Patent law scholars argue over how much time and money the Patent & Trademark Office (PTO) should spend on pre-grant review of patent applications. Likewise, they argue over the degree to which patents should be given a "presumption of validity" once granted. But what if patent law scholars completely ignored the details of the PTO's review, and simply assumed that patents are generally properly and efficiently granted? What if courts in patent infringement cases treated PTO review as a mere formality and focused exclusively on the appropriate scope of patents and whether they have been infringed in the marketplace? We would probably conclude this to be error. Given the time and public money expended each year on the process of patent examination, not to mention the role of published patent specifications in establishing and providing public notice of granted rights, it is worth paying attention to the administrative procedure through which patents are created and preserving the relevance of the PTO's analysis through doctrines like presumption of validity and prosecution history estoppel.
Yet, according Rebecca Tushnet's new article in the Harvard Law Review, Registering Disagreement: Registration in Modern American Trademark Law, foundational trademark law and scholarship suffer from a similar form of tunnel vision. Not only do they place inconsistent weight on the PTO's assessments during registration for establishing substantive trademark rights, but they do not apply a consistent vision of what the role of trademark registration actually is. Read more at the jump.
Foundational critiques of trademark law, Tushnet writes, focus on the proper scope of trademarks, largely as determined by courts at the enforcement stage, but "tend not to address the role of registration." (3) Modern trademark law scholars argue vigorously about fundamental scope issues such as whether trademark law serves the goal of lowering consumers' search costs or exists mainly to reward producers' investment in unique marks. But this debate over substantive scope neglects the separate role of registration as an administrative procedure through which rights are defined and recorded. (8)
To some degree, this difference makes sense. Patents issued by the PTO represent the full universe of patent rights, at least until they expire or are invalidated in litigation or post-grant administrative review. But as a registry of substantive rights, trademark registrations are both over and under inclusive. Trademarks are created in the marketplace through consumer use and protected by state and federal statutory and common law to the extent that they result in a likelihood of confusion. Thus, registered trademarks are merely a subset of the universe of trademark rights. (11) On the flip side, the trademark registry contains plenty of "deadwood" in the form of improperly granted registrations that can be invalidated later during litigation. (2) So within this "mixed registration" system, we would not expect courts or scholars to treat trademarks issued by the PTO as representative of an actual right or to consistently give them a "presumption of validity" in the same way as a patent. We also would not expect the PTO to expend as much effort perfecting the process of trademark review: if trademarks are, at the end of the day, determined by usage, then why bother?
But, Tushnet points out, maintaining the federal registration system requires "substantial government and private resources." (5) This does not even count the cost of maintaining a separate system of state registrations. Moreover, registered trademarks are ostensibly there to inform businesses and the public about what marks are taken. As Tushnet notes, law & econ scholars defend registration as a way to "provide cheap notice to potential adopters" of a mark; simultaneously requiring actual use, they contend, serves to "avoid rent seeking behavior by squatters." (3, n. 9, discussing Landes & Posner's view). So it seems odd and inefficient that the PTO's prior assessments during registration do not consistently supply a framework for courts' validity and infringement analysis at the enforcement stage. As Tushnet puts it, "[a]pplicants and the PTO spend much time and effort crafting the equivalent of an exquisitely detailed origami crane; rather than considering the details, courts then ask the equivalent of 'is this paper folded?' and move on." (5)
The situation is still more complicated, Tushnet asserts, because sometimes the registration system does significantly affect the boundaries of the actual trademark right. Tushnet identifies two competing accounts of the role of registration. In one account, the "procedural version," registration has no independent meaning, and serves merely to record the actual exclusive rights created in the marketplace, whose boundaries are determined by the degree to which infringing uses are confusingly similar. But in the other account, which she calls the "substantive version," registration represents a meaningful subset of protectable marks regardless of consumer confusion. The analogy in the patent context might be if courts paid little attention to the patent's actual claim language in calculating the scope of the patentee's exclusive right, and instead relied merely on whatever rough outline or working model of the invention that the patentee submitted to the PTO at filing.
Perhaps the clearest example of trademark law's inconsistent treatment of registration along this spectrum is the doctrine of incontestability. By statute, after a mark has been registered for five years, the registrant may file a declaration of incontestability, which becomes evidence of the validity of the mark and prohibits invalidation for mere descriptiveness. (35-36, citing 15 U.S.C. § 65; Park 'N Fly, Inc. v. Dollar Park & Fly, Inc., 469 US 189, 196 (1985) ("Mere descriptiveness is not recognized ... as a basis for challenging an incontestable mark")). The opportunity to gain incontestability suggests that registration provides a substantive benefit that is not conferred by usage with a chance of confusion, and that registration of a mark alone creates a scope of exclusivity. At the same time, however, the Supreme Court has held that a registered, incontestable mark that is merely descriptive and lacks secondary meaning, and therefore not likely to confuse consumers, cannot support a successful claim for trademark infringement. (38, discussing KP Permanent Make-Up v. Lasting Impression I, 543 US 111 (2004) (holding a plaintiff claiming infringement of an incontestable mark still must show likelihood of consumer confusion in the marketplace as part of her prima facie case)). This suggests that, in fact, actual usage and consumer confusion are the benchmarks for determining trademark rights, and that the incontestability concept is "functionally meaningless." (15).
If this seems highly inconsistent and confusing, this is Tushnet's point. The confusion, she asserts, does a disservice to the operation of trademark law because it allows the law to shift selectively and sometimes arbitrarily between serving separate goals: avoiding actual consumer confusion, on the one hand, and serving other "non-confusion" goals, on the other, such as inducing investment in marks or managing relations among trademark owners by clearly allocating rights. (14, 74). It may take some time to sift through the problems Tushnet has identified. I am very grateful to have encountered this article before preparing to teach trademarks for the first time this coming spring.