Friday, February 19, 2021

WIPIP SESSION 7.A. — Trademarks

Jeanne Fromer & Barton Beebe, NYU School of Law

The Future of Trademark Depletion in a Global, Multilingual Economy: Evidence and Lessons from the European Union

TMs transcend boundaries b/c brands transcend boundaries; even small businesses are often looking beyond borders to other countries. EU is good for study b/c it has a system made of 27 countries with 24 language, 15% of global economy, 450 million consumers. Can help us understand how rights assertion is working in a global, multilingual system.

Spoiler: things are bad. EU TM runs in parallel w/national system. All or nothing: you need to show entitlement throughout EU; if a term is generic in one country, then no EU TM is allowed. EUIPO doesn’t engage in relative grounds review: doesn’t look for confusing similarity; 3d party oppositions are the only check. When assessing confusing similarity in an opposition, one pronounced difference is doctrine of translational similarity (known in US as doctrine of foreign equivalents): mark can be confusingly similar if it translates; it’s common in the EU for people to speak 2 different languages.

Registration can occur w/o use; no examination for use; 5 year grace period for use. TM clutter is widespread.

Registration rate is really high, not surprising given lack of relative examination.

Selection: not more of a pool to choose from in a global market, but less of one, because if a mark means something bad in one place it won’t work for a global branding strategy. Irish Mist ran into this in Germany, where “mist” means manure. Also, mark may be taken in a jurisdiction, which happened to TJ Maxx in the UK, which then had to switch to TK Maxx/Milky Ways varying by jurisdiction. And there’s a limited set of words that are very same/similar across multiple languages b/c of cognates, loan words, onomotopoeia. English is the dominant language in registrations; it’s the language of business.

Studied EU using a bunch of databases; found incredible levels of depletion, worse than in the US despite fewer marks being registered in EU. Most common words are claimed unless they mean something negative, and when we translated registrations into the 5 major EU languages the level of depletion got worse. Likewise, for close similarity, the congestion is worse than in the US. Similar results for pronounceable syllables. Marks are getting longer over time b/c more is taken.

German: it’s only 40% depleted in common terms, but when you account for translational similarity it goes up to 80%. Most multilanguage terms are taken (REPUBLICAN is the one that’s intelligible in all 5 languages that’s not taken; strong meaning makes it unattractive).

Crowding has increased over time across classes. Oppositions are going down over time. Very few companies do most of the oppositions. Compared to 2(d) refusals, just really low. Oppositions that do proceed tend to succeed b/c there is so much crowding. As opposers increase, chance of success increases, until inflection point where crowding is so great that confusion w/any given mark becomes unlikely.

What to do? (1) More enforcement of use requirements to get marks off the register, whether up front or examination after a grace period. (2) Ex officio review. EU does a bad job of putting existing registrants on notice; there are a lot of unrecorded settlements so you can’t tell what’s really going on. (3) Translational similarity should be revisited. We can train consumers; consumers would eventually learn to distinguish translations. (4) Fee structure: should charge more for more valuable words.

Lisa Ramsey: Do EU lawyers care about clutter? In INTA they don’t seem to.

Barton Beebe: There’s no interest at all in increasing use requirements/shrinking use to 3 years by major EU lawyers b/c they tend to represent the big brands. But maybe they don’t realize how bad things have gotten for SMEs especially.

Fromer: more recognition academically; EU classes are broader/cover more things than US registrations.

RT: But can you really train consumers?

Fromer: not convinced that the consumers are translating now—need evidence on that in order to defend the doctrine.

Beebe: consumers could probably be taught about clearly different languages. Translational similarity may be a special case where consumers react differently [v. my example of the UNC and University of Wisconsin cases in the US where the PTO said that consumers had stubbornly continued to perceive the marks as indicating a single source despite over 100 years, in the former case, of uncontrolled use].

Irine Calboli: Single market: freedom of goods means that there’s no way to stop confusingly labeled goods at the border if you’re giving a European market. So the consideration goes beyond TM law/translational similarity.

