First-tier Tribunal (Tax)

[2016] UKFTT 788 (TC)

TC05515

 

Appeal number: TC/2013/09345 and TC/2013/09347

 

CUSTOMS CLASSIFICATION - multi functional machines - Kip/Hewlett-Packard C-362/07 and C-363/07, ECLI:EU:C:2008:710 and Hewlett-Packard C-361/11, (ECLI:EU:C:2013:18 applied - appeal allowed

 

 

FIRST-TIER TRIBUNAL

TAX CHAMBER

 

 

 

	BROTHER INDUSTRIES UK LIMITED -AND- BROTHER UK LIMITED	Appellant
	- and -
	THE COMMISSIONERS FOR HER MAJESTY'S	Respondents
	REVENUE & CUSTOMS

 

 

 

TRIBUNAL:

JUDGE Barbara Mosedale

 

 

 

 

 

Sitting in public at the Royal Courts of Justice on 21-23 June 2016

 

 

Mr F Mitchell, Counsel, instructed by PwC Legal LLP, for the Appellant

 

Mr J Shields, Counsel, instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents

 

© CROWN COPYRIGHT 2016

DECISION

 

1.               The dispute between the parties arises out of repayment claims submitted by the appellants in respect of their importation of multifunction machines ('MFM's) in 2005 and 2006.

2.               The appellant considers that they should have been classified under Heading 8471 of the Combined Nomenclature ('CN') with a nil duty rate; whereas HMRC considers that they were rightly classified under CN Heading 9009 with a 6% duty rate and the repayment claims were therefore rightly refused. The claims were for some £413,439.33 it total. The decision under appeal was dated 30 August 2013.

THE FACTS

The evidence

Mr Thomas

3.               The appellant called one witness, Mr Andrew Thomas. Mr Thomas had worked with computers since leaving school in 1984. He joined a Brother company in 1988 working as technical support on printers and faxes; in 1995 he worked as senior network specialist, advising colleagues in Brother on the capabilities of, amongst other things, brother printers and MFMs. His role developed and by 2005 he was involved in product planning with responsibility for developing printers and MFMs. Currently he decides the specification and design of printers and MFMs that Brother sells in Europe (which means adapting Japanese machines for the European market).

4.               A company in the appellants' group had been in dispute with HMRC over the classification of their MFMs before. In the case of Brother International Europe Ltd (2008) C00248 Mr Thomas had given evidence. In this appeal some of the MFMs in issue were the identical units to some of those at issue in the earlier appeal. He confirmed he stood by the evidence he had given earlier and accepted the summary of it recorded in the published decision.

5.               I found Mr Thomas to be a consistent and reliable witness and accepted his evidence.

HMRC witnesses

6.               I mention in passing that an HMRC officer, Mr Harris, submitted a witness statement. In it he disagreed with various matters stated by Mr Thomas in his witness statement but without actually putting forward evidence of his own. So, as Mr Harris' statement comprised opinion, it was agreed Mr Harris would not be called and the Tribunal would not accept the contents of his witness statement as evidence: the opinions he expressed could be adopted as submissions and his challenges could be put to Mr Thomas in cross-examination.

7.               Another HMRC officer, Mr Welland, also submitted a witness statement. This gave evidence on the procedural background to the appeal and the appellant accepted it was accurate and did not require him for cross examination. Therefore, I found the contents of this witness statement reliable, but in practice it was not relevant to any of the issues I was called on to decide and I do not refer to it again.

The exhibit

8.               The exhibits: the appellant had been able to find one of the (now old) MFMs at issue in this appeal (MFC-7420) together with the printer on which it was based and bring it to the Tribunal hearing. Mr Thomas demonstrated by removing screws from the MFM that it was possible to take off the automatic document feeder ('ADF') and scanner to reveal the printer module. I mention this at §12 below.

Findings of Facts

The MFMs

9.               The MFMs in issue all looked a bit like a small photocopier, only capable of scanning and printing A4, and somewhat larger than a standard small office printer. The expected market for the machines were home workers and small offices.

10.            The MFMs were manufactured by Brother by modifying a printer in order to build in three extra modules. This meant, for instance, that the parts common to both printer and MFM were the same, for instance the paper trays and toner cartridge (which fitted in the printer module of the MFM) were identical and interchangeable between the MFM and the printer on which it was based.

11.            The first extra module was a fax module which was quite small and was inserted into the printer component without increasing its external dimensions. Not all the MFMs at issue in this appeal had a fax module. The second module to be added was a scanner module which was basically a squarish component a few inches high, containing a glass flat-bed, which fitted on top of the printer, instead of the hard plastic cover which would normally cover a printer. Then on all the MFMs at issue in this appeal (save one basic entry model, DCP-7010) an ADF was fitted on top of the scanner. This was a hinged component of about the same size as the scanner and which obviated the need to place individual sheets on the platen glass. In other words, adding the scanner and ADF modules to the printer unit created a larger machine which looked like a small photocopier, complete with hinged lid to access the platen glass.

12.            I find that once the 'extras' were removed from the MFM and apart from the hard top which covered the printer, the MFM and the printer on which it was based would look, and in practice were, virtually identical. In practice, of course, once the machines were manufactured, the plastic mouldings were necessarily different and a printer could not actually have a scanner and ADF added to it to convert it to a MFM, and while an MFM with the ADF and scanner removed might be capable of functioning as a printer, it would lack the hard plastic top a printer would normally have.

13.            The appellant's case was that its MFMs were a printer with added extras.

Functionality of the MFMs

14.            Some 16 machines were in issue. Although similar to each other, I have set out a summary of their individual features that indicates the distinctions between them in an appendix at the end of this decision.

15.            The functionality of the modules was as follows:

(1) Printer module - designed to print output from three sources of electronic data:

(a) if connected, as it was designed to be, to a computer, it could print electronic data sent to it from that computer. In other words, it operated as a printer, dedicated or networked.

(b) irrespective of whether it was connected to a computer, it could print electronic data sent to it from the scanner. In other words, it operated as a standalone photocopier.

(c) irrespective of whether it was connected to a computer, it could print electronic data sent to it from the fax module. In other words, it operated as a standalone fax machine receiving faxes.

(2) Scanner module: designed to convert hardcopy documents into electronic data and capable of sending the data to three destinations:

(a) irrespective of whether it was connected to a computer, it could print scanned documents - this is item (b) above. The combination of scanner and printer modules enabled it to operate as a standalone photocopier.

(b) irrespective of whether it was connected to a computer, it could fax scanned documents. In other words, it operated as a standalone fax machine sending faxes.

(c) if it was connected as it was designed to be to a computer, it could send a scanned document to that computer in the form of an electronic file.

(3) automatic document feeder module : the ADF was designed to automate the presentation of documents to the glass flat bed of the scanner. It could be used whether or not connected to a computer. It facilitated the use of the scanner for any one of its three capabilities mentioned above.

(4) fax module : designed to fax documents and to receive faxed documents. To operate it had to be connected to a phone line. It could receive data for outgoing faxes from two sources and could send data from incoming faxes to two destinations:

Outgoing

(a) - this is point (3)(b) above- irrespective of whether it was connected to a computer, the fax module could fax electronic data sent to it from the scanner and the machine could operate as a standalone fax machine;

(b) if connected to a computer, the fax module could fax electronic data sent to it from the computer;

Incoming

(a) - this is point (1)(c) above - irrespective of whether it was connected to a computer, it could receive a fax and send the electronic data comprising the incoming fax to the printer module for printing;

(b) if connected to a computer, it could send the electronic data comprising the incoming fax to the computer as an electronic file.

 

16.            All modules in appeal had printer and scanner modules and could therefore print, scan and copy. There was no photocopying module: the ability to photocopy arose because of the combination of scanner and printer. Most but not all MFMs had the additional fax functionality described above. And while some of the MFMs' functionality existed irrespective of whether it was connected to a computer, it was clearly designed to be permanently attached to a computer/network because it would lose so much of its functionality if it was not.

The micro-processor

17.            In reality, the above explanation might be thought misleading because what really happened was that the machine itself contained a micro-processor and it was that micro-processor which channelled the data from the various input sources to the various output destinations. It would be wrong to envisage the various components all acting independently and communicating with each other: the micro-processor, albeit very small, was the hub of the machine, receiving and sending electronic data from and to the appropriate parts of the MFM or to the attached computer/network.

18.            For example, as I have said, the fax module was connected to a telephone line. Data comprising the incoming fax was passed to the MFM's micro-processor. That micro-processor then sent the data to its destination, whether it was the printer module of the MFM or to a computer connected to the MFM. Similarly, when sending a fax the micro-processor received the data from the two possible sources (an attached computer or the scanner module) and sent that data to the fax module to send down the telephone line.

