed whether or not the product in question is prepared outside the scope of Chapter 2 of the Combined Nomenclature (CN).

When classifying the chicken to CN heading 1602, the chicken was excluded from Chapter 2 because of the sugar content.

…

As no analysis of the goods was carried out in relation to BTI GB 120015154, the BTI was reviewed.

Based on the evidence provided both at the time of your BTI application and at this review stage, I have concluded that there is no reason to exclude the Brazilian, frozen boneless chicken breast preparations from Chapter 2 and that there is sufficient evidence to suggest that classification to heading 1602 is incorrect.”

HMRC now accept that sugar (as opposed to salt alone) can constitute seasoning for the purposes of AN6(a) so that if the effect of the brine solution was to make the seasoning “clearly distinguishable by taste” then the Product is classifiable under Chapter 16. But the dispute on classification now centres on how that fact or condition falls to be established and indeed what is meant by the phrase “clearly distinguishable by taste”. Invicta say that the condition or standard is met and can be established for the purposes of AN6(a) if on a comparison with an untreated sample of the same Product the treated sample is clearly distinguishable by taste. The Leatherhead Report showed that the samples which they tested met this standard. But HMRC contend that this is the wrong approach and that the requirement for the Product to be clearly distinguishable by taste has to be established on a tasting of the seasoning alone and not simply by comparison with an untreated control sample from the same source.

As formulated the issue has become therefore one of principle in terms of what the AN6(a) test requires to be shown in order for the Product to be classified under Chapter 16. HMRC’s position is that the seasoning is only “clearly distinguishable by taste” if the Product has a perceptible flavour created by its seasoning which can be detected by testing the Product on a stand-alone basis. This, Mr Bremner says, is consistent with classification under the CN being determined by reference to the objective characteristics of the Product as presented for customs clearance. It avoids the difficulties inherent in identifying and obtaining suitable control samples; provides for ease of verification; and promotes legal certainty. The draftsman of AN6(a) must be taken to have been concerned with the inherent qualities of the Product itself and not simply with whether it is different from another product. Although the Leatherhead test did show that there was an appreciable difference in taste between the control sample and the seasoned Product, it did not prove that on an objective basis the seasoning was “clearly distinguishable by taste”.

It is obvious that the danger in adopting a purely comparative method of tasting flavour is that it may be possible to detect some difference in taste between a seasoned and an unseasoned product even though on a blind tasting the seasoned product might not appear to be seasoned at all. The use of the phrase “clearly distinguishable by taste” (my emphasis) in AN6(a) suggests that the minimal (although) appreciable difference in taste that might be revealed on a comparative tasting of that kind would not be sufficient to meet the standard prescribed by the 6(a) test. But the word “distinguishable” (or even the French text “perceptible”) does suggest some form of comparison with the Product in its normal state and if that is right then a comparative method of testing is not ruled out provided that what it establishes is that the seasoning is, objectively speaking, clearly distinguishable by taste.

By the end of the appeal Ms Shaw QC did not contest the need for the Product to meet the minimum objective standard prescribed by AN6(a). Her contention was that at the time when the Product was presented for clearance there was no regulation which laid down a single method of testing and that the Leatherhead test did in fact establish that the seasoning was clearly distinguishable by taste. The relevant findings of the FtT are contained in the summary of the conclusions of the report set out earlier and in [164]-[165] of its decision:

“164.

In our view, the seasoning of the Product was clearly distinguishable by taste. We saw no justification for the suggestion that seasoning had to make the chicken taste different from the normal taste of cooked chicken. Seasoning does not need to be so obtrusive that it masks or changes the underlying flavour. Seasoning is generally intended, in our view, to enhance or accentuate the underlying flavour of meat in order to make it more flavoursome and appetising to the taste. It is true that in some cases seasoning can be so strong that that it overpowers the underlying flavour of the meat or that it imparts a different flavour of its own, but we do not think that it is a necessary for it to do so in order to be "clearly distinguishable by taste."

165.

Mr Stokes' evidence was that he did not consider that chicken in its natural state would pick up the salt and sweet characteristics of the Product. Furthermore, we also accept Mr Stokes’ evidence that consumers expected chicken to taste as it did after the tumbling process, indicating that the Product had a significantly different taste. Mr Stokes’ evidence was that the samples that were sent to Leatherhead for the purposes of testing were of the same specifications as those in respect of which the repayment claims were made. The Leatherhead report stated clearly that the Product had significant differences in taste and aftertaste as regards saltiness, sweetness and savouriness (“umami”) from the control sample. This, in our view, established that the Product was seasoned in a way that the seasoning was clearly distinguishable by taste. The report’s methodology of comparing the Product against a control sample seemed to us an eminently sensible approach to adopt. Indeed comparing the taste of the Product with a control sample was the approach suggested by HMRC and the actual methodology used was finalised after discussion with HMRC by Mr Stokes.”

