DECISION

1. This appeal relates to a tariff classification HMRC made for a product imported by the appellant from the United States known as a “Network General InfiniStream Unit” (“the InifiniStream”). HMRC have classified the InfiniStream under the tariff classification 9031 80 38 90. The tariff heading 9031 includes measuring or checking instruments, appliance machines (not specified or included elsewhere in Chapter 90). The appellant contests the classification and argues it should be a non-duty classification 8471 50. Heading 8471 refers to automatic data processing machines and units thereof. Alternatively the appellant argues the classification is 8471 70 50. The sub-heading 8471 70 refers to storage units and 8471 70 50 refers to hard disk drives.

2. The appeal was filed out of time and there is a preliminary issue as to whether the Tribunal should give permission to extend the time to appeal.

3. There is also an appeal against a duty demand, the determination of which is at least in part dependent on the outcome of the dispute on classification.

Permission to appeal out of time

4. The appeal was lodged on 27 July 2011. The chronology leading up to that is set out below.

Chronology

5. On 27 February 2007 the Appellant applied for a Binding Tariff Information (BTI) under commodity code 8471 70 50 90 in respect of “Network General InfiniStream Units”.

6. On 4 June 2007, HMRC issued the BTI reference GB 116330705 classifying the product under 9031 80 38 90, which attracts a duty rate of 4%. There was a notice on this which mentioned there was a right to appeal against this BTI. No time limit was stated.

7. On 27 August 2010, HMRC issued a C18 Post Clearance Demand Note, specifying that charges had been underpaid by £182,772.31 on products under BTI GB116330705, and another BTI not subject to the present appeal, BTI 116331114. The letter stated:

“If you do not agree with any decision issued to you there are three options available. Within 30 days of the date of the decision you can either:

- Send new information or arguments to the decision maker,

- Request a review of the decision….

- Appeal direct to the Tribunal who are independent of HMRC

If you opt to have your case reviewed you will still be able to appeal to the tribunal if you disagree with the outcome.”

8. On 1 September 2010 HMRC wrote to the appellant. The letter included a schedule by which the duty underpaid was calculated. It also stated:

“With effect from 1 April 2009 there has been a change to the Review and Appeals/ Tribunal procedures for HMRC. We are initially required to advise you of any decisions made and allow you the opportunity to put forward any explanations and or further evidence to support a change or revocation of this decision.”

9. The letter also gave web links to further information on appeals and reviews on the HMRC and Tribunal Services website.

10. On 21 September 2010, the appellant wrote to HMRC requesting a review of the duty demand, stating that in late 2007, Network General Inc. were acquired by NetScout Inc. and merged their product ranges; the “InfiniStream” name was retained but the functionality changed, and in subsequent imports the appellant used community code 8471 70 50 90. It recognised that it should have re-submitted an application for a BTI for a NetScout InfiniStream, and did so in an enclosed application.

11. On 24 September 2010 Mr Jones wrote to Mairead Hutchinson at HMRC to notify her that the appellant would be sending new information or arguments to her by the end of that September.

12. On 1 October 2010 Mr Simmonds accordingly wrote to Ms Hutchinson. In particular he sought to highlight the alleged similarities between Niksun units (given BTI Reference GB116322311 8471 70 50 90) and the InfiniStream product. The letter asked HMRC to consider removal of the duty owed in the light of the information provided.

13. On 22 October 2010, HMRC upheld the classification for the Network General InfiniStream. Ms Hutchinson’s letter again set out 3 options: 1) provide new information/ arguments 2) request a review within 30 days of the date of the letter 3) appeal to Tribunal. The letter stated again that if the appellant opted for review they would still be able to appeal.

14. On 2 November 2010 Mr Jones e-mailed Ms Hutchinson to say that the appellant would take option 1 and send in new information and arguments.

15. On 30 November 2010, Mr Simmonds wrote to HMRC to respond to the observations made by HMRC in its letter of 22 October 2010.

16. On 24 December 2010, HMRC upheld the classification of the product on reconsideration. The letter set out the same 3 options stating:

“If you are not happy with this decision you still have the same 3 options available; within the next 30 days you may…”

17. On 17 February 2011, Mr Simmonds after apologising for the delay in responding gave further information and suggested that a different code, 8471 50 was appropriate. The letter mentions that Mr Simmonds had discussed HMRC’s evidence and findings at length with the board of directors of the appellant.

18. On 8 March 2011, HMRC maintained its classification. Ms Hutchinson’s letter stated a review of the letter could be requested within 30 days of the letter. The letter also stated the appellant could appeal direct to the Tribunal.

19. On 23 March 2011, HMRC issued a civil penalty of £1,500 against the Appellant for failing to declare the correct particulars on its Customs declarations for import.

20. On 6 May 2011, the Mr Simmonds wrote to the National Review and Appeals Team at HMRC to request that the decision to apply a binding tariff of 4% duty should be reviewed.

21. On 13 May 2011, HMRC wrote to inform the appellant that the request for a review was out of time, the request not having been made within 30 days of the 8 March 2011 decision and no exceptional circumstances having been given. The letter advised that the Appellant might wish to consider a late appeal to the VAT and Duties Tribunal under s16 (1D) Finance Act 1994 (“FA 1994”).

22. The appellant wrote again on 23 May 2011 explaining the delay in replying to the 8 March 2011 letter, and asking for a reconsideration, to which HMRC replied in similar terms on 25 May 2011, refusing the review as the request was out of time. HMRC’s letter indicated if the appellant was unhappy with the decision refusing a request for late review that the appellant could appeal to the VAT and Duties Tribunal under Section 16 (1D) FA 1994 and gave details of the Tribunal’s Service website. The reasons for the delay stated in the appellant’s 23 May 2011 letter are discussed further below at [29].

23. Further correspondence took place between the appellant and HMRC in which HMRC maintained that it would not review the decision because the request had been made outside the 30 day period. The letter informed the appellant that it might wish to appeal under s16 (1D) FA 1994. In particular Mr Simmonds e-mailed HMRC on 26 May 2011 (attaching communications with the importer dated 25 May 2011) and again on 2 June 2011 restating the reasons for the late submission of the request for review and asking for further details as to which link on the Tribunal’s website to follow to obtain the correct documentation. Ms Warn of HMRC replied on 10 June 2011 with precise instructions on how to get to the appeal form on the website.

24. On 4 July 2011, HMRC emailed the appellant stating:

“With reference to our telephone conversation on Friday, I have had a quick look at the Tribunals website. The web site has guidance notes on “Making an Appeal” and the form that is required for you to complete. I am sorry I am unable to help you any further, but you can contact them on 0845 223 8080 for advice.”

25. On 27 July 2011, the appellant submitted its Notice of Appeal, seeking to appeal a decision of “4 July 2011”, observing that there was no appeal against a penalty, that there had been a review by HMRC on 25 May 2011 and the period for a review had ended. The reasons stated for a late appeal referred to other business commitments, and that board approval was required to proceed with the appeal.

Legal framework for appeal and time limits

26. Rule 20(4) of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009 (“the Tribunal Rules”) as amended by SI 2010/2653 (with effect from 29 November 2010) provides:

“(1) A person making or notifying an appeal to the Tribunal under any enactment must start proceedings by sending or delivering a notice of appeal to the Tribunal.

….

(4) If the notice of appeal is provided after the end of any period specified in an enactment referred to in paragraph (1) but the enactment provides that an appeal may be made or notified after that period with the permission of the Tribunal –

(a) the notice of appeal must include a request for such permission and the reason why the notice of appeal was not provided in time; and

(b) unless the Tribunal gives such permission, the Tribunal must not admit the notice of appeal.”

