IPSC: Trademarks in Society

Jonathan Masur, University of Chicago Law School, Measuring the Value of Trademark Distinctiveness: Evidence from the Market for Bordeaux Wines

Branding is important; confusion is the core justification for TM to allow consumers to get what they want and firms to recoup quality investments. So we don’t want too much bunching of highly similar marks. At the same time, generic/descriptive names are close to the linguistic core and therefore powerful for marketing b/c of association w/product core. So you get bunching there: POLAR and ARCTIC coolers etc. There are also brand names that are hard to remember b/c they’re so far from the linguistic core—DAUSROOB coolers. Marketers think you want to pick something in the middle. But there isn’t a lot of empirical evidence, so we wanted to test whether distinctive brand names generate price premiums.

Testing hypotheses: distinctively named wines will command higher prices; holding wine quality constant, distinctively named wines will command higher prices; price differential will be greater for higher-quality wines b/c greater reward for good product (could imagine the alternate hypothesis, where wine market is segmented where best wines are purchased by most knowledgeable consumers who make finer distinctions).

Took out all the French articles, words like chateau, then computed similarity w/ten nearest neighbors. Result: highly similar wines (wines w/many near neighbors)—expected price $13.44; highly dissimilar, expected price 27.39—100% price premium. But control for wine quality. A wine rated 90: highly similar neighbors (0.83): expected price 21.68. Dissimilar: 25.02.

Interaction term b/t similarity and rating is negative and significant. Low rated (82), price premium 2.21; high rated (93), price premium for distinctiveness 9.59, which is slightly higher in percentage terms and also obviously absolutely higher. No meaningful distance costs for being too unique. Don’t see price benefit for low priced wines that might have gotten a boost for similarity to their high priced neighbors.

So why don’t they switch?

Congestion is a problem and TM should push brands apart.

RT: Are there any Amazon style GOTUFK names? Can you test this hypothesis with Amazon brands for coolers/gloves?

A: Red Bicyclette is probably one of the most distinctive Bordeaux names, aimed at American audiences. [That’s nowhere near as far away as GOTUFK.] The innovators generally have French names and designs on the bottles.

McKenna: you’re using brand and TM interchangeably; encourage you not to do that b/c marketing literature distinguishes b/t those for the reasons you just gave—packaging and icons matter. Uniqueness might be a better term than distinctiveness (or differential distinctiveness). Many Bordeaux wines are quite clustered in branding while having distinctive trademarks. If you switch, you might lose benefits of conveying you’re a Bordeaux wine.

A: yes, we’re using it to mean “distinctive from” not “TM distinctive.”

Stacey Dogan: congestion is a problem: for whom? For the wine seller who wants to charge a greater price premium that is not related to the relative quality of its neighbors. From a business perspective, sure, that’s a problem for the seller, but is it a problem from a social perspective?

A: the thesis is that the winery can charge more b/c people more easily remember that they liked it.

Dogan: a premium even after controlling for quality: what does that mean?

A: if you liked two wines a lot, and one of them was easier to remember, you could go for that first wine again, driving up the price, but that price is driven by consumer preference.

Felicia Caponigri, Marquette University Law School, Pantone's Color Monopoly?

Early stage project; wading into antitrust/IP interface. Pantone’s story: before 60s, color production was a mess—different printers would mix reds differently. Some color standardization systems existed, but Pantone really standardized. But why do Pantone’s colors cost so much? Pantone’s color books cost up to $9000; swatches $5. 15-25% more for a solid spot color than a dot color. Is Pantone color even a product? What services does Pantone provide (like color matching, color recipes) that enable it to corner the market?

April 2025: Pantone cancelled its agreement w/Adobe & created a separate digital platform at $15/month required to digitally communicate using Pantone colors.

Booklets are tangible products; but can also check color online. Pantone targets two demographics: fashion designers and graphic designers/product packaging designers w/two different systems. Fashion designers: materials like cotton. Product packaging: core Pantone matching.

Pantone uses ©, TM, a strange licensing system, and contract to cement the idea that they own the standardization of color.

Many © registrations for nondramatic literary works—matching system booklets. In 1968 successfully enjoined a company creating similar color sheets with bands of colors. Court allowed © in selection and rearrangement. There were differences b/t Pantone & D’s sheets—different numbers of colors and sheets—but still found substantial similarity. Pantone has been using this one case to assert © protection.

Trademarks: in the chip display (for printing ink/other forms of color generation), swatches design (for formula scales, bulletins, promoting the study of color), and PANTONE MATCHING SYSTEM. But it seems to offer licensing for specific colors (though in practice it might be color + word Pantone). Color matching: one on one contracts with brands like Valentino for a specific color, or color of the year.

Licensing agreements waive any contest of Pantone’s rights.

They say “we’ve created the best system.”

Legal questions: Originality? Is Pantone claiming © in the color matching process, not just the books? The one court said there was only © in the booklet, but what’s the difference. What value is there in Pantone (co)branding colors from the “real world”—finding Pantone matches for the ocean? Artists like Stuart Semple reverse engineer things like Tiffany Blue and then sell the paint. Are the licenses Pantone uses too restrictive? Is Pantone a monopolist or just an innovator? Is it possible to compete?

