You cannot compare apples and oranges, the old say goes, although in the trademark world we can, and truth to be told, we usually find them to be “similar” products.
But what about cars and watches? And what about cars and clothing, footwear and headgear? Are they “similar” products? At first sight, it would seem a non-starter, as cars seems objectively and undeniably quite different from watches and clothing.
However, “it ain’t necessarily so“, said the General Court (GC) in its decision T‑177/20, of June 2, 2021, Erwin Leo Himmel v. EUIPO, which annulled the EUIPO Board of Appeal (BOA) decision that the opposition was ‘manifestly unfounded’. How it did, it is explained more in detail below.
Spoiler Alert. The GC did not actually say that “cars” are similar to “watches” and/or “clothing etc.”, see at §56: “Since, owing to its error, the Board of Appeal did not specifically examine the relevance and then, if necessary, the impact of that criterion in its assessment of the similarity between the goods at issue, the Court cannot itself give a ruling on that issue“.
However, the GC held that the BOA had been wrong to rule out, as a matter of principle, the relevance of the “market practice” criterion in its examination of the similarity of the goods. In addition the GC expressly rejected EUIPO’s arguments that, first, the so called “Canon criteria” must always be taken into account in order to assess the similarity of goods or services – since in the GC’s view additional criteria may be relevant (depending on the particularities of the case) and , second, that these additional criteria are ‘weak’ and therefore cannot, by any means, outweigh the lack of similarity with regard to the Canon criteria.
Whether of not EUIPO will appeal to the Court of Justice is to be seen. It might, because whether the Canon criteria should always be taken into account or not, and what their value is in a balancing assessment seem to be issues that are quite significant with respect to the unity, consistency or development of EU law.
In addition, as better explained below, the consequences of the GC’s finding that even a slight similarity between the goods concerned would have required the BOA to ascertain whether a high degree of similarity between the signs could have given rise, in the mind of a consumer, to a likelihood of confusion as to the origin of the goods are significant. Of course, if there is a (slight) similarity the BOA is required to take into account the similarity of the signs, because they are “cumulative”. The problem arises when the BOA finds there is no similarity whatsoever (as it did). To which extent can the GC go to re-evaluate and re-assess the factual conditions that have already been examined (twice) so as to justify the “slight” similarity, and how “slight” is enough?
If not appealed, it will be still interesting also to see whether, on remand, the BOA, applying the CG’s indications, will still find that, all things considered, it cannot compare apples and oranges… or cars and watches/clothes. So, this is not over yet and stay tuned for further developments. Meanwhile, for an interesting look at the background of this dispute, check at https://www.autocar.co.uk/car-news/motor-shows-geneva-motor-show/who-has-right-use-hispano-suiza-name
For all interested in a more detailed analysis of the case, here it comes.
A EUTM application for “Hispano Suiza”, to designate “cars” in class 12 received an opposition by Erwin Leo Himmel (hereinafter Mr. Himmel) owner of the earlier EUTM registration for HISPANO SUIZA, designating “horological and chronometric instruments” in class 14 and “Clothing, footwear, headgear” in class 25. The basis was art. 8 (1)(b), i.e., classic likelihood of confusion.
Unsurprisingly, the Opposition Division rejected the opposition on the ground that, there being no similarity between the contested goods and those covered by the earlier mark, there could be no likelihood of confusion.
No better luck had the appeal, where Mr. Himmel did not even try to argue the “classic” tests of similarity, but rather threw all its chips in a rather different line of thoughts.
First, he argued there is a ‘market practice’ consisting, for car manufacturers, in extending use of their well-known brand to other goods, such as clothing, watches and other accessories (possibly, but Mr. Himmel had neither argued, nor proven that its earlier EUTM registration was a “well known” brand). Second, the general public is aware of that practice (possibly, but says who? where is the evidence thereof?). Third, the general public will thus assume that use of the same brand for the goods at issue indicates that they originate from the same undertaking (possibly, but reality check shows many identical “brands” coexisting on the market because … they designate different goods/services which seems to show that the general public is not in the habit of “assuming” that use of the same brand for the goods at issue indicates that they originate from the same undertaking), therefore those goods will therefore be perceived as being ‘similar’ .
The BOA was somewhat unimpressed and also dismissed this line of arguments by holding that Mr. Himmel was trying to substitute a new criterion of similarity, namely ‘market practices’, for the criteria established by case-law, namely nature, intended purpose, method of use, potential competition and complementary relationship, even though Mr. Himmel had not denied that when the latter criteria are applied there was no similarity whatsoever.
At this point anyone would have just given up (and get a life somewhere else…). Anyone, but Mr. Himmel, who undeterred by the BOA’s decision and findings, decided to bring the case up to the General Court, alleging infringement of, first, Article 46(1)(a) of Regulation 2017/1001, in conjunction with Article 8(1)(b) thereof, and second, the obligation to state reasons.