Barton: Spain gave a national registration to Matratzen for mattresses, even though it’s generic in Germany.

Alex Roberts: Say more about charging more for more valuable marks. How do we do that/does that discriminate against smaller entities?

Fromer: any changes can’t be prospective only or new entrants will be harmed. We have to recognize we’re not starting from scratch. One possibility: whether at initial stage or at renewal, charge more based on some metrics of value such as being multilingual.

Irene Calboli: Interaction with failure to function?

Beebe: The future of the TM system is crowding. The registering agency/system overall in EU has given up on the problem of crowding. So what? They’ve heard the argument: Maybe it’s good to have barriers to entry b/c TMs are just artificial product differentiation. This argument failed in the US/law and econ and might fail in the EU too.

Jake Linford, Florida State University College of Law

An Information Theory of Bad Faith Trademark Use

Intent is crucial to many P wins, but many scholars criticize this as irrelevant to consumers who can’t see the bad faith adoption. Sheff says: maybe both infringement and dilution are proxies for when we think sellers are manipulating consumers’ bounded rationality. Intent may be irrelevant still, he suggests, but Linford says that may not be right. Intent could help us rightsize info to make sure sellers aren’t overly opportunistic. We should care about bad faith. Second entrant should have to account for why it ended up with a mark that is highly similar to an existing mark. Information theory: noise or interference that makes it harder for info to be received as intended.

What counts as bad faith? In some circuits, failure to do a search; becoming closer to P’s packaging over time; intentionally dragging out litigation. Although good faith doesn’t preclude a confusion finding, it should probably work for the D like evidence of bad faith works for P to negate the ratcheting effect. More leeway for good faith behaviors may make the market more efficient for consumers. Conducting a search/seeking advice of counsel should be counted in favor.

Mark Lemley: to do this we need a very clear definition that is significantly narrower than existing definitions. Right now intentional reference to the TM is often considered bad faith. Also: should ask same Qs of P’s good faith: was there evidence of confusion or were you suing from fear of competition?

Betsy Rosenblatt: Model jury instructions are terrible, as are standard proposals. Bad faith is bad, but even lack of bad faith might also be bad; says nothing about good faith! Why do we want balance? The argument is: people who are trying to confuse might be better at it, but she doesn’t think there’s any evidence that’s true.

Carys Craig: Intent may be discounted in the legal articulation, but bad faith may be driving more of the results than we think, neglecting actual confusion considerations. Good faith parties may end up winning anyway for doctrinal reasons.

Roberts: doctrinal narrative that producer is really powerful. Thus if trying to fool, likely to be successful. Thus even if we’re punishing bad faith it coincides with this narrative of producers being likely to succeed at what they do.

A: Henry Smith’s idea of equity as an overarching attempt to police against opportunism.

RT: was going to talk about Smith’s idea of equity but I think it might be why your idea struggles against the multifactor test. In Smith’s idea, bad faith/fraudlike behavior is a precondition to entering into the equitable realm, not part of a multifactor test. If we had double identity + unfair competition, and opportunism of some sort was a precondition for entering the unfair competition realm, then that could work, but Smith thinks that multifactor tests like the one in TM are evidence of an unproductive collapse of law and equity that just creates a big muddle.  

Beebe: bad faith seems to overwhelm a multifactor test cognitively; it’s a bludgeon. Circularity is an issue too, but its presence seems to blow out any other factors.

Calboli: Bad faith internationally/in the history?

Lisa Ramsey: if the underlying subject matter is itself attractive (inherently valuable speech), then the TM owner/claimant is the one who’s free riding.

Brad Biddle, Arizona State University College of Law (with Jorge Contreras & Vigdis Bronder)

Certification (and) Marks – Understanding Usage and Practices Among Standards Organizations

Some mismatch in definition of “certification” marks—some organizations deliberately avoid that term. TM registration much more common than certification mark registration in this space. Many SSOs don’t apply for TMs at all. Conventional wisdom doesn’t match real world practice. For reformers: focusing on certification mark rules won’t matter if most certification happens in TMs. Lawyer-driven? Feeling of greater freedom under TM rules?

