WIPIP 2022, Session 2 (TM and a bit of copyright)

Jim Gibson, Against Reputation

In contracts, defamation, TM, conceptions of reputation are misguided. Assumptions about accuracy, reliability and value of reputation are blithe/unsubstantiated. How is reputation constructed in law? Regard that consumers have for a quality of a firm and its products. Heart of the paper: looking at how consumers obtain and process reputation information; how firms cultivate reputation; and how info intermediaries attempt to solve some of the problems/failures of reputation marketplace.

Specific reputation: firsthand impressions. We only tend to remember bad (or good) especially when forming moral judgments; we tend to remember good stuff about competence; we’re not good at remembering experiences in general. Our impressions once formed resist new contradictory information. Not dynamic/responsive!

General reputation: second-hand impressions. Those aren’t actually shared very much. The average Amazon product w/at least 1 review has 2 reviews (median). But if what is shared is a good representative sample, that would be ok—but what is shared is unrepresentative. Bimodal distributions with J or U shaped curves—they only share really good or really bad experiences. But Consumer Reports and other forced reports have a bell curve distribution. We’re also really bad at interpreting aggregated data—people pay attention to total number of negative reviews on eBay and not to the percentage thereof.

Firms don’t just sit back and let consumers form reputations. Traditional advertising can change firsthand consumer impressions, both pre and post experience—telling consumers coffee isn’t bitter makes their impressions of bitterness change, etc. Ad volume has changed though—3000 ads/day which is 100x 1970s (though estimations of that are hard to verify). But repetition works so it’s more effective in changing consumer perception. Third-partying: making it look like info is coming from disinterested third party rather than from the firm. Influencers, fake reviews, etc etc. B/t 1/3 and 1/6 of reviews may be fake online; and even small differences matter—half a star in Yelp is worth 20% increase in reservations. Dark patterns. If it’s cheaper to compete on subjective brand characteristics than on other metrics of quality, then producers will do that. Some say that this is real value if it exists in consumers’ minds, which may be folded into our conception of quality, but if we do that it needs to be a conscious decision about what kind of consumer economy we want—brand premium for subjective branding experience.

Info intermediaries: Certifiers can’t keep up with the volume of goods. They don’t improve reputation so much as marginalize on it—you’re no longer relying on past encounters. What about aggregators?  Suffer from problems of general reputation and targets of fake reviews. Both groups are firms themselves and face same need to cultivate reputations of their own and same pressures. Revenue models almost always end up dependent of firms whose products are reviewed; Consumer Reports is the very rare exception. Aggregators are often purchased by the industries that they are supposed to rate—Rotten Tomatoes is now owned by NBC Universal, GoodReads by Amazon, TripAdvisor by Expedia.

Reputation’s deficiencies are disconnected from the legal story we tell ourselves.

Mark McKenna: Nature of the way that we encounter marks in the environment we live in now. Are there legal changes that could make reputation work, or should we reimagine what law should be like? Could pick one place to say a more realistic account of reputation would mean X.

A: struggling with this.

Jeanne Fromer: often people say that TMs aren’t as necessary any more b/c of aggregators, but the areas of law may matter much more in a system where there are lots of people trying to game reputation. The reach/extent of it is greater b/c Amazon filters what we see in the first place.

RT: Capitalism at large also has effects on individual reputation: Pervasive misapprehensions about sponsorship, like misapprehensions about who sponsors big sports events—consumers are just convinced it’s Nike pretty much no matter what; Beebe et al’s work on dilution that shows that a putatively diluting product can destabilize consumers’ evaluations of all products they encounter soon thereafter. Courts lack ways to appreciate that. Your project could fit into the role of probabalistic judgments in law generally: maybe we just should be honest that we are willing to sanction particular interventions in the info market b/c we think they’re wrong (or that it’s not worth taking the risk of this particular effect on reputation).

Legally: Falwell v. Hustler and its application in lower courts—anti-evasion principle.

Betsy Rosenblatt: This should be a book! Focusing on changes that make law both more and less important.