19.            The same was true of the scanner module. It scanned hardcopy documents and passed the electronic data to the micro-processor., which converted it and passed the electronic file to the printer or attached computer.

20.            The same was true of the printer module. The micro-processor received data either from the scanner or attached computer and sent it to the printer for printing.

21.            The micro-processor was not a module added to the printer during manufacture to convert the printer into an MFM. The micro-processor was a part of the printer module: a standalone printer module had to have the micro-processor in any event in order to operate as a printer (to receive electronic data from the attached computer and convert it to information the printer could print). What was added at some point during production of the MFM was extra software to run the micro-processor so that it could carry out the additional functions of the MFM as outlined above. It involved no physical change to the printer unit.

Alternatives

22.            If a user did not purchase an MFM, to obtain virtually the same functionality they would have to buy a standalone printer, standalone scanner, standalone photocopier and (for the MFMs with fax) a standalone fax machine. As that would involve having three or four machines rather than one, the machines would take up more space, more electric power, more connection cables, and would require more consumables, such as toner cartridge, than an MFM.

23.            I accept that having an MFM involved some small compromise on functionality over having three or four individual machines: a standalone scanner manufactured by Brother would be likely to have a higher resolution and higher speed than the scanner module in the MFMs. The printer module in the MFM, on the other hand, had the identical functionality with the printer model on which it was based (with the one exception mentioned below relating to the control panel).

Control panel

24.            The MFMs all had a control panel which comprised an LCD screen and control buttons. Brother's standalone printers did not have control panels, merely buttons which flashed to indicate matters like low toner.

25.            Where the MFM had a fax module, a control panel was vital to enable the user to check the inputted fax number. But even the models without fax modules had control panels so they were clearly seen as vital to the photocopying function as well. The control panels would include buttons limited to the copying function while other buttons related to the machine's other functions. The control panel was utilised for the printer to display error messages like 'low toner' whereas, as I have said, a standalone printer would merely have flashing buttons.

26.            The evidence was therefore clear that while the control panel was co-opted for the printer module, it was installed because of the existence of the other modules and photocopying functionality.

Relative sophistication of modules

Printer speed

27.            The appellant's advertising literature stated that the printer and photocopier operated at the same speed. As Mr Thomas explained, that is only true if looking at the speed at which the MFM could print multiple copies of a document which it had already scanned. It is the nature of the process that copies of different documents take much longer to print out than printing a multi-page document from a computer, because each document has to be scanned separately.

28.            The printing speed of the various MFMs in this appeal varied depended on their specification: high-end MFMs printed at a faster speed than the low-end ones. This is recorded in the appendix at the end of the decision: in summary all machines printed at 20 pages per minute, except five which printed at 28 p/m and one which printed at 10 p/m.

Colour

29.            The machines could only print or photocopy in black and white. However, all bar 3 machines could scan in colour, which meant that colour copies of originals could be sent to the attached computer/network albeit they could not be printed out colour on the MFM.

Printer resolution

30.            The MFMs had more choice of resolution for the printing than for the photocopying and the best printing resolution was better than the best copying resolution. Because the scanner's resolution was the same as the printer's, it was possible to achieve the best printing resolution for a photocopy by scanning the original to an electronic file on the attached computer, and then printing that file. But if operated as a standalone photocopier, the best resolution for copying was lower than the best resolution for the printer of which the MFM was capable.

31.            HMRC's case was that it was inherent in copying process that the resolution would not be as good as printing: but I accept Mr Thomas' evidence recorded in the previous paragraph that this was not so. It would have been possible for the photocopying process (with better software) to be at the same resolution as the printer process: the fact that it was not indicates that the machine was more of a printer with added photocopier than one where the printer and photocopier were of equivalent standard.

Quantity printed

32.            The MFMs all had a paper feed tray capable of taking 250 sheets of A4 (bar one which only took 200). All had a manual paper feed slot, most of which were capable of only taking a single sheet of A4 but some of which could take up to 50 A4 sheets at a time. Some of them could have an optional extra paper feed tray added but this was irrelevant to the appeal in that the MFMs were imported without the optional extra and customs duty classification could therefore not be affected by it.

33.            In practice, the printer could therefore print 250 sheets without needing to be re-filled. The scanner/copier on the other hand could only scan/print without needing to be re-filled as many originals as its ADF would hold. As indicated in the appendix below, the ADFs only held 20-50 pages at a time (depending on model). One model did not even have an ADF.

34.            The appellant's case was that this made the MFM impractical as anything other than an occasional photocopier and it was primarily a printer. I agree that if measured by output the MFMs were more efficient as printers than scanners or photocopiers.

35.            I also accept the appellant's case that the manual document feed was a functionality that was more likely to be of use to the printing module than scanner module: the obvious purpose of photocopying is to reproduce the original whereas the obvious purpose of the manual document feed was to allow printing on 'special' paper such as headed paper. It would be unusual to photocopy onto anything other than plain paper. It was, in any event, functionality built into the printer module and not added for the other modules/functionality.

Duplex copying function

36.            Duplex meant that the machine could print double-sided: it was a mechanical functionality built into some of the printer modules. Those MFMs with duplex functionality could print or copy double-sided.

37.            I accept Mr Thomas' evidence that it was possible for non-duplex machines to print double-sided by the operator giving the computer the instructions to print every other page and then manually inserting the printed pages into the paper feed tray and instructing the computer to print the missing pages: it would be much more time consuming and difficult to achieve double-sided photocopies with a non-duplex machine. I don't consider this point, however, relevant to the issues in this appeal as it seemed to me to be inherent in the different nature of printer and copier.

Exclusive functionality?

38.            The printer module has some functionality not possessed by the photocopier function: it could watermark prints and some of the MFMs could produce A5 booklets.

39.            In reality of course, this additional functionality could be utilised for any scanned document by creating an electronic folder and sending it to the attached computer, and then printing it from the computer. But it could not be utilised by the standalone copier functionality. HMRC pointed out, as I accept, that a 'watermark' function or booklet function were not really necessary for photocopying where the object is to reproduce an original.

Relative sizes of components

40.            I accept that the printer module was much larger than the scanner or ADF modules. The fax module was so small that it fitted inside the printer module.

41.            HMRC's point was that the size of the printer module was to a large extent dictated by its need to be physically large enough to hold the toner cartridge and to hold a stack of, and print, A4 paper. This need to accommodate A4 paper also explained in part the size of the scanner and ADF.

Relative costs of modules

42.            I find that the printer module was the most expensive. It was always more than 50% of the cost of the MDM and in some cases was up to 71% of the cost. This was perhaps not surprising: it was the largest module, contained the paper feed tray and micro-processor. While HMRC pointed out that the appellant's costings ignored certain costs, I accept Mr Thomas' evidence that had they been included, the comparative cost of the printer module would have increased and not decreased. The appellant did not, for instance, include development and support costs, but I accept Mr Thomas' evidence that these were always higher for the printer module because of the constant need to ensure that the printer was able to interact with all possible kinds of PCs and operating systems.

Developmental history

43.            Brother started out as a manufacturer of printers and created its first MFM in 1995 due to market demand. It achieved this by adapting its printers to include fax and scanner modules and thereby creating something with above described functionality. It never manufactured standalone copiers.

44.            The development history of MFMs in general is that they became possible to manufacturer when indirect photocopying became possible because indirect photocopying only requires a scanner and printer. Originally photocopiers were analogue and operated like a camera: they would take an image based on exposing the original to light, and reflecting it to create a negative. But digital copying was then developed: the original image is scanned to create a digital file and then printed. The MFMs were indirect copiers, as they utilised the scanner and printer to create copies of originals, and did not actually 'photocopy'. The CJEU in Rank Xerox C-67/95 held that 'photocopying' included indirect copying.

THE LAW

General principles

45.            The parties were agreed on many aspects of the law. They were agreed that the Combined Nomanclature ('CN') published each year by Council Regulation is binding. They were agreed that a product can only be classified under one CN code and that in the case of the MFMs at issue in this appeal there were only two CN codes which were potentially applicable.

46.            The CN heading 8471 (contended for by the appellant) read in both 2005 and 2006 as:

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.

