This appeal was about the correct customs classification of ‘mobility scooters’ (and parts thereof) that had been imported by the appellants. The competing customs classifications were (1) “vehicles principally designed for the transport of persons” (heading 8703 of the combined nomenclature (“CN”)) or (2) “carriages for disabled persons” (heading 8713). The latter was free of duty; the former attracted a duty of 10%.

The importations by the first appellant were made between 12 December 2016 and 2 October 2018. A “C18” (post clearance demand note) sent with a letter dated 9 January 2020 demanded £979,063.63: £815,901.84 customs duty and £163,161.79 import VAT. A second C18, sent with a letter dated 13 April 2020, demanded £12,562.05: customs duty of £10,782.86 and import VAT of £1,779.19. HMRC’s statement of case accepted that the relevant C18 should be reduced by £125.10 of customs duty and £25.02 VAT.

The second appellant’s appeal related to:

The second appellant also appealed against HMRC’s refusal, in a letter dated 20 April 2021, of its refund application in respect of importations of certain mobility scooters. The amounts claimed were £178,344.44 customs duty and £35,668.89 import VAT. These importations took place between 12 February 2018 and 28 September 2020: the second appellant had declared the scooters under headings 8703 and 8708 but then said that it was wrong to do so and so sought repayment of the duty paid.

We had a main trial bundle of 4,378 pages, consisting of pleadings, witness statements, correspondence leading up to the decisions under appeal, Tribunal decisions and directions, applications, further documentary evidence (model brochures, technical specifications, and shipping documentation), and EU Commission documents. Correspondence between the parties was in a separate 215 page hearing bundle.

There were four witnesses: Jonathan Hearth (managing director of the first appellant) and Adam Williams (a director of product management at the second appellant), who gave evidence about (inter alia) the appellants and their businesses, about the models of imported mobility scooters, and about the question of whether the imported mobility scooters were “for” disabled people, all from their perspectives; and HMRC officers Samantha Rzepala-Lewis and Ben Key, who gave evidence about the ‘compliance history’ i.e. interactions between the appellants and HMRC, leading to the making of these appeals.

We set out in this section the underlying law whose application was, broadly speaking, uncontroversial (in the Discussion section below, we address in greater detail certain aspects of the law whose application was somewhat in dispute).

Most of the importations in question took place prior to 31 January 2020 i.e. while the UK was an EU member state. Some took place later in 2020, during the “implementation period” when EU law continued to apply to the UK.

The following propositions, based on the European Union (Withdrawal) Act 2018, were not in dispute in the appeal:

Goods imported from outside the EU are classified, for customs duty purposes, under the CN set out in Annex 1 to EC Council Regulation 2658/87. The CN uses an eight-digit numerical code to classify products. The first four digits are referred to as headings; eight-digit level numbers are referred to as subheadings. The CN is amended annually and reproduced in the UK tariff.

It was agreed that the imported mobility scooters fall within the following parts of the CN:

The relevant CN headings and subheadings in this appeal are:

As was stated by the CJEU in Case C-198/15 (“Invamed CJEU”) at [18] and at [22]:

The following was said by the Court of Appeal in HMRC v Honeywell Analytics [2020] EWCA Civ 243 at [127] and [130]

Mobility scooters were the subject of earlier litigation which culminated in the decision of the Court of Appeal in Invamed v HMRC [2020] EWCA Civ 243 (“Invamed CA”). The Invamed litigation concerned mobility scooters imported between 2004 and 2007 (i.e. before the issuance of an EU Commission ‘classification regulation’ in 2009 – more on that below) . The Court of Appeal, overturning the decision of the Upper Tribunal in [2018] UKUT 305 (TCC), reinstated the Tribunal’s decision ([2016] UKFTT 775 (TC)), which, following a reference to the CJEU (Invamed CJEU)), classified the mobility scooters in question to heading 8713.

(Both appellants were also parties in the Invamed litigation; and eight of the models of mobility scooter at issue in this appeal bore the same names as models at issue in the Invamed litigation (four of the first appellant’s ‘Rascal’ models; and four of the second appellant’s models: Pearl, Little Gem, Little Star and Sapphire 2).)

