WIPIP part 2, TM doctrine (and false advertising)

Christine Haight Farley, American University Washington College of Law

Sleeping Treaty: The Pan-American Trademark Convention

TTAB’s 2000 Belmont case: British-American Tobacco v.  Philip Morris—cancelled an incontestable mark on the ground that it violated the Pan-American Convention, which is self-executing. That was a huge deal, and yet few people noticed or reacted to it.  1929 regional convention dealing with TMs and unfair competition, 10 countries ultimately ratified it.  Provision was that owner of a mark protected in one state who knows of use in another state shall have the right to oppose use or registration on proof that the registrant had knowledge of the existence and continuous use of the mark in any contracting states. Huge exception to territoriality!

Since then, a few other cases, in which constructive knowledge satisfied the Board. Held by SCt to be self-executing in a 1940 case.  In a timeline of self-execution, comes after Missouri v. Holland, before backlash to self-execution after WWII; period during which we were comfortable with self-execution. Two circuit court cases assert that claimants have to use section 44 of the Lanham Act as a vehicle to assert rights.  §44 previously specifically named the Pan-American Convention as covered; now the language is more general. Lanham Act drafter: he was concerned about Latin American friends pointing the finger at us as hypocrites—want to see statutory evidence of compliance.  Unfair competition: clear self-executing language for states without legislation.

Chapter on geographic indications: clear and broad—ever indication that doesn’t actually correspond to the place of manufacture (etc.) shall be considered deceptive.

Ladas, 1929: because we had the TMs, protection of foreign marks was really protection of US nationals. Foreigners registered very few marks of their own.  US was looking for market for its goods.

Initially Latin American conventions were meant to be substitutes for Paris Convention, but direction changed after WWI.  1923 Convention, still in force between Brazil and US, has similar provision.

Be careful what you wish for: US never really expected Belmont case.  Recent Free Trade Agreements—we haven’t seen anything like these GI or unfair competition provisions, nor have we seen them in the TPP (or even references to existing rights!). But these rights are still being asserted by US companies in the know.

David Welkowitz: discussion is predicated on the self-executing nature of the treaty.  Is there anything that prevents Congress, after a self-executing treaty is signed, from deciding that it’s not self-executing/saying our statute is all-encompassing?

A: would need a strong reason to think there was a conflict with the Convention.  There’s no evidence of that, and §44 supports a complementarity reading.  More disuse than repudiation. Nothing about GIs in §44, where there’s another set of rights not covered at all by §44.

RT: Now that I’ve read the original Lanham Act legislative history, I’m really interested in this rhetoric that we were or might be creating an independent federal unfair competition cause of action through the Lanham Act by referring to our international obligations.  There was much debate about that, and it now no longer makes any sense to modern TM lawyers because we’ve grown up assuming that of course it wasn’t like that.  Similar to what the original understanding of incontestability and defenses thereto might have been, now forgotten because confusion has eaten everything.  A history of forgetting?  Compare to histories of discovering causes of action?

Compare to discourse in copyright policymaking about how our TRIPS etc. obligations mean we have to accept that we already have a right of making available/can’t change copyright laws in any ways that favor users?

Irine Calboli: The US wants others to follow certain rules.  US thereby locks itself in, preventing differences in our own laws because of US lobbies/USTR captured.  Businesses dictate the laws—claims that we’re bound by int’l law are actually about policy laundering, because we did it to ourselves.

Deborah Gerhardt, UNC School of Law, Kevin McGuire (UNC), & Mark McKenna, Notre Dame

An Empirical Study of False Advertising Claims Under the Lanham Act (presented by Gerhardt)

Not much empirical work has been done.  What’s going on with prudential standing?  What’s going on with different media when challenged?  Other attributes of ad that may affect result—if it’s comparative.  What about the type of falsity? 

Looked at opinions issued after 1988 revision became effective.  Looked for standing issue or substantive false advertising question, about 1380 opinions (ultimate set will be smaller). Started by coding cases that just deal with 1 ad or 1 analysis on group of statements or ads. Bigger challenge: coding for multiple ad cases.  849 single-statement analyses.

