IPSC Breakout Session II: Trademarks, Advertising & Consumers
Relying on Reputation
Reputation: what rational consumers would use to decide what products/services to buy. We shouldn’t expect that info to be widely available to consumers w/o some help, and yet the law assumes that the market for reputation works enough to rely on it.
Public goods problem: each of us has our own reputational experiences w/firms. We can generally rely on it but it’s anecdotal: we want aggregate information. My review externalizes my benefit but without reward, so we’d expect underproduction. Reputational aggregators can help. But their interests may not be aligned with consumers’; no movie has ever had less than 3 stars on Fandango, and Yelp sells its services to businesses rather than to consumers.
IP rights: find a party interested in creating/disseminating the good & give them exclusive control. The firm: They don’t have the externalities problem b/c they internalize benefits of buying. But the firm’s interest is in creating positive/bottom-line maximizing info, not accurate representations of quality, price etc. Consumers are manipulated into believing untrue things simply through repetition. If a firm tells consumers that its coffee isn’t bitter, that affects consumer perception even though consumers know the info comes from self-interested firm.
Courts think that info about firms will solve problems with boilerplate contracts. Reputational constraints as keeping firms from enforcing onerous terms. If reputation doesn’t work that well, reliance is misplaced.
TM law: core infringement dovetails well w/reputation approach, but expansions don’t—dilution by tarnishment; lack of control over assignments/licensing; ways in which mark itself is viewed as having a reputational value qua reputation as opposed to something that provides accurate information.
Defamation: Really about the truth, not about reputation. I can destroy reputation by revealing the truth. Bill Cosby: invested a lot in reputation as America’s dad, which has since been destroyed. Defamation has it right.
RT: Many people enjoy providing reviews/perceive it as a positive. Why isn’t that enough? Do those people who enjoy providing reviews differ in some systematic way from those who don’t?
A: Spite is a powerful motivator; what about good experiences? There’s probably a relatively low level of crowdsourcing that’s necessary; eBay is probably a really good reputational system b/c it’s costless to share. People are more attentive to recent ratings than aggregate ratings. It may not be enough.
Lisa Ramsey: there are fake comments; sometimes there are verified reviews. I also look for number of comments/negative comments. It’s so easy to comment now. [I think this is a matter of leveraging reciprocity norms; Ramsey mentions guilt.] I feel good when I reply to surveys about performance—I’m performing a public service.
A: True, but that feedback often just goes to the firm—helpful, but not fully solving the problem.
Michael Madison: Doesn’t buy the tragedy of the commons framing. There are all sorts of info problems, including overproduction, reciprocity problems (seller worried about giving poor rating out of fear of getting poor rating), asymmetry, information that is out of date (right to be forgotten), etc. Account for what’s going on in the world by adding more of the complexities. Push back on reputational information as a “good.” [Arrow’s information paradox.] As if there were a simple definition of what that means; as if there is a natural social interest in “accurate” reputation as opposed to collection of opinions. Precisely because we have all these mechanisms to share opinion, how much of that is the right amount is a contestable thing. As you frame this question, which is an important one, you need more dimensions to the model to capture some of the looseness and fluidity of what’s going on rather than treating reputation as a species of good that was produced by manufacturing.
Going Native: Can Consumers Recognize Native Advertising? Do They Care?
David Hyman & David Franklyn
A couple of small empirical studies and armchair empiricism; lots of regulatory interest. Lots of argument by invective (John Oliver) and blanket denials and media angst. Regular ads as control; native ads as treatment. If consumers don’t do a good job ID’ing them, can we change the disclosures to improve that? What do labels even mean to consumers?
Online survey of 1000 respondents, who saw 18 images/2 videos. We tested things like Forbes content by Fidelity aka “ForbesBrandVoice” which doesn’t have “paid,” “ad,” or “sponsored.” The Onion also has paid content: Woman Going to Take Quick Break After Filling Out Name, Address on Tax Forms: small “sponsored,” actually for H&R Block. NYT’s Netflix article w/very tiny “paid ad,” and Atlantic’s native ad for Scientology that looks exactly like Atlantic article except for “sponsored content.”