27. As of 1 April 2009, the relevant provisions of FA 1994 can be summarised as follows:

(1) A decision by HMRC in relation to any customs duty as to whether or not anything is charged with such duty, or as to whether a person is liable to pay any amount charged or the amount of his liability is a “relevant decision” (s13A(2)(a));

(2) If HMRC notify a person of a relevant decision, it must offer the person a review of the decision (s15A(1)), although that section does not apply to the notification of the conclusions of a review (s15A(2);

(3) If HMRC have offered a review under s15A, and the person has notified its acceptance of the offer of review within 30 days, HMRC must review the decision (s15C(1));

(4) If a person has not notified HMRC within 30 days, then HMRC must still review the decision if it is satisfied that the person had had a reasonable excuse for not accepting the offer within the time allowed, and made the out of time request without unreasonable delay thereafter (s15E);

(5) If an application is made for a review out of time under s15E, an appeal must be made within 30 days of the date on which HMRC decide not to undertake a review (s16(1D)(b)(ii)).

(6) Under s16(1F), an appeal may be made after the end of the period specified in subsections 1 and 1B “if the appeal tribunal gives permission to do so”.

28. HMRC pointed out that the appeal, which was lodged on 27 July 2011, was over 30 days after the original BTI decision of 4 June 2007, the duty demand of 27 August 2010, the reconsiderations of 22 October 2010, 24 December 2010 and 8 March 2011, and the decision of 13 May 2011 not to undertake a review on a late request.

29. On behalf of the appellant Mr Jones explained that he and Mr Simmonds had been allocated responsibility within the appellant company for handling the dispute and appeal. They were both directors at the appellant company and were having to handle the matter outside their normal day to day working. Mr Jones said they were not familiar with the relevant processes and had followed the process as set out to them by HMRC. They had understood there to be 3 different levels of escalation: (1) dealing with Ms Hutchinson in HMRC’s office in Belfast, (2) escalation to HMRC’s review section in Southend, and (3) appealing to the Tribunal. They had always thought an appeal to the Tribunal was the last stage and that they might have been able to resolve the matter through dealing with Ms Hutchinson or Ms Warn. The reason their correspondence was late was their business year end was at the end of May and they were busy trying to “close the year”.

30. Mr Jones also argued that if HMRC were going to make an issue about the appeal being out of time, then the appellant ought to have been told about this sooner. There had been no mention that HMRC were going to take a point on this in their Statement of Case, the first the appellant knew that a point was being taken on timing was in HMRC’s skeleton argument filed before the hearing.

31. In our view the 4 July 2011 “decision” referred to in the appellant’s Notice of appeal was clearly no such thing; it was simply an e-mail from HMRC giving further details as to where further guidance on making an appeal could be found.

32. Under the legislation on time limits which applied to the situation where HMRC had been requested to undertake a review out of time under s 15E FA 1994 the relevant date for the appeal time limit to start running is the date HMRC decided not to undertake a review. This date was 13 May 2011. Under s16 (1D)(b)(ii) FA 1994 the appeal was to be made within 30 days of that date. The appellant ought therefore to have filed its appeal no later than 12 June 2011. Instead it filed the appeal on 27 July 2011.

The Tribunal’s discretion to extend time to appeal

33. We were referred by HMRC to Sir Stephen Oliver’s comment in Ogedegbe v HMRC (2007) UK FT 364 (TC) that “[the power to extend time for making an appeal] will only be granted exceptionally” and it was submitted by HMRC that it was difficult to envisage this case as being one which was suitable for exceptional leave to be granted.

34. We do not regard Sir Stephen Oliver’s comment as suggesting that the test for the Tribunal in exercising its discretion is whether the case is one which is suitable for exceptional leave.

35. The comment was recently considered by the Upper Tribunal in its decision of O’Flaherty v HMRC [2013] UKUT 161 (TCC) (published after the hearing in this matter) where Judge Berner stated at [36]:

“It is in my view important that this comment should not be thought to provide a qualitative test for the circumstances the FTT is required to take into account. It should properly be understood as saying nothing more than that permission should not routinely be given; what is needed is the proper judicial exercise of a discretion, taking account of all relevant factors and circumstances.”

36. We consider the approach to exercising the Tribunal’s discretion to grant permission is as described in the Upper Tribunal decision of Data Select Ltd v HMRC [2012] UKUT 187 (TCC). There Morgan J set out the questions which courts and tribunals ask themselves when they are asked to extend a relevant time limit :

(1) what is the purpose of the time limit?

(2) how long was the delay?

(3) is there a good explanation for the delay?

(4) what will be the consequences for the parties of an extension of time?

(5) what will be the consequences for the parties of a refusal to extend time.

37. The discretion in that particular case was to extend time pursuant to the 30 day time limit in making an appeal to the Tribunal pursuant to s 83G(1) VATA 1994 but it is clear the questions posed by Morgan J describe principles of wider application (see O’Flaherty at [28]) and we see no reason why we should not take the approach described by Morgan J to the statutory enactment in issue here.

38. We have considered the purpose of the time limit in promoting certainty and finality and that the delay is in the order of 6 weeks. The explanation given by the appellant, perhaps in seeking to answer the periods of delay mentioned in HMRC’s skeleton argument does not specifically address the delay between 12 June 2011 and 27 July 2011.

39. However, we observe that the appellant’s correspondence in this period is consistent with the tenor of Mr Jones’ and Mr Simmonds’ explanations that the appellant was dependent on HMRC for explanations on the next steps and time limits and is also consistent with Mr Simmonds’ account of having to run decisions in relation to the appeal through the appellant’s board.

40. In relation to HMRC’s explanation of time limits we note that in contrast to the preceding correspondence where time limits e.g. for review are set out, the 13 May 2011 letter, while mentioning the relevant statutory provisions for appeal does not set out the time limit for the appeal. That is not a requirement of the legislation but given the context of previous letters it does in our view offer an explanation as to why Mr Simmonds had not respected the relevant time limits as promptly as he had generally done in relation to the appellant’s previous correspondence with HMRC. We accept that a decision to appeal in this matter was something that Mr Simmonds wished to raise with the board of the appellant (and which he did further to his e-mail of 21 July 2011 which referred to a review by the board) and this may have meant he could not act as speedily as he might have wished.

41. The need for board input, or the extent to which the directors handling the appeal were occupied with their day to day responsibilities may go some way to explain the delay but we do not think it can amount to providing a sufficiently good explanation. The appellant in this appeal is the company. The company, we were told, had 42 employees and a turnover of £17 million in 2012. It had no insignificant resources at its disposal. The company was responsible for organising what resources it would deploy to deal with the appeal and for ensuring the timing of its board reviews and meetings were such that it could comply with the time limits.

42. As for the appellant’s complaint that HMRC should have alerted it sooner to the fact that HMRC would be taking a point on the appeal being out of time, we do not think this is relevant to the issue of whether the time limit for appealing should be extended. Rather, the underlying issue this raises is whether it was unfair to hear the matter when we did on the grounds of lack of notice. While it would have been preferable for notice to have been given that the point would be taken sooner than it was, we do not think any delay resulted in any significant prejudice to the appellant. The point was raised in HMRC’s skeleton which was filed 14 days in advance. The directors best able to speak to the handling of the matter, Mr Jones and Mr Simmonds were both present at the hearing and were given the opportunity to give evidence as to the reasons for the delay and to make submissions on the point.

43. Turning to the respective prejudice to the parties of granting or not granting permission to extend time, the consequences of not granting permission would be that the appellant would not be able to contest its appeal and would be liable for duty in a significant amount. The issue in the appeal relates to a product used in the information technology industry which has a number of detailed technical specifications which must be considered against the relevant detailed descriptions in the combined nomenclature. Having considered the parties’ submissions in outline we think the appellant’s case is at least arguable. The consequences to HMRC in granting permission are that an issue that would have otherwise have been final on 12 June 2011 had to be reopened. On the other hand the circumstances of the case, namely the focus on the objective characteristics of the device and its classification, the length of delay of around 6 weeks mean that issues of evidence becoming unavailable or stale are not significant. The balance of prejudice in such circumstances is we think in the appellant’s favour.