Invamed CJEU decided that:

Invamed CA concluded as follows:

The interpretation of tariff headings is assisted by explanatory notes produced by the World Customs Organization (known as HSENs) and by the EU Commission (known as CNENs). As was said in Invamed CJEU at [19-20], these notes are “are an important aid to the interpretation of the scope of the various tariff headings but do not have legally binding force”; their content “must therefore be in accordance with the provisions of the CN and may not alter their meaning”.

The HSENs to heading 8703 contain the following text:

From 4 January 2005, the CNENs in respect of heading 8713 contained the following text:

Commission Regulation (EC) No 718/2009 of 4 August 2009 (the “2009 Regulation”) entered into force on 28 August 2009. The annex to the 2009 Regulation describes a four-wheeled mobility scooter and a three-wheeled mobility scooter. The text of the 2009 Regulation is appended to this decision.

An oft-cited summary of the law relating to classification regulations was provided by Lawrence Collins J in VTech Electronics (UK) Plc v HMRC [2003] EWHC 59 (Ch):

Invamed CA referred to the 2009 Regulation at [22]:

We note that in Hewlett Packard BV itself (the case cited by Lawrence Collins J), it was held that “in the interpretation of a classification regulation, in order to determine its scope, account must be taken inter alia of the reasons given” ([20]).

Another CJEU case, Anagram International Inc (Case C-14/05) cast light on the process of applying a regulation by analogy. In that case – about toy balloons – the classification regulation described the balloons as comprising plastic foil on the exterior of which an aluminium layer is bonded. The product in question in the case was the other way round. The court said at [33]:

It was also reiterated in Anagram that the application by analogy of a classification regulation to similar products “facilitates a coherent interpretation of the CN and the equal treatment of traders” ([32]).

We make findings here about the objective characteristics, or design features, of the imported mobility scooters, so far as relevant to the primary legal issue in this appeal – whether those mobility scooters were “for” “disabled” persons, as those concepts are interpreted in the binding legal authorities (principally, Invamed CJEU and Invamed CA).

Whilst the oral evidence was helpful in providing background and in raising questions to be considered in making our factual findings (and in deciding what factual findings were material to the legal issue before us), it was not, on the whole, determinative; this is because, as just mentioned, the legal authorities clearly require us to make findings on, and base our legal decision on, the objective characteristics of the imported mobility scooters; the opinions of the witnesses – and, indeed, any information they might provide that was not evident from the objective characteristics of the mobility scooters themselves – was not of any direct assistance to us in that exercise (as we say, it was indirectly of some assistance, as it gave us things to think about in making our objective determination).

The imported mobility scooters came in 29 different models, 18 models of the first appellant and 11 models of the second appellant. All the models bore the core design features of a ‘mobility scooter’, which we would articulate as follows: battery-powered, relatively slow-moving (in practice, walking speed or thereabouts) vehicles with one seat, a low (i.e. near the ground) floor (or ‘platform’), a steering device, three or four wheels, and no covering. Their dimensions were: not much wider than the person sitting on the single seat; and not much longer than required to accommodate the seat, the steering device, and the wheels. These core design features can be readily apprehended, visually, by looking at the photographs elsewhere in this decision (which are not of the imported mobility scooters themselves, but are of other ‘mobility scooters’, which have the same core design). There was a range of size amongst the imported mobility scooters, from the shorter and narrower (which tended to be the slower, the somewhat less comfortable, the less sturdy, and so the less appropriate for outdoor use), to the wider and longer (and somewhat faster, more comfortable, more sturdy, and so more appropriate for outdoor use); but all conformed to the core design as just set out.

We were presented with a good deal of evidence about the variants that could be seen within this core design, amongst the various models; for example, exactly how wide or long they were; exactly how large the wheels were; whether the tyres were pneumatic or made of other material; whether the maximum speed of which the scooter was capable was 4 miles per hour or up to 8; additional features such as a basket or a collapsible covering; whether or how easily the scooter could be taken to pieces for storage or transportation in (say) a car; whether ignition was by key or by switch; and whether or not there were armrests. However, for reasons that will become evident in the Discussion section below, we do not consider that these variations (within the core design features) are material to the legal issue before us; and so we decline to needlessly lengthen (and, in so doing, obscure the core reasoning of) this decision by elaborating, or making further detailed factual findings, on them.