James Grimmelman: Pantone sells a children’s book which claims to have samples of different colors. The Pantone numbers are real but the names are made up, like “pickle green.”

RT: legal actions/registrations outside the US?

A: haven’t looked at that yet; EU requires you to cite to a color identifier for your EU registration and it offers Pantone as one of the options.

McKenna: Legitimate business interest in selling standardization that guarantees people they can recreate colors in the same way. So maybe sort among types of uses.

Masur: Is Pantone asserting rights against someone making a yellow or a purple? If Pantone says that purples too similar to its purples infringe, that’s a problem.

A: Not doing that, except that (maybe) printers can’t use the word Pantone w/o permission

Alex Roberts: use of licenses in literature more generally? [I was thinking about the tarot copyright presentation yesterday; the methods of claiming ownership seem very similar!]

Betsy Rosenblatt: is there a way to claim a taxonomy that isn’t dodgy under IP law?

A: maybe not!

Sari Mazzurco, Southern Methodist University Dedman School of Law, Source & Solidarity

Intersection of labor law and TM. Union marks are characteristically different than TMs and Congress knew it when it protected certification/collective marks under the Lanham Act. Aim: To improve coherence and legibility of union disputes.

Starbucks, Trader Joe’s, and Medieval Times sued their labor unions for TM infringement. TJ made a standard source confusion claim based on Trader Joe’s United sale of buttons, mugs and bags. Dct found no likely confusion, but TJ might bridge the gap and start selling mugs and buttons. The fact that it’s possible to navigate directly to TJ United’s shop which didn’t directly criticize TJ, and the similarity of logos might make that shaky. The court didn’t seem to give weight to the fact that TJ United was a union.

But unions are specifically about workers. Meant to achieve industrial democracy through worker participation in business decisionmaking. Goals are economic and political—working conditions, voice, and self-determination. Union names and labels mirror political/economic character, w/emphasis on political. Early names referred to the trade and to unionization. Crafted to be broad, cover the entire target constituency.

Union labels are a bit different. Original conception: physical labels applied by union member to product if and only if employer met union conditions. Signaled made w/union labor. Value to labor unions was different than the value of TMs—weapons deployed in warfare against employers to rally consumers to buy only union-made goods, putting pressure on employers to bargain. Distributed novelties and paraphernalia to consumers to promote this.

Courts don’t seem aware of these differences; they apply standard TM conventions and doctrines that don’t fit well and treat unions as profit-motivated transactors or speakers; treat union labels as source signals; evaluate sponsorship/affiliation confusion as if it’s possible for employers to sponsor unions, which is illegal in the US; and they treat confusion in a standard way.

Past precedents: inconsistent results based on fundamental incoherence: Save Brach’s case—barred labor union’s use of Brach’s logo in logo used to protest closure of Chicago plant b/c protest was a “service”; not much consumer confusion reasoning. Another court found that parody version of saloon logo used to protest management wasn’t a commercial use but an expressive one.

Courts don’t seem to recognize the different character of labor unions and their marks, but Congress did by recognizing collective/certification marks. Those don’t distinguish one manufacturer’s products from another; they are used to advance noncommercial objectives; union members could use them only as long as they were part of the union. Some courts held them categorically ineligible from TM protection; unions thus pressed for sui generis state law protection; all states adopted union label laws by 1900. Applied only to unions, protected against unauthorized use or counterfeiting; required registration; criminal and civil penalties.

Much debate in drafting of Lanham Act of where to put union marks; ultimately included in both collective and certification marks. Understood that union marks weren’t trademarks or service marks—didn’t consider unions as providing goods or services to members. Despite common practice of distributing merchandise to publicize the union label. Second, unions didn’t participate at all—ABA, Nat’l Ass’n of Manufacturers, and USAG did and talked about them. But Third, they’re the only institutions included specifically in the definitions of collective and certification marks. Other groups were understood to be covered by the loose “association.” Fourth, the Lanham Act was enacted in a period of hostility to labor unions—but Congress never questioned whether they should be covered.

Recovering labor solidarity as a purpose has potential to affect Lanham Act doctrine.

Fact that union marks aren’t source signals should support a presumption against likely confusion. Intent to deceive should be required. Plus employers shouldn’t be able to drag unions through litigation—courts should use NFU for unions or adopt a new Rogers-type labor use defense asking whether a union is a bona fide union and is the use explicitly misleading.

Jeanne Fromer: History is about union marks as sword rather than shield. How do you analytically make the leap to “when can unions use the employer’s mark”? Could TM use be part of that?

A: Congress understood that their marks don’t stand for source, but that ought to affect confusion analysis.

McKenna: a bunch of the reasons that early courts gave for not calling these marks are no longer true. That can explain why NAM would want them recognized—the understandings of what made something a mark were just too limited.