44. On balance, taking account of the length of delay, although the appellant has not provided a particularly good explanation for that delay our view is that it is fair and just that the Tribunal should exercise its discretion to extend the time for the appellant to file its appeal to 27 July 2011.

Law

Applicable law to appeals against classification decisions

45. As summarised in Outside in (Cambridge) Ltd trading as Lumie v The Commissioners for Her Majesty's Revenue and Customs [2011] UKFTT 441 (TC), the applicable law to classification decisions is as follows.

46. The EU's customs tariff (“the Tariff”) applies to the import of goods across the external borders of the EU. Classification of goods for the purposes of the Tariff takes place in accordance with the Combined Nomenclature (“CN”). The CN is set out in Annex 1 to Council Regulation 2658/87 (“the CN Regulation”) and is updated annually by way of Regulation. The CN is based on the worldwide harmonised commodity description and coding system (“HS”) drawn up by the Customs Co-operation Council (now the World Customs Organisation) (see further Intermodal Transports BV v Staatssecretaris van Financien (Case C 495–03) [2006] 1 CMLR 32 at paragraph 4). Having been effected by way of Regulation, the CN is directly applicable in the UK. The CN provides a systematic classification for all goods. The aim of the CN is to ensure that any particular product falls to be classified in only one place within the nomenclature. This aim is aided by the general rules for the interpretation of the CN (“GIRs”).

47. The GIRs, like the CN, are set out within the CN Regulation. The GIRs provide (among other things):

“The titles of sections, chapters and sub-chapters are provided for ease of reference only; for legal purposes, classification shall be determined according to the terms of the headings and any relevant section or chapter notes and, provided such headings or notes do not otherwise require, according to the following provisions.

...

2. (a) Any reference in a heading to an article shall be taken to include a reference to that article incomplete or unfinished, provided that, as presented, the incomplete or unfinished article has the essential character of the complete or finished article. It shall also be taken to include a reference to that article complete or finished (or falling to be classified as complete or finished by virtue of this rule), presented unassembled or disassembled.

2. (b) Any reference in a heading to a material or substance shall be taken to include a reference to mixtures or combinations of that material or substance with other materials or substances. Any reference to goods of a given material or substance shall be taken to include a reference to goods consisting wholly or partly of such material or substance. The classification of goods consisting of more than one material or substance shall be according to the principles of rule 3.

3. When by application of rule 2 (b) or for any other reason, goods are prima facie classifiable under two or more headings, classification shall be effected as follows:

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

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

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

...

6. For legal purposes, the classification of goods in the subheadings of a heading shall be determined according to the terms of those subheadings and any related subheading notes and mutatis mutandis to the above rules, on the understanding that only subheadings at the same level are comparable. For the purposes of this rule the relative section and chapter notes also apply, unless the context otherwise requires.”

48. In addition to the GIRs, the explanatory notes to the CN (“CNENs”) and the explanatory notes to the HS (“HSENs”) also provide interpretative assistance (see for instance BAS Trucks BV v Staatssecretaris van Financien [2007] (Case C-400/05) at [28]. Unlike the GIRs, neither the CNENs nor the HSENs are legally binding (Intermodal at [48]).

49. These proceedings concerned the validity of HMRC's classification of the InfiniStream product. Under s 16(6) of FA 1994, it is for the appellant to show that the grounds on which its appeal is brought are established.

The competing classifications

50. HMRC gave the product the BTI classification 9031 80 38 90. The particular headings and sub-headings of the classification are set out below:

51. Chapter 90 covers:

“optical, photographic, cinematographic, measuring, checking, precision, medical or surgical instruments and apparatus; parts and accessories thereof”

52. The relevant headings and sub-headings under Chapter 90 are:

9031 Measuring or checking instruments, appliances and machines, not specified or included elsewhere in this chapter; profile projectors

9031 80 – Other instruments, appliances and machines

- - Electronic:

9031 80 38 ---Other

---- 9031 80 38 05 For use in civil aircraft

---- 9031 80 38 10 Acceleration measurement device for automotive applications comprising one or more active and/or passive elements and one or more sensor, the whole contained in a housing.

---- 9031 80 38 90 Other

53. The appellant argues the product is an automatic data processing machine and the classification should be 8471 50

8471

Automatic data-processing machines and units thereof; magnetic or optical readers, machines for transcribing data onto data media in coded form and machines for processing such data, not elsewhere specified or included:

…

8471 50 00

Processing units other than those of sub-heading 8471 41 or 8471 49 whether or not containing in the same housing one or two of the following types of unit storage units, input units, output units.

54. Headings 8471 41 00 and 8471 49 00 concern other automatic data-processing machines comprising in the case of 8471 41 00 machines “comprising in the same housing at least central processing unit and an input and output unit whether or not combined” or in the case of 8471 “other, presented in the form of systems”.

55. Alternatively the appellant argues the product is a data storage device and should be classified under 8471 70 50 90

8471

…

8471 70 - Storage units

…

-- Other

--- Disk storage units

---- Other

-----Hard disk drives

----- For use in civil aircraft 8471 70 50 10

----- Other 8471 70 50 90

56. Under GIR 1 classification shall also be determined according to any relevant section or chapter notes. The Chapter notes to Chapter 84 include the following Note 5 which provides:

“(A) For the purposes of heading 8471, the expression ‘automatic data-processing machines’ means machines, capable of

(1) storing the processing program or programs and at least the data immediately necessary for the execution of the program;

(2) being freely programmed in accordance with the requirements of the user;

(3) performing arithmetical computations specified by the user; and

(4) executing, without human intervention, a processing program which requires them to modify their execution, by logical decision during the processing run.

(B) Automatic data-processing machines may be in the form of systems consisting of a variable number of separate units.

(C) Subject to paragraph (D) and (E) below, a unit is to be regarded as being a part of an automatic data-processing system if it meets all of the following conditions:

(1) it is of a kind solely or principally used in an automatic data-processing system;

(2) it is connectable to the central processing unit either directly or through one or more other units; and

(3) it is able to accept or deliver data in a form (codes or signals) which can be used by the system.

Separately presented units of an automatic data-processing machine are to be classified in heading 8471.

However keyboards, X-Y coordinate input devices and disk storage units which satisfy the conditions of paragraphs (C)(2) and (C)(3) above, are in all cases to be classified as units of heading 8471.

(D) Heading 8471 does not cover the following when presented separately, even if they meet all of the conditions set forth in note 5(C) above:

(1) printers, copying machines, facsimile machines, whether or not combined;

(2) apparatus for the transmission or reception of voice, images or other data, including apparatus for communication in a wired or wireless network (such as a local or wide area network);

(3) loudspeakers and microphones;

(4) television cameras, digital cameras and video camera recorders;

(5) monitors and projectors, not incorporating television reception apparatus.

(E) Machines incorporating or working in conjunction with an automatic data-processing machine and performing a specific function other than data processing are to be classified in the headings appropriate to their respective functions or, failing that, in residual headings.

57. The following cases were referred to:

Outside in (Cambridge) Ltd trading as Lumie v HMRC [2011] UKFTT 411 (TC)

Intermodal Transports BV v Staatsecretaris van Financien (Case C-495-03) [2006] 1 CMLR 32

BAS Trucks BV v Staatssecretaris van Financien [2007] (Case C-400/05)

Igekami Electronics (Europe) (Common Customs Tariff) [2005] EUECJ C-467/03.

Evidence

58. We had a lever arch file containing correspondence between the parties and which included data sheets and marketing material and internet articles about the product.