Respondents do much worse identifying native ads than regular ads. Native: 37% thought it was paid content. Regular: 81% ads. 49% thought native ads were unpaid v. 12% for non-native; remainder unsure.
What if we tweak the label? In-your-face “paid ad” bar across the Fidelity ad: those who saw “paid ad” went to 56% understanding from 40%, suggesting that modifications to labeling can make a difference, though still well short of understanding for regular ads.
Finally, asked whether respondents thought each label was paid, unpaid, don’t know. Paid ad/paid content/this content was paid for by/paid post/ad all did better than 80%; sponsored did 79%, sponsored content and sponsored post 76%; brand voice/brand publisher/presented by in the 60s and partnered content/partner was at 57%. Written by: 23%. So many labels are bad and the diversity is not helping anything.
Limitations: representativeness of our ads? Of our sample population? Used reputable organization but replication should be tried. Validity of our approach: we threw out people who went through too quickly or didn’t answer our attention question, but this isn’t in the wild, and we’re open to alternative methodologies if they exist.
Consumers do much worse at identifying native ads than “regular” ads, regardless of brand, platform, architecture, and labeling. If they’re trying to hide the fact that they’re running ads, mission accomplished. Evidence of two-way blurring; even modest tweaks materially improve recognition; does any of this rise to the level of deception?
RT: two-way blurring?
A: if you give people unpaid editorial content, a material number think it’s actually paid, 35%. A little variation across examples. From consumer protection perspective, that’s not something the FTC would much care about.
Barton Beebe: will consumers develop sophistication about this? People have learned about movie sponsorship. [armchair empiricism?]
A: Possible, but the variability is much higher. Movie = single thing that people see together, but native ads have huge variation and narrowcasting. Harder to police given speed of emergence.
Lisa Ramsey: 4% of people thought it was unpaid when it said paid—are there certain kinds of people who are making that mistake?
A: people who participate are over 18, relatively good cross-section of community. Disconcerting that people don’t always recognize “paid ad” as ad, but that’s how people are—they don’t look at labels.
McKenna: consistent w/TM surveys!
Laura Heymann: what would deception mean here?
A: FTC says: you don’t have to prove actual deception; the more you blur the lines, the easier FTC enforcement action would be. We did ask whose content you think this is, advertiser or platform or both—w/native ads, pretty substantial #s thought it represented both the advertiser’s viewpoint and the platform’s, though that’s not the same as deception.
Q: wide variety of examples—they may involve different relations between the platform and the sponsor. Sometimes the content was written by the platform but paid for by someone else v. written by advertiser and posted on the platform. Does that make a difference?
A: A lot of the media companies have set up in-house operations for native ad campaigns to be sold to advertisers. We don’t see a difference based on that issue in people’s perceptions. The development of in-house units probably increases the likelihood of deception b/c they can frame the content to be exactly like the publication they’re working for.
Naming and (Re)Claiming
What makes something a name and how much do audiences assume that a name has connotative and denotative functions? The meaning of a name comes from the public. You can announce your name as whatever you like, but it’s not unless a sufficient number of people adopt it for you, even if you use it. Same for a TM—recognition of a mark as a mark/name is what gives it existence. Richard Craswell’s great article on sports nicknames—crowdsourced. Research on prison nicknames: bestowed upon people, not given to themselves. Inherent in this process is recognition that a lexical unit is functioning as a referent and not as a description.
Referring: repetition and reinforcement. If meaning of a term becomes fixed, is it apt to say that a name has truth value? We don’t expect a man named “Rich” to be wealthy; we recognize that as his name. A “rigid designator”: our beliefs about qualities of subject aren’t relevant to the referential process. It doesn’t matter if someone loses a limb; they’re still the same person with the same name; so too if it turns out they plagiarized a paper. TM embodies that in some way—TM doesn’t require TM owner to surrender TM if it changes ingredients, suppliers, employees who make the product.