Jessica Kiser, Gonzaga University School of Law

The Reasonably Prudent Consumer of Alcohol

Sees a trend of more refusals in alcohol/wine/beer. Some dumb ones; even as PTO recites rule that there’s no per se overlap between beer and wine. A couple of theories: is PTO more likely to find confusion between beer and wine than between two similar wine marks b/c it thinks wine consumers are more sophisticated? One possibility: this is a problem here b/c it is a really crowded space—a beer company might release ten new names a year, even if they aren’t in use two years later. Alcohol has been found related to cigars because people consume them together; Harlequin romance has a wine line so now publishing may be related? Does wine, in fact, go with everything?

Alexandra Roberts, University of New Hampshire Franklin Pierce School of Law

Mark Talk: Making Secondary Meaning

McCarthy: Trier of fact, like lexicographer of modern slang, must attempt to find out the meaning to the consuming public. Yet almost all the evidence they consider is indirect, proxies for understanding.  The assumption is that if the producer puts in the work, it will happen. These principles have been around since even before the Lanham Act—length of use, advertising spend. Yet a lot has changed in those 75 years—internet destabilized active producer/passive consumer binary. Can look in spaces like Reddit to see how consumers are actually using a term: “boy brow” turns out to refer almost always to Glossier makeup product, while “no-poo” refers to general practices of not using shampoo/products that people do use instead and not to the registered mark for a particular non-foaming cleanser. Celine used mark talk on the internet to show secondary meaning for its robot face bag.

Is this a one way ratchet? 6th Circuit, in case about DJ Logic, said that FB fans, likes, followers, celebrity followers could have substituted for sales evidence (but it wasn’t). Astroturfing is a risk. Thinking about how to formalize this.

Eric Goldman: relatedly, weaponization is a risk: get people to talk about a competitor’s mark as generic. Google results may be harder to game than the direct evidence.

Lisa Ramsey: Might also shed light on failure to function!

Linford: project to identify fame: large corpora and data analysis? That’s what you’re trying to do too. WIPO often considers internet results to be sufficient evidence in domain name disputes; maybe that could be a model.

Lemley: this will be more complicated for genericness b/c a number of uses are likely to be ambiguous. Knows of existing ML project: build a context engine to parse the use of terms that are possibly being used as a mark v. generically.

Calboli: maybe many of these shouldn’t be registered anyway—generating new rights that might not be worth protecting. Especially with trade dress.

A: maybe should focus more on word marks. [Hmm, I think it might actually be more useful for trade dress, but that might depend on your priors about how likely it is that people really recognize various supposed trade dresses.] But wants to make space for niche uses: can be a well-known mark among consumers of particular product.

RT: preservation through transformation: if we do this because it gives us more direct evidence than we could easily acquire before, secondary meaning will become a different thing than it is now. That is neither good nor bad in itself. Will also accelerate quantification (isn’t it kind of wild that 75 years after the Lanham Act we have no idea whether secondary meaning has to be the dominant or most common meaning because we so rarely have direct evidence?). Compare to what happened with rise of control arms in surveys and effect on what percentage is sufficient to infringe—unrecognized by courts, they increased the standard by looking at confusion net of control; also what happened in 1-800 v. where clickthrough rates substituted for a survey.

A: Ideally it gets closer to what we have said it is all along, tied to the goals of TM law.


Jane Shay Wald said...

I enjoyed everyone's thoughts. However, I found this startling: "What counts as bad faith? In some circuits, failure to do a search;" In what circuits has failure to do a search resulted in a finding of bad faith? What cases?

Jane Shay Wald said...

PS to my last comment about bad faith from failure to search -- to clarify -- in the last 20 or so years....seems like a concept that found some favor in 2 cases in 2d and one in the 7th in the '90's but didn't catch on.