[2011] UKFTT 834 (TC)

TC01670

Appeal number: TC/2010/2568

CUSTOMS DUTY – what is the test for minimum protein content in high quality wheat – which samples relevant for test – decision wheat not high quality wheat – what rate of duty applies to medium quality wheat – decision in principle a retrospective application for preferential rates can be made

FIRST-TIER TRIBUNAL

TAX

FRONTIER AGRICULTURE LTD Appellant

- and -

THE COMMISSIONERS FOR HER MAJESTY’S

REVENUE AND CUSTOMS Respondents

TRIBUNAL: Mrs B Mosedale (Tribunal Judge)

Mr J Stafford (Tribunal member)

Sitting in public at 45 Bedford Square, London WC1 on 24 November 2011

Mr P McNeil, import administrator of the Appellant and Mr M Adams, both for the Appellant

Mr M Fell, Counsel, instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents

DECISION

1. Frontier Agriculture Ltd appeals against a review decision of HMRC dated 12 February 2010 upholding the imposition of a C18 demand note dated 4 September 2009. The C18 reclassified wheat imported by the Appellant from code to 1001 9099 12 to code 1001 9099 20 with the consequent increase in import duty of £89,256.80.

2. The UK is a member of the European Union. Under the EC Treaty the UK is obliged to apply the customs rules of the European Community and this is enacted into our law by the European Communities Act 1972 s5. The tariff of duties and the classification of goods for the tariff is therefore a matter of Community law.

3. Under Community law, there are various import classifications for grain. The tariff classification for “wheat and meslin” is 10 01 with a number of sub-categories. Under “other”, “high quality common wheat and spelt” are categorised as 1001 90 99 12. Medium quality common wheat and spelt is categorised as 1001 90 99 20.

4. The question for the Tribunal was whether the wheat imported by the Appellant was correctly classified to the code for high quality wheat with a nil rate of duty; and if not, to what duty is it liable?

The facts

5. To a large extent the facts were not in dispute: the parties had an agreed statement of facts. Where the facts were in dispute we set out the reasons for our findings of fact. We find as follows.

6. The Appellant imported 1,075 metric tonnes of organic Kasakh origin milling wheat. It paid well over the normal threshold price for high quality wheat. This shipment was one of a number from the same supplier. The other shipments have been imported by the Appellant as high quality wheat and accepted as such by the UKBA. It is only in respect of this single shipment that a C18 has been issued.

7. The wheat was expensive as it was organic and had to meet the Soil Association’s requirements for organic wheat. It was important to the Appellant that it was of the right protein content for high quality wheat as not only had it to meet the standard required to qualify for the 0% duty rate, it had to meet the standard required by its customers who would reject it if it was not of sufficient protein content. It was therefore a term of the Appellant’s contracts with both its supplier and its customers that the wheat had sufficient protein content. To ensure that it did, the Appellant had the wheat tested before purchase, before loading and after docking.

8. Sample 1 was taken before the Appellant agreed to buy the wheat. The sample was taken some months before the grain arrived in the UK but was tested by the same method (the Kjeldahl method) as HMRC used and tested by the same organisation (Campden BRI) that tested the HMRC sample (see below) yet it yielded a much higher protein result (13.72%).

9. Samples 2 & 3 were taken at loading in Trabzon, only two weeks before the wheat arrived in Ipswich. HMRC did not accept that the testing of these samples was necessarily accurate as they did not know the conditions under which it was tested. Nevertheless, we find that the conditions under which it was tested were likely to be reasonably comparable to those undertaken by Campden BRI because the protein content of the wheat was of real commercial importance to the Appellant and it would not rely on a company to carry out the tests if it did not trust the result: exaggerated protein content would not have helped the Appellant when buying the wheat.

10. Sample 2 was tested by the Kjeldahl method and yielded a result of 13.48% protein content which was somewhat lower than Sample 1. Sample 3 was tested by the Dumas method some weeks after the sample was taken. It yielded a result of 13.64% protein content which was some 0.16% higher than the Kjeldahl test on Sample 2.

11. On docking at Ipswich, the grain was partially loaded onto lorries when SGS (acting for the Appellant) noticed that the shipment appeared to be infested with grain beetles and weevils. After five hours, involving negotiation with the supplier who had warranted it free of insect infestation, Frontier Agriculture decided nevertheless to accept the shipment and discharge recommenced.

12. At the time of discharge, SGS took a bulk sample from the imported wheat. SGS provided an UKBA officer with 10kgs of this grain which the UKBA officer separated into 3 samples. Each sample was placed in an air-tight bucket. One bucket was placed in a sealed bag and marked infested and given to the Appellant’s shipping agent as the trader’s sample. It was not fumigated. This was sample 5 in the below chart.