It seems to me that the FtT was right to say that seasoning can be of many kinds not all of which produce a distinctive flavour of their own independent of whatever may be the natural flavour of the product which is seasoned. The purpose of some seasoning (such as the brine solution in this case) will be to enhance the natural flavour of the product in its original state. It is not suggested that a product seasoned in this way is not included in the description of “seasoned meats”. In those circumstances the FtT was entitled to conclude on the basis of the findings in the Leatherhead Report that the seasoning was “clearly distinguishable by taste” unless those findings were rendered inadmissible by the terms of the CN including AN6(a) or by some other relevant regulation. In its decision (at [40]) the Upper Tribunal rejected the suggestion that it could be inferred from what is said in the Leatherhead Report that the seasoning would have been found to be clearly distinguishable by taste had the Product been subjected to a stand-alone test. I do not disagree with that but the issue for us on this appeal (as it was for the Upper Tribunal) is whether a finding that the Product had reached the prescribed standard of distinctive taste is admissible if it is based on a comparative test of the kind which Leatherhead carried out.

The essence of Invicta’s case is that until 2014 it was open to an importer to establish the requirements of AN6(a) by either a stand-alone or comparative test. Although not estopped by this fact, we know that the use of a comparative test in this case was approved at the time by HMRC. Whichever method is used it must be established that the seasoning is clearly distinguishable by taste. But, as explained earlier, that finding was open to the FtT unless the method of testing on which the findings were based was impermissible. The stated requirement that the seasoning should be clearly distinguishable (i.e. perceptible or discernible) by taste does not, Invicta submit, say anything about how that fact is to be established.

AN6(a) was considered by the CJEU in Gijs van de Kolk-Douane Expéditeur BV vInspecteur der Invoerrechten en Accijnzen, AmersfootCase C-233/88 [1990] ECR I-265 where the validity of the Note was challenged (also in relation to the import of seasoned chicken breasts) on the ground that it was not in conformity with the criteria set out in the CN. The argument focussed in part on whether the criteria set out in AN6(a) of “seasoning ... clearly distinguishable by taste” was consistent with the case law of the CJEU which requires goods to be classified on the basis of their objective characteristics and properties at the time of clearance. The CJEU said in its judgment:

“11.

The note at issue purports to specify what is to be understood by seasoned meat or meat offals within the meaning of the abovementioned Explanatory Notes of the Customs Cooperation Council. In order to do so it lays down criteria for classification based on the sensory analysis of the goods.

12.

Those classification criteria comply with the case-law of the Court of Justice according to which, in the interests of legal certainty and ease of verification, goods must be classified on the basis of the objective characteristics and properties of products which can be ascertained when customs clearance is obtained (see, inter alia, the judgment of 16 December 1976 in Case 38/76 Luma v Hauptzollamt Duisburg ((1976)) ECR 2027, paragraph 7).

13.

In order to apply criteria such as those set out in the note at issue, there are objective techniques of sensory analysis which have recently been developed and for which national and international standards have been laid down, for example, Standard DIN 10954 in the Federal Republic of Germany and Standard ISO 4120, which the International Organization for Standardization, Geneva, submitted to its member committees in 1982. As the Commission has pointed out, those methods of analysis allow, in particular, the goods as presented for customs clearance to be accurately assessed for the four basic flavours - sweet, acid, salt and bitter - which can be detected, even at very low levels, by a statistically significant population.

14.

It is true that in its judgment of 17 March 1983 in Case 175/82, cited above, the Court took the view that a criterion as subjective as taste may not be used to assess the seasoning of meat and that Heading 16.02 of the Common Customs Tariff must be interpreted as meaning that it also includes poultry meat to which salt and pepper have been added even if the pepper can be detected only microscopically.

15.

However, it must be pointed out that that judgment was delivered in different circumstances from those in the present case; there was no provision in a regulation on the interpretation of the Common Customs Tariff and at the time of the national authorities' inspection of the goods Standard ISO 4120 had not yet been devised.

16.

In those circumstances, it must be concluded that in laying down the abovementioned criteria, the Council did not infringe the principle of legal certainty and did not otherwise exceed the discretionary power conferred on it in that field.”

Ms Shaw points out that ISO 4120 referred to in the judgment as one of the objective methods of sensory analysis available to enable goods such as the seasoned chicken breasts to be accurately assessed in accordance with AN6(a) is in fact a comparative method of testing sensory differences between two samples. Although the Court was not directly seized of the issue we have to consider, its reliance on ISO 4120 is inconsistent, she submits, with HMRC’s position on this appeal.

The Upper Tribunal was not impressed by this. It accepted Mr Bremner’s submission that the CJEU in van de Kolk was concerned only to make the general point that advances in sensory analysis meant that even a criterion based on taste would comply with the need for the classification of imported goods to be based on the objective characteristics of the products when presented for clearance. The reference did not go further than that. The Upper Tribunal was more impressed by Mr Bremner’s argument that for goods presented for clearance to have to undergo a comparative form of testing with untreated products would be unworkable in practice. There would be issues about whether the control sample could properly be said to be comparable with the seasoned Product and, in the case of meat such as poultry imported from a large country such as Brazil, variants in taste according to the type and place of origin of the chicken would be inevitable.