59. We heard evidence from Mark Attridge, the HMRC Review Officer for the decision under appeal and David Harris, HMRC officer and Technical Team Leader of the HMRC team responsible for classifying goods within the Electrical, Mechanical, Medical and Scientific Sector of the Customs Tariff. The evidence was cross-examined by the appellant. In the course of their submissions Mr Simmonds, (the appellant’s Technical Director) and Mr Jones (the appellant’s Finance Director) mentioned matters which amounted to evidence and upon which HMRC were given the opportunity to ask questions.

Facts

Background to appellant

60. The appellant company was founded in 1988 as an importer and reseller of IT networking and network tools and equipment. It employs 42 staff in two UK locations and had an annual turnover of £17 million in 2012. As described by Mr Simmonds and Mr Jones the unique selling point of the appellant is in understanding customer problems in relation to the customer’s network. Its strength was in taking the best bits of other products and putting them together in a way which added value for the customer.

The InifiniStream

61. We were taken through what was a termed a data sheet for the InfiniStream and which carried the copyright date of 2006. As well as containing certain technical specifications the data sheet for the InfiniStream also included statements which were more in the nature of marketing claims. We took this into account in evaluating the evidential value of the document but were satisfied that we were able to accept the document accurately reflected the technical specification of the product and further that to the extent the data sheet highlighted certain features it also gave insight into the purpose for which the product was intended to be used. The intended use of the product as set out below was also generally consistent with web articles reviewing the InfiniStream product in industry press written by Ellen Messmer of Network world on 10 February 2003, and Cameron Studevant of eWeek on 14 April 2003 which were also put before us.

62. The data sheet included the following:

Product overview

“…network performance issues can negatively impact an enterprise’s overall efficacy, as these problems tend to occur randomly. Analysing these events can be enormously time-consuming especially when trying to recreate or pinpoint specific scenarios. Reviewing terabytes of data to find “the needle in the haystack” often proves fruitless. The Sniffer InfiniStream platform makes it easy to look back-in-time to examine network performance, observe traffic trends, isolate anomalies, perform deep packet analysis, and generate summary reports.

Designed for continuous 24x7, line-rate stream-to-disk usage, Sniffer InfiniStream captures up to months of granular network data for 10/100/1000 Ethernet, 10 Gigabit Ethernet (10GbE), WAN and ATM networks. This comprehensive data view facilitates quick problem resolution. Combined with the industry leading Sniffer Expert analysis and protocol decodes, performing retrospective analysis is painless. With Sniffer InfiniStream you can select the data time frame to ensure only relevant transactions are examined.

The Sniffer Inifinistream Family

Sniffer InfiniStream is a network performance hardware and software solution that enables users of today’s high-speed multi-topology networks to capture up to months of data in order to monitor, measure, manage and resolve high-impact intermittent network problems. Available with up to four terabytes of storage Sniffer InifiniStream provides best-in-class streaming capture performance and flexible data mining for:

- Real-time analysis

- Back-in-time analysis

63. The features highlighted on the data sheet were “10 GbE Network Support”, “Real-time alerting” and “Quick Select Statistics”. The latter feature was stated to provide a graphical way to locate anomalies. narrowing down the view to the appropriate time frame (down to one second intervals) to present only the applicable variables necessary to pinpoint and resolve the performance problem. Other features referred to “optimising WAN performance”, “MPLS Quality of Service” (MPLS stands for Multiprotocol Label Switching), “Type of Service / Class of Service” statistical support, “Advanced protocol decodes and expert analysis”. The data sheet stated “The Sniffer InfiniStream platforms serve as instrumentation to collect and analyse network data.”

64. The product comes in a range of sizes and capacities but carries out the same essential functions. It is designed to be placed in a data centre equipment rack with dimensions ranging from 48.3cm x 61cm x 13.3cm and weight of 43 kg to dimensions of 42.6cm x 77.2cm x 4.26cm and weight of 16.3 kg.

65. The operating system is Red Hat LINUX 9 and the application software Apache 2.040, net-snmp-5.09. The Motherboard has a Xeon processor. It also has RAID (redundant array of independent disks) storage sub-system and system drives including either DVD/CD-RW or DVD-ROM drives.

66. We had the opportunity to ask further questions of Mr Simmonds, who as Technical Director, was familiar with the product. From his responses we made the following further findings.

67. When the device starts up it starts up from power on, it has a collection and collation programme. It makes decisions about which traffic it monitors. It performs “packet slicing”. This means it will look at packets of data either completely, or partially, and decide whether or not the data is needed. It stores data intelligently in order to maximise storage capacity. It manages its own disk space and performs statistical analysis deleting ongoing data. It does not record everything because there is no point in recording everything for example if certain information is encrypted and no way of decrypting it, then there is no point in storing it.

68. As part of the installation programme the customer can configure and alter it as appropriate to meet the customer’s requirements.

69. The device can be configured to carry out calculations for example round trip time of a particular piece of data.

70. The device connects to the network and is analogous to a device which taps a telephone by recording in that it records network traffic onto an internal hard drive. The drive is described as a “leaky bucket” in that it is part of its design that data drips out. The device can be configured according to time or volume of traffic or a mix of the two.

71. Once information is in the product the user can search for a specific piece of information for example the user could search for a particular e-mail. The user can replay and revisit a scenario which let to the network issue instead of waiting to see whether it will happen again.

72. The product does not generate network traffic.

Appellant’s arguments

73. The appellant’s arguments are set out in more detail in the discussion section below.

74. In summary, the appellant argues the product is an automatic data processing machine as identified by the four characteristics listed in Note 5(A) to Chapter 84 at [56] above.

75. In relation to the Note 5(E) to Chapter 84 which states that where a device has a more specific function it must be placed under the more specific heading the product fulfils a number of differing roles and functions which cannot be specifically and individually classified.

76. While EC Regulation 129/2005 classifies network analysers under heading 9031 the InfiniStream is not a network analyser because it does not meet a number of the characteristics set out for network analysers in the Regulation.

77. The US Government have agreed that Harmonised Tariff 8471 50 is appropriate. The US manufacturer and exporter of the equipment uses a non-duty attracting code to export the equipment.

78. Because of HMRC’s classification the product attracts duty. The appellant will lose out to competitors because its prices are as a result 4 % higher. Further, HMRC have not taken action to enforce the classification in relation to competitors and other manufacturers of the product.

Respondents’ arguments

79. The Respondents’ arguments are similarly set out in more detail in the discussion section below.

80. The product is not an automatic data processing machine. It does not meet the requirements of Chapter 84 Note 5(A) in that it is not freely programmable. Even if Note 5(A) were satisfied the specific function of the product, as shown by the evidence and the data sheet for the product, is one of analysis. In line with Note 5(E) and the Igekami decision the product must be classified under this more specific function.

81. EC Regulation 129/2005 which sets out characteristics for network analysers does not do so exhaustively. The intrinsic properties of the product are such that it falls within heading 9031 “Measuring or checking instruments…”.

82. In relation to any possible classification by the US the Intermodal case establishes that what is decisive is the objective characteristics of the goods not how others may classify them.

Discussion

Approach to classification

83. Determinations as to the classification of particular goods under the Combined Nomenclature are made in accordance with the EC customs code established by way of Regulation 2913/92 (“the EC Customs Code Regulation”). Such determinations are made by way of BTIs by customs authorities (Article 12 of the EC Customs Code Regulation).

84. As regards the proper approach to such classification, HMRC drew our attention to the decision in Intermodal where the ECJ stated the position in the following terms (at [47]):

“According to settled case law, in the interests of legal certainty and ease of verification, the decisive criterion for the classification of goods for customs purposes is in general to be found in their objective characteristics and properties as defined in the wording of the relevant heading of the CN and of the notes to the sections or chapters.”

