1st Breakout Session: Trademark Theory
Harms, Benefits, and Justifications in Trademark Law
© and patent are often considered to be about incentives (getting benefits), but TM law in the general view is about preventing harm/misinformation, at least in theory; misinformation can harm both producers and consumers. Quibbling about whether consumers or producers are law’s intended beneficiaries, but all in agreement about the harm based narrative. In her view, despite its recent caricature in the literature, the economic view of TM, appropriately conceived, is perfectly consistent with this harm based narrative and doesn’t call for intervention when misinformation is irrelevant.
Slew of TM scholarship in last decade has criticized law’s breadth and expansion by demonstrating that current law targets behavior that is simply not harmful. Dogan & Lemley, Tushnet, etc.—what courts are treating as harm isn’t really harm. But the more she thinks about it, the more she realizes that the harm narrative simply doesn’t work as a comprehensive description of TM law in the US, and hasn’t for a long time. Courts turn to the trope of harm, and contrast it w/patent and copyright, but many features of system make no sense in a law that aims primarily to prevent injury either to producers or consumers. Not a utilitarian view that preventing misinformation is the best way to promote competition. Instead, a moral sense that TM owners and not some third party deserve the results of their labor, absent some justification. Unfair to profit from deliberate use of someone’s protected mark, unless exception is shown. Benefits-capturing rather than harms-preventing approach. Courts/Congress are sometimes explict but other times turn to pretext about harm.
Post-sale confusion and counterfeiting: quite explicitly finds any confusion among anyone sufficient, even if confused person would have no purchase interest; case law is explicitly benefits-capturing. Dilution law as applied also has a benefits-capturing approach. The statutory preamble discusses harm, but the factors listed are all focused on evocation. Merchandising/confusion as to association/confusion in unrelated markets. Initial interest confusion. Courts just don’t like someone making money from evoking a TM.
All of these doctrines involve deliberate use of mark. And, to the extent it’s found to infringe or dilute, the conduct is seen as lacking in social value. Deliberate free riding: instinctive sense of unfairness. Absent justification, courts distrust attempts to benefit commercially from use of someone else’s mark. Even in early TM opinions, see courts troubled by uses w/no explanation or justification but to take advantage of another’s reputation. Decisions that refused to enjoin such conduct even in early days tended to involve cases where there was reason to worry about enforcement. E.g., use of a geographic term—might be some confusion, but also important to preserve term for others to use truthfully. This is a justification. Showing either your specific use or this type of use generally deserves recognition.
In today’s benefits-capturing TM world, the action is in demonstrating the social value of unauthorized use. At least where a party is deliberately using another’s mark for commercial reasons, party has a soft burden to show that its use/type of use is worth preserving. Courts do shape the doctrine to respond to those concerns. Other values courts have responded to. Call to action: instead of waving our hands about absence of injury, we should think about what kinds of uses of marks have social value and begin demonstrating the social value of uses we think shouldn’t be prohibited. E.g., study showing that elderly people comply better w/drug regimens when they have the same shape/color drug over time.
RT: Is evoking benefits anything other than good/inevitable litigation strategy? Compare §230 rhetoric or even a classic criminal defense. The harm/benefit distinction has been challenged by realists. Playing that out here, it may just be standard for defendants who are protected even by a harm rule to remind courts/courts to remind readers of the reasons for that rule--§230 couldn’t be more explicit that it sacrifices the interests of some people who are indeed harmed to the greater service of protecting “the internet,” and courts and defendants nonetheless remind us all the time of the good purposes Congress had. First Amendment implications: free riding alone may not justify speech regulation, per Alvarez. I also caution against the (very common; I’m sure I’ve done it myself) use of the idea of an “instinctive” sense of unfairness. Did it seem unfair when your kid learned words from you or learned to write? Unfairness is socially constructed. (That we are hardwired to find some things unfair is plausible to me. But not which things, at least in this context.)
Dogan: courts are making fairness determinations within things like dilution—sex exceptionalism in what counts as dilution. Agrees that “unfairness” is socially constructed and evolves over time; some chicken and egg in terms of legal standards driving what counts as unfair versus changes in nature of brands over the course of 20th century not driven by the law.
