AU TM roundtable, part two

Interesting how much we talked about territoriality in various ways here.

Paper 5: Greg Lastowka, “Nominative Fair Use Still Doesn't Make Sense”

He’s taught nominative fair use a bunch of times, and doesn’t enjoy it because it doesn’t make sense.  It’s gained new prominence because of its inclusion in the TDRA; many have a favorable attitude towards it as a counterpart to classic fair use.  To be clear, fair use is a good thing; being able to refer to others’ marks in expressive use is a good thing. The problem is that nominative fair use hasn’t done well in serving that purpose. In fact it seems to have been counterproductive, eclipsing prior common-law developments that provided those shelters and did so in a way that effectively precluded those doctrines from showing themselves—not for a skilled lawyer, but for a more average defendant’s lawyers.  Grand Theft Auto: court eventually gets around to a First Amendment defense; but in other cases court sees only NFU as a possible defense and rules for the P because that doesn’t apply.

Fair use carves out breathing room as an exception to the scope of TM.  NFU makes no sense as a substitute for likely confusion, since that’s about the scope of the right and not a defense.  (I don’t think this is right—it’s about a normative statement that no actionable confusion is likely as a matter of law.  I agree that it’s weird to talk this way about likely confusion, but that’s the fault of the wacky concept of likely confusion we have now, especially since it’s materiality-free.)

There were lots of pre-New Kids cases finding for the defendant but they weren’t called NFU.  The question was whether they were fair uses carving out space—privileged despite confusion—or situations where there was no confusion and that’s why the use was fair.  Counter-history: defenses might have developed differently and better for defendants in the absence of this overarching concept.

Summit Media: Bella’s jacket case.  Defendant should’ve won that one.  And (I think this is what he thinks, but I’m extrapolating) NFU distorted that by making it possible to find D took too much of the mark (which was in a jacket that D had itself created!).

McGeveran: the idea is, show us that you should be on the NFU track and we’ll give the D a different, meetable burden—not a burden to disprove confusion.  True, some 9^th Circuit courts won’t comply with how the test should work, but that’s kind of true of the 9^th Circuit generally.  NFU is flawed but better than the alternative of not having it.

Lastowka: the alternative universe to New Kids could have a more robust First Amendment defense, as the district court in New Kids initially did.  Post-Tabari, it’s the plaintiff’s burden to show the absence of the three factors once the D makes the initial showing.  If the P shows D used too much, does it automatically win on likely confusion?  Tabari does seem to suggest that.

McGeveran: though in actuality the judge is going to consider whether confusion is likely.

Lastowka: Grand Theft Auto case rejects NFU, then goes to Rogers because they’re not the same mark.

McGeveran: there’s a difference between identifying problems with nominative fair use, with which he agrees, and saying that a world without it would be better for defendants. Not yet buying that second part.

Lastowka: does nominative use eat up the territory of referential uses?  Grand Theft Auto case says it doesn’t because Rogers remains available, but then how much sense does it make to have NFU in the first place?  (Expression v. traditional advertising may have something to do with the differences here—NFU is problematic perhaps because it spans both types of uses, whereas Rogers is for expression that isn’t standard advertising.)

Ramsey: maybe separate out comparative advertising, advertising for spare parts/used goods/accessories, etc. and have narrower doctrines.  Or go even bigger and have a trademark use requirement.  Can the things covered by NFU be dealt with in other ways? 

Lastowka: the best way to define scope of TM is to look for material confusion as to source.  He’s with creating fair uses that allow material confusion as to source when First Amendment values or competition dictates that leeway is required.  As an umbrella doctrine, though, the third step of NFU can be read as a restatement of the likely confusion requirement.  Why do we need a requirement of “need” to refer to mark?

Discussion—“need” really means “is it a referential use, given that you’ve chosen to talk about the P”?  (With the exception of the terribly reasoned Hyundai/Louis Vuitton case—though this does make Lastowka’s point that the way the test is framed invites error.  That argues more for restating the test than getting rid of it, though.)

McGeveran: Rogers and NFU are, today, pretty similar in how they work. 

Lastowka: but as a substitute for likely confusion, NFU seems dangerous.  If it has interpretations and requirements that aren’t consistent with likely confusion, and the D flunks the NFU test, then it seems to be liable even without the standard likely confusion analysis.

McGeveran: higher burden on P, though, and cheaper to litigate.