Glynn Lunney: why does reputation matter? Perfect information would make reputation unnecessary. Why are people hiring fake reviews when consumers pay attention only to number of negative reviews? Is that rational of consumers or of fake review producers? [Gibson points out that some hire fake negative reviews of competitors or even of themselves to try and get negative reviews taken down as inauthentic.] Using defamation’s falsity standard may be help.

Gibson: Defamation is his foil b/c it actually uses falsity as a guardrail to prevent it from constraining most effects on reputation.

Jeanne Fromer, Against Secondary Meaning

Booking.com’s Super Bowl ad leaned in to the literal meaning of the name; Off-White has been going back & forth with PTO over “for walking” for shoes, and PTO said it was descriptive. Off-White is trying to show secondary meaning. Conceptual relationship b/t putative mark and the relevant goods. Does adding quote marks make it “sarcastic” and thus protectable? Vagisil: opposed registration for vagi-formative mark for product for vaginal use. Confusion turned on secondary meaning/strength, and court found that there was likely confusion despite conceptual relation of vagi- to product.

Protecting relatedness of term to product harms competition; narrows consumer attention in similar ways. Might misdirect consumers from other possible products.

This is in significant part due to secondary meaning’s centrality to TM, and other concepts would do better at serving TM’s goals. Right now, it plays a huge role in both protectability and scope. If it’s not perceived as a mark it’s not serving TM’s goals. But courts have developed this further w/Abercrombie spectrum. Secondary meaning is relevant to all Abercrombie categories. Dangers of enshrining secondary meaning:

Ease of establishing it under current doctrine. Expensive to show it by surveys; most show it by circumstantial evidence of advertising and sales. Courts are varyingly strict about whether the ads/sales helped established secondary meaning, but often less strict. Intentional copying by D often leads to presumption of secondary meaning.  Registration + 5 years gives you incontestability, even if it shouldn’t have been deemed inherently distinctive in the first place. Tom Lee et al.’s coauthors: easy to manufacture secondary meaning w/contextual clues in packaging.

Why not just raise the bar? Enormous distributive consequences for businesses w/deep pockets that can afford to run surveys, also show more ads/sales than smaller businesses. TM should not raise barriers to entry based on TM law, so giving big businesses more access to rights in descriptive terms is distributively improper.

But there is a classification problem—marks that have been found fanciful include SEGWAY, PERCHVILLE for a fishing competition, MIRACLE SUIT for slimming bathing suit, TELMEX for Mexican telecom company, JAYCEES for Junior Commerce, and many more. RADIO SHACK found to be arbitrary; K2 for skis which is the second highest mountain in the world; MATERNALLY YOURS for maternity clothing; SEVENTEEN for the teen magazine. Suggestive: CITIBANK, DIAL-A-MATTRESS, POM for pomegranate juice. Courts get this wrong a lot, presuming that secondary meaning exists.

Law underemphasizes coexistence of primary and secondary meaning—it assumes that the primary meaning ceases to exist once secondary meaning is acquired. Proposals to use secondary meaning to protect generic terms forget that there was a primary meaning.

Legit businesses have incentives to develop secondary meaning generally, regardless of what TM law says (except, she says, for businesses that want to copy and businesses that adopt descriptive marks to show up in people’s searches—a legit business strategy). Money spent on most expensive ads is often for arbitrary and fanciful marks.

Rather than trying to solve all these problems w/secondary meaning, we should use primary meaning as gauge for secondary meaning. When the association is too close, don’t allow protection in the first place. VAGISIL is a good example: vagi- means related to the vagina. In the actual decision, 66 other products in the same category started w/ the same letters b/c they didn’t have many practical alternatives [but the court discounted that b/c it called that “generic”], whereas STARBUCKS is conceptually distant from coffee and NIKE’s association with Greek goddesses is also too distant. Replace/collapse Abercrombie spectrum.

Harder to think about what to do about scope. Secondary meaning is connected to likely confusion. Sliding scale of secondary meaning—courts do this anyway without saying this. Zatarain’s Fish Fri=found that small survey was enough to establish secondary meaning, while Tender Vittles from around the same time had Purina spend $4 million in one year and become market leader and still not enough for secondary meaning. We might also draw from reverse confusion, where courts say that they’re willing to accept very little secondary meaning for a term for a successful reverse confusion case. Requirement of correlation to business footprint would require more from the Coca-Colas of the world.