Lisa Ramsey: First Amendment analysis may support your call for greater use of empirical evidence, looking for evidence of real harm/support of narrow tailoring.
Dogan: is a pragmatist/instrumentalist. We can move incrementally. If we successfully demonstrate social value in a handful of cases, over time we can generalize.
Fair Music and False Advertising
Topic: Advertising that makes claims about how musicians are compensated by companies. Sellers compete at least partially on how well they compensate musicians. CD Baby purports to offer the most “organic, free-range, grass-fed” music in the world. Lion’s share goes to artists. Magnatune makes similar claims: 50% goes straight to musicians, not to lawyers or labels. Bandcamp makes aggregate claims about how much musicians have made through the site. Artist compensation claims; practice contrast claims; moral positioning claims—each relating to consumer concerns about artist treatment.
Artist compensation: claims about amount or percentage of sales that go to musicians; aggregate revenue claims. Practice contrast claims tend to be more general, comparing to general music industry practices—more revenue than through online retailers. Moral positioning: even more general, and encourage consumers to see buying music through these sites as part of a moral stance towards artist compensation. Do good with your consumption!
This is an example of process advertising, recognized in other areas like fair trade chocolate, cruelty-free makeup. Long and complex production chain for music, like other goods; hard for consumers to verify these claims. Gov’t agencies and private parties may police these claims under federal and state laws.
Artist compensation claims are fact-based, not opinion/puffery. Concrete basis for comparing how various online stores pay musicians. Auditing could glean the truth. Consumers are suspicious about unverifiable claims and want to buy from sources that do treat artists well. Next part of project: there are always issues of whether consumers have cognitive limitations in evaluating process claims. Spotify has started making fair music claims; advertise how much they’ve paid in royalties, but contractually barred from disclosing percentages/allocations.
Eric Goldman: claims about artists, musicians, labels: there is a difference between paying a label and paying the musician. How does that affect claims?
A: still refining that, but one of the problems w/these types of claims is how you do define the participants. (RT: Indeed, that seems to be the problem with Spotify’s advertising—disclosing how much they paid in royalties is misleading about how much artists get.) Maybe need clearer explanations of full chain of payment. CD Baby and Bandcamp tend to be companies that have music aggregator business with lots of artists signed up.
Q: Number claims are great for consumers, and easy for FTC etc. to go after. What about the other two? Are they problematic? They seem associational/tongue in cheek.
A: most actionable, but the other claims are more encouraging consumers to care more about these issues—telling consumers why they’d choose to buy from this seller.
Q: 50% of nothing is nothing—micropayments. Spotify/streaming is almost nonexistent for some independent artists. 50% might oversimplify in a way that is important. Can advertising law reach that?
A: there have been some fair trade certification projects launched in the music space—music venues; standards for what would constitute fair music. Whether it’s misleading is an important question.
RT: (1) Is there reason to think there is a false or misleading advertising problem in this market? (2) What generalizable conclusions? Encourage writing that next part and making it part of this paper. If process claims are available, and relevant to consumers, in many more contexts than previously realized, among other things that has implications for the First Amendment treatment of advertising regulation—compare the claims made in the Nike v. Kasky dissent in the California Supreme Court, or the dissent in the recent DC Circuit country of origin labeling en banc opinion.
A: working through those issues & how preferences develop and how advertising law ought to respond. RIAA’s negative publicity led to boycott/anti-record label reactions. Now that’s died down but there are more legal sites making fairness claims.
Trademark and the Problem of Agency Costs
Similar starting point from Dogan. Consumer protection/producer incentive to develop repeat customers/economic efficiency are traditional TM rationales. The core case on which everyone tends to agree is fraud on the consumer. The consumer is not only confused but also makes a purchasing decision she would not otherwise have made. Confusion is just one of the predicate occurrences required to reach core passing off TM case. But is fraud the right word? Gets at the notion of consumer injury.
Wants to double down on purist approach, not ready to be pragmatist—try another approach to the harm issue. If we really think core TM harm is fraud on the consumer, why do we give the producer the right to sue? Interests must converge somehow, at least coincidentally. Whether that’s historically accurate or not, he might not care whether we have always been at war with Eastasia; the fact of the matter is that it’s the TM owner who litigates. That makes sense b/c individual consumers don’t have enough incentives to sue and producer can aggregate fraud. Makes producer an agent of the consumer. (Class action theory—why do we hate class actions so much but love TM owners? TM owner is seen as having respectable monetary motivations.)