 

47.            The CN heading 9009 (actually applied to the goods at the time and contended for by HMRC) read in both years as:

9009	Photocopying apparatus incorporating an optical system or of the contact type and themo-copying apparatus

 

48.            While, in order to obtain a duty rate, an imported good had to be assigned under an 8-digit code, the dispute between the parties was at this 4-digit heading level. In other words, they were agreed that if the goods came under 8471, then they would be classed as 'printers' under the sub-heading 8471 60 and sub-sub-heading 8471 60 40:

8471           8471 60   8471 60 40*

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. .... -Input or output units, whether or not containing storage units in the same housing: .... ---Printers

 

*note that in 2006, while the wording was in so far as relevant unchanged, the code for printers was revised to 8471 60 20, still with a nil duty rate.

49.            And if HMRC were right, the parties were agreed that the MFMs in issue should have been, as they were, classed under 9009, then they were photocopiers using an indirect process, which were classifiable under 9009 12 00:

9009         9009 12 00

Photocopying apparatus incorporating an optical system or of the contact type and themo-copying apparatus: -Electrostatic photocopying apparatus ... --Operating by reproducing the original image via an intermediate onto the copy (indirect process) ....

 

Rules for classification - objective characteristics

50.            Both parties agreed that the Upper Tribunal set out the correct approach to classification in Barrus Ltd [2013] UKUT 449 (TCC) where the Upper Tribunal said:

[41] In our view the following principles can be derived from the authorities we have reviewed:

(1) the decision 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...

(2) the relevant criteria must be apparent from the external characteristics of the goods so that they can be easily appraised by the customs authorities...

(3) By the examination of the external characteristics the main purpose of the product must be inferred. It does not matter if there are other purposes for the product...

(4) the CNENs and HSENs should be used as an aid to interpretation as can specific classification regulations, but the latter only in relation to products identical to those specifically classified...

(5) Marketing materials and a product's targeted use are not to be taken into account...

51.            The requirement to consider the item's objective characteristics was set out by the CJEU at §14-15 of Holz Geenen (C-309/98) and I consider the MFM's objective characteristics below at §§101-109.

52.            So far as (5) was concerned, the appellant submitted that marketing material could not be used in so far as indicated targeted use of the product but it was legitimate to refer to it to ascertain the specification of the product. I agree with what Tribunal in V-Tech [2016] UKFTT 43 (TC) said at [50] that marketing material was not excluded just because it was marketing, but because it did not disclose objective characteristics. In so far as it did disclose objective characteristics, it could be referred to. So I agree that in so far as marketing material disclosed the specification of the MFMs, it could be referred to as disclosing objective characteristics and some of the information summarised at §§27-42 above was taken from such material.

GIR 1 - CN headings and notes

53.            Having ascertained the objective characteristics, as stated in Barrus, they had to be measured against the terms of the headings in the CN, and the notes to the CN. This requirement is the first of the General Rules for Interpretation ('GIR') of the CN, which are contained in the Council regulation establishing the CN.

54.            Both parties were agreed that the GIRs must be applied in order and GIR 1 was relevant. It provided:

GIR 1 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 relative section or chapter notes and, provided such headings or notes do not otherwise require, according to the following provisions

Chapter 84 Note 5

55.            The terms of the headings are set out above; the only applicable 'relative section or chapter notes' were agreed to be Note 5 to Chapter 84, which so far as relevant provided as follows:

(B) Automatic data-processing machines may be in the form of systems consisting of a variable number of separate units. Subject to paragraph E below, a unit is to be regarded as being a part of a complete system if it meets all of the following conditions:

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

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

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

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

(D) Printers, .... which satisfy the conditions of paragraphs (B)(b) and (B)(b) above, are in all cases to be classified as units of heading 8471.

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

 

56.            There was a lot of discussion about the implications of these notes. Firstly, heading 8471 applied to units of automatic data-processing machines ('ADPs'). It was enough if the MFMs were units of an ADP: they did not have to comprise a complete ADP. A unit of an ADP had to meet conditions (a), (b) and (c) of Note 5(B). There was no issue surrounding conditions (b) and (c): the MFMs were clearly connectable to a separate ADP (ie the user's computer or computer network) and clearly able to both accept and deliver codes which could be used by the user's computer or network. So far as (a) was concerned, the MFMs were not 'solely' used in an ADP system: they were capable of independent operation in that they could photocopy (and those with a fax module could also send and receive faxes) without connection to any other machine. So the question was whether they were 'principally' used in an ADP system.

The decision in Kip/HP1

57.            That this is the correct analysis is apparent from the CJEU's decision in the joined cases of Kip Europe SA and others C-362/07 and Hewlett Packard International SARL C-363/07 (' Kip/HP1'). By way of background, the summary of the facts in that case show that Kip's MFMs were quite different to HP's and those at issue in this appeal. Kip's were intended for businesses which used plans: the machines were for large documents, much larger than A4. They comprised a printer module, a scanner module and an integral computer (and not merely a micro-processor).

58.            HP's MFMs, however, were more similar to those in this appeal. They did not include a computer but had a printer module and a scanner module and were capable therefore of copying as well. They were, as the machines in this case were, intended for home use and SME businesses. HP was regarded by Brother as a competitor in the market for the types of machine at issue in this appeal but Mr Thomas was unable to say without more information whether the particular machines in question in this appeal were seen as being directly in competition with the models at issue in the HP1 case.

59.            The CJEU gave its ruling on the basis of GIR 1 (ie relying on the notes to the chapter) as they said as follows:

[40]...irrespective of their intended end-use, all the machines at issue in the main proceedings are characterised by the fact, firstly, that they perform printing and electronic scanning functions in connection, directly or over a network, with automatic data-processing machines and secondly, that the copying function which they have is used autonomously.

...

Accordingly, those machines are likely simultaneously to meet the three requirements laid down in Note 5(B)(a) to (c) to Chapter 84 of the CN for them to be considered units forming part of an automatic data-processing system, capable of connection to the central processing unit and receiving or supplying data in a form usable by that system.

[43] ...although it is true that the machines at issue in the main proceedings are not of the kind used 'solely in an automatic data-processing system' within the meaning of Note 5(B)(a) to Chapter 84 of the CN, the fact remains that they are likely to be considered of the kind used 'principally' in such a system within the meaning of the same note.

Note 5(E)

60.            I have included Note 5(E) above as there was some discussion in the hearing of its relevance. The Advocate General in Kip/HP1 said that Note (E) only applies if the sole function of the MFM is one which is not data processing and it seems the CJEU approved this view. Therefore, as all the MFMs at issue in this appeal performed functions (such as printing) for which they had to be connected to an ADP system, Note 5(E) does not apply.

Note 5(C)

61.            But what did Note 5(C) mean? It seemed to mean that if the MFM fulfilled conditions (a)-(c) of Note 5(B) then it must be classified to 8471 and could not be classified to any other part of the code, such as 9009. In Kip/HP1 the CJEU said:

[56]'...if the copying function performed by the machines at issue in the main proceedings is secondary in relation to the printing and electronic scanning functions, they must be considered units of automatic data-processing machines within the meaning of Note 5(B) to Chapter 84 of the CN which, by application of Note (C) to that chapter, if they are presented in isolation, fall within heading 8471...'

Bearing in mind, as I have said, that the MFMs would fulfil the conditions of Note 5(B) if they were principally used in an ADP system, the CJEU here appears to be saying that if the MFMs at issue in that appeal were principally used in an ADP system, in other words, that their use as a copier was secondary to their use as a printer/scanner, then the MFMs must be classified to 8471.

62.            That suggests all I need determine is whether the MFMs were principally used in an ADP system (ie principally as a printer and scanner) in order to determine this appeal. If they were principally used as printer/scanner they are classified to 8471; if they were not, they cannot be classified to 8471 and will fall to be classified to the next most appropriate heading which the parties were agreed was 9009.

The relevance of GIR2

63.            The problem with that analysis is that at [49] of Kip/HP2 the CJEU stated that GIR3(b) applies if the MFMs are not principally used as units of an ADP system. GIR 3 applies where there is more than one potentially applicable chapter heading. So how did the CJEU arrive at that conclusion in the face of what Note 5(C) actually says?

64.            There is no explanation in the CJEU's decision. Mr Mitchell suggested that the printer module of the MFM would be classifiable under 8471 due to Note 5(D). My view is that the explanation must lie in the GIRs. As the parties agreed, they must be applied in order. GIR2(b) provided:

"...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 [GIR] 3."