The following aspects of the brochures for the models of mobility scooter in question have, in our view, some relevance to their objective characteristics:

In what follows, we have assimilated the parties’ submissions, to the extent we have accepted them, into our reasoning; where we have rejected a significant and material argument made by a party, we have sought to explain why, though without always identifying the provenance of the argument. We mean no disrespect to the parties’ counsel (all of whom were excellent) in adopting this approach (which rather ‘anonymises’ their individual submissions in the hearing) – but we think it makes for a clearer and shorter decision.

The primary question in this appeal, based on heading 8713 as interpreted in Invamed CJEU and Invamed CA, is whether the imported mobility scooters, judged by their objective characteristics or design features, are intended for use solely by persons with a non-marginal limit on their ability to walk.

We start by observing that having a non-marginal limit on one’s ability to walk, covers a range of forms of disability. At one end of the range is someone unable to walk, at all. But included in the range is someone who walks so slowly, or requires so many significant breaks in the course of walking, that their ability to walk can fairly be said to be non-marginally limited; the same can be said for someone who can walk but in doing so endures significant pain or discomfort (such that, in an extreme case, they would require medical treatment, or extended rest, afterwards): such a person could also fairly be said, depending on the precise circumstances, to have a non-marginal limit on their ability to walk. In practice, all these measures of ability (or disability) – duration of walking, speed of walking, need for breaks, degree of pain and/or discomfort – may well combine or coalesce in any one person’s non-marginal walking disability. We note this because it is important to keep this spectrum of forms of disability within “non-marginal limitation” in mind, when considering who the mobility scooters are solely designed for. For example, it would be a mistake only to think of the needs of someone at, or near, the “unable to walk, at all” end of the spectrum.

In our view, is clear from their design features that the imported mobility scooters are designed for persons with a non-marginal limit on their ability to walk (note that at this stage of the analysis we have not put the word “only” before the word “for” – we will turn to that shortly): a relatively small, relatively slow, one-person vehicle clearly enables such persons, without assistance from someone else, to get from A to B (where A and/or B may be indoor locations inaccessible to large vehicles like cars or buses) where, otherwise, due to their disability, they would simply not be able to do so, or doing so would cause them significant pain, or take them significantly longer than the time taken by those without such limitation. So, for example, many of the imported mobility scooters would (judging from their design) enable such persons, without assistance from someone else, to move around environments where the ‘norm’ would be to walk: indoor spaces (e.g. homes; shops; restaurants); and outdoor public spaces normally reserved to pedestrians (e.g. pavements; parks; squares).

Clearly, the imported mobility scooters do not enable the user to fully replace the walking function e.g. they do not go up or down stairs, or across very uneven surfaces. However, this does not detract from the fact that (judging from their design) the scooters enable those with a non-marginal limit on their ability to walk to move, or ‘mobilise’, independently, in ways that they would otherwise not be able to do (or, would not be able to do without excessive pain, or without taking an excessive period of time).

The point was made to us that the design features of some of the larger mobility scooters did not indicate design for (solely) indoor use; rather, their features indicated that they could be used, for example, to travel on roads, at a fairly slow speed (compared to other road-users like cars and motorcycles), for a modest distance (perhaps, to the local shops); and that there they might be “parked up” whilst the user goes on foot (perhaps into a shop). It does not seem to us that this kind of possible use means that these larger mobility scooters are not, objectively, designed for those with a non-marginal limit on their ability to walk: for persons in this category who can walk very short distances (i.e. pop into a small shop), or who are able to walk slightly longer distances slowly, or with breaks, or with some pain, this sort of use would be welcome and appropriate.

A similar point can be made about the fact that some of the imported mobility scooters provided their users with more than just mobility (getting from A to B) – they had facility for carrying goods (in, say, a basket). To our minds, this is no indication that (objectively) the mobility scooters are not designed for the those with a non-marginal limit on their ability to walk; it is perfectly consistent with use by such persons in, say, getting to the local shops.

We would also make the wider point that, just because mobility scooters provide mobility at somewhat greater speeds than walking, or for greater distances than an average person may be willing to walk, does not indicate that they are not, objectively, designed for those with a non-marginal limit on their ability to walk. The impact of walking-disability is not limited to inability to get from place A to place B wholly on foot – it extends to inability to get from place A (home) to place B (local bus stop) and therefore inability to get to place C (the shops a few stops away by local bus). People with such limitation face barriers in using buses or trains or other public means of transport; and so, for them, a vehicle which, to some extent, enables them to do, independently, what others can do by, say, taking a local bus, is also welcome and useful.