Classic passing off case: not much agency cost; producer and consumer interests converge, b/c consumer and producer are harmed by the same conduct. Agency cost literature: monitoring costs, bonding costs (how do you pay agent to do your business and not its own?), and residual loss (sometimes the agent does its own thing anyway). Not a problem for passing off, but as one gets farther from that, the interests start to diverge.
Rights without fraud, and fraud without rights. Rights without fraud: TM owner wants to sue but it’s difficult to show consumer fraud in first instance. We all know the litany: merchandising rights/sponsorship or affiliation claims etc. Fraud without rights: consumers arguably suffer deception and make different purchase decisions, but producers can’t or don’t bring suit. Lots of examples in licensing—including bankruptcy (e.g., Hostess’s Twinkie TM, which is valuable and thus is used to pay off creditors, and some courts will accept that without caring that a new producer will now make the Twinkies as long as the assignment says the right words about goodwill; security interests). Also concurrent use through settlement; abandonment through disuse. Laches.
Mark McKenna: Agents for which consumers? They aren’t an undifferentiated mass. One difficulty w/some examples is that some consumers may have problems, but solving them would create problems for other consumers. Fraud without rights: one basic challenge to thinking of TM owners on behalf of consumers—vindicating what interest? There may be other consumer confusion not clearly within ambit of TM law. Why not? Why does TM law have a boundary from advertising law?
A: maybe agency analysis would help us sort out the types of consumers. This is sort of an unfair competition paper, about misinformation in the market generally, but theory isn’t limited to consumer deception over marks.
Eric Goldman: This discussion has taken place in the false advertising context. You should decide whether you are mirroring the false advertising discussion or whether there is something special about TM—fraud w/o rights might be an example.
RT: Other entities who specifically act as agents in the sue/not sue decision. Class action comparison might be fruitful—there are a lot of cases about what counts as harm to the class. Also perhaps copyright troll/NPE literature? TM law need not have a boundary from advertising law! The basic principles could be the same! Then registration/ownership would just be their own issues.
Laura Heymann: What does it mean to be an agent in some contexts but not others? Not acting as an agent when a TM owner changes its formula/hires new employees/changes the mark.
Semantic Shift and Trademark Capacities/The Death of De Facto Secondary Meaning
Under the doctrine of TM capacity/de facto secondary meaning, a term deemed generic due to its etymology/history can’t be a mark even if consumers recognize it as such. Part of the theory is that this doesn’t happen often. But literature about semantic shift—new meanings added to words/old meanings lost—suggests otherwise. Prototypical category gets narrowed to single member—some people say it’s the most common semantic shift, but at least it happens all the time. So we need to rethink the idea that it’s not common.
Replace w/primary significance test—if consumers see it as source signifier, treat it as such.
COMPUTER for computers: very hard to imagine a shift happening right now in the modern economy where people see it as signifier for one producer. But Babbage’s “analytical engine” was a predecessor: at one point in time it was the only word we had for the product now known as a computer. It must be the case that ANALYTICAL ENGINE would be protectable for computers now, despite its etymology.
Hotels.com: Even w/evidence of secondary meaning, TTAB said .com didn’t add anything for hotel services. Chocolate fudge soda. Kiss for chocolate candy: etymology was that they were standard for tiny candy. But unpublished PTO opinion said 80% recognition made it a mark. SINGER for sewing machines had TM protection, lost it, then regained it. CODE for sealant? In 16th century, “code” meant sealing wax, but nobody knows it now. Etymologically it shouldn’t be allowed for sealants, except obviously it should.
Built in trust of incumbents in current doctrine—worry they’ll distort consumer protection. TM incapacity looks like functionality, but that’s wrong. Enforcement costs/risks of false positives—given how often this happens, that’s probably wrong. Also, which groups of consumers do we care about? Assumption has been that consumers who understand the term as generic will be in a bind b/c they won’t have access to new entrants. If that’s right, then properly functioning TM significance survey should find them. One pushback: we shouldn’t trust survey evidence. In that case, we’re just living in a world of marketplace dominance. (??)