So, because the MFM comprised more than one module, GIR 2(b) meant that GIR 3 applied if a part of the product would (if presented separately) be classed under one heading, while another part (if presented separately) would be classified under another. So even if GIR 1 did not result in the MFMs being classified under 8471, it was possible that GIR 3 could result in such classification. The Notes, such as Note 5(C) were irrelevant to GIR 2 and GIR 3. It would have been helpful if the CJEU had clarified this. However, the appellant and HMRC did not suggest this analysis was wrong. They accepted that if the MFMs were not principally used as a scanner/printer, then classification would be under GIR3. That is how the CJEU proceeded in KIP/HP1 and in HP2.

Section XVI Note 3

65.            Before moving on to consider GIR 3, I mention one other complication. Chapter 84 of the CN is within Section XVI of the CN. That section has its own notes. Note 3 of Section XVI provides:

3. Unless the context otherwise requires, composite machines consisting of two or more machines fitted together to form a whole and other machines designed for the purpose of performing two or more complimentary or alternate functions are to be classified as if consisting only of that component or as being that machine which performs the principle function.

 

66.            In Rank Xerox (above) the CJEU said at [28] that Note 3 did not apply to MFMs as they were potentially classifiable under a heading not covered by that section note; this seemed to be repeated by the CJEU in Kip/HP1 at [47] (second half) where it restricted the application of Note 3 to assigning products between the sub-headings within 8471. Both parties expressly stated that they did not rely on Note 3 as having any application in this hearing. It seems, as the Tribunal in V-tech said at [43-45], that Note 3 only applies to assign a product to sub-headings within 8471 and cannot be used, and was not used by the CJEU in Kip/HP1, to assign a product as between different chapters. GIR3 applies where there is more than one potentially applicable chapter heading.

The application of GIR3

67.            Both parties having accepted that if the MFMs were not, according to their objective characteristics, principally used in connection with an ADP system, they would nevertheless be potentially classifiable under 8471.

68.            It was HMRC's position that they were also potentially classifiable under 9009; while Mr Mitchell did not at first accept that, by the end of the hearing he accepted that an application of GIR1 meant that MFMs could also fall into 9009. In other words, the MFMs' objective characteristics meant 9009 potentially applied.

69.            The effect is that the MFMs (if not classified to 8471 under GIR 1) were potentially classifiable under 8471 or 9009. Where there is more than one potential classification, GIR 3 applies.

GIR 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:

(a) ....[not relevant]

(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.

(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.

70.            Applying GIR3(b) requires the Tribunal to identify the essential character of the MFM. But GIR3(c) recognises that it may not be possible to identify the essential character, and if that is the case, the MFMs should be classified by the highest of the potentially applicable chapter headings. In this case, there was no dispute that if I determined that GIR3(c) applied, the MFMs would be classified under 9009.

71.            The CJEU in KIP/HP1 summarised this at [48]

[48] However, if the copying function of the machines at issue in the main proceedings is of an equivalent importance to that of their other two functions, those machines could not be considered units of automatic data-processing machines because they do not meet the condition laid down in Note 5 (B)(a) to Chapter 84 of the CN, that is to say, that they be 'of a kind solely or principally used in an automatic data-processing system'.

[49] In such a case, machines made up of different components, that is to say either a printer module and a scanner module, or a printer module, a scanner module and a computer module, should be classified, by application of GIR 3(b), according to the module which, of those two or three modules, is identified as determining their essential character, provided such identification is possible. If that is not the case, in accordance with GIR 3(c), they are to be classified under the heading which occurs last in numerical order among those which equally merit consideration.

Summary

72.            In summary, the authorities appear to dictate that I should:

(1)           First determine the MFMs' objective characteristics;

(2)           Then determine from their objective characteristics whether they were principally used in an automatic data-processing system (GIR1);

(3)           If not, determine which of their components gives them their essential character (if any) and classify them by that. (GIR 3(b))

(4)           Otherwise, they are classified under 9009 (GIR 3(c)).

73.            Much the same was said by the CJEU in Kip/HP1:

[50] It follows that each of the machines at issue in the main proceedings should be classified under heading 9009 only if it is apparent, on the basis of its objective characteristics, that it is not of a kind used principally in an automatic-data processing system, since the copying function is of an importance equivalent to that of the other two functions and that it proves impossible to determine which , of the printing module or the scanner module ... give it its essential character.

74.            HMRC's case, however, was that the CJEU's later decision in Hewlett Packard Europe BV C-361/11 ("HP2") effectively superseded the CJEU's analysis in Kip/HP1 and meant that any MFM of a type similar to those in HP1 and HP2, such as those in this appeal, should always be classified under 9009, and 8471 was simply irrelevant. I move on to consider this but, as it leads on from this, I consider what the CJEU said in Kip/HP1 about Regulation 400/2006.

Regulation 400/2006

75.            This regulation had effect at the time of the importation of the MFMs at issue in Kip/HP1 and those in this appeal . It classified the following to Heading 9009:

A multifunctional apparatus capable of performing the following functions:

-scanning

-laser printing

-laser copying (indirect process)

The apparatus, which has several paper feed trays, is capable of reproducing up to 40 A4 pages per minute.

The apparatus operates either autonomously (as a copier) or in conjunction with an automatic data-processing machine or in a network (as a printer, scanner and a copier).

The explanation given in the regulation for this classification was:

The apparatus has several functions none of which are considered to give the product its essential character.

76.            The last question in Kip/HP1 was whether this regulation was valid. Its validity depended on whether it changed the duty rate of the products to which it applied as the Commission had no power to do that.

77.            The CJEU ruled the regulation to be valid, but, by taking into account the explanation incorporated in the regulation, that it only applied where none of the MFM's functions gave the product its essential character [61].

78.            Regulation 400/2006 therefore had no impact on the above analysis at §82 as the Regulation only applied if none of the MFM's modules gave it its essential character, in which case GIR3(c) would have applied to put the MFM under 9009 in any event.

79.            Regulation 400/2006 is relevant as the CJEU refer to it in HP2, as I explain below. The appellant also says it is relevant as it tells this tribunal about the machines in Kip/HP1: they had to have fallen within Reg 400/2006 because otherwise Reg 400/2006 would have been irrelevant to the CJEU's decision and the CJEU would not have been referred to it. I consider this below at §§90-96.

The impact of HP2 on the above analysis

80.            HP2 concerned MFMs which were imported in 2009, some years after those at issue in this appeal. With effect from 1 January 2007, in other words after the period at issue in this appeal, the EU Commission had (further) amended the CN with Regulation 1549/2006.

81.            The referring court stated that printers of the sort at issue in HP2 could have been classified under 8471 60 20 prior to 1 January 2007 but the effect of regulation 1549/2006 was that it was no longer possible to classify MFMs under 8471 60 20 (0%), and caused them to be classified under 8443 31 91 with a duty rate of 6%. Indeed, it is clear that the effect of the 2009 regulation was to delete 8471 60 20 and 9009 and to create a new code 8443 31 91 at 6% if the MFMs copied at a rate of 12 sheets per minute or more.

82.            As the CJEU recognised, the EU Commission had no power to change duty rates so if the MFMs at issue in the appeal would have attracted a 0% duty rate before 1 January 2007 but a 6% one after that date, then the EU Commission had exceeded its powers.

83.            So, although the case was actually about the classification of machines imported in 2007, in practical terms the case was about what would have been the proper classification of those machines if imported before 1 January 2007, one side arguing for 8471 and the other for 9009. It is therefore highly relevant to the machines in this appeal. Without any discussion of its reasoning, the court held:

[47] The [CJEU] has also held that, since it classifies machines capable of performing printing, electronic scanning and reproduction operations under 9009 12 00 on the ground that none of the functions corresponding to those operations can be regarded as giving those machines their essential character, without, in principle, requiring all machines having those three functions to be classified as photocopiers, the relevant Commission regulation was valid (see, to that effect, [Kip/HP1] paragraph 62.

[48] In the present case, the characteristics of the printers at issue in the main proceedings, as referred to in paragraphs 25 and 26 above, indicate that, similarly to the machines at issue in [ Kip/HP2], they perform a number of functions, namely scanning, printing, copying an in some cases fax, none of which can be regarded as giving them their essential character.

[49] In those circumstances, if the printers at issue in the main proceedings had been imported before 1 January 2007, then following the case law they would have been classified under CN 9009 12 00.

[50] Consequently, in classifying multifunctional printers such as those at issue in the main proceedings under CN 8443 31 91 the commission did not amend the 6% customs duty applicable to them. The commission did not, therefore, exceed the powers conferred on it....

HMRC's case was that these paragraphs in effect superseded or overruled Kip/HP1. Are HMRC right?

