By Sarah Burstein, Professor of Law at the University of Oklahoma College of Law
Converse v. ITC and Sketchers, New Balance, et al. (Fed. Cir. 2018) [ConverseITC]
As Professor Crouch already pointed out, the majority here held that “[i]n the context of trade-dress infringement, . . . accused products that are not substantially similar cannot infringe.” In doing so, the majority cited a perfectly good and persuasive authority, Versa Prods. Co. v. Bifold Co. (Mfg.), 50 F.3d 189, 202 (3d Cir. 1995). The discussion could have just stopped there.
But no. The majority, in a decision written by Judge Dyk, went on to add this bit of dicta:
We have applied an analogous requirement in the design-patent context, where infringement cannot be found unless an ordinary observer would perceive that the “two designs are substantially the same.” Egyptian Goddess, Inc. v. Swisa, Inc., 543 F.3d 665, 670 (Fed. Cir. 2008) (en banc) (quoting Gorham Co. v. White, 81 U.S. 511, 528 (1871)).
Hold up. “An analogous requirement?” I’m honestly not sure what the majority thinks is analogous here, other than the use of the words “substantially” and “similar” in these two areas of law (which are already rife with faux amis).
So, just for the record: The tests for design patent and trademark infringement are different.
A design patent is infringed when the defendant’s goods are too visually similar to the claimed design. (Yes, this is true even under Gorham v. White, as I explained at p. 177 here.) A trademark is infringed when consumers are likely to be confused as to the source, sponsorship, or origin of goods or services. Thus, factors like labeling, marketing, etc. are (at least in theory) relevant to the question of product-design trade dress infringement. But they’re not relevant to the question of design patent infringement.
It’s true, as the majority points out, that visual similarity is one element of the various likelihood-of-confusion tests. To the extent that these judges are suggesting that they’d like to import the (very high) degree of visual similarity required for a finding of design patent infringement into the trade dress infringement test, that would be a wise and welcome addition to the law. But I’m not sure that’s what they’re trying to say.
Reading the whole paragraph in context (as reproduced below with the citations omitted), it seems like they might be trying to import the anticipation-infringement symmetry rule from patent law into trade dress law. Here is the entire paragraph for context:
The likelihood-of-confusion analysis for determining infringement turns in part on the similarity of the accused products to the asserted mark. We described earlier that, in the invalidity determination, marks that are not substantially similar cannot be considered. In the context of trade-dress infringement, we also hold that accused products that are not substantially similar cannot infringe. We have applied an analogous requirement in the design-patent context, where infringement cannot be found unless an ordinary observer would perceive that the “two designs are substantially the same.”
I’m not aware offhand of any trademark cases that stand for the proposition that there must be symmetry of this type (though I haven’t researched this precise issue and would be happy to hear from anyone who has seen such cases). In any case, this seems like the type of situation where asymmetry would actually be a good rule. We might want product-design trade dress claimants to establish more visual distance from other users in the field than we would allow them to capture in any eventual trade-dress scope.
In the end, I’m not sure what the majority was trying to do here. But this decision could do without the design patent dicta. At best, it’s harmless surplusage. At worst, it unhelpfully conflates two dissimilar infringement standards and could cause unintended mischief in the future. So, Federal Circuit, when it comes to conflating design patent and trade dress infringement – can you just not?