TM acquisition is simply a form of language change. It’s simply a way of processing how language changes—Apple has meaning for computers when it didn’t for computers before. Need to find new name is important cause of semantic changes. Invent torpedo: need to explain what it is. Borrow name from manta ray! Occasional meanings put forward by any speaker may change into usual meanings, or fail. (“Stop trying to make fetch happen!” Or, for TM: stop trying to make Bing happen! It’s not going to happen.)
Language changes to resolve ambiguity and handle semantic overload. Sometimes by narrowing or changing meaning. Hound used to be any dog, now it’s a species. Wife used to be woman of humble rank, now it’s married woman.
All sorts of semantic shifts where Abercrombie spectrum matches up nicely to what we see in linguistics. Outlier is de facto secondary meaning, and it shouldn’t continue to exist.
If it’s correct that consumers see term as having source significance we’re introducing error into the language when we call it generic.
David Welkowitz: more a process matter, not just distrust of surveys: worry about consumer surveys as proxy for actual knowledge given how TM uses that proxy in general. Troubled by another expansion that says “whatever consumers say they are confused about is what should be,” regardless of a variety of social interests. Asking the consumer is not always the right solution even when you want to know what consumers think. Source of much creeping expansion. Part of it is “how many consumers”? In other areas of TM we allow much less than 50% to be a controlling factor—why would you set a different line here? Isn’t it troubling to let a small number set social policy? A manipulated misperception of the law troubles him.
A: why 50%? In part because of what he sees when he looks at the cases. 50% seems a clearer cut. Equally credible in acquiring secondary meaning in a formerly generic term. Restatement says it’s appropriate only when the old meaning is completely wiped out; restriction also occurs when the dominant meaning coexists with a less dominant meaning. Do we distrust consumers/lawyers enough to discard surveys? More sanguine about that—empirical study about how often a survey pushes the outcome of the case, Shari Diamond & David Franklyn—shows the surveys only matter in close cases. At the extremes it doesn’t.
RT: I’d think that’s why you’d worry about surveys. It’s a close case, and surveys may be having too much influence if you distrust the framing of surveys. I think there’s a tension between saying “linguistics matters a lot” and then turning to litigation surveys, which we know are carefully crafted to produce particular results and often do so, because language matters. Surveys in general aren’t a bad form, but the reason I like linguistic research more than litigation testimony is that, while the empirical work carried out by linguists surely has biases/aims of its own, those aren’t correlated with a desired litigation outcome. So research about how language generally works is more credible, as a rule, than a litigation survey.
A: the project began because he did want to know generalizable rules about language change. Other types of evidence of secondary meaning—advertising—are equally troubling. (RT: Dictionaries/news uses are not in the control of the TM claimant and are less troubling.)
Heymann: corpus analysis—dictionaries, Google.
McKenna: Lisa Oullette’s paper about this.
Trademark Functions and Trademark Rights
Double identity: identical sign/identical goods or services—confusion isn’t directly required, though perhaps it’s presumed. Also, for nonsimilar goods or services when there’s confusion. These are mandatory under TM directive. Also (discretionary in EU, but mostly implemented) use of mark w/a reputation (dilution); use of a mark without due cause that takes unfair advantage of, or is detrimental to, the distinctive character or the repute of the trademark—covers free riding, blurring, tarnishment.
TM owner is entitled to protect uses that may affect any function a TM can fulfill—quality, communication, and advertising, even when function of indicating origin is not affected at all. Double identity rule doesn’t even contain safeguards of antidilution law.
The use must be in the course of trade, in relation to goods/services, for the purpose of distinguishing goods/services (TM use). Directive Art. 5(5) This doesn’t affect protection against uses for purposes other than distinguishing goods or services—this implies that the rest of the Directive focuses on purpose of distinguishing goods or services. So what does that mean?
BMW/Deenik (ECJ 1999): it means signalling origin from a particular source, origin as such. That’s the “essential” function of a TM, to guarantee identity of origin to consumer and distinguish that product from products w/another origin. TM use: is public perceives as use to indicate origin, even if public knows it’s false; or when use is liable to affect the origin function, whether or not consumers are confused; a use liable to affect any TM function, not just the origin function.