84.            What did the CJEU mean by what they said in [47]? They referred to [62] of Kip/HP1, which is the second part of the decision where the CJEU considered the validity of Regulation 400/2006. In my view, [47] of HP2 is intended as a summary of what the CJEU ruled at [61-62] of Kip/HP2. The 'it' referred to in line 1 is Regulation 400/2006. So [47] concerns 400/2006 and not the later 1549/2006. Nothing in [47] therefore detracts from what was said in Kip/HP1: on the contrary, it repeats it.

85.            What of [48]? At first glance this seems to be saying that the MFMs in Kip/HP1 had various functions none of which could be regarded as giving them their essential character. The difficulty with that interpretation is that that is not what the CJEU said in Kip/HP1, where they did not rule on classification but indicated that a 8471 classification was likely (see [40-43] cited at §59 above). So it seems [48] must be read as if the words 'similarly to the machines at issue in [ Kip/HP1]' only qualifies the immediately following phrase 'they perform a number of functions, namely scanning, printing copying and in some cases fax'. That qualification is then clearly right: both cases did concern MFMs which could undertake those functions. But that means the next phrase says nothing about the MFMs in Kip/HP1 but is a statement which was only intended to relate to the MFMs in HP2: in conclusion [48] means that the CJEU decided that 'none of which' functions of the machines in HP2 'can be regarded as giving them their essential character'.

86.            What the CJEU appears to have meant is that the analysis in Kip/HP1 still held good and confirmed that if an MFM had components none of which gave it its essential character, then it was properly classified to 9009 before 1 January 2007. And in its opinion, the machines in HP2 were such machines, properly classifiable to 9009 before 1 January 2007.

87.            It follows that [49] and [50] are therefore statements limited to the MFMs in HP2 and say nothing about the MFMs in Kip/HP1. Because the HP2 machines were 9009 (charged at 6% before 1 January 2007), the Commission could place then in 8443 31 91 after 31 December 2006 charged at 6%.

88.            My conclusion is that HMRC are wrong to say that HP2 superseded the analysis in Kip/HP1. I comment that if the CJEU had wanted to depart from its analysis in Kip/HP1, it should have said so explicitly: on the contrary, it mentioned its earlier decision with approval. So the Kip/HP1 analysis still applies but HP2 provides some guidance on the type of machines where the CJEU considered no individual module gave it its essential character.

The machines at issue in HP1 and HP2

89.            The 'Kip' machines are not relevant: no one suggested that they were similar to the MFMs at issue in this appeal. As I have said at §79, it seems an obvious inference that the HP machines at issue in Kip/HP1 were machines which fell within the definition of machines to which Reg 400/2006 applied.

HP1 machines

90.            The appellant says by implication the HP1 machines were different as HMRC never suggested that the machines at issue in this appeal were caught by Reg 400/2006. But HMRC's view, or supposed view, on the applicability of Reg 400/2006 is neither here nor there. In any event, the proposition is fallacious because the CJEU's judgment in Kip/HP2, summarised above at §78, really meant the Regulation was irrelevant to classification as the ruling was that it only applied if the MFM fell into GIR3(c) under the application of normal rules.

91.            Reg 400/2006 is somewhat ambiguous about the MFMs to which it applies and the CJEU do not clarify it. What did it mean by 'capable of reproducing up to 40 A4 pages per minute'?

92.            HMRC's interpretation was that it applied to any machine other than one which could produce more than 40 A4 pages per minute. In other words, it was intended to catch low-end machines that might be used at home or in small offices.

93.            On the other hand, the 'capable' might indicate that the machine had to be able to do at least 40 pages per minute under some conditions and that a machine which could never reproduce more than, say, 12 pages a minute, was not caught. Moreover, a purposive interpretation would suggest that, as its effect was to classify MFMs as photocopiers, it must have applied to high-end MFMs as it would be bizarre to categorise slow-copying machines as photocopiers but allow for the possibility that fast-copying machines could be classified as something else. And if it was intended to capture low-end as well as high-end machines, why not leave reproduction speed out of the definition?

94.            However, unsatisfactory as the definition is, I am persuaded that HMRC are right on the basis that, had the Commission meant only to capture high-end machines, that could have been achieved by omitting the words 'up to', so I conclude they did intend the Regulation to apply only to low end machines, would could reproduce at any rate, as long as that rate did not exceed 40 pages per minute.

95.            I note in passing that more facts about the HP1 machines were recorded at [51-52] of the Tribunal's decision in Xerox Ltd [2010] UKFTT 527 (TC) (not to be confused with the decision of the CJEU in Rank Xerox already mentioned). These facts were apparently taken from the French tribunal decision subsequent to the CJEU decision in HP1 and are not apparent from the CJEU decision. Xerox records that the HP1 machines had no ADF, no sorter and the printer comprised 80% in value of the machine. A lack of an ADF is incompatible with an ability to reproduce more than 40 pages per minute and that supports HMRC's case on this and my interpretation in the preceding paragraph: however, no reliance can be placed on these 'facts': those facts were not recorded in the CJEU decision and cannot be said to have formed a part of its decision.

96.            The machines in this case had printing speeds of 20-28 pages per minute (with one machine at 10 p/m). Therefore, it seems that the MFMs at issue in this appeal may have been within Reg 400/2006, although without knowing their reproduction speed I cannot be certain. In any event, it does not affect the outcome of the appeal for the reason given at §78.

HP2 machines

97.            But what were the particulars of the MFMs at issue in HP2? These are set out in [§25-26] of the CJEU's decision which reads as follows:

[25] According to the referring court, those printers, which are intended for use in households and also small- and medium-sized businesses, combined a laser printing module and a scanning module. They possessed scanning and printing functions when connected directly or through a network to an automatic data-processing machine. They also had a copying function, which could be used independently from an automatic data-processing machine. Some of the printers at issue also had a facsimile (fax) function.

[26] Printing and copying was done through the same printer unit, which was the principle component of the printers in question. Although the copying function required the document to be scanned first, the printing speed was identical for the printing and copying functions. The referring court also mentions that the different paper trays used on the printers in question could be used for both copying and printing. The sheet feeder could be used for copying, scanning and fax functions.

98.            Much of what is said here is true of the machines at issue in this appeal: the only noticeable distinction is that the HP2 machines appeared to have more than one paper tray ('...different paper trays....') whereas those in this appeal had only one.

99.            The HP2 machines copied at at least 12 pages/minute: this is clear because Regulation 1549/2006 applied and that was one of its criteria.

100.        I accept the appellant's case that without precise printing and copying speeds, details about resolution, and so on, it is not possible to determine whether the HP2 MFMs were so similar to the machines at issue in this appeal that they were competitor machines. As the CJEU decided that they fell into 9009 (or at least they would have done if imported in 2006) the appellant's case is that their machines were distinct from those at issue in this appeal. HMRC does not agree. I do not know: I can only apply the principles as outlined at §72 to the objective characteristics of the machines at issue in this appeal.

Application to facts in this appeal

Ascertaining objective characteristics

101.        Objective characteristics of a product can include its use if it is inherent in the product: Ikegami Electronics (Europe) GmbH C-467/03. In Honda at [30] the CJEU said much the same thing when they said it did not matter what a product was actually used for, but what it was designed to be used for.

102.        The use of the MFMs was as a printer, scanner, copier and, where there was a fax module, as a fax machine. These uses were inherent in the design.

103.        The appellant's case was that the MFMs were principally intended for use as a printer with additional, but less important, other functionality. Whether or not that was the intention behind the MFMs' design, what matters is what is objectively ascertainable about the product. They could in fact be used for all four functions (or three in the case of the non-fax module machines).

104.        I have set out those objective characteristics at §§14-42 above.

105.        In summary, in all 16 MFMs the printer module was the largest in terms of size and cost. It was a somewhat better printer than it was a photocopier in that (a) the printer had the better resolution (§§30-31) and (b) it could print more pages without stopping than it could photocopy originals (§§32-35). As it seemed to me that both of these functionalities were not inherent, in other words, that Brother could have improved the performance of the photocopier relative to the printer, I considered them significant.

106.        I did not consider that the other differences in comparable functionality, such as watermarks, booklet printing and manual duplex (§§36-39) were important distinctions for reasons recorded above: they were really inherent in nature of the different modules rather than a design choice. So far as the control panel was concerned, this was necessary for the fax, scanner and copier functionality but not really for the printing functionality, although it was co-opted for it (§§24-26). Therefore, the presence of the control panel did not really seem to me significant as it was inherent in the nature of some of the modules.

107.        It seemed to me that the printing and scanning functions were equivalent. The resolution of both was, it seems, the same (§30) and while, on one hand, the MFMs could print more pages without stopping than it could scan originals, on the other hand, most could scan in colour while the printers could only print black and white.