85. Accordingly, the decisive issue is what are the product’s objective characteristics and properties?

Tribunal’s jurisdiction – objective characteristics of appliance / intended use

86. We note from [23] of Igekami

“…according to the Court’s case-law, the intended use of product may constitute an objective criterion in relation to tariff classification if it is inherent in the product, and such inherent character must be capable of being assessed on the basis of the product’s objective characteristics and properties…”

87. So, the intended use of the InfiniStream is an objective criterion if the intended use is “inherent in the product”.

88. In examining the InfiniStream’s objective characteristics, while we took into account reviews written about the product in industry news articles we found the descriptions of the product and its intended use as set out in the product data sheet to be of the most assistance. We were able to ask Mr Simmonds questions on the technical terms referred to in those materials and in so far as Mr Simmonds’ responses and submissions contained matters of evidence we found him to be a knowledgeable and credible witness. We also found Mr Attridge and Mr Harris to be credible witnesses. However, we take account of the fact that much of their evidence amounted to either giving an opinion on whether the product’s characteristics met certain classifications based on materials (e.g. the data sheet) which we had before us and upon which we were tasked with reaching our own view, or was evidence which sought to recount and explain HMRC’s decision making process. Accordingly that evidence, although no doubt helpful in seeking to explain to the appellant how HMRC had come to the decision it did, and in explaining HMRC’s interpretation of the nomenclature, was of less use to us in ascertaining what the objective characteristics of the product were.

The Automatic Data Processing machine argument (8471 50)

90. The appellant says the product is an automatic data processing machine as identified by the four characteristics listed in Note 5(A):

(1) Storing the processing program or programs and at least the data immediately necessary for the execution of the program;

(2) being freely programmed in accordance with the requirements of the user;

(3) performing arithmetical computations specified by the user; and

(4) executing, without human intervention, a processing program which requires them to modify their execution, by logical decision during the processing run.

91. HMRC say that the InfiniStream does not fulfil the above characteristics as it is not “freely programmed” and that even if it is the application of Note 5(E) means that it should be put under heading 9031 given the InfiniStream has a more specific function.

Note 5(A) to Chapter 84 - Can the InfiniStream be freely programmed in accordance with the requirements of the user?

92. The appellant says the device is customised for customers and freely programmable. The analysis provides results to the local screen and it provides filtering of useful /non-useful information. The device generates statistics once data is captured.

93. HMRC say the reference to “freely programmed” means more than saying the product is freely programmable within the confines of its function.

94. HMRC’s witness, Mr Harris, said he doubted that the product was freely programmable and thought this characteristic was “more for products such as a desktop PC”. But as pointed out above Mr Harris is a witness of fact, and was offering his opinion on the interpretation of what “freely programmed” might mean. While we may consider that by way a submission it does not amount to evidence on whether the product was freely programmed or not.

95. We see no basis for restricting the plain words in the way suggested. We note in Igekami that although there was no discussion specifically on the point, the equipment there (close circuit surveillance equipment which was not anything like a desktop PC), was accepted as prima facie falling within the Automatic Data Processing machine heading, and Note 5(A) was referred to [para [9]]. Mr Simmond’s submissions on “freely programmed” are consistent with the data sheet which refers to alerts “that can be triggered on a broad-range of performance metrics” and which also states “With Sniffer InfiniStream you can select the data time frame to ensure only relevant transactions are examined”.

96. We accept the product fulfils characteristic 2) (freely programmed in accordance with the requirements of the user.)

97. We did not understand there to be a dispute as to whether the other characteristics in Note 5(A) (1), (3) and (4) were satisfied but in any event having considered those we think the InifiniStream satisfied those characteristics too.

98. We discuss below the relevance of there being similar products described as network analysers ( a product which e.g. provides information on the performance of networks by monitoring network activity) which are classified by a particular EC Regulation (EC/129/2005) under heading 9031. We note that the explanation for both the products referred to in the Regulation falling within heading 9031 mention that the products are excluded from heading 8471 by the application of Note 5(E) to Chapter 84. In our view this rather suggests that those products would otherwise fall within the definition of “automatic data processing machine” for the purposes of heading 8471. The fact that an EU wide regulation implies that similar but not identical products would be capable of falling within heading 8471 suggests to us that the InfiniStream is also capable of falling within this heading.

Note 5(E) to Chapter 84 –is the InfiniStream a “machine incorporating or working in conjunction with an automatic data-processing machine and performing a specific function other than data processing”?

99. Following from the above, although we did not receive any submissions on the point we do not think there can be any dispute that the InifiniStream is “a machine incorporating or working in conjunction with an automatic data processing machine”. The question is whether it performs a specific function other than data processing.

100. The appellant says, in relation to Note 5(E) to Chapter 84 (which states that where a device has a more specific function it must be placed under the more specific heading), that the product is user-programmable to fulfil a number of differing roles and functions which cannot be specifically and individually classified. It is, they say, a “leap of faith” to say the product has just one function.

101. HMRC say the product is excluded by Note 5(E). It is not enough that the LINUX system allows for programming because the specific function of checking network traffic overrides the programming function. They refer to the Igekami decision where a digital recording machine used for surveillance “must be regarded as performing a specific function, going beyond automatic data processing” (at [20] of that decision). HMRC say the evidence before us in terms of what we have heard and in the data sheet / press articles (including the reference to the term “sniffer”) shows that the specific function of the product is the specific function of network monitoring and analysis.

102. HMRC also point out that EC Regulation 129/2005 classified similar network analyser products under heading 9031 and used Note 5(E) as a basis to do this. The product here should similarly be excluded from heading 8471 applying Note 5(E).

103. The appellant says the product has multiple functions comprising storage capacity, freely programmable features, filtering of useful and non-useful information.

104. In response HMRC say EC Regulation 129/2005 anticipates multiple functions and sets them out. They argue that just because a device can do a number of things that should not preclude it from having a specific function.

Our views on application of Note 5(E) and whether InfiniStream “performing a specific function…”

105. We accept the device can be described as performing various functions but in our view it is relevant to consider the purpose for which those functions are performed. To the extent the product has a data storage function this is not an end in itself but is in order to allow analysis of the data. Similarly the filtering of data is not an end in itself but is a means to ensure efficient data storage with a view to the product then allowing the data to be analysed.

106. In relation to the appellant’s argument that the product performs a number of functions it is no doubt possible to take a function and sub-divide it but we do not think that such a level of dissection of function is contemplated by Note 5(E). We note that it is inherent in the characteristics set out in Note 5(A) that there may be different functions e.g. functions stemming from free programmability, storing data. Those different functions cannot then be reasons for not applying Note 5(E) otherwise Note 5(E) would never have any application.

107. We note from the facts of the Igekami decision at [20] that the specific function was “the storage of analogue signals corresponding to pictures or sounds transmitted by external sources, the digital conversion of those signals, their compression and their reproduction on screen”. That description was capable of being broken down into separate actions but the fact that it was did not stop those components in their totality being viewed as a specific function.

108. Further we note that Note 5(E) which applies to heading 8471 instructs that the machine performing a specific function is to be classified in the heading appropriate to the respective function “or failing that, in residual headings.” This contemplates that even if there is only a residual heading elsewhere which captures the specific function, then that residual heading is to be preferred to heading 8471. This suggests to us that the reference to “specific function” in Note 5(E) must be read in the context of what other headings are available, including residual headings and that a high degree of specificity is not necessarily required in order to come to the conclusion that the machine performs a specific function other than data processing.

109. Taking account of the data sheet for the product and the way the product is marketed we think the product has a specific function of network monitoring and analysis which going beyond the function automatic data processing. We therefore agree with HMRC that Note 5(E) applies.

Relevance of EC Regulation 129/2005 – does it exhaustively set out the characteristics of network analysers which fall under heading 9031?

110. Both parties’ submissions refer to Regulation 129/2005 which provide a description of two network analyser products and which requires their classification under heading 9031. It is accepted by both parties that the InfiniStream does not have all of the characteristics referred to in the Regulation. The appellant uses this as a basis to argue heading 9031 is not therefore relevant. HMRC on the hand say the fact the InfiniStream does not fully meet the characteristics of the network analysers mentioned in the Regulation does not prevent the InfiniStream from falling under heading 9031 as the Regulation does not specify exhaustively what network products may fall under the 9031 classification.