13. As Campden BRI, which carries out the tests on behalf of HMRC, would not accept the samples until the infestation was treated, the UKBA officer took the other two samples for fumigation. Fumigation took 7 days and required the buckets to be opened. After fumigation the two samples were resealed and despatched to Campden BRI. One bucket split in transit, although the bag remained sealed. This was sample 6 in the below chart and for the time being was left untested. Campden BRI ran the protein test on the contents of the intact bucket. This was sample 4 in the below chart.

14. As the wheat was organic, unlike the samples, it could not be fumigated. It was instead treated with diatomaceous earth. We accepted Mr McNeil’s evidence (which was unchallenged) that the Appellant sold the wheat to its customers who have not complained about the protein content of the wheat and who have paid for it as high quality wheat.

15. Samples 7 & 8 were taken from the cargo of wheat held by the Appellant and did not form part of the samples taken by HMRC. HMRC do not accept that the testing of these samples was necessarily accurate as they do not know the conditions under which it was tested. Both tests were carried out using the Dumas method and show similar results for protein content of 13.39% and 13.43%.

16. At a pre-hearing review on 6 August 2010, the Judge directed that the two untested samples, one being the spare Customs sample (sample 6) and the other being the trader’s sample (sample 5), should be sent to an independent laboratory for testing.

17. The results of the testing of all the samples is as follows:

When and where sample taken	When and where sample tested	Sample number	Test used	Test result*	Adjusted result Dumas to Kjeldahl
13 May 2009 Purchase of wheat in Turkey	Campden BRI	1	Kjeldahl	13.72	-
24 August 2009 Loading at Trabzon	SGS Istanbul 24 August 2009	2	Kjeldahl	13.48	-
24 August 2009 Loading at Trabzon	ATC Maidenhead 12 November 2009	3	Dumas	13.64	13.48-13.18
9 September 2009 Docking at Ipswich	Campden BRI 9 September 2009	4	Kjeldahl	13.12	-
	Premier 22 October 2010	5	Dumas	12.9	12.74 – 12.47
	Premier 15 November 2010	6	Dumas	13.5	13.34-13.05
Stock still held at Ipswich 4 December 2009	SGS Thurrock 8 December 2009	7	Dumas	13.39	13.23-12.94
Stock still held at Ipswich 4 December 2009	Frontier Agriculture 21 December 2009	8	Dumas	13.43	13.27-12.98

* the test results are adjusted for moisture content at 12%

Expert Evidence

18. We heard expert evidence from Mr Garstang. Mr Garstang holds a BSc in Agriculture and has worked for ADAS UK Ltd since 1967 and now holds the position of Principle Consultant. In that time he has been involved in the sampling and analysis of grains and feedstuffs and has been a principle adviser to MAFF and a High Court witness for DEFRA. His qualifications as an expert on the matters on which he was asked to opine were not questioned and we accepted him as an expert in these matters.

19. Mr Garstang’s evidence was that Dumas and Kjeldahl are two internationally accepted methods of testing the protein content of wheat, He gave evidence that studies have compared the two methods and the fairly consistent results of these studies are that the Dumas method will give higher readings than the Kjeldahl method, with a report in 2003 suggesting that the Kjeldahl test result will be about 0.43% lower than Dumas test result. This evidence was not challenged and we accept it as grounded in authority.

20. Mr McNeil’s submission was that the Kjeldahl method was outdated and the Dumas method is now used commercially. He also considered the Dumas method to be more accurate. Mr Garstang’s evidence was that the Dumas method was indeed preferred by the industry as it was much faster (it took a few minutes compared to a few hours) and required a smaller sample. It was also his unchallenged evidence that the Kjeldahl test is the one used by the International Association for Cereal Chemistry (“ICC”) because the ICC considered it the best test for identifying the point at which nitrogen becomes protein.

21. All parties were agreed that, now, HMRC allow traders to use the Dumas method to test the protein content of wheat. Mr Adams even suggested that HMRC’s change of policy was due to the concerns the Appellant raised with their industry body over this particular case. In any event, it was clear to us that HMRC did not allow the Dumas test to be used at the time of importation of the wheat in this case. It was HMRC’s case that even now the use of the Dumas test was concessionary and HMRC reserved the right to use the Kjeldahl test in cases of difficulty.