108.        I find that the three MFMs which could not scan in colour were different to the other MFMs. One in particular (MFC-7225N) was distinct in that it had a sophisticated fax module. At the same time it had one of the smaller ADFs and slower print speeds. The other two (MFC-8220 and MFC-9070) were also distinct in that while both had an ADF, there was no access to the platen glass so they could only scan/copy using their ADF. Similarly to MFC-7225N, these had one of the smaller ADFs and slower print speeds; indeed MFC-9070 only printed at 10 p/m. None of these three machines could have an optional extra paper feed tray. Objectively, the fax element was relatively more important in these machines (particularly with MFC-7225N) than the fax element of other machines.

109.        I have already mentioned that DCP-7010 was also distinctly different to the other machines in that it lacked an ADF. It was one of the machines which lacked a fax module too. So it comprised only two modules, a printer and scanner module.

Ascertaining whether MFM is principally used in an ADP system

110.        So, as I have said, the first test is whether the MFM is principally used in an ADP system. I take that to mean 'used' in the sense that data is either or both sent to or from the MFM to an ADP system. It stands to reason that the MFMs were most likely to be permanently wired to an ADP system as some of their functions required such connection (there was no suggestion of an intermittent wireless connection). So when 'used' as a photocopier it was not used in an ADP system even if in practice at the time it was connected to an ADP system; when 'used' as a printer or scanner, however, it was used in an ADP system.

111.        What does 'principally' mean in this context? The CJEU in Kip/HP1 said:

[44] In the present case, it is apparent from the description of the characteristics of those machines that most of the functions which they perform, that is to say, printing and electronic scanning, can be used only in connection with an automatic data-processing machine. Accordingly, those machines are likely to be of a kind used principally in an automatic data-processing system.

...

[46] It is therefore for the referring court to assess, taking into account the objective characteristics of those machines such as the print and reproduction speeds, the existence of an automatic page feeder for originals to be photocopied or the number of paper feeder trays, whether the copying function is secondary in relation to the other two functions or whether, on the contrary, it is equivalent in importance.

[47] If the copying function is secondary in relation to the other two functions, those machines should be considered units of automatic data-processing machines without the meaning of Note 5(B) to Chapter 84 of the CN......

112.        That guidance to the national court is what matters: it is not actually relevant what the national court actually decided, although it seems clear that the French Tribunal decided that the HP machines in Kip/HP1 were within 8471 by application of GIR 1 (see [51-52] of Xerox Ltd (above)).

113.        However, the FTT in Xerox Ltd did not consider that GIR1 led to classification under 8471 of the machines at issue in that case:

[100] the appellant's case that the machines were designed primarily to place marks on paper and that therefore printing is the machines' principal function seems to me to be misconceived, because, of course, printing is essential to physical output from both the data-processing and the digital copying functions.

[101] for the reasons given above, I agree with HMRC's submissions that the machines are, by reference to their objective characteristics, genuinely multifunctional and that it is not possible to conclude that the copying function is subsidiary to the data processing function or vice versa.

The MFMs in Xerox Ltd, did it seems have an ADF (see [86]), and the print resolution was better than the copier resolution (see [23]).

114.        The CJEU in Kip/HP1 (see [46] cited at §111 above) highlighted various factors to be considered in answering the question of whether the MFMs were principally used in an ADP system and they were:

(a)           print and reproduction speeds,

(b)           the existence of an automatic page feeder for originals to be photocopied or the

(c)           number of paper feeder trays

115.        The purpose of considering these was fundamentally to see if objectively the copier was significantly less sophisticated than the printer/scanner such that it could be said the MFM was principally used as a printer/scanner. The list was not exhaustive as the CJEU said 'such as' so I will consider those three factors and other relevant factors too when considering the relative sophistication of copier versus printer/scanner.

Speed

116.        It is inherent in the two processes that photocopying must take longer than printing as it requires the original(s) to be placed on the platen glass or in the ADF. But the printing element of the digital copying process in the machines at issue in this appeal took exactly the same amount of time as printing a document sent to the machine from a computer (§27).

ADF

117.        Where there is no ADF, each sheet to be copied must be placed successively on the platen glass manually. That is a time consuming process and may well indicate that an MFM without an ADF is one where the copying is secondary to the printing, as indeed the French Tribunal found in HP1. It follows that the presence of an ADF does not indicate that copying was secondary to the printing. The machines at issue in this appeal all had ADFs bar DCP-7010 (see §11 and §109).

Multiple paper trays

118.        The parties were divided on the relevance of multiple paper trays. HMRC thought it irrelevant; the appellant though multiple paper trays would indicate the MFM was a copier, but a single paper tray would indicate it was primarily a printer.

119.        It is not immediately obvious to me why the CJEU in Kip/HP1 thought multiple paper trays were more consistent with a copier than with a printer: I think that by itself having a single paper tray is at best a weak indication that an MFM is principally a printer/scanner.

120.        Nevertheless, all the MFMs at issue in this appeal had only one paper tray (together with a manual feed slot and in some cases the ability to purchase an optional extra paper tray).

Resolution and capacity

121.        Resolution was not mentioned by the CJEU as a relevant factor, but as I have said, theirs was only an indicative list. A significantly poorer copying resolution to printing resolution would, in my view, be an indication the MFM was principally a printer/scanner. As I have also said, another indication is paper tray capacity versus ADF capacity as that also indicates a higher capacity printer than copier.

122.        As I have said, both these factors are present in the MFMs in this appeal and indicate that the printer component was better than the photocopying functionality.

Other specifications

123.        I find that some functionality, such as duplex where machines had duplex functionality, and sorter, was available in both printing and copying modes. Some functionality was limited to the printer, such as the ability to watermark. There was some limited functionality that existed only for the copier, in particular some of the buttons on the control panel related only to the copier. These differences, as I have said, appeared to be inherent to the different processes and are therefore of not much significance.

Conclusion on whether MFMs were principally used in an ADP system

124.        The objective characteristics of the MFM which lacked an ADF (DCP-7010) were such that I am satisfied it was principally used in an ADP system. This is because its copying function was very limited compared to its printer function. The three indicative tests posed by the CJEU were all in favour of the copying function being subsidiary to the printing function: the printer resolution was better, there was no ADF and only a single paper tray. No other functionality suggests an alternative answer, in particular it did not have a fax module that could, like the photocopying function, operate independently of an ADP system.

125.        If the FTT in Xerox Ltd correctly recorded the facts of the HP1 machines, then DCP-7010 was similar to an HP1 machine and like those machines it should have been classified by application of GIR 1 under 8471.

126.        DCP-7025, DCP-8045D, DCP-8060 and DCP-8065DN: The remaining four machines without a fax module all had ADFs, although the ADFs supported the scanner function as well as the copying function. These MFMs had three uses: as a printer, scanner and copier. Two of those uses required the machine to be in an ADP system. It is also true for these machines that their printer resolution and capacity was better than when functioning as a photocopier (in other words, when not in an ADP system) and these differences were not inherent in the nature of the printing and copying processes, but there because the printer was built as the highest specification module. It also seems relevant that it was only the scanner which operated in colour: again that was higher specification than the copier function.

127.        I have to decide whether 'the copying function is secondary in relation to the other two functions or whether, on the contrary, it is equivalent in importance.' Two out of the three uses were uses in an ADP system, and the use (printing) within the ADP system was by design at a higher specification than use (as a copier) outside the ADP system; the scanner was also of higher specification than the copier as it was colour. The MDMs' objective characteristics point to the single use of copying outside an ADP system being of secondary importance to the other two uses combined which took place within an ADP system. These MDMs in my judgment should be classified to 8471.

128.        I am aware that I have reached a different conclusion on that to both the VATTR in Brother International Europe Ltd [2008] C00248 and the FTT in Xerox, but the facts presented to me appear to have been different than in those cases (and Brother did not have the benefit of the CJEU decision in Kip/HP1).

129.        That leaves the MFMs with a fax module. As I have said at (§108), three of them were distinct. I will deal with those first. Firstly, the scanner was of lower specification than those of the other MFMs as, like the printer and copy function, it was only black and white. It was also of lower specification for MFC-8220 and MFC-9070 as there was no access to the platen glass and scanning could only be done using the ADF. I find that the scanner was a less important use of the machine then with the other MFMs, although the lack of access to the platen glass affected the specification of the copier as well. All of them were at the slow end of printer speed and paper capacity, MFC-9070 having a particularly slow printing speed of 10 sheets/minute. MFC-7225N had a sophisticated fax module. At the same time, for all three of these MFMs the printer's resolution and capacity were better than those of their copier.