If names are rigid designators, we might return to In re Tam and ask whether a name can offend or disparage, qua name. A word can do so when used to describe others, and research on slurs would say that a slur is offensive b/c it is uttered by a person toward another person, not necessarily the word itself. This is Tam’s entire point: the ability to reclaim a word means that whether that word has communicative/truth value depends on the identity of the speaker. One reason offered by court to refuse to approve a name change was b/c of speech burden imposed on others who’d have to use the name in exchanges w/others. This is a strategy employed by Starbucks customers in reverse: give name as “black lives matter” so barista will have to say that out loud.
Slurs convey expressive content: conveying attitude of speaker. The idea of reclaiming is to deprive the word of meaning and transform it into a rigid designator, same as when McDonald’s reclaims MickeyD’s. Are names truth? Pursuing America’s Greatness v. FEC: can independent political committees use candidates’ names in the names of their committees or projects? DC Circuit finds that’s content based discrimination, and it’s not enough to be able to use the candidate’s name in the content of the project/webpage. That’s what she’s interested in—what do we mean when we say that something functions as a name? Are they sometimes true, never true, etc.? Bands named “Cheap Girls” w/no women performing in them: is that deceptive? Is it a signal? Would the Slants be disparaging or not depending on the ethnicity of its members?
Relevant doctrines: When we allow TM to reemerge from genericity: Singer, Best Buy (once it argued in court that the phrase was generic!); public use doctrine allowing TM owner control over a term that the public uses to identify the TM owner but the TM owner doesn’t; abandonment; §2(a) bar on deceptive marks. ORGANIK for clothing not made from organic fibers: if people use that to tell what the clothing is made of, then they’re seeing it as informational device and not as a rigid identifier and therefore it isn’t functioning as a TM should.
Madison: how much do you need to invest in framework of rigid identifiers, partly b/c there’s an enormous literature around that. Tam case may be too rich; what about the dispute over the ownership of “Stephen Colbert,” the “character” from CC’s “The Colbert Report.” Captures the themes you’re talking about w/o triggering disparagement concerns.
A: two different things embodied in the same person!
Q: Whose perspective are we taking? What happens if they intend it as a name, and it’s understood as a name by some people and not by others.
Ramsey: slogans can be identifiers, but they don’t have to be. How do you think about them? “Make America Great Again” is not the same as “Trump.” Also, Lee et al.’s paper on how context is key—you don’t know if a name is a mark until you see how it’s presented.
A: there has to be an initial moment associating the name with the thing; they don’t exist in the abstract.
[Still marshalling thoughts about this. I think the answer is often that a TM can be a rigid identifier and convey lots of information, including w/Best Buy and ORGANIK—in fact, the aim of creating a rigid identifier may often be to convince consumers that certain characteristics are embedded in the product being sold no matter what the facts are, the ingredients are, etc. That’s why TM owners like descriptive marks; that’s why VIAGRA invokes vigor, Niagara, etc. Thus, making a binary may not be hugely helpful.]
Why Does Trademark Law Protect the Strong More Than the Weak?
Barton Beebe & Scott Hemphill
Strength, scope, and competition in TM law: Not a purely original argument—we’ve all been thinking this forever, but people haven’t systematically presented this basic argument except in a Judge Rich opinion in 1988, suppressed by J. Rader very quickly thereafter.
Standard view: there’s a positive relation b/t strength and scope. This is wrong, at least in part. Rationale for standard model: strong marks are more easily activated in consumer’s mind; consumers will infer that junior use originates in senior user; infringers can be assumed to target stronger marks; consumers approach more famous marks with less consumer care—just asserted by Fed. Cir.