Unsurpisingly, 2d Circuit is biggest, a lot in 9^th, 3d Circuit also prominent.  Number of decisions increased over time.  Natural persons not a big driver.  Spike in 2006 (why? Looking for theories).

Plaintiffs win 15% of the time over all decisions (not all in dataset are final; win means win on some sort of dispositive procedural posture, like D win on motion to dismiss; P win on motion to dismiss doesn’t count as a win).  1990, P win rate was 40%, but dropped sharply and now hover 10-15% for past decade.  40% are decided at motion to dismiss stage.  Need to see how that compares to other IP contexts.  Only place where Ps win more than Ds is permanent injunction stage.

Ps have standing most frequently when both parties’ products are mentioned; mentioning either also increases standing finding, but not as much.  But mention of P’s product doesn’t increase overall wins very much.

Text by itself, image by itself doesn’t seem to have different win rate, but with both text and image there seems to be a higher P win rate, up to 25%.

Many cases, courts don’t even consider falsity. Only about 42% do. When they do, they find falsity 40% of the time.  Longer litigation = increased chance of P win.

P win rate higher when statements found true but misleading versus literally false—though this may change with additional, multiple statement data.

Conclusions (preliminary): increasing, at greater rate since 2005. Prudential standing at issue in 18% of cases.  At trial, Ps win 48-50% of time.

My Qs: is there any interaction between literal falsity and media type? My confusion: Did you code for arguing literal and misleading in the alternative, and which if any the judge accepted or rejected? The judge can really change the course of the case by adopting one theory over the other.

Lemley: one possibility with spike in 2005 is that Westlaw has become more inclusive—could look for some similar cause of action and see if you see a similar spike.

A: we took a look at the Administrative Office data on filings, and the spike wasn’t nearly as high, but of course that’s not directly comparable.

Felix Wu: one story consistent with data is that parties are only willing to litigate to SJ motions.  Either the D wins or SJ is denied, but P might be getting really good settlement—Ps might be doing better than 15% because of unwillingness to go to trial.

A: no matter how you count it, the importance of the motion to dismiss stage is huge.

McKenna: plausible, but P success rate isn’t that high after SJ.  Vast majority of cases fall out before that.  If you have a 50% win rate for going to trial, D isn’t necessarily that scared, except of the expense of continuing the case.

Lunney: Qs of selection bias—low win rate suggests suit is for some other reason than prevailing in litigation.  In terms of literally false v. misleading: that is counterintuitive, but clients who can’t afford the litigation pursue literal falsity claim but can’t afford a consumer survey—so this may indicate difference in Ps. 

[RT: my thoughts here: this is why the falsity by necessary implication finding is very interesting—relatively high win rate.  Falsity by necessary implication is a tool judges can use to move a case from “misleading” to “false” and help Ps avoid a survey. It is very much a matter of judicial choice.]

A: can check that.

Eric Goldman: Could have spun finding as Ds win 50% of time. Sounds less interesting.  Middle ground of undecideds—what are those?  Are these really P wins that will settle?  Measuring P wins in court is useful but incomplete because of other battlegrounds/D changing ad and mooting the case. 

A: Ps win 24% of the time when there’s a dispositive winner.

Q: re internet ads as affecting spike?

Michael Grynberg, DePaul University College of Law

Thick Marks Thin Marks

Mark might be strong but receive weak protection based on nature of claim, such as nominative fair use.  Courts adjust thickness of protection to resolve cases involving nontraditional source meanings of TMs. Urge for some courts to give thick protection stems from difficulty understanding overlap in meanings TM might embody.  Tracks some accounts of property rights as solution to info management problem, but also explains why property can never really get the job done.

Courts calibrating thickness: Example—Maker’s Mark case from 6^th Circuit, sued Cuervo for use of red wax seal on tequila.  There was a ton of distinguishing context making confusion extremely unlikely—different spirits; Cuervo brand was premium; distinguishing labels; no evidence of confusion. Court waved that away, speculating consumers might perceive some connection—less obvious harm becomes easier to prove than traditional harm.  Louboutin: 2d Circuit finds mark protected, but not infringed.  Disposed of case without giving us clean doctrine.