130.        The overall impression was that for all three of them, particularly MFC-7225N, was the fax module was significant. As I have said, the fax module could be used independently of an ADP system: it could send and receive faxes without any connection to an ADP system. Its functionality was increased if it was connected to an ADP system as that meant faxes could be both sent and received electronically rather than requiring a hard-copy to be printed off, but it wasn't necessary. I cannot be satisfied that these three MDM's objective characteristics point to its use outside an ADP system (sending and receiving hardcopy faxes and photocopying) being of secondary importance to its use within an ADP system (printing/scanning/electronic faxing). I do not consider they should be classified via GIR 1 as 8471.

131.        The remaining 8 MFMs were of a muchness with the 4 MFMs without a fax module discussed at §126 above, except of course that they did have a fax module. In particular, their scanner was colour and the scanner or copier could be used via the platen glass or ADF, while the printer was of higher specification than the copier. Its printing and scanning functions could only be used in an ADP system, its photocopying outside and ADP system, and its fax function half-and-half. Was its use outside an ADP system objectively secondary to its use within an ADP system? I find the addition of a further non-ADP functionality (fax) is enough to mean that with these 8 MFMs I am not satisfied that their use outside an ADP system was secondary to their use within an ADP system and therefore they should not be classified via GIR 1 as 8471.

132.        The appellant points out that the MFMs' copying function exists solely because of software added to its CPU. Putting aside those with fax capability, without that software, the machines could only print and scan, both functions requiring connection to an ADP system. It is odd, says the appellant, if the reason the machines fail to fall into 8471 is merely because of added software. My view is that it may or may not be odd: the test is not whether the functionality would not exist without a piece of software but whether the resulting MFM functioned principally when connected to an ADP system, and I have not been satisfied, in respect of those MFMs with fax modules, that it did.

133.        Therefore, as explained above at (§72), via GIR2, I must move on to GIR3 to consider those MFMs with fax modules. The below discussion therefore only applies to those MFMs with fax modules; nevertheless, in my view what I say below would apply as much to those categorised to 8471 under GIR1 had I not already allowed the appeal in respect of them.

Ascertaining a product's essential character

134.        The test from the Notes to Chapter 84 (GIR1) is different to the test under GIR3: the former relates to 'function' while the later relates to 'essential character'. As the CJEU in Sony Computer Entertainment Europe Ltd T-243/01 pointed out:

[124] Moreover, according to the clear terms of GIR 3(b), it provides for the classification of mixtures and composite goods according to the material or component which gives them their essential character. It does not provide for the possibility of classifying mixtures or composite goods according to the function which gives them their essential character.

The question is which (if any) component gives the product its essential character.

135.        How is 'essential character' determined? The Tribunal is entitled to refer to the explanatory notes to the Harmonised System of the WCO ('HSEN') but these are not binding (see [41] of Barrus cited at §50 above). HSEN 8, in respect of GIR3(b), states:

The factor which determines essential character will vary as between different kinds of goods. It may, for example, be determined by the nature of the material or component, its bulk, quantity, weight or value, or by the role of a constituent material in relation to the use of the goods.

136.        I had evidence about all these characteristics in relation to the MFMs at issue. Needless to say the parties did not agree on the outcome from this test. The parties were agreed, as I am, that the essential character has to be determined from the product's objective characteristics. Nevertheless, objective characteristics do not necessarily have to be immediately apparent (see [62] of V-tech which both parties considered to be correct. What of the product's developmental history?

Developmental history

137.        While a product's classification depends on its ascertainable objective characteristics, the appellant suggests that nevertheless sometimes the developmental history of product is relevant to classification and for this it relies on what the CJEU said in Kloosterboer Services (C-173/08) where it said at [34]:

....It is, moreover, not disputed that, before goods such as those at issue in the main proceedings were developed, the cooling of computer processers was exclusively effected by heat sinks. Adding fans to those heat sinks did not fundamentally change their properties, but merely improved their effectiveness by increasing their cooling capacity.

This was said by the CJEU when determining essential character (see [33-36]).

138.        As I understand it, the appellant's suggestion is that it is relevant that it developed the printer before it added a scanner to the unit to make it an MFM (§43-44). I cannot agree. The CJEU's comments in Kloosterboer seemed to be aimed at heat sinks in general and not the particular heat sinks in issue: here, however, different MFMs may have different developmental histories but there is no reason to suppose that those different developmental histories are objectively apparent from the MFMs themselves. In particular, there was no evidence that there would be an objective difference between an MFM developed by a manufacturer of printers such as Brother which was of the same specification as an MFM developed by a manufacturer of photocopiers such as Xerox. So I do not consider development history gives rise to an objective characteristic in this case.

The relevance of function

139.        The appellant points out that the GIR3 test looks at components and not functions. Which component gives the item its essential character?

140.        However, I agree with HMRC that a product's essential character may be dictated by its function. Where the CJEU refers in [48] of HP2 to function in the context of the GIR3 test, it seems likely that they were eliding the two elements of the test: working out essential character from function and then working out if that essential character was given by a component. This also happened in Reg 400/2006 where the drafters said 'the apparatus has several functions none of which are considered to give the product its essential character.' (cited at §85 above). Nevertheless, the comment from the CJEU in Sony, cited above at §144, makes clear that one has to be careful to recognise that GIR 3(b) looks at components and not functions.

141.        That function may give a utilitarian product its essential character seems a matter of common sense: it is also apparent from HSEN 8 referred to above at §135 where it mentions 'the use of the goods' and from numerous cases cited by HMRC:

142.        In Turbon C-250/05 the question was the classification of ink cartridges whose function was to supply ink to ink-jet printers. The product comprised two components: the cartridge and the ink. The CJEU said that as the purpose of the product to be classified was to supply ink to a printer, the essential character of the product was given by the ink. This was the case even though the ink was useless unless delivered in the cartridge: [20-23].

143.        In Sportex R-253/87 the question was the classification of sheets made of carbon glass and resin which were intended for the manufacture of tubes. The essential character of the sheets was found to be their flexibility because that was essential for their use to make tubing. That essential character was conferred by the resin: [9].

144.        In VauDe Sport C-288/99 the question was the classification of a child carrier made from textile and supported in a frame. As it was possible to carry a child in the textile body without frame but not vice versa, it was held that the textile gave the product its essential character: [26-28].

145.        In VOBIS C-121/95 the question was the classification of a computer housing which had a power lead and disk drives and cables (but no computer): the CJEU held that its essential character was determined by relative value and intended use in an ADP system. They held that therefore the disc drives gave it its essential character: [21-23]

146.        In Sony (above) the question was the classification of Sony's Playstation2. Its essential character was given by the fact it was intended to play video games [127].

147.        The appellant criticises HMRC's reliance on all these cases as in all of them the product had only one use. With the possible exception of Playstation 2, none of them were multifunctional. However, I think the cases do show it is legitimate to look to use to determine the essential character of a utilitarian item, although other factors may be relevant, such as relative value as in VOBIS, or other factors such as bulk, quantity, and/or weight.

148.        The appellant also suggested that whether a module or component was capable of independent operation must be relevant to essential character. It suggested this, no doubt, because the printer module could theoretically operate without the scanner or fax module being attached (although it was clearly not intended to do so and it would take, as I have said, some labour to physically separate them and the result would be without the normal plastic housing over the top - §§8 & 12). I do not, in any event, agree with the proposition in law. Many components of a product might be capable of independent operation (eg a CPU or disc drive in a computer) but I see no logical reason why that factor would affect the essential character of the product in issue.

149.        The appellant also relies on the FTT decision in Xerox Ltd (above) but HMRC say it applied the wrong test.

Xerox Limited

150.        This FTT decision preceded HP2 but followed Kip/HP2. It considered MFMs of similar functionality to those in this appeal, although it is clear that the machines in that appeal had multiple paper feed trays. Its conclusion was that they were classifiable to 8471.

151.        As I have said, the FTT applied Kip/HP2 but rejected the GIR 1 route on the basis that the MFMs were not principally used in an ADP system. It then applied GIR 3 and concluded:

[112] I am faced with a choice of approach. On the one hand, I could adopt the reasoning of the [VAT Tribunal] in Brother International and conclude that the characteristic properties of the machines in this appeal...were printing, scanning and copying, that is, they are multifunctional or 'expressly designed for multiple purposes'. On the other hand, I could adopt the reasoning of the [French Tribunal in HP1] and conclude that the essential character of the machines is given by the printing module which has the most parts, is largest in terms of volume, requires the most maintenance and receives the most use.