I accept that these are potential difficulties but they seem to me to go to whether the particular comparative test proposed is acceptable rather than to whether a comparative method of testing is ever permissible. In the case of seasoned meat whose classification depends on taste, the testing process cannot by its very nature be conducted by a customs officer at the precise moment the Product arrives for customs clearance. Some organised process of testing will always be required. There is nothing in AN6(a) which in terms dictates the form which that should take and problems about verification and the availability of suitable control samples in particular cases does not in my view justify the adoption of a general rule that a comparative system of testing should be impermissible. As the FtT pointed out, seasoning can include products (as in this case) designed to enhance the existing flavour of the imported meat and it is difficult to see how they could be fairly tested for distinctive taste even on an objective basis without some form of comparison against the untreated Product. So far as it goes, the decision in van de Kolk seems to support that view.

What is common ground on this appeal is that as a result of the Commission Implementing Regulation (EU) 1362/2013 of 11 December 2013 (“the CIR”) the prescribed and therefore only permissible method for the tasting of poultry meat in order to establish whether the seasoning is clearly distinguishable by taste is the stand-alone method. Part II of the Annex to the CIR sets out the instructions on the method of testing in which the test sample is “a representative portion of the meat destined for consumption”. It is not a comparative test involving an untreated control sample. The CIR applies to imports after 8 January 2014. Ms Shaw, whilst accepting the effect of the CIR, relies upon it as indicating that its purpose was to regulate a previously unregulated position about methods of testing in which other methods such as comparative testing were not only used but also permissible. The CIR is, she emphasises, not merely clarificatory for the future but, more importantly, also regulatory. Its purpose is apparent from recital (2) which states:

“In order to ensure that customs authorities apply a uniform approach for the purposes of customs classification, it is necessary to lay down methods for determining whether or not uncooked poultry meat is seasoned within the meaning of Additional Note 6(a) to Chapter 2 of the Combined Nomenclature.”

On this issue both sides seek to gain some support for their arguments on the permissible methodology under AN6(a). In response to Ms Shaw’s submission that the CIR was the starting point for the regulation of testing, Mr Bremner says that the CIR supports HMRC’s case about the pre-January 2014 position by confirming that a stand-alone test is the only permissible method of testing for the purpose of AN6(a) and by providing clarification that this methodology is the only one which is compatible with the approach to classification enshrined in the CN. The reference in the recitals to AN6(a) confirms the need for a uniform approach to achieve a fixed rather than variable standard. The fact that customs authorities may in the past not have adopted a uniform approach to testing is not, he says, to be treated as confirmation of the legality of that practice.

To support his submissions about the effect of the CIR Mr Bremner referred us to the Opinion of Advocate General Jacobs in Deutsche Nichimen GmbH v Hauptzollamt Düsseldorf Case 201/99 [2000] ECR I-2701 in which the CJEU was asked to consider whether satellite receivers had been included in the definition of television receivers in heading 8528 of the CN before Commission Regulation No. 884/94 came into effect and (it was accepted) did clearly so include them. In [42] of his Opinion the Advocate General touched on the question whether assistance in relation to the earlier period could be obtained from the treatment established by the new regulation. He said:

“Further support may be derived from the subsequent classification under that heading by the Commission in Regulation No 884/94. Whilst a regulation classifying goods under a particular tariff heading or subheading, being of a legislative nature, cannot have retroactive effect, I pointed out in my Opinion in Siemensthat the form of such regulations, which generally state (as is the case with Regulation No 884/94) that classification is determined by the provisions of the general rules for the interpretation of the nomenclature and by the wording of the relevant headings and subheadings, suggests that the legislature takes the view that the classification enacted in fact follows from the legislation already in force.”

The Upper Tribunal considered that it obtained no real assistance from the CIR in relation to the legal position before it took effect. But I think that the recitals are of some assistance, although clearly not determinative. Although the CIR is clearly intended to introduce a measure of regulation to taste testing for the purposes of AN6(a), any possible inference that the content of the CIR represents the correct application of Chapter 2 of the CN as qualified by the AN depends on how one reads recital (2). It seems to me that the requirement to lay down a uniform method of testing for the future indicates that customs authorities were until then free to adopt or rely upon different methods of testing in the past. The recital does not indicate that this was incompatible with the proper application of AN6(a) and I see no reason to give it that meaning or to draw that inference from the other background materials I have referred to. In short, the question whether the seasoning was clearly distinguishable by taste could, until the coming into effect of the CIR, be determined by either a stand-alone or a comparative method of testing provided that the process used conformed to proper testing standards. It is not part of HMRC’s case on this appeal to criticise the methodology adopted by Leatherhead if otherwise legally permissible.

I would therefore allow Invicta’s appeal and restore the decision of the FtT.

I agree.

I also agree.