Impulse to give thick protection is driven by fact that courts have difficulty isolating meanings a TM might embody.  Automotive Gold: makers of car accessories using TMs as complementary good.  Court rejects functionality defense because it would be death knell for TM—consumer demand is difficulty to “quarantine” from reputation function of marks.

We might see this move as a propertization of TMs in response to info management problem, as per Henry Smith, who tries to justify strong rights to exclude as part of a modular system.  Tries to apply that to © and patent.  In modular system, intense interactions between modules are hidden and communications between modules are limited.  In a car, the brake system operates independently of car—can deal with brakes without having to deal with entire car.  If you are in a parking lot and you see a car, you don’t need to know much about ownership status other than that right to exclude means you shouldn’t mess with it; owner can attend to more complex Qs of efficient use.

This might explain some of what judges are thinking, since judges can’t sort out TMs meaning—what does association really mean? How seriously do we take the dilution story? Judges mask those interactions behind the property signal.  They’re applying source confusion language, but the moves are unsatisfying.

The other side of modularity is that there are times it doesn’t work—stakes are too high; other interests implicated; interaction between property right and potential users needs to be more nuanced. This is the governance model. TM is necessarily marinated in this because standard of delineation of property right is effect on consumers—3d parties—and their interests are supposed to be paramount, along with competitors’ interests. Info management is simultaneously appealing and impossible, leading to push-pull between thick and thin protection.

Example: keyword cases.  9^th Circuit began with ridiculously thick protections.  Changed because courts got wise. But they’ll always oscillate.

Calboli: why not say thin protection/thick protection, not thin mark/thick mark?  Sees no competitive problem with Automotive Gold.

A: Maker’s Mark and Automotive Gold are both approval cases.  Could say latter is a lot closer to actual sponsorship, but the 9^th Circuit is equally broad in its language.  His point is what the 9^th Circuit does in the opinion, which is to say that it can’t separate the 2 functions (communicative and branding) of the mark.

Farley: would robust rule of aesthetic functionality deal with many of your problems?

Wu: is your notion of thick or thin derived from the cases/describing the cases, or prior to the cases? 

Linford: judges seem uncomfortable with forfeitures (aesthetic functionality)—though that creates a chicken/egg issue with the concept of property.

Zahr Said, University of Washington School of Law

Interpretive Complexity in Copyright and Trademark: Comparing Substantial Similarity and Likelihood of Confusion

Judges in copyright cases routinely make decisions about where to locate their interpretive authority—a judge has to decide whether interpretive authority rests primarily in the text, in the author, in the audience/reception/critic/expert, or in judicial intuition. These are interpretive pressure points, and they’re a feature of copyright rather than a bug.  But the choices aren’t transparent, though they’re necessary.  Copyright realism: judges avoid choice, or behave as if they’re compelled to decide one way.  Usually outcome-driven analysis.  Conclusion: judges might have too much discretion over whether and how to make interpretive decisions, creating confusion in copyright law.

Substantial similarity analysis: interpretive pressure points consistently unacknowledged, but matter a great deal to outcomes.  Broken in several places. 

Compare to TM.  Most TM scholars love to hate likely confusion analysis. As with ©, TM requires substantive comparison of two things—P’s prior existing thing and D’s allegedly infringing thing. Must compare in order to allocate rights in those things.  Comparison must find interpretive authority—how do we know that consumers might be meaningfully confused?  Actual confusion (audience); risk of confusion high because owner says it (author); surveys/experts.  TM asks how we know what we know about these marks.

From ©’s perspective, even this broken test seems like an improvement.  Different elements of the Polaroid test appeal to audience (strength of mark, actual confusion); text (similarity of marks); authorial intention (D’s intent); experts (actual confusion, buyer sophistication); judicial intuition (quality).  Though they aren’t consistently applied, judges at least have to say how they know what they know. 