The arguments at the General Court
Mr. Himmel criticized the BOA for neither having considered all factors relevant to the comparison of those goods, in the light of the circumstances of the present case, nor addressed all relevant arguments put forward in support a finding that the goods at issue are similar.
In Mr. Himmel’s view, the so called “Canon criteria” are explicitly listed as “inter alia”, which shows that the enumeration of those factors is purely indicative and that there may be other relevant factors in addition to or instead of them, as is the case here.
According to Mr. Himmel, if the BOA had assessed the similarity in the light of the criterion relating to the “market practice,” it would have concluded that the goods at issue are similar, because actual and potential buyers of clothing and accessories bearing a trade mark that is also used for cars, generally intend to convey that they own a car from a particular manufacturer or that they have an interest in, or admiration for, cars from that manufacturer, and therefore to associate those various product categories.
EUIPO, in defending the BOA decision, agreed that the case-law does not provide an exhaustive list of relevant factors. However, the same case law nevertheless establishes some factors that should always be considered in order to assess the similarity of the goods or services concerned, namely the Canon criteria.
Depending on the particularities of the case, ‘additional’ factors, such as the distribution channels, the relevant public or the usual origin of the goods or services might also be of relevance, but
Second, EUIPO disputed the relevance of the particular “market practice in the car industry”. Market practices might be of certain relevance in the evaluation of a link under Article 8(5). However, they do not constitute a relevant, let alone decisive, factor in the assessment of the similarity of goods under Article 8(1)(b).
Third, EUIPO asserted that a restrictive approach should be taken in regard to the additional criterion relating to the usual origin of the goods to avoid diluting it. EUIPO took the view that, while it might not be uncommon for car manufacturers to license their car brand for various other products, the public is well aware that this is usually done in the context of ancillary merchandise produced and the public is fully aware that the market for clothing and watches differs fundamentally from the market for cars and will therefore generally not expect the goods at issue to be manufactured and offered by the same undertaking.
The GC findings
The starting point of the GC’s evaluation was the omnipresent interdependence principle, i.e., that a low degree of similarity between the goods or services may be offset by a high degree of similarity between the signs, and vice versa (although the vice versa was never the object of a CJEU case but rather it was a dicta in a case concerning two highly similar trademarks for goods having a low similarity…. see case C‑39/97, at §17 and §19).
Then the GC went on saying that as the likelihood of confusion presupposes both that the marks at issue are identical or similar and that the goods or services which they cover are identical or similar, these conditions are cumulative. So far so good, nothing really new.
From this rather mundane starting point, however, the GC decision-making process went somewhat “astray”. Indeed, one would expect that at this point the GC would analyze whether the two “cumulative conditions” existed. After all, if the conditions are “cumulative”, the absence of one would inevitably lead to the conclusion that a likelihood of confusion was impossible, as a matter of law.
Instead, the GC took issue with the fact that the BOA had found that there was no likelihood of confusion on the part of the relevant public based solely on a comparison of the goods at issue, without examining the other conditions, including, in particular, the similarity between the signs, and without carrying out a global assessment of whether there is a likelihood of confusion taking into account all the relevant factors.
The GC remark, however, is somewhat debatable and does not appear to be at all justified by the GC’s comment that even a slight similarity between the goods concerned would have required the BOA to ascertain whether a high degree of similarity between the signs could have given rise, in the mind of a consumer, to a likelihood of confusion as to the origin of the goods. Of course, if there is a (slight) similarity the BOA is required to consider the similarity of the signs, because they are “cumulative”. The problem arises when the BOA finds there is no similarity whatsoever (as it did). To which extent can the GC go to re-evaluate and re-assess the factual conditions that have already been examined (twice) so as to justify the “slight” similarity, and how “slight” is enough?
In the present case, and notwithstanding both EUIPO Opposition Division and the BOA had found that “cars” and clothing/watches are not similar, it seems hard to deny that the GC went to a great length and to do so it paid a lot of attention and deference (perhaps too much, given that it did without any real factual evidence) to the ‘market practice’ criterion.
Starting with the observation of case law origin that ‘all the relevant factors relating to the goods or services should be taken into account’ and that ‘those factors include, inter alia, their nature, their intended purpose, their method of use and whether they are in competition with each other or are complementary’, the GC agreed with Mr. Himmel that the list of criteria is not exhaustive and was in fact it was supplemented by the addition of other criteria (including the usual origin of the goods concerned , their distribution channels and the fact that the goods at issue are promoted by the same specialized magazines facilitate the perception by the relevant consumer of the close connections between them and strengthen the impression that the same undertaking is responsible for the production of those goods.
Therefore, the GC held that the BOA had erred in law, by ruling out, as a matter of principle, an assessment of the similarity of in the light of the market practices criterion (which the applicant had proposed). The GC held that it “cannot be ruled out that criteria other than those referred to by the Board of Appeal in paragraph 17 of the contested decision, namely, besides the Canon criteria, the distribution channels and the fact that the sales outlets are the same, may be relevant in assessing the similarity of goods or services in general and of the goods at issue in particular“(at §46).