Mark McKenna: you made a powerful argument that it was a mistake for us to have abandoned technical TMs. Secondary meaning was not a condition for TM status in the past, but it started out as a tool for whether you could get more limited unfair competition remedies even though you don’t have a TM. Scope problem goes largely away if you don’t call these TMs. What you’re arguing for is a system we used to have.

A: partly yes and partly no. Abercrombie has outlived its usefulness. Courts are too formalist about it. We could keep building a Ptolemaic model or we can chuck it and start again. A lot of the older cases get

RT: Why isn’t this against inherent distinctiveness? Or against descriptiveness? Is it also against genericity?  Also another group of people who don’t use secondary meaning are Amazon sellers—Grace McLaughlin’s note.

A: Piece is against the Abercrombie spectrum—both too broad and too narrow to say against secondary meaning.

Sari Mazzuco: Abercrombie as heuristic to simplify process of establishing protection. Shifting away from heuristics to different empirical approach would have different set of distributional problems. Roughshod rule of thumb approach has problems too.

A: not saying anyone should have to show secondary meaning—it would be conceptual relatedness that is key. Though it would have distributional effects in terms of choosing a “good” protectable mark in the first place—people with money would have a better chance at that.

Leah Chan Grinvald: Does social media make it easier to establish secondary meaning with virality? Doesn’t descriptive fair use preserve the primary meaning already? Why not just get rid of inherent distinctiveness and require all marks to have acquired distinctiveness?

A: on the one hand, it’s too easy to establish secondary meaning, especially for businesses w/deep pockets. Her worry is that if it’s too easy, there’s a competitive advantage to picking marks w/strong conceptual relatedness; regardless of distributional effects it’s wrong to center secondary meaning. Even if it becomes easy for a small business, they get a persistent competitive advantage that TM shouldn’t be enshrining.

Glynn Lunney: Why have courts made these moves? Don’t make booking.com worse than it already is—it doesn’t really say that generic terms can be protected if they have secondary meaning [it only says that if it has secondary meaning it can’t be generic]

A: Internet = everything is on steroids; old problem w/ new breadth.

McGeveran: this really is Against Abercrombie—reputation and scale of protection is a problem. Big companies with more to spend on marketing get a competitive advantage; is TM law cool with that? Current answer: yes.

A: TM shouldn’t amplify intrinsic advantages.

Justin Koo, Communication to the Public - The US Versus the EU Experience

What approach should we have? If you look like previous infringers, is that enough to make you an infringer/vice versa?

YouTube decision in EU: what lessons from what would have happened in US?

Internationally, it’s exclusive right to communicate to public by wire or wireless, including access at a time/place of individual choosing. Covers both push (broadcast) and pull (on demand) technologies.

Problem: easy to say that someone has uploaded something w/o authorization and is liable. But what about platforms? Who is doing the communication to the public? Bottom line: shift liability, in practice,  away from people committing the infringements to platforms?

US approach: incorporated into existing rights of distribution, public performance, public display. That is internationally compliant but how is it built out in the courts?

Key US cases: Netcom, Perfect 10, Aereo, Giganews, Goldman v. Breitbart, Polyvore. US test: indirect infringement where there is knowledge; volitional test; server test; safe harbors.

EU: direct infringement where there is knowledge of infringing content (Pirate Bay, despite not hosting) or where service provider gives access to works uploaded by users (subject to limits). Court focused on whether YouTube looked like the Pirate Bay—was that an act of communication to the public? Court fudged and said YouTube wasn’t liable b/c they didn’t look & feel like an infringer; they had Content ID and honored notice and takedown and had other worthy uses.  Is that a good enough test? [Sure seems more natural for an indirect infringement analysis.] Perhaps motivated by the fact that the EU safe harbors are a lot more limited than in the US. Pirate Bay was probably indirectly liable in the US. Does it make a difference whether you use direct and indirect liability from a practical perspective?

[sorry, had to run out]