Use of the mark to refer to the TM owner’s own goods: BMW/Deenik case was a garage that repaired BMWs. ECJ found TM use: Advertiser used the mark to identify source of goods in respect of which its own services were delivered, and that counted. But Adap Opel/Autec (ECJ 2007) found that Autec’s scale model cars weren’t making TM use of Opel mark for motor vehicles, because Autec was not selling motor vehicles.
That didn’t end the question, though, because in a comparative advertising case the ECJ said using the competitor’s TM to refer to competitor’s goods is TM use because the advertiser intends to identify its own goods/services by comparison to others.
Use likely to affect the origin function: Arsenal case, where people saw football marks as badge of loyalty, not origin. Sort of post-sale confusion narrative, where people might see the goods after they were purchased. Liable to jeopardize the guarantee of origin and interfere with the essential function of the mark.
Other functions of the mark than the essential: quality, communication, investment, advertising—use in factor in sales promotion, or commercial strategy, or use of mark to acquire/maintain reputation and attract consumers. So now there aren’t very many limits on what counts as an actionable use.
Problems: legal uncertainty; protection for functions typically given to marks w/reputation is now granted under the main protections for all marks, interfering w/more balanced approach. EU Commission has proposed new directive that would cover only interfering with origin function, but Parliament voted against it. Fear is that regional exhaustion may be harmed.
Ramsey: what’s the solution? TM use—apparent there’s been creep in that definition.
A: still thinking about that. Even if we accept limitation to origin function, we still need to know what that means.
Dogan: fits with her talk—at least Europeans are honest. What is the more balanced approach of the protection for marks with a reputation? Language is “lack of due cause” but has that concept been developed and due causes elaborated, or is that just a hope?
A: more like a hope. It’s for national courts to decide on specific cases.
The Use and Abuse of Certification Marks
Attempt to get cool kosher restaurant/nightclub. Called Jezebel to reflect “cool” biblical rebel vibe. Want more powerful certification with national recognition, so they talk to OU, one of the two big national/int’l certifiers. OU says it will only certify if (1) they change the name of the restaurant and (2) they remove the racy artwork in the bathrooms. Thinking this important economically, they reluctantly decided to do so. Changed name: to “The J Soho.” Went out of business months later.
Another story, not entirely dissimilar: movie ratings. Allegations for years that the standards are applied in inconsistent ways. Henry: Portrait of a Serial Killer sued the MPAA—didn’t think they deserved the X rating, because R ratings went to similar movies, especially produced by non-independent movie studios. Court thought there might have been truth to that, though it lacked jurisdiction. (Had to release it unrated, which many theaters wouldn’t show; marketplace death sentence.)
These are stories where certifiers with market power can do great harm. Greater worry than w/TMs because certifiers have control over a whole industry. The community for which it matters arguably would care about whether OU certified Jezebel, and that’s a legitimate concern, but the problem is that there are no articulated standards despite the promise of neutrality. The standard is vague and adjustable—then it’s used to keep out competitors. Everything OU says officially is that it certifies based on food; then when this story broke it claimed that certification also covered “ambiance.”
Ramsey: consider literature on GIs, which raises similar concerns. Many certification marks do face competition or possible competition if there’s abuse.
Goldman: combination of certification and market power is the issue. Refusal to deal under antitrust might deal with that. Are there specific antitrust considerations for the treatment of certification marks in particular? Collective marks also provide a comparison.
Q: maybe difference between standards and scope. OU might be reasonably required to identify what it is they certify (just food? Whether you open on the Sabbath?) even if we can’t control their standards. Employment law/public access—maybe we’re particularly concerned if arbitrary actions are likely to be discriminatory. (MPAA’s ratings definitely have gender components—a woman getting sexual pleasure gets you NC-17 very easily.)
Sheff: Is this really different from other TM antitrust concerns?
A: certifier wants to reach into many different spaces—the alternative is 10,000 different certifications for “organic” and consumers have no idea what to believe. Conventional TMs pose less of that issue. (Though if we adopt Linford’s suggestion, maybe …)