22. Mr Garstang’s evidence was the two methods tested nitrogen content as that had a direct correlation with protein content. His evidence was that some pests would eat the protein in the wheat and cited a report where nitrogen content reduced by 0.5% in a month in stored infested wheat. The degree of damage to protein content done by pests would vary between type of insect (not identified in this case) and how long the grain was left untreated. It was his evidence that fumigation itself would not affect the protein content of the wheat.

23. We accepted Mr Garstang’s evidence as he was an expert and he gave credible explanations for the views he held. We find that the Kjeldahl test (which was not disputed) is the test used by the ICC.

24. We accept the Appellant’s evidence (which was not challenged) that the protein content of wheat could vary throughout a cargo load. It would not necessarily be homogenous throughout. Mr Garstang agreed with this evidence and explained that that was the reason tolerances in the measurements were allowed.

Decision

The necessary protein content of wheat

25. Commission Regulation (EC) No 1249/96 (as amended) deals with, amongst other matters, the objective grading criteria and tolerance rates to be used in grading imported cereals for the purpose of the common customs tariff duties. It provided at Article 3 that:

“The quality standards to be met on importation into the Community and the tolerances allowed shall be those shown in Annex I”

26. The version of Annex I of Regulation 1249/96 in force at the time of the events in this appeal was the one contained in the amending Commission Regulation (EC) 2104/2001. In so far as protein content was concerned, it was identical to the previous versions of Annex I. It provided that high quality wheat would have a minimum protein content of 14% (with a moisture content of 12% by weight); medium quality wheat should have a minimum protein content of 11.5% (with the same moisture content). The permitted tolerance was 0.7%. In effect, wheat would be high quality wheat if it had a protein content of 13.3% or above at 12% moisture level.

The test for protein content in wheat

27. The first question to be resolved is what is the proper test to be used to determine protein content. Regulation 2104/2001 (referred to in paragraph 26 above) specified that in respect of the quality of wheat including the minimum protein content:

“The methods of analysis laid down in Article 3 of Commission Regulation (EC) No 824/2000 … are applicable.”

28. Article 3(5) of Commission Regulation (EC) No 824/2000 provided that:

“the standard method for determining the protein content of ground common wheat shall be that recognised by the International Association for Cereal Chemistry (ICC), the standards of which are laid down under heading No 105/2: ‘method for the determination of the protein content of cereals and cereal products;’

However, Member States may use any other method. In such a case, they must furnish the Commission with evidence of recognition by the ICC that the method in question gives equivalent results;”

29. This Regulation was repealed and replaced by Commission Regulation (EC) no 687/2008 with effect from 1 July 2008. It contained the same provision at article 5(e) and a provision that references to the repealed Regulation 824/2000 should be read as references to this Regulation.

30. We find that at the time the ICC approved method to determine the protein content of wheat was the Kjeldahl method. This is because Mr Garstang’s evidence was that the ICC approved method of testing is the Kjeldahl method and the Appellant did not challenge this evidence.

Was Sample 4 a representative sample?

31. HMRC’s case was that Sample 4 was the HMRC sample and the only one which HMRC consider matters. It was taken at the right time (at arrival in the UK) under the conditions dictated by the EU and tested to the appropriate standard by Campden BRI using the Kjeldahl method. The sample yielded a result of 13.12% protein content which was below the level required for high quality wheat.

32. The Appellant’s case is that the bulk sample drawn by SGS and/or the sample taken by the bulk sample and tested by HMRC (samples 4, 5 & 6) were not representative of the cargo of wheat as a whole and that this is evidenced by the rather different results of other tests of the protein content of that cargo. The results for all the other samples are above the 13.3% minimum protein quality. It is the Appellant’s case that, even if the infestation did affect the protein content, it could not have had such a dramatic affect and the real reason for the discrepancy was that the sample was not representative.

33. As part of its original case, the Appellant originally claimed that the sampling process must have been flawed possibly because the sample was mixed up with samples of the low quality wheat which was shipped on the same vessel but in different holds. The Appellant confirmed at the hearing that it no longer maintained this part of its case and we consider it no further.

34. Commission Regulation (EC) No 1249/96, referred to above, provided at Article 6 that:

“Representative samples shall be taken of every consignment of …common wheat of standard high or medium quality by the customs authority of the importing Member State, the provisions of the Annex to Commission Directive 76/371/EEC applying for determination of protein content, …..as definited (sic) in Council regulation (EEC) no 2731/75….”

35. Directive 76/371 established Community methods of sampling and was replaced by Commission Regulation (EC) no 152/2009. In summary, the EC set down the sampling methods which had to be used to determine protein content.