[113] I conclude that the printing module does give the machines their essential character. I accept that (to quote [the VAT Tribunal] in Brother International) 'the products are not printers which happen to copy, for example, but machines which are expressly designed for multiple purposes' but I consider that this reasoning ignores the fact that the printing process is essential to copying, which is a 'scan-and-print' process. I accept that not all the uses of the machines involve printing - information can be scanned in the IIT and transmitted to a remote computer - but it seems to me that objectively a sufficient proportion of the processes performed by the machines involve printing to enable me to conclude that the essential character of the machines is that they 'place marks on paper' ....

It doesn't seem to me that Xerox Ltd applied the wrong test. It looked at function and decided the MFMs' essential character was given by its function of making marks on paper. But would it be right for me to similarly conclude that the essential character of the MFMs was the printed output (or putting marks on paper, as the appellant phrased it)? HMRC urge on me the analysis from the earlier Brother International (above) that the MFMs' essential character was the fact it was multi-functional in nature. I reach my conclusion on this below.

Ascertaining the component (if any) which confers the essential character

152.        Having identified the product's essential character, GIR3 requires the tribunal to decide whether that character is conferred by a component. That is usually ascertained from the 'removal' test, whereby a component is theoretically removed to see if the product retains its essential character (and not whether it retains its function). This is apparent from cases referred to above.

153.        For instance, In Turbon, removal of the ink would rob the product (the ink cartridge) of its essential character (of delivering ink). It did not matter that robbing it of the cartridge would similarly make it useless. VauDeSport is a less extreme case as it appears the textile would function as a child carrier even without the frame

154.        In Sony (above), the CJEU referred to HSEN 8 at §125 and said:

[126] [the interpretation of GIR 3(b) set out in §124] is also supported by the case-law of the CJEU, according to which, in accordance with [GIR 3(b)] 'it is necessary, in carrying out the tariff classification of a product, to identify, from among the materials of which it is composed, the one which gives it its essential character. This may be done by determining whether the product would retain its characteristic properties if one or other of its constituents were removed from it.

 

155.        Much the same was said in Kloosterboer Services BV C-173/08. But as Sony itself shows, the removal test is not conclusive. In that case, the essential character of the product was its ability to be used for video games: if the CPU was removed, the product would not work. Nevertheless, the CPU did not confer its essential character because it was just a CPU, which was a necessary component in any computer and not just one devoted to video games: [120-127]. It may simply be that this case should be seen as one where the essential character was conferred by software on the CPU and as software is not a physical component it can't fall within GIR3(b).

156.        The 'removal' test was applied in the recent Tribunal decision of V-tech, already mentioned. Mr Mitchell relies on the removal test in support of this case and largely relies on its application in V-tech. In that case, to paraphrase, the Tribunal ruled that the product (which looked like a wristwatch but which had many functions including a camera as well as timekeeping) would lose the greatest part of its appeal if the camera was removed, whereas it would remain largely functional if the time-keeping function was removed. While Mr Mitchell relied on this conclusion, he pointed out V-tech in his view misapplied the test which looks at whether a module is removed and not whether a function is removed.

157.        The appellant points out that in this appeal, if the printer module is removed, both the printing and photocopying capacity is removed, and all that is left is a scanner module and (in some cases) a fax module (although the ability to print faxes is lost). If the printing function confers the MFM's essential character, then I would agree that removing the printer component would remove that essential character.

158.        However, as I have said, the mere fact that removing the printing module would leave an MFM barely able to function at all, would not of itself mean that the printer module conferred the MFM's essential character.

Conclusion on essential character

159.        I find that it is relevant that the printer element is:

Largest element by size

Most expensive module

Better resolution then copier

Higher paper capacity than ADF

160.        I also agree that the printer was essential for three of the MFMs four functions (printing and copying and standalone fax machine). The printer was only irrelevant when, while attached to an ADP system, it produced 'soft copy' scans, and was used as a 'soft copy' fax machine.

161.        I think it fair to say the MFMs main or essential character was its use to produce hardcopy output (or 'make marks on paper' as the appellant says).

162.        HMRC's view was that the MFMs' essential character was that they were multifunctional. The problem with that view is that it does not convey any character at all. Many very different products are multifunctional: a teasmaid might be multifunctional (it tells the time and makes a pot of tea); the product in V-Tech was multifunctional (a watch and camera amongst others). But the functions are all very different to those of each other and an MFM. It seems to me that what HMRC really mean is that, in their view, the MFMs' multifunctional nature is such that they do not have an essential character. That is clearly a possible outcome under GIR3(b) as it refers to 'in so far as this criterion is applicable'. Either way, if HMRC are right to say that the MFMs' essential character is their multifunctionality, or if HMRC really mean by this that the MFMs do not have an essential character, classification would be under GIR3(c).

163.        But, on balance, taking all the relevant facts into account, I agree with the appellant and with the Tribunal in Xerox, for the reasons given at §§159-160, that the essential character of these MFMs was conferred by their ability to produce hardcopy output ('make marks on paper'). It is only if the printer component was removed that the MFM would cease to have this essential character and therefore the printer component conferred that character. Under GIR3(b) all the remaining MFMs at issue in this appeal should also be classified to 8471.

Are GIR1 and GIR3(b) the same test?

164.        Before concluding, I mention HMRC's submission that a finding in their favour on GIR1 must mean that they win on GIR3(b): in other words, it was their view that the test from Chapter 84 Note 5(B)(a) whether the product is 'of a kind...principally used in an ADP system' amounts, in this case at least, to the same test as GIR 3(b) which is to classify by 'the material or component which gives [the product its] essential character'.

165.        However, while not without overlap, the tests are not the same and the CJEU has clearly stated, in a case involving MFMs, that they must be applied in order, which it would not be necessary to do if they were the tests were the same: see Kip/HP1 at [48-49] cited above at §71.

166.        Is it possible for the product not to be 'principally' used in an ADP system despite its printer module, but nevertheless be given its 'essential character' by the printer module?

167.        It seems to me that the answer is yes. This is because, so far as an MDM is concerned, GIR1 looks at use within an ADP system compared to use outside an ADP system: in carrying out that comparison, the printer is on one side of the divide and the copying function on the other side of the divide. Then the GIR3 test looks at essential character and if that character is printing, then the printer and copier are on the same side of the divide as both rely on the printing functionality. So it is possible for some MDMs to fail to qualify as 8471 under GIR1 but to qualify under GIR 3(b) (as some have done in this appeal).

168.        My conclusion is that all the particular MFMs at issue in this appeal should be classified under 8471 and the appeal is allowed.

 

 

169.        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.

 

 

BARBARA MOSEDALE

TRIBUNAL JUDGE

 

RELEASE DATE: 28 November 2016

 

 

Appendix.

 

All machines had a printer module and scanner module.

 

	Print Speed Pages/ minute	Colour scan?	Copy function	ADF (sheet capacity)	Fax module	Duplex function	250 sheet paper tray plus single sheet manual feed slot
DCP-7010	20	ü	ü	r	r	r	ü
DCP-7025	20	ü	ü	ü (35)	r	r	ü
DCP-8045D	20	ü	ü	ü (50)	r	ü	ü plus†
DCP-8060	28	ü	ü	ü (50)	r	r	ü plus†‡
DCP-8065DN	28	ü	ü	ü (50)	r	ü	ü plus‡
MFC-7225N	20	B&W	ü	ü (20)	ü (sophisticated)	r	ü
MFC-7420	20	ü	ü	ü (35)	ü	r	ü
MFC-7820N	20	ü	ü	ü (35)	ü	r	ü
MFC-8220*	20	B&W	ü*	ü (30)	ü	r	ü plus†
MFC-8440	20	ü	ü	ü (50)	ü	r	ü plus†
MFC-8460N	28	ü	ü	ü (50)	ü	r	ü plus†
MFC-8840D	20	ü	ü	ü (50)	ü	ü	ü plus†‡
MFC-8840DN	20	ü	ü	ü (50)	ü	ü	ü plus†‡
MFC-8860DN	28	ü	ü	ü (50)	ü	r	ü plus‡
MFC-8870DW	28	ü	ü	ü (50)	ü	ü	ü plus‡
MFC-9070*	10	B&W	ü*	ü (20)	ü	r	ü (paper tray limited to 200)

 

*these machines' copy function was limited in sense that machine had no scanner glass so could only scan/copy sheets fed through ADF.

 

plus† = optional extra additional 250 paper tray

 

plus‡ = manual feed slot capable of taking up to 50 sheets rather than standard one sheet.