Empirical scholarship suggests that these factors often don’t control. But the existence of the factors allow us to diagnose what courts are doing—departures from methodology are more visible because of the steps the court is taking that it isn’t. We have a clearer version of the “ought.”

Lemley: everyone thinks the discipline they’re not in has got to be more organized, and he wonders if this is going on here. Transparency seems a good thing, but would you also get a substantive alteration in the way we make decisions and would that be good?  Likely confusion makes a particular choice about audience (individual consumer). You could make that choice in copyright, but a test that focused on some of the things likely confusion does would get different, and probably broader, results. Maybe what you want is some properly designed multifactor test, but the real work is in designing the test.  Does the analogy to confusion end up biasing the test towards results that we won’t like (market substitution rather than technical substitution—competition as harm).

McKenna: others criticize courts for rigidly marching through factors. The factors are not the “ought.” The factors aren’t just used to judge similarity, but some result. In ©, similarity is itself infringement. That’s why people criticize the factors—in certain kinds of cases the factors don’t have much to do with the result.  What’s the ultimate Q we want to answer?  (See: why nominative fair use was developed.)

A: does feel it would be better for judges to have less discretion.

McKenna: then TM is not your example.

A: rigidity is also transparency, though.  Transparency without inflexibility?

Lydia Pallas Loren: substantial similarity is about “at what point have you copied too much?”  If you’re asking if this is illicit, we have very different biases about what’s illicit, incredibly context-dependent. Likes idea of transparency.

David Welkowitz, Whittier Law School

Fault Lines in Default Judgments

Defendant’s profits are available as measure of damages for willful infringement (requirement differs by circuit); TM counterfeiting, if willful—statutory damages multiply by a factor of 10.  Indications that under §35(a), while it’s not supposed to be penalty, willful infringement increases the likelihood of enhanced damages. Willfulness therefore plays an important role; people who don’t have resources to contest suit may not fully realize what they’re doing when they default. Because damages aren’t a sum certain, default judgment means damages must be proved up. But the evidence accepted is often lacking (e.g., willful because P sent a C&D and D didn’t stop). Sometimes courts find that default means that we accept the well-pleaded allegations as true, and the allegations are that D was willful.  Some play in that, but plausibility doesn’t seem to be a big barrier.

Alternative: accept allegations as true to the extent of showing liability, not damages. TM claims don’t necessarily require willfulness (cybersquatting aside), and intent is of course relevant to courts’ assessments. But realistically, willfulness isn’t required. In most cases, should be part of damages, not readily accepted on complaint’s allegations.  Found one court that realized this, but found independent evidence of willfulness.

Many courts say that the fact of default makes you a willful infringer (also said in copyright cases).  Why?  Courts just think: good people don’t default.  Questionable. Also: bankruptcy says that willful & malicious debts are not dischargeable in bankruptcy.  Usually a default judgment isn’t issue preclusive, but some states don’t follow that rule.

McKenna: encourage you to think more broadly about willfulness—not many litigated cases find bad faith in our false advertising set, so it might not be a huge number. Does it show differences?

Loren: is there wiggle room after default and even if willfulness is pled? 

A: what happens in TM cases is that if the complaint pleads willfulness, the tendency of courts is to accept that after default.  He thinks this is a bad result.

Linford: you have an intuition that we’re catching lots of people we shouldn’t be catching—aren’t these folks likely to be bad actors?

A: maybe.

Calboli: no need to prove willfulness in copyright/patent for big damages—many Ps are international corporations and expect that treatment. Property concept.

A: TM cases dealing with statutory damages tend to cite copyright cases.

Said: Pam Samuelson & Tara Wheaton on statutory damages in copyright—mere knowledge of infringement constitutes malice.

RT: many of these defaults are from immigrants.  Sociological literature on counterfeit sellers in the US: many come from places where legal system is corrupt; may perceive police crackdowns on counterfeits as ordinary shakedowns; moral force of default is different, at least in the absence of some educational attempt.

Farley: define what we mean by willfulness: a big deal. Does default plus willfulness findings come up in other areas?

A: willfulness comes up a lot in TM, not sure about bigger picture.