36. We find samples had to be taken and tested in accordance with the method dictated by Regulation 1249/96 which meant that HMRC as the “customs authority of the importing Member State” had to take the sample and carry out the test. We also note that Article 3 required the quality standards to be met “on importation” and only samples 4, 5 & 6 were taken on importation and, of these, only sample 4 was tested relatively contemporaneously.

37. In conclusion, the only sample on which the question of the wheat’s status as high or medium quality wheat could be determined was sample 4.

38. Article 6 of the Regulation 1249/96 did require that HMRC take a “representative” sample. It is the Appellant’s case that Sample 4 was not representative. Nevertheless, irrespective of whether that sample was in fact representative, we find as a matter of law that Annex 1 of Commission Regulation (EC) 152/2009 deems it to be representative as long as it was taken in accordance with the methods set out in that Regulation. This is because Annex I provides:

“Samples thus obtained [ie in accordance with the Regulation] shall be considered as representative of the sampled portions.”

39. As the Appellant no longer impugns the conditions under which the sample was taken it cannot make out its case that the sample was not representative: it is deemed to be representative. Therefore, its appeal against reclassification to code 1001 9099 20 must fail, as sample 4 did not have a protein content of 13.3% or above.

Was Sample 4 actually unrepresentative?

40. Nevertheless, although on our understanding of the law it is not relevant, as it was argued and as the case may go higher, we consider the Appellant’s case that, due to no one’s fault, Sample 4 must have been unrepresentative because it was so out of line with the earlier results for protein content of the same cargo of wheat.

41. We consider that two qualifications were necessary before we could consider the results shown in the table.

42. Firstly, we consider that in order to fairly compare the sample test results, bearing in mind Mr Garstang’s evidence that the Dumas test will in general deliver higher results than the Kjeldahl method, we need to compare like with like. We will therefore look at the Dumas test results as converted by Mr Garstang into results in his opinion likely to have been given by a Kjeldahl test on the same sample.

43. Mr Garstang worked on the basis of two variations. Firstly, he had noticed that samples 2 & 3 were taken at the same time and one was tested by the Dumas method and one by the Kjehdahl method. The difference between the results was 0.16%. He therefore applied a 0.16% reduction to the Dumas test results on other samples to arrive at an approximate Kjeldahl figure. In the table above the higher of these two approximations is the one arrived at by this method (shown in the last column).

44. The lower of the two approximations shown in the last column of the table is that arrived at by applying the 0.43% reduction given by an independent study to which we refer in paragraph 19.

45. Secondly, we consider that sample 5 should be ignored. Sample 5, taken at the same time as samples 4 & 6, resulted in a very low protein content out of line with the results for all the other samples. Mr Garstang’s evidence was that insect infestation was likely to lead to a drop in protein level within a few weeks, although he was unable to be more precise without knowing the particular type of insect in question. Sample 5 was not fumigated and left for a year before it was tested. As Mr Adams said, the bugs ate the protein. HMRC also relied on Mr Garstang’s evidence as it was their case that the insect infestation accounted for the drop in protein level of the entire cargo during shipment. We too accept Mr Garstang’s evidence. So we find that Sample 5 is an unreliable guide to the protein level of the cargo at the date it arrived in the UK.

46. Once the results shown in the table are considered in the light of these two qualifications, we find that although the protein content of the wheat before purchase and prior to loading is much higher (samples 1-3) than sample 4, that of samples 7 & 8, show lower protein levels than sample 4.

47. We find that the likely explanation is not that the sample was unrepresentative but rely on Mr Garstang’s evidence that the protein level was likely to have dropped in shipment. It is consistent with what we know. The eggs of the weevils must have hatched after loading which would explain why the Appellant’s agent did not notice the infestation at loading and why the protein content remained high (samples 1-3). However, once hatched, the weevils consumed protein in the wheat. This is consistent with the visible infestation two weeks’ later at docking. So at docking the protein content of the wheat was significantly diminished (samples 4 & 6). There was a measurable but slight further drop (samples 7 & 8) in protein content of the main cargo after this date which we presume was due to a delay in treating the wheat. For the one sample which was not treated (sample 5), the protein content had a further significant drop in level.

48. We also considered the Appellant’s case that it sold the wheat as high quality wheat to its customers, who have not rejected it. However, we find that the industry uses the Dumas method of testing and the Appellant’s undisputed evidence is that samples 6, 7 & 8 were (just) high quality wheat on the Dumas test, which is the test by which its customers would have tested the wheat. Therefore, this evidence is consistent with a finding that samples 6, 7 & 8 were not high quality wheat on the Kjeldahl test taking into account our finding of fact (paragraph 19) that on the whole the Dumas test delivers higher results than the Kjeldahl test.