The conclusion reached by the GC seems logical and correct. However, saying that “something” cannot be ruled “out”, does not necessarily translate in can be ruled “in”. This is equivalent to say “aint’ necessarily so”, without saying why. Which is precisely what the GC did.
The GC recognized that the BOA had set out the relevant criteria for assessing the similarity of the goods, while stating that other factors may also be taken into account. However, the GC criticized the BOA for having failed to assess the similarity of the goods at issue in the light of those other factors. Which does not seem to be true.
The BOA had expressly rejected the criterion of the market practice relied on by the applicant, because it had held that it was an attempt to substitute a new criterion of similarity – market practices – for those established by case-law, namely the nature, intended purpose, method of use, competition and complementary relationship. In other words, the BOA had recognized that if there are a (overwhelming) number of criteria which, when applied at the case at issue, all exclude any finding of similarity (even slight), the possibility -better, the probability, that just one single theoretical (because not even adequately proven) criterion might lead to establish a (slight) similarity are so low as to become irrelevant.
So how did the GC come to a differing conclusion? With a little bit of stretching of imagination…. The GC built its argument on paragraphs 21 and 23 of case Hesse v. OHIM, C‑50/15, saying that the CJEU had indicated that it “cannot be ruled out that a relevant criterion may be capable of being the sole basis for the existence of a similarity between the goods or services, despite the fact that application of other criteria would indicate that there is no such similarity” at §48.
Actually, in case C-50/15 the CJEU never said ” despite the fact that application of other criteria would indicate that there is no such similarity. Furthermore, in case C-50/15 the (relevant) criterion was that of ” complementary nature of the goods” which the CJEU has already indicated to be among those which are “relevant” (see at Sunrider v OHIM, C‑416/04 P, §85) , while and admittedly, whether the market practice criterion is “relevant” has never been the subject of a CJEU decision (and despite the GC’ efforts, not even, really, of any earlier GC cases, see also infra). Thus, when the GC later used the reference to case C 50/15 also to counter EUIPO’s arguments that, first, it is apparent from the case-law that the Canon criteria must always be taken into account in order to assess the similarity of goods or services, whereas additional criteria may be relevant depending on the particularities of the case, and, second, those additional criteria are ‘weak’ and therefore cannot, by any means, outweigh the lack of similarity with regard to the Canon criteria, the GC stated a “rule” without laying down any real and proper foundation.
Truth to be told, the GC was aware to be treading on thin ice, because on the one hand it also added, to further substantiate its position that the existence of a certain market practice may constitute a relevant criterion for the purpose of examining the similarity between goods or services : ” in any event, it is impossible to assess the impact of an additional criterion, such as market practice, on the analysis already carried out on the basis of the factors arising from [Canon], before that criterion is properly examined. The Board of Appeal carried out no such examination. Which, however, is not true….
On the other hand, the GC also tried to show that the market practice criterion had previously been taken into account in assessing the similarity of contested goods or services. However, the cases cited by the GC are only partially (to be generous) relevant to prove the relevance of the “market criterion” as an “autonomous criterion”.
For instance, in case T-562/14, the Court had first held that the products were sold in the same places and were complimentary, and only as third reason, it stated “it is common knowledge that providers of restaurant or catering services often make their own pastries or chocolate products. Conversely, bakers or pastry chefs have developed catering services and snacks, which include, in particular, pastries and chocolate products“, which objectively seems different from saying that the public believes car manufacturers also make clothes/watches.
Similarly unconvincing are the references to other earlier cases, like T-387/20, where, according to the GC, it had been held “important to assess whether consumers would consider it usual for the goods at issue to be sold under the same trademark, which normally implies that a large number of producers or distributors of those products are the same”.
The problem with such a sweeping argument is that it is an over- generalization because it was drawn from specific circumstances of that particular case, where, most of all, evidence had been submitted to prove that “the goods being compared may be sold together in shops specialising in interior design, which offer coordinated goods making it possible to arrange a home in such a way as to achieve overall harmony (see, in this regard, for example, the evidence presented by the intervener during the administrative proceedings before EUIPO, concerning the sales on an e-commerce site of goods intended for ‘bedrooms’ and ‘decoration’). That fact is likely to facilitate the perception by the relevant consumer of the close connections between them and to strengthen the perception that the same undertaking is responsible for the production of those goods” (at §59).
Even making the best efforts, how such a case may support the “relevance” of the “market criterion” as an alternative, independent and autonomous factor in assessing similarity of goods, seems at best, dubious, if at all.
Please note the GC also held that, by not stating the reasons why criteria relating to the usual origin of the goods or the fact that the producers and distribution channels are the same were not taken into account, the BOA had failed to set out all the facts and legal considerations that are crucial for finding that the goods are not similar, with the result that the contested decision is vitiated by a failure to state reasons. Of course, whether application of those criteria would have led, in the present case, to a finding that the goods at issue are similar is still the real question. Let’s see what the BOA, on remand, will say.
© BUGNION S.p.A. – July 2021