But are the mobility scooters intended solely, or uniquely, for those with a non-marginal limit on their ability to walk, judging by their design features?

We find that they are, essentially because those able to walk (which we mean to include those with only a marginal limit on their walking ability) clearly have better alternatives, in whatever mode or environment the mobility scooter might, realistically, be used:

The repeated references in the marketing material to the scooters providing “independence”, “freedom”, and ease of life, are further evidence that, objectively, the imported mobility scooters are both for those with a non-marginal limit on their ability to walk, and exclusively for them (as this kind of “messaging” to the consumer would be, at best, confusing, and, more likely, quite off-putting, to the walking-able).

We note that our conclusions about what the imported mobility scooters are “for” are generally in line with those of the Tribunal in Invamed (following receipt of the ruling in Invamed CJEU), as follows:

We conclude that the imported mobility scooters fall within heading 8713.

Our starting point here is that the imported mobility scooters are not identical to the vehicles in the 2009 Regulation; the question is whether the classification regulation is to be applied by analogy.

We find that, if we looked only to the left hand column of the 2009 Regulation, we would have concluded that the imported mobility scooters were similar to those in the regulation: the regulation vehicles bear the core design features of ‘mobility scooters’, as we have found them. We do not find it helpful, as the appellants invited us to do, to find distinctions in the fine details, or specifications, of the regulation vehicles, as these are not in our view material differences.

However, the case law is clear that it would be wrong to pay attention only to the left hand column: the ‘reasons’ column must be taken into account in determining the scope of the regulation. The classic case illustrating this principle, Hewlett Packard BV (the authority cited by Lawrence Collins J in VTech), was about a machine that could perform the functions of both a fax machine and printer; the relevant CN provided that, with multifunctional machines, classification was to follow the component which performed “the principal function”; and the classification regulation considered in the case stated in the ‘reasons’ column that the fax function was the principal function of the machine considered in the regulation. Following AG Mischo, the CJEU rejected interpreting the classification regulation as meaning that all machines that could operate both as fax machines and printers (per the description in the left hand column of the regulation) must be treated as having the fax function as their principal function (the ‘reasons’ column). Rather, the correct interpretation is that the regulation only applies if the fax function is in fact the principal function (see [22] of the CJEU’s decision). We see this as an illustration of AG Mischo’s dictum that great care needs to be employed when seeking to apply a classification regulation by analogy to a non-identical product.

Here, the ‘reasons’ column in the 2009 Regulation says that classification under heading 8713 is excluded “as the vehicle is not specially designed for the transport of disabled persons and it has no special features to alleviate a disability”. The HSEN and the CNEN are then referred to. It seems to us the situation is akin to that in Hewlett Packard, in that the ‘reasons’ column of the regulation makes a statement that echoes one of the key factual variables in deciding the correct heading: in Hewlett Packard, it was a statement that the fax function was the principal function; in the 2009 Regulation, it is a statement that the vehicles are not specially designed for the transport of disabled persons. What Hewlett Packard tells us is that one cannot infer from such a statement in the “reasons” column about a factual characteristic of the product in question, that every product like the one in the left hand column, inevitably has that factual characteristic; rather, one is to infer the reverse – that, to fall within the regulation, the product must in fact have the factual characteristic mentioned in the ‘reasons’ column. In our case, we have found that the imported mobility scooters do not have the factual characteristic of being “not specially designed for the transport of disabled persons” (interpreting “disabled” in line with Invamed CJEU); we therefore conclude that, for this reason, and exercising the “great care” required, the imported mobility vehicles are not sufficiently similar to the vehicles in the 2009 Regulation, such that the regulation should be applied here by analogy.

We note that when this point came up at the hearing, Mr Pritchard submitted that there was a material difference between our situation and that in Hewlett Packard, in that, there, the statement in the ‘reasons’ column (about the fax function being the principal function) was about an objective characteristic. We are not swayed by this submission, as, in our view, a statement about which function of a product was the “principal” one, and a statement about what sort of people a product was specially designed for, are equally statements about the product’s objective characteristics.