Farley: why is it a higher burden on P?  Tushnet says NFU is a way to eliminate survey evidence.  But the second and third prongs are openings for the P to suggest that there’s confusion.  Tushnet and McGeveran say we don’t have to think about what the audience thought, but we want to know what the audience thinks.  (I think that’s actually not true—we do care about D’s intent a lot; we also create rules that allow us to skip evidence about a particular mark or use because we think the average presumptive result will be more accurate, as in Wal-Mart.)

Carroll: Kozinski is a common law lawyer—he likes these tests.  He created NFU in the context of mushrooming of factors in TM generally, as with likely confusion; then the legislature gets into the act with the TDRA.  Stripped to the basics, what is the legal conclusion that’s relevant?  What types of evidence are relevant?  These factors point to types of evidence. But it’s not enough to identify those types; we have to think about how litigation actually happens and burdens Ps & Ds. How do we sort sheep and goats and get rid of some cases early?  There’s a rigidity to the NFU factors that lacks the necessary suppleness—trial courts tend to be mechanical because they don’t want to get reversed, especially not for failing to consider factor 8 of a 10-factor test.  Functionally: we should figure out the most effective switches to turn liability off.

Lastowka: what Kozinski was trying to do was create breathing space but when he spells out the factors he does create this rigidity.  To a trial court judge, factor three looks like a likely confusion test.  It doesn’t sound like a broad test.

Lunney: could do the Rogers move and say “needs to be explicit,” instead of factor three. 

Lastowka: Rogers is explicit that we tolerate some risk of confusion to protect speech.  (And for comparative advertising etc., then, we tolerate some risk of confusion to protect competition.)

Farley: maybe an interesting comparison to what Kozinski did with reverse confusion: he explicitly eliminated certain factors from the multifactor test; Tushnet says NFU is a way of doing the same thing (specifically, survey evidence/evidence of actual confusion) for other types of uses but maybe it needs to be done more explicitly.

McGeveran: your thesis seems to be: a parallel universe without NFU would be better.

Lastowka: District court in New Kids used the First Amendment.  Wouldn’t that have been good?

McGeveran: Rogers is getting better, but was in hibernation at the time.  So it’s a tricky thing to make these counterfactual arguments.  Not seeing why this particular innovation made things worse.

Lastowka: Jardine: we turn it into a substitute likely confusion test and put the burden of disproving confusion on defendant.  That leads directly to KP Permanent.

Lunney: that burden-on-defendant rule already applied in the 9^th Circuit for descriptive fair use—preexisted New Kids.

RT: why is comparative advertising okay?  Why is advertising used goods ok?  Is it because it’s not confusing (an empirical claim) or because it’s important?  What nominative fair use might have precluded was a real account of that (though Louboutin suggests courts aren’t good at giving an account of where TM stops anyway; they just rule that it does stop).  Here’s an example of a use of nominative use that deals ambiguously with confusion.

Paper 6: Glynn Lunney, “Trademarks and Consumer Welfare: The Unrepresented Interest”

Even if courts generally get it right, what goes on outside the courtroom is more significant—big settlements are problems.  Congress basically got the Lanham Act right, but over the years courts progressively made it worse.  Every once in a while Congress intervened, often to recognize how wrong courts had gotten it—trade dress is an example (“symbol” was put in the Act in 1943 to exclude trade dress, but now it’s been read in the reverse fashion).  He’s never seen a trade dress case where consumers are better off with a monopoly on the design; word marks are different.  1962 housekeeping amendments: courts said it meant that any confusion at all was actionable, but that wasn’t true—it was an attempt to standardize four different statements of likely confusion previously in the statute, and to encompass potential consumers, and the “housekeeping” characterization made clear that it wasn’t supposed to be a radical expansion. But that didn’t stop courts from construing it as such.

Cycle of trade dress protection: every 30 years the SCt has to come along and stop lower courts. Kellogg; Sears/Compco; then the Court’s own Two Pesos mistake has to be corrected by Wal-Mart.  What’s going on?