Commission Regulation 129/2005

111. The recitals to Commission Regulation 129/2005 provide where relevant as follows:

Whereas:

(1) In order to ensure uniform application of the Combined Nomenclature annexed to Regulation (EEC) No 2658/87, it is necessary to adopt measures concerning the classification of the goods referred to in the Annex to this Regulation.

(2) Regulation (EEC) No 2658/87 has laid down the general rules for the interpretation of the Combined Nomenclature. Those rules apply also to any other nomenclature which is wholly or partly based on it or which adds any additional subdivision to it and which is established by specific Community provisions, with a view to the application of tariff and other measures relating to trade in goods.

(3) Pursuant to those general rules, the goods described in column 1 of the table set out in the Annex should be classified under the CN codes indicated in column 2, by virtue of the reasons set out in column 3.

112. Article 1 provides:

The goods described in column 1 of the table set out in the Annex shall be classified within the Combined Nomenclature under the CN codes indicated in column 2.

Annex 1 Column 1

3. A network analyser, consisting of an analyser module, a capture memory and an interface to an automatic data processing (ADP) machine, in a single housing.

The analyser is designed to provide information on the performance of networks by monitoring network activity, decoding all major protocols, and generating network traffic.

The ADP machine is not presented with the analyser.

9031 80 39

Classification is determined by the provisions of General Rules 1 and 6 for the interpretation of the Combined Nomenclature, Note 5(E) to Chapter 84, Additional Note 1 to Chapter 90 and by the wording of CN codes 9031, 9031 80 and 9031 80 39.

The analyser, performing a specific function by means of the analyser module, is excluded from heading 8471 by application of note 5(E) to Chapter 84.

The analyser is specifically designed for analysing the traffic in a network and not for measuring or checking electrical quantities, thus being excluded from heading 9030.

4. A network analyser consisting of a central management bus, an analyser module, an automatic data processing machine, a monitor and a keyboard, in a single housing.

The analyser is designed to perform the following functions:

— analysing the operational state of existing networks and network products,

— simulating traffic and fault conditions into existing networks and network products,

— generating network traffic.

9031 80 39

The analyser, performing a specific function by means of the analyser module, is excluded from heading 8471 by application of Note 5(E) to Chapter 84.

The analyser is specifically designed for analysing the traffic in a network and not for measuring or checking electrical quantities, thus being excluded from heading 9030.

Additional note to Chapter 90

1. For the purposes of subheadings 9015 10 10, 9015 20 10, 9015 30 10, 9015 40 10, 9015 80 11, 9015 80 19, 9024 10 11, 9024 10 13, 9024 10 19, 9024 80 11, 9024 80 19, 9025 19 20, 9025 80 40, 9026 10 21, 9026 10 29, 9026 20 20, 9026 80 20, 9027 10 10, 9027 80 11, 9027 80 13, 9027 80 17, 9030 20 91, 9030 33 10, 9030 89 30, 9031 80 32, 9031 80 34, 9031 80 38 and 9032 10 20, the expression ‘electronic’ means instruments and apparatus which incorporate one or more articles of heading 8540, 8541 or 8542 but for the purposes of the foregoing, no account shall be taken of articles of heading 8540, 8541 or 8542 which have solely the function of rectifying current or which are included in the power pack of instruments or apparatus.

113. The headings 8500, 8541 and 8542 referred to in the note above refer to the following: heading 8540 refers to “Thermionic, cold cathode or photocathode valves and tubes (for example, vacuum or vapour or gas filled valves and tubes, mercury arc rectifying valves and tubes, cathode ray tubes, television camera tubes), heading 8541 refers to “Diodes, transistors and similar semiconductor devices; photosensitive semiconductor devices, including photovoltaic cells whether or not assembled in modules or made up into panels; light-emitting diodes; mounted piezoelectric crystals:”, and heading 8542 refers to “Electronic integrated circuits”.

114. Commission Regulation (EC) No. 1179/2009 which came into force in early 2010 subsequently made amendments to the codes in Regulation 129/2005. The recitals to the amending regulation refer to the need to amend Regulations concerning the classification of goods because they make reference to codes which no longer exist or are no longer valid. Accordingly per Article 1 of Regulation 1179/2009 Annex 1 point 184 the references in Regulation 129/2005 to 9031 80 39 were amended to refer to 9031 80 38.

115. The appellant says the product does not meet the characteristics set out in EC Regulation 129/2005 because:

(1) There is no interface to an external Automatic Data Processing machine provided in the appliance

(2) The appliance does not generate network traffic

(3) The Automatic Data Processing machine is present in the appliance housing.

116. HMRC’s argument is that they are not saying that falling within Regulation 129/2005 is only way for products to be viewed as having the function of a network analyser. They argue that if the characteristics mentioned in the Regulation were meant to be exhaustive they would have been written into Annex 1 to Council Regulation 2658/87 (through the annual process for updating the CN Regulation described at [46] above). Article 1 of the Regulation 129/2005 says goods shall be classified “within” the Combined Nomenclature under the CN codes. The Regulations set out items which Member States must find to be classified within the classification but do not preclude other products falling within the relevant heading. The intrinsic properties of the InfiniStream are that it falls within heading 9031.

117. We accept HMRC’s analysis of the non-exhaustive nature of Regulation 129/2005 is correct and this is supported by the recitals to the Regulation which provide an aid to the interpretation of the Regulation. Recital 1 refers to it being necessary “to adopt measures concerning classification of the goods referred to in the Annex to this Regulation” (emphasis added). The starting point is defining certain goods and then specifying where they are to be classified within a broader system of classification. In other words the Regulation is concerned with putting certain goods into a certain category and is not about defining the boundaries of a category as applied to goods more generally. As Mr Donmall put it just because a banana is a fruit it does not mean something else cannot also be a fruit on the basis that it is not a banana. By the same logic just because a certain product as described in Regulation 129/2005 falls within the specified heading does not mean another product cannot also fall within that heading.

118. The analogy is helpful in explaining what is meant by an argument about whether rules are exhaustive or not but we are conscious that it is a restatement of the point which is being argued about rather than a reason for reaching a conclusion on the point.

119. To be clear, the reasons why we accept HMRC’s view on the non-exhaustive status of the Regulation as being the better one is firstly because of the way in which the recitals explaining the purpose of the Regulation have been drafted and secondly because of the existence of other legislative mechanisms for updating the CN codes definitively in the form of annual updating of the CN Regulation by means of Regulation.

Interpretative value of Regulation 129/2005

120. The question arises, even if it is accepted as we do that Regulation 129/2005 is non-exhaustive, as to what the interpretative significance, if any, of the Regulation is when considering products which do not fall within its terms.

121. One approach would be to say that because the InfiniStream does not fall within Regulation 129/2005 we should put the Regulation to one side and look again at the headings in the CN. The other is to take account of the Regulation. There is then a disagreement. The appellant’s argument is that the regulation sets out fundamental attributes of a “network analyser” and if those attributes are not present as with the product here then the product cannot be a “network analyser” and cannot be said to be performing network analysis as a specific function. HMRC disagree the attributes in set out in Regulation 129/2005 are fundamental in this way.

122. Because the Regulation is not exhaustive there is the possibility that there are products which are network analysers but which are not covered by the Regulation. But, even so there must come a point where a product is so different from the products set out in the Regulation that it may no longer be viewed as a network analyser or as a measuring or checking device such as to justify it being place under heading 9031. Accordingly in our view it is relevant to consider whether the attributes, the lack of which results in the InfiniStream falling outside the Regulation, are so inimical to the relevant heading in the CN that the InfiniStream falls outside of the heading.