Assuming identical marks, a product dimension (x), and a percentage of consumers confused (y), a curve describes consumers likely to be confused by use on competing goods, related goods, or even unrelated goods. Consumer population is diverse. Maximum acceptable percentage of consumer gives us the scope of the right at issue. Increased strength of a mark may push up the curve and broaden the scope of the mark, but the standard model assumes the shape of the curve remains the same as it’s pushed up.
Now assume identical products, a mark dimension (z), and percentage of consumers confused (y), same thing. A 3D distribution will describe the proportion of consumers likely to be confused by junior use of mark identical, similar, or dissimilar to senior mark on identical, similar, or dissimilar goods.
Standard model assumes that increased strength increases likely confusion and increases scope of mark. Other assumptions: The relation between strength and confusion/scope is always positive—except in the limited cases of reverse confusion and, controversially, parody. Strength operates the same way in identical mark, different goods cases as it does in different mark, same good cases. Case law under either situation is interchangeable. Inherent/acquired strength operate in the same way, and cases about inherently distinctive marks can be cited in cases about secondary meaning.
Judge Rich: BVD v. Body Action Design (Fed. Cir. 1988): BVD v. B.A.D. for men’s undergarments. The fame of a mark cuts both ways. Better known it is, the more readily the public becomes aware of even a small difference. BVD would trigger the observer to notice at once that BAD, with or without periods, is a different symbol. See also Jiffy v. Jordan (CCPA 1973) (Rich, J., dissenting). A few other cases have made similar observations. A few articles also question the correctness of the claim, but Jerre Swann defends it.
Kenner Parker Toys v. Rose Art (Fed. Cir. 1992) (Rader, J.) Play-Doh v. FUNDOUGH for toy modeling compounds. TTAB applied BVD and dismissed opposition. Rader says: free riding justifies the positive relation—competitors want to snuggle as close as possible to a famous mark. Fame could never be a liability in TM—while scholars might debate as a factual proposition whether fame heightens or dulls the public’s awareness of variances in marks, the legal proposition is beyond debate. If investors forfeit legal protection by increasing fame, the law would then countenance a disincentive for investments in TM.
But we can solve this problem: begin w/assumption that TM’s guiding purpose is to promote competition; consumer protection is underneath. Free rider justification isn’t persuasive w/r/t continuously positive relation b/t strength and scope/confusion. Identical marks, there is a positive relationship b/t strength and scope/confusion. But for similar marks, competing products, there is a local maximum. The 3D distribution may rise, but also changes shape. Proportion of consumers confused may increase, but scope of protection narrows. Follows a fortiori that superstrong marks would also be less protected for similar marks, similar products and similar marks, dissimilar products.
Heymann: does sponsorship/affiliation confusion complicate this? E.g., Boy Scouts and Peewee Scouts.
Beebe: Best I can do for now is to fall back on the idea of TM as both prescriptive and descriptive; can prescribe a condition in which consumers learn. It may be that the association challenge is devastating for some things, but similar marks/similar products and similar marks/dissimilar products might be different b/c of consumer understandings about famous brands. If you saw something like the Apple logo on a very low-end tech product, you might reject the idea that it came from Apple.
McKenna: This isn’t really an empirical rule, but modern doctrine’s obsession w/confusion requires it to pretend that it is. Your argument at the end seemed nonempirical. You need to clarify: is this empirical? If so: Look at brand extension literature. Also: when you talk about similarity, note that people have hard time w/specificity of visual marks but may do better on word marks. [Though there is also a huge difference b/t production and recognition worth thinking about, since it’s recognition that TM law is generally concerned w/.] Then if it’s empirical, think hard about different forms of confusion.