49. Our conclusion is that the results for sample 4 were not an aberration. And (were it relevant) the Appellant’s case that it was therefore an unrepresentative sample cannot succeed. So we are satisfied that not only was the sample taken by HMRC deemed to be representative, it was in fact representative.

What of sample 6?

50. When assessing whether the wheat was of high quality wheat should we have regard only to sample 4 or is sample 6 also relevant? Sample 6 was taken by HMRC but tested about 12 months later at an independent laboratory using the Dumas method. Sample 6 was from the same bulk sample as sample 4, but tested over a year later though we were not given any evidence to suggest that protein content may have deteriorated bearing in mind that it had been fumigated with sample 4. The results of the test of Sample 6 under the Dumas method were a “pass” with 13.5% protein content at the right moisture content level.

51. We find that this sample was not treated in accordance with the Regulation (see paragraph 36) in that (a) it was not tested as quickly as possible and (b) it was tested in an independent laboratory by the Dumas method. In our view, for the reasons explained in paragraph 36, Sample 6 is therefore irrelevant.

52. Nevertheless, even if we were to accept the Appellant’s case that it was relevant, we consider that the result would have to be adjusted (for the reasons explained above in paragraph 19 and 43) to an estimated Kjeldahl method result. Such an adjustment gives a result in the region 13.34- 13.05% protein. Even if this result were to be given equivalent weighting with the result on sample 4, an averaging out of the Kjeldahl estimated results would be a “fail” at 13.195%. An average of the sample 4 and sample 6 results is also a “fail”. So even if we had found the test results of sample 6 to be relevant, which we do not, we would still find that the wheat was not high quality wheat at the time of importation.

Import duty applicable to product actually imported

53. The product actually imported was wheat with a protein content at import of less than 13.3%, but well in excess of 11.5% which was the minimum protein requirement for medium quality wheat.

54. Regulation 1249/96 provides at Article 6(3) that:

“If the analysis results show the imported wheat to be of a lower standard quality than entered on the import licence the importer shall pay the difference between the import duty applicable to the product shown on the licence and that on the product actually imported. ….”

55. Therefore, the Appellant should pay the import duty applicable to medium quality wheat. It is the Appellant’s case that, as the wheat was imported from Turkey, it is entitled to a preferential tariff of only €12 per tonne. HMRC’s view, however, is that such a tariff only applies where the Appellant had an import licence for medium quality wheat (whereas of course they held an import licence for high quality wheat) and therefore the default is that they must pay the normal duty on medium quality wheat at €95 euro per tonne. It is on that basis that the C18 was issued in the sum of £89,256.80.

56. Although it was not drawn to our attention at the hearing, we note that the Community Customs Code (contained in Council Regulation (EEC) No 2913/92) provides at Regulation 20 that the import duties shall be determined by both the normal import duties and the preferential tariff measures, and further, at Article 20(4) provides:

“Without prejudice to the rules on flat-rate charges, the measures referred to in paragraph 3(d), (e) and (f) [the preferential tariff measures] shall apply at the declarant’s request instead of those provided for in subparagraph (c) [the normal import duty] where the goods concerned fulfil the conditions laid down by those first-mentioned measures. An application may be made after the event provided that the relevant conditions are fulfilled.”

57. This Article therefore permits the Appellant to claim an applicable preferential tariff rate retrospectively. Our decision in principle is therefore that the Appellant is liable to pay the duty on the basis it imported medium quality wheat, but that the rate of that duty depends on whether or not its importation fulfilled the relevant conditions for the preferential tariff, which it is entitled to claim retrospectively.

58. We were not given sufficient information to determine whether or not the Appellant is actually entitled to the preferential tariff rate so this will have to be agreed between the parties, and if they are unable to agree on this, they must notify the Tribunal and we will reconvene to determine the issue.

59. This is a very unfortunate case for the Appellant. The Appellant’s contract was for high quality organic wheat and they paid a high price accordingly. What they got was infested wheat and they were put to the trouble of treating it. Far more significantly, the infestation reduced the protein content and took the wheat below the quality necessary for the 0% duty rate. We were told that the margins on wheat of about €2 per tonne mean that if any duty at all is payable the shipment is uneconomic. HMRC stressed that they imply no criticism of the Appellant. It was simply HMRC’s case that, through no one’s fault, the shipment did not meet the minimum protein content for high quality wheat, and we agree.

60. This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party. The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice.

First-tier Tribunal (Tax)