Extent to which characteristics of the products set out in Regulation are met by Inifinistream

123. HMRC do not accept the functionalities of simulating traffic or fault conditions into network traffic or that of generating network traffic are needed to fall within Regulation 129/2005 and in any case they say they do not need to establish the product is exactly the same to fall within 9031 80 38.

124. HMRC say it does not matter whether the automatic data processing machine is within the product or whether it is presented outside of it, given both are possible under codes 9031 80 38 and 9031 80 39 as set out in Regulation 129/2005. The issue of whether the product generates traffic, although present in the descriptions for both the devices mentioned in paragraph 3 and 4 of Article 1 of the Regulation, does not make a difference as to whether the product can perform the underlying nature of monitoring and analysing network performance. This attribute and other points of difference between the InfiniStream and the products mentioned in the Regulation do not establish the product is not a network analyser.

125. The appellant says HMRC is wrong to pick and choose between attributes set out in Regulation 129/2005. The items which are common to paragraphs 3 and 4 in the Regulation are that the automatic data processing machine is contained in a single housing and it generates traffic. If these attribute were not material they could have been omitted. It must have been thought that these attributes were relevant to defining a network analyser otherwise why would they have been specified? A number of the attributes specified are missing from the InfiniStream e.g. the need to generate network traffic. It should not therefore be treated as a network analyser.

126. Mr Harris who had examined the appellant’s application accepted it did not generate network traffic. It was not in dispute that the product did not generate network traffic. There was also no evidence before us which suggested the product simulated fault conditions.

127. In relation to conditions which are not common to both paragraph 3 and 4 which describe products which both lead to classification under the same 9031 code in our view it is difficult to see how the absence of one of those would mean a product could not be within 9031. The absence of simulation of traffic and fault conditions is not therefore something which means a device could not be capable of falling within heading 9031. The generation of network traffic characteristic is present in both paragraph 3 and 4 of Article 1 of Regulation 129/2005 and bears further consideration.

Significance of whether device generates network traffic

128. After the hearing it came to the Tribunal’s attention that there had been CJEU case in which issues relating to classification of network analysers were raised (Case C-227/11 DHL Danzas Air & Ocean (Netherlands) BV v Inspecteur van de Belastingdienst/Douane West, kantoor Hoofddorp Saturnusstraat ). The Tribunal asked the parties to obtain an English translation of the decision and invited written submissions which were received from the appellant and HMRC on 30 April 2013 and 1 May 2013 respectively.

DHL Danzas

129. The relevant parts of the case were helpfully summarised in the Respondents’ submissions which we largely set out below.

130. In DHL Danzas, the referring court in the Netherlands, the Rechtbank Haarlem, was seeking to determine an appeal relating to customs duties payable in respect of 53 different models of network analyser in a period between 1 February 2005 and 30 September 2006 [27] and [28]). DHL Danzas contended that the network analysers should have been classified under 9030 40 00. The customs authorities in the Netherlands had classified under 9031 80 39, the same classification as given by the Commissioners to the InfiniStream network analyser in this case.

131. At [32], the referring court had explained that it was only concerned with the classification of so-called ‘active’ analysers.

132. The referring court explains that the various network analysers are divided, for the purposes of their classification in the CN, into two categories, that is so-called “passive” analysers which simply measure, display and analyse data, and so-called “active” analysers which can in addition generate data traffic and advise on solutions to problems found. It states that it has no doubts as to the CN classification of passive network analysers, which, moreover, are not covered by Regulation 129/2005.

133. The referring court referred two questions:

“Must the active network analysers [of the type J6801B] be classified under heading 9030 40 or under heading 9031 80?

Is Commission Regulation (EC) No 129/2005 of 20 January 2005 invalid because in that regulation the Commission incorrectly classified the network analysers mentioned in points 3 and 4, namely, under CN code 9031 80 39, instead of under code 9030 40?”

134. As to the second question, the CJEU found that the question of invalidity of Regulation 129/2005 in the light of a subsequent Classification Opinion of the HS Committee simply did not arise, because that opinion was adopted in September 2010, which was after the date of the imports in the referred case [44]. It noted, however, that such a classification opinion did not have legally binding force, and would be merely an interpretative aid [44].

135. The CJEU found in respect of the first question that it was a question of fact for resolution by the referring court [52], stressing that the question was whether “the network analysers in question in the main proceedings have the very purpose of carrying out checks on electrical quantities”. This was pursuant to its having observed, at [51], in respect of heading 9030:

“More specifically with regard to CN heading 9030, the Court has repeatedly held that this heading only covers, according to its very wording, instruments and apparatus for measuring or checking electrical quantities, that the same interpretation, based on the purpose of the apparatus in question, must also be applied to define the content of this heading and that only apparatus whose very purpose is to carry out checks on electrical quantities can be regarded as apparatus for checking such quantities (Case C-218/89 Shimadzu Europa [1990] ECR I-4391, paragraphs 9 to 11, and Case C-108/92 Astro-Med [1993] ECR I-3797, paragraph 8).”

136. HMRC understand that the referring court has subsequently found that, contrary to Danzas’ case that the classification was 9030, the classification for the network analysers was 9031. HMRC further understand that this decision has in turn been appealed by DHL Danzas.

137. The appellant argues that the InfiniStream is “passive” in that it does not generate network traffic and as such that it should not be classified under 9031 80 39 or 9031 8038. It notes that although the alternative code in contention in DHL Danzas was different (9030 40 as opposed to 8471 70 50 90) both attract 0% duty rather than the 4% duty which 9031 attracts.

138. HMRC highlight that the distinction between network analysers which generated data traffic from those that did not was one that was made by the referring court, not the CJEU and the CJEU did not comment on it. The CJEU did not consider the question of whether the generation of data traffic was a necessary element for a device to come within the scope of regulation 129/2005 still less 9031 80 39.

139. They disagree in any case that the Inifinistream would fall within the Dutch referring court’s definition of a “passive analyser” as the data sheets demonstrate the product monitors network performance and proactively issues alerts, inspecting network traffic and decoding protocols.

140. While they say DHL Danzas makes it clear that unless the domestic court considered that “the very purpose [of the machine was] of carrying out checks on electrical quantities” it would be classified under 9031 80 39, HMRC argue the InfiniStream cannot be considered to have “checks on electrical quantities” as its “very purpose”. On the contrary its purpose is checking networks with proactive functionality.

Our views

141. We do not think we need to make a finding on whether the device is “passive” within the definition described by the Dutch referring court. It is sufficient to note our finding of fact that the InfiniStream does not generate network traffic. That means it cannot fall under the scope of Regulation 129/2005. But, as discussed above, Regulation 129/2005 does not exhaustively describe all network analyser devices that can fall under heading 9031. So, the fact the InfiniStream does not generate network traffic is not fatal to a 9031 classification.

142. As pointed out by HMRC the CJEU did not consider the question of whether generation of data traffic (which we equate as generation of network traffic) was a necessary element for a device to come within the scope of 9031 80 39. While the referring court appears to have come to the view that if an analyser was “passive” this would resolve which of headings 9030 or 9031 were relevant the CJEU did not comment on this. In any case there is nothing to suggest the CJEU’s statement as to devices falling under 9030 only if “the very purpose [of the machine was] of carrying out checks on electrical quantities” would not also be relevant to devices which did not generate network traffic. Such devices would not inevitably fall into 9030 or out of 9031. In our view there is nothing in DHL Danzas which suggests that a device which does not generate network traffic cannot nevertheless fall under heading 9031.

143. For the sake of completeness we note the appellant does not seek to argue the InfiniStream is a device whose very purpose is to carry out checks on electrical quantities. Even it had, we do not think that there is anything in the evidence before us that would support such an argument. While the appellant’s observation on 9030 also being a code which does not attract duty, the issue of duty is not relevant to classification which as stated above at [85] is to be the product’s objective characteristic and properties.