We are also persuaded by an alternative line of argument put forward by the appellants (but one which, as we will explain, in essence invokes the same basic principle about employing great care in applying a classification regulation by analogy), based on a line of CJEU authorities starting with Case C-51/16 Stryker EMEA Supply Chain Services BV. That case was a preliminary ruling by the CJEU on two matters: (1) the interpretation of a CN heading (could a certain medical implant screw, solely intended to be inserted into the human body for treatment of fracture etc, be classified under heading “X”?) and (2) whether a classification regulation (classifying a particular medical implant screw to heading “Y”) was valid. The CJEU concluded on the first question at [57], classifying the medical implant screws at issue to heading X, based on their characteristics and method of manufacture, which meant they were to be distinguished from “ordinary” goods. The CJEU then went on to the second question:

As context for this, we note the following explanation of what the CJEU does in a preliminary ruling, taken from Invamed CJEU at [16-17]:

Case C-24/22 PR Pet BV was similar to Stryker in that it was a preliminary ruling about (1) the interpretation of a CN heading, and (2) whether this had any implications for the validity of certain classification regulations. Having answered the first question, the CJEU noted the similarity between the products in the classification regulations and the goods in issue; it then said:

In Case C-227/17 Medtronic GmbH, the referring court had suggested three possible classifications for a product. At [54] the CJEU concluded that the product could not be classified to one of these possible headings if it was established that the product was not “intended principally” for a certain use (the treatment of fractures). At [56], the CJEU noted that the referring court had asked whether the product in issue should be so classified on the ground that the product consists in part of something similar to a product that was the subject of a classification regulation. The CJEU then said:

It seems to us that these authorities are saying that where the CJEU has, through a preliminary ruling, given “all the information necessary to classify a product under the appropriate CN heading”, then application of a classification regulation to that product by analogy is (1) not necessary; and (2) not possible. It seems to us we are in just such a situation, given the ruling in Invamed CJEU. HMRC argued that our circumstances are different, in that we are not in a situation where the CJEU has been asked, at the same time, to opine on the interpretation of the CN (heading 8713) and to comment on the validity of a related classification regulation. However, we are not persuaded that this difference in circumstances makes any difference to the underlying principle: if the CJEU has spoken as to how to go about classifying mobility scooters (and it clearly has), it is unnecessary to go through the (painstaking) process of applying a classification regulation by analogy.

HMRC also argued that to apply the Stryker line of cases here, would be to disregard the statement in VTech at [22], that where products are similar, the classification regulation is to be followed unless and until declared invalid by the CJEU. However, Stryker and Medtronic (CJEU authorities that did not exist when VTech was decided) each clearly acknowledge the general principle behind following classification regulations (consistent interpretation, equal treatment, etc); yet in each case the CJEU immediately went on to state that, in the given circumstances, application of the regulation by analogy was unnecessary/impossible. This seems to us very clear guidance by the CJEU on the point and so, to that extent, stronger authority than Vtech at [22].

Although the Stryker line of authorities is a different line of reasoning than the one we initially used (based on insufficient similarity), it seems to us that they both reflect a common principle: the need for great care when applying a classification regulation by analogy.

We are of course aware of what was said in Invamed CA at [22] about the 2009 Regulation, had it applied, being “binding and definitive” – but this clearly obiter statement was, equally clearly, not attempting to encapsulate the whole of the law as regards applying classification regulations by analogy, in a single sentence.

We also acknowledge that, in 2021, the Commission issued a further classification regulation, 2021/1367, that classified another vehicle (which had all the core design features of a ‘mobility scooter’) to heading 8703. However, this makes no difference to our analysis: due to its timing, it does not affect the classification of the imported mobility scooters; and our analysis does not turn on what the Commission may or may not have intended, whether in 2009 or in 2021: in turns on the correct approach, in law, to applying the 2009 Regulation by analogy.

It follows from the foregoing that, in our view, the 2009 Regulation does not affect the classification of the imported mobility scooters to heading 8713.

The conclusions reached above mean that the appeal succeeds: the imported mobility scooters are to be classified to heading 8713.

We acknowledge that the appellants pursued further detailed arguments as to the invalidity of the 2009 Regulation (in the event that we had decided, contrary to the above, that the regulation did affect the classification of the imported mobility scooters), and the powers of the Tribunal to treat the 2009 Regulation as invalid, post-Brexit. HMRC opposed these arguments. However, in the light of our conclusion that the 2009 Regulation does not affect the classification of the imported mobility scooters to heading 8713, it is unnecessary for us to resolve those arguments, and we refrain from doing so.

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.