Are you an entrepreneur who sells t-shirts for dogs? If you want to register your logo as a trademark with the USPTO, you’ll need to include in your application an acceptable description of the goods you sell. But you’re in luck: if you pull up the Trademark Office’s ID Manual and type in “dog shirts,” you’ll get “Animal clothing for {specify type of animal, e.g., pigs, dogs, etc.}, namely, {indicate specific item, e.g., shirts, skirts, scarves, etc.}.” You can fill in the information in brackets—“Animal clothing for dogs, namely, shirts”—and proceed with confidence that the examining attorney will accept your description, because it already has the USPTO’s seal of approval.
But if you’re a psychologist who specializes in gender-affirming therapy and wants to register your service mark, your search of the ID Manual won’t be as fruitful. The Manual contains zero identifications relating to gender-affirming healthcare. You’ll have to craft your own description of services and hope it passes muster. The ID Manual is a searchable database that includes more than 70,000 pre-approved identifications of goods and services for trademark registrations. But what it excludes may be just as important as what it contains. In Discrimination in the Trademark ID Manual, Professor Rachael Dickson explains that while the USPTO created the ID Manual to offer a helpful guide for trademark applicants and attorneys, it also serves as “a powerful gatekeeping device—one that quietly but systematically disadvantages entire industries.”
And Dickson should know: she’s not only a law professor, she’s also a trademark lawyer, a former USPTO examining attorney, and something of an ID Manual vigilante. Over the last several years, she’s submitted 43 suggestions for inclusion in the ID Manual, many related to precisely those categories of goods or services she argues the Manual disfavors. In addition to engaging in that administrative process and cataloguing the results, Dickson submitted FOIA requests to gather information about how those submissions were received and why most were declined. Unsurprisingly, the results came back heavily redacted.
The USPTO publishes and updates the ID Manual regularly to provide applicants with options that purport to encompass the full range of goods and services trademark applicants might sell. Applicants or their lawyers can search the Manual, select the best fit, and rest easy knowing that any description they find there is preapproved and unlikely to generate pushback. Even better, they’ll save money by choosing an ID from the Manual. Applicants who don’t select their descriptions of goods and services from that pre-approved set pay a $200 surcharge per class, increasing their filing costs by fifty-seven percent.
Those who use descriptions from the Manual also save time in the examination process. Applicants who go off-list require custom descriptions, which usually involves back and forth with the examining attorney that extends the timeline to registration and requires more billable lawyer hours to resolve. The USPTO explicitly advises applicants that choosing an ID from the Manual means they’ll reach the finish line more quickly.
The ID Manual is not mentioned in any statutes, and those outside the world of trademark registration are unlikely to know about it. It has never been the subject of a law review article until this one. “And yet,” Dickson argues, “for applicants seeking to register trademarks, the ID Manual often functions as de facto law.” The USPTO’s choices about what to include and what to leave out have broad impact, and they affect different applicants differently: “the USPTO’s administration of the Manual systematically disadvantages trademark applicants offering lawful but socially disfavored goods or services,” Dickson claims, such as those relating to reproductive healthcare, gender-affirming services, or drugs and other vices.
Dickson’s evidence highlights the importance of ID Manual entries for applicants, as where the addition of several identifications for medical marijuana in 2010 (later removed) correlated with the filing of at least 86 related applications in a three-month period. The same pattern has played out with the addition of identifications in other categories, such as NFTs, condoms, sex therapy, and intra-uterine devices. (Use of IUDs dates back to at least 1969, but the term wasn’t added to the ID Manual until 2009).
Years after the Supreme Court struck down the prohibition on registration of marks deemed scandalous or immoral, mark owners in fields from sex work to cannabis often find the Manual holds no easy-to-locate identifications that fit their goods and services directly, leaving them to craft custom descriptions. Dickson argues the Manual also disadvantages pro se applicants and applicants offering goods designed for or marketed to members of minority religious and ethnic groups.
Dickson’s activism and her research reveal inconsistencies in the ID Manual and the USPTO’s treatment of proposed additions, and perhaps some prudishness, though not necessarily conscious discrimination. But she argues that the stated reasons for refusing proposed identifications reflect agency bias, and some of her evidence supports that theory.
As Dickson notes, courts in recent years “have shown an increasing willingness to scrutinize trademark doctrines that impose viewpoint-based harms.” Her research suggests the ID Manual might not withstand such scrutiny.