144. Returning to the question of whether generation of network traffic is a necessary requirement for falling under heading 9031, it seems to us that generation of network traffic is a means by which the network may be analysed. There is nothing to suggest within the terms “measuring or checking” that if a product is able to measure or check a network by means other than itself generating traffic that it would not nevertheless be measuring and checking the network. While the feature of generating network traffic is we think essential for the device to be a “network analyser” within the scope of Regulation 129/2005 the fact a product misses that attribute but nevertheless has the function of monitoring network traffic is not precluded from falling under the heading 9031 which refers to measuring or checking instruments.

145. The products mentioned in Regulation 129/2005 are similar but not identical to the InfiniStream. The fact that such products are placed under heading 9031 is, we think, a strong indicator that that heading is also the correct one for the InfiniStream. We note that Regulation 129/2005 refers amongst other things to Note 5(E) by way of explanation for why the device mentioned there is appropriate for the particular heading 9031. This indicates it was thought that the heading of 9031 “measuring or checking instruments…” was a more specific function than data processing.

Application of GIRs

146. We did not receive specific submissions on the application of the GIRs but we note that the opening paragraph of them provides that classification shall be determined according to the terms of the headings and any relevant section or chapter notes. The Notes 5(A) and 5(E) discussed above are binding.

Data storage argument – classification 8471 70 50

147. Although the appellant did not develop this argument at the hearing it was mentioned in their correspondence in relation to the appeal and we considered it. The appellant’s argument in relation to this heading was that the product simply stored network data on an array of hard disk drives. It was then a separate and subsequent step for users to retrieve and analyse the stored data.

148. HMRC argue the function cannot be one of data storage. The data sheet refers variously to proactively issuing alerts, resolving problems with network traffic, undertaking decoding of protocols and packet slicing.

149. The product undoubtedly does store data; the data sheet mentions the InfiniStream’s storage capacity. However, the product’s function as shown by its data sheet indicates to us that the characteristics of the product, including the way in which it is intended to be used, go beyond data storage. The data storage is “in order to monitor, measure, manage and resolve high impact intermittent network problems” and the focus of the data sheet is very much on the network monitoring aspects and features of the product. We did not have evidence before to support the appellant’s submission that the measuring, checking or analysis was performed separately and that submission was certainly not borne out by the emphasis placed on monitoring and analysis in the product’s data sheet. Further we noted the evidence that the Inifinistream stores data intelligently. It does not store data (e.g. encrypted data) which is not susceptible to analysis. The intelligent way in which data is stored with a view to the analysis of the data supports our view that the function of the product is to analyse networks.

Relevance of US coding

150. The appellant says the US manufacturer and exporter of the equipment uses a non-duty attracting code to export the equipment:

(1) The US manufacturer exports under CN number of 8471 50 and has not been requested to alter the code

(2) The US Government have agreed that Harmonised Tariff 8471 50 is appropriate.

151. At the hearing the appellant handed up a chain of e-mail correspondence between Mr Simmonds and the US manufacturer, Netscout, relating to the commodity code for the InfiniStream used by Netscout. The chain contained an exchange of views between 15 December 2011 and 16 December 2011 and a further set of e-mails in the period 10 July 2012 to 19 July 2012 following Mr Simmond’s request for any correspondence Netscout had received from the US Government which confirmed “HTS 8471 50 0150”. Netscout said they used an outside consultant to classify their products and did not have anything from the US Government. They subsequently attached an HTS classification from the US Export Assistance Center from 1999.

152. The attachment was a copy of a fax from the US Export Assistance Center which stated that arrows were placed next to codes Census thought best suited the product based on Netscout’s description. It was not explained to us who Census was but we understand the reference to “Census” may refer to the US Department of Census, and that they would be interested or involved in the tariff code matters for the purpose of maintaining trade statistics. The header of the fax referred to the date 21 October 1999 “from US Dept. of Commerce”.

153. HMRC were given an opportunity to consider the documents and no objection was taken to their admission in evidence.

154. HMRC submitted the correspondence was an insufficient basis to make a finding as to whether the US authorites responsible for classification had made a tariff classification akin to something like the process for BTI. Even if the US authority had made such a classification HMRC say that according to Intermodal the decisive criterion is the objective characteristics of the goods not how others have classified them. The difficulty with looking at how other authorities have classified products is what is to say the other authority is correct?

155. We agree there is insufficient material before us to say the relevant US authority has made a tariff classification. We accept however that the US manufacturer uses a classification under heading under 8471 as opposed to 9031.

156. Even if there was evidence as to the classification made by a competent US authority it is debatable what relevance this would have to the issue before us. While we do not think it would be irrelevant in the way HMRC seem to suggest, taking account of the decision in Intermodal, at best it would be a matter to take account of. It certainly would not be determinative. This is on the basis that in Intermodal determinations of other Member State authorities were not determinative although they would be reason for the court or tribunal considering classification to take “particular care” in its assessment. That being the case in relation to other Member State authorities it would be odd if the Tribunal were required to give any greater weight to determinations by authorities from non Member States.

157. Further it is not clear what if any differences there would be in the classification used by the third country and in particular whether they would have anything similar to the classification Regulation 129/2005 to take account of. At its highest, the most that can be said is that if the US authority had given the classification as suggested, this would be something which would cause us to take particular care over our assessment of classification (the level of care that we consider we have taken in relation to classification in any event). A US authority classification would not however be something which would be determinative of our classification. In the same way that HMRC’s stated view as to the correct classification does not necessarily establish that such classification is correct, the views of a third country authority or for that matter another Member State authority do not of themselves establish what is the correct classification.

Competitive disadvantage / actions against competitors

158. The appellant argues that because of HMRC’s classification, which results in 4% duty on the InfiniStream, their prices have had to be 4% higher. They have, they say, lost out to competitors because of this. They also query why no enforcement action has been taken to subject their competitors to the same classification. The appellant also queried the extent to which any ambiguities about classification had been raised at European meetings that HMRC had attended and how such ambiguities had been dealt with at a worldwide level.

159. The issue before the Tribunal is the determination of the classification of the InfiniStream according to the product’s objective characteristics based on evidence that has been provided to us of those. The issue of what action or lack of action HMRC have taken in relation to competitors and any resulting disadvantage the appellant may have suffered if they are subject to the duty but others are not are matters relating to HMRC’s conduct and for which other avenues of recourse are in principle open. The issue is not however relevant to the classification of the InfiniStream. It is not within the remit of this Tribunal hearing an appeal of this type to deal with the exercise by HMRC of it enforcement powers in relation to others. The conduct of HMRC in relation to its attendance or involvement in European or worldwide meetings on classification is also not relevant to the issue before us.

Conclusion

160. Although we accept that the InfiniStream prima facie falls within heading 8471 (automatic data processing machine) and also that it does perform a function of data storage, the device in our view does have a more specific of measuring and checking networks. It is similar, but not identical, to other products which are required by Regulation 129/2005 to be classified under heading 9031 80 38. The InfiniStream therefore falls under heading 9031. The appellant has not satisfied us that HMRC’s classification 9031 80 38 90 is incorrect and therefore that classification is upheld.

Duty demand

161. In relation to the appeal against the duty demand while the determination of that is in at least some part dependent on the outcome of the decision on classification there was an outstanding issue over whether some element of the demand related to non InfiniStream products and over the amount of certain refunds.

162. The parties have asked us to reserve the determination of amount in the duty demand appeal. While we have decided the issue on classification in HMRC’s favour we accordingly reserve the issue of amount on the duty demand and if this cannot be agreed the parties have permission to revert to the Tribunal on that issue.

163. 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.

TRIBUNAL JUDGE

SWAMI RAGHAVAN

	AXIAL SYSTEMS LIMITED	Appellant

	- and -

	THE COMMISSIONERS FOR HER MAJESTY’S	Respondents
	REVENUE & CUSTOMS

First-tier Tribunal (Tax)