A: still requires squeezing/reorientation of understanding what unions are. Merchandizing for political/labor purposes is different than merchandizing for your company. [Donald Trump would disagree and even non-Trump judges are pretty likely to agree with him.]

Rosenblatt: Marks of all sorts do non-economic things. Union marks may do both collective/certification things, but they may also do service things. It would be weird under current doctrine to say that what unions do now would not qualify as a regular service.

A: and that’s a departure from labor unions’ understanding of what kind of organization they are. They’ve adapted. But that’s not what Congress intended, and leads to weird results in court. [But they aren’t distinctively weird—there are lots of bad TM cases.]

[RT: The history is great; should be front and center. But: History doesn’t show “labor solidarity” as a purpose of the Lanham Act; if anything, shows intent to assimilate unions into the larger economic scheme. Unions weren’t there to articulate a distinctive vision of why they were protected. Why should infringement of a TM by a union require separate analysis; why would it matter whether the union itself owns a TM versus a collective/certification mark?

The fall of sectoral bargaining and the rise of employer by employer bargaining also changes things in TM-relevant ways. What kind of mark is Amazon Labor Union? I think “highly descriptive” is the best answer, even though it incorporates and refers to a mark that is inherently distinctive for the relevant services. This might connect up with the idea that nominative fair use is actually a kind of descriptive fair use (as the 6th Circuit has sometimes held) where you are using the TM to talk about your relationship to the TM owner. It is a descriptive use b/c it is descriptive of the relationship existing between the two. That could let you say something really interesting about TM defenses more generally.]

Zaneta Robinson, Wake Forest University School of Law, Language Preservation with Trademark Registration

Early stage. At least 50% of world languages are in danger of extinction or serious endangerment by 2100. Pessimistic view: 90-95% will become extinct or seriously endangered. Roughly 7000 languages spoken now; could go to 300-600. That’s terribly alarming. More endangered languages spoken in/around NYC than have ever existed anywhere else, but likely never that many in any single place again.

Thesis: current TM examination process supports linguicide. Words are at the center of language and TM law. Forcibly lost when gov’t actors reject marks derived from minority languages through a registration process that quickly represses them. Doctrine of foreign equivalents is the bad guy. Creates inequities in registration. DFE: Common modern language that the ordinary American purchaser is likely to stop and translate. It is reflexively applied. There is no guidance as to when it should apply or not. TM examiners can use any internet/database evidence. End up with wild decisions based on newspaper articles or other random evidence. Nigerian immigrants wanted to use a name to pay homage to Ibo heritage; there was a prior English mark that was similar, and so Office Action was possible. There are references to how many people in the US that speak an “African” language, so purchasers would stop and translate. But what is a “common modern language”?

Office actions citing DFE and likely confusion: examples include endangered/potentially endangered languages like Inukitut, Sicilian, Western Frisian, Persian, Scottish Gaelic; also cited Latin, Afrikaans, Hindi, Japanese. Connotation is overused—translate to English, then compare English to translated mark. But that makes a similarity finding much easier, ignoring sight and sound.

What about registration as a tool for preserving languages or at least words from dwindling languages? There are minimal language protections in the US; SCt struck down the only provisions that protect culturally offensive uses. What has been successful with other communities? Hawaiian, Hebrew, Maori. Special rules for in-group members v. third parties like Disney (Hakuna Matata)? Dwindling language communities: notice, seek consent from community as from living individual, repository? PTO should consider having an endangered languages registry as it does for Native insignia.

Fromer: Is the lever to push on that the term would be understood? There’s a tension here between protecting language and encouraging its broader use. Protecting language shrinks the ability to use it. Not clear the solution is TM protection if you care about protecting vibrancy of language. Might want to look to computational linguistics—researchers have been working on developing corpora for underresourced languages to preserve them and make them usable with AI etc.

A: getting words into circulation is valuable. Has article on corpora coming out.

Rosenblatt: Matal, Commanders—a lot of discussion of reclaiming derogatory terms in the hope that would prevent others from using them—the opposite of what you’re talking about here. The response was: that won’t stop other people from using them b/c TM restrictions only do certain things. That’s also true on the flip side; you need goods & services to go with these marks. What pieces of the language are you preserving and who decides?

A: the relevant language community.

Q: Cf. patent space and traditional knowledge: is protection in the form of exclusivity what we want, or is it that we don’t want the wrong people to have exclusivity? TM corpus as a body that we could think of as preserving something, like we think of patents as a corpus preserving knowledge.

[RT: following Fromer: most TMs fail and don’t make an impact on the linguistic communities to which they’re directed, so not a good place to exercise leverage. More broadly: preserving languages and preserving words from them are different things; a word preserved as a TM would be commodified in a way that doesn’t seem like it’s part of a living language. Specifically, wouldn’t DFE point in a different direction for descriptiveness/genericness—encouraging applicants to disparage their own language as extinct or endangered—than it does for confusion? Maybe this is a congestion argument: seems like you agree that we should not consider most translations to be too similar—apply a culturally sensitive version of stop and translate that doesn’t just ask “do a lot of people know this language?”]