TM provides important incentivizing functions that make it more like © or patent than conventionally assumed, and we should care about that. Traditional story: TM are instrument of commerce, protected under Commerce Clause, not incentive for creation/innovation. TM as red-headed stepchild of IP family. Three traditional functions: source identification, advertising, and quality guarantee. Neglected incentivizing functions: creation of original marks; hedonic benefits of loving TMs; socially desirable behavior [work hard so you can afford your Louboutins, or at least your Nikes]. [This is like claiming that securities law exists to encourage people to work hard and make more money. It has a level of generality problem, among others. Also query whether all incentives to work hard and make more money, if those are the same thing, are socially desirable; see, e.g., drug money, Wall Street finance, Donald Trump.]
Expansion of TM law over time beyond source confusion to affiliation confusion, dilution, initial interest confusion, post-sale confusion.
TM as creations: often the most valuable asset of a business, e.g., Starbucks. [Deven Desai would say this is conflating a TM with a brand.] Companies invest $ in infusing brands w/a personality: SoulCycle isn’t just a name, it’s a lifestyle, also Harley Davidson. [Though note that these lifestyles are usually shared w/ other brands as well—they’re not creative/unique even if it takes intelligence, luck, and even creativity to come to “symbolize” that lifestyle. Or you could just be like PBR and be adopted by hipsters for a time.] Takes on life of its own. TMs can be included in art exhibits.
TMs as hedonic goods: experience of goods = happiness, just as via © or patent. TMs are inherent parts of products; there is no defensible distinction b/t “real” and “artificial” experiences of marks. [Though the research on body image etc. suggests that people who spend a lot of time thinking about their presentation to the world end up unhappier and also poorer off in other ways, so I’m not sure that conclusion holds here even if it might re: your perception of a car’s value.]
TMs as socially desirable behavior incentives: People use branded goods as self-rewards; I can have Starbucks b/c I did such a good job today. Possible correlation in US b/t working longer hours and desire for brand consumption. Use of branded goods to overcome bias and achieve community status—anti-aristocratic b/c you can become more equal by wearing certain things or appearing a certain way. Wearing a nice suit to interviews. [A branded suit? Tressie McMillan Cottom has a great essay about this, but TMs don’t seem to be required.]
These incentivizing functions have increased in importance over time. This brings TM closer to copyright and patent, and needs to be considered in making policy.
Gibson: what implications are there for TM scope? If we had a law that focused on source confusion and not dilution etc., would people not find alternative hedonic pleasures/ways to reward themselves for working hard?
A: given the state of the empirical data, it would be speculative to say. Doesn’t know whether we need dilution to get those things. This is the law we have. Not necessarily enforced strongly, but might still provide a deterrent. There are also 1A implications w/dilution. But we can’t divorce TM law from other social policy as others have done, where we pretend that only confusion is at issue.
Madison: Timing of the shift in case law in relation to this?
Manta: thinking about how we could measure brand development—going from TM as boring, traditional function to powerful. Measuring strength of marketing, penetration of advertising.
McKenna: circularity: you say we have to account for how people interact socially w/brands, but that’s the result at least in part of how the law has treated brands. Even if we wanted to preserve hedonic value, consumers would likely find it somewhere else, and if we don’t know that, the policy payoff is unclear. Also, there are other people who get hedonic value from access to brands, or access to similar brands; if we do a whole hedonic calculus we have to consider them too. If we want to consider incentives to work hard, TM would not be the place to start—tax, education policy would be more appropriate.
Manta: yes, utilitarianism is hard, but we have to start somewhere. This matters too. Hedonic value of access to brands, knockoffs—but there are also people who don’t care much about brands. [?] In this room, we probably chose our jobs b/c we cared less about money than making partner, but others chose differently. Some criticisms of dilution have disregarded these incentives, and have assumed that brands are just about false needs.
Ramsey: even if the empirical research shows that people are doing what you say, should we care? The costs of dilution protection are also higher—protecting these values has different tradeoffs than protecting against confusion over source.
Manta: there haven’t been economic discussions about dilution’s harm/benefits. It’s all highly theoretical, speculative. I don’t know which way it cuts, but that doesn’t mean it’s not a real effect. Skeptical that dilution has had more than a marginal impact on speech.