Law and economics scholars argue that judges are trying to adopt efficient rules. Many judges don’t think they’re doing that. Even if they were, do they really have the right information given the subset of activity they see?  Only a few persist to litigation, and it’s a biased sample. Even if the rule is efficient for that sample, it won’t be efficient for the broader population.  Nonmotivational law & economics approach: if you get an efficient rule, parties will settle and stop suing because individual interests are always limited.  So only if there’s a larger efficiency problem will one party have an incentive to stick it out rather than settle.  Difficulty here is that if only one party has a continuing stake in the legal rule, the rule will eventually favor that party almost inevitably.  Will spend more on litigation: e.g., Louis Vuitton.  D only cares about this one case and wants to get out inexpensively.  Looking at TM, you see these things.  One bad decision: Boston Pro Hockey—and we’re stuck with it; Ds drop out/let it stand.

Good decisions like Pagliero get distinguished/challenged until they tend to become relatively less important, even though the interests on the other side (I want exclusive rights in a pretty design) are not strong.  May be an information issue: SCt tends to do a better job, and that may be because you tend to get relatively good, information-rich amicus briefs.  Opportunity to bring forth the consumer interest.  Not sure what to do about the problem. Consumers are too dispersed to benefit from direct consumer standing.

My comments: Some Ds are repeat players: P&G actually has a much greater interest in a balanced TM law than New Kids on the Block does.  Big portfolio = much less likely to like state law dilution, for example. (Fame is fine.)  Possible theory: There has also been a corrosive spillover from right of publicity in particular, where Ps aren’t ever likely to be Ds? 

Paper says: “In any given case, a trademark owner usually has more to gain by prevailing than a group of would-be defendants who want to enter a market that the trademark owner is seeking to control.”  Perhaps consider expressive uses as a potential area where Ds are developing an understanding that they have to fight just as hard?  This may even fit into the paper’s idea that Ds are less likely to fight when the gains will be dissipated by competition—a D with a copyright has something to protect, e.g., a video game. Google is separate kind of repeat player, also expressive, also with its own IP to protect.

Solutions: as always, consider false advertising law for consumers.  Tort law: make TM owners responsible for torts of licensee.  You might stop getting some claims to control markets.

Lunney: there definitely are repeat Ds—Google, Wal-Mart—one reason Wal-Mart came out the right way was that this was Wal-Mart’s business model and they wanted to maintain the ability to copy.  Might have had stronger interest in pursuing the litigation than Samara Bros.

McGeveran: Expressive uses do involve categories of Ds have noticed that they need to fight.  Cases that have reached decision almost always often involve big entertainment companies (or Ds represented by Public Citizen).  Doesn’t share Lunney’s belief that trade dress was a wrong turn.  Even if you’re less pessimistic about TM than Lunney, the paper has something to say, but might want to consider how to speak to that audience.

Laura Heymann: Does information-sharing matter?  People post C&Ds online—do they fight harder?  Do they educate each other?  Transaction costs of litigation—reputation etc. may also matter.

Lunney: some clients see PR as a benefit and not a cost.  Suing may be bad for your reputation or good; being sued may be good for your reputation.

Calboli: UK jurisprudence—lawyers can be disciplined for bringing bad cases.  Even though TM is considered property, you don’t see the nutty cases we see in the US; lawyers are more conservative.  Confusion standard in the US, but dysfunctional litigation/favors deep pockets.  Commercial defamation as a way to respond to bad claims?  Some way to get into the plaintiff’s pocket, or the attorney’s pocket.

Lunney: Hard to say should we adopt UK system wholesale, but the contrast may help prove that private interests do influence legal development; hard to imagine that economics works differently though different incentives may be present.

Heymann: consider insurance coverage—another party with a long-term interest.

Lunney: We don’t have a general design right because auto insurers said that it would increase the cost of auto repairs and went to Congress to block it.  Most insurers today are cost-plus businesses, so may not have much to say about the TM rule.  (I note that most will deny coverage for claims in this field.)

McGeveran: is issue preclusion in the model?  If it’s a case where the result in case #1 will be preclusive against P in future, P’s stake will be even higher.

Lunney: often room to argue. Losing Chewy Vuiton doesn’t deter Louis Vuitton from arguing that the next case is infringing, and perhaps gives it more incentive to do so to limit the rule of Chewy Vuiton.

McGeveran: so are there ways to harness doctrine to be more preclusive in ways that put more P skin in the game?

Bartholomew: areas of TM that “force” policing of the mark—did TM owners actually like this?

Lunney: Yes, TM owners were in favor—they only wanted abandonment in the bill, not genericity. They wanted the rule to be that if there was policing there could be no genericity; policing was in their control, but consumer understanding isn’t always.

Creativity is efficiency: when producers can experiment with “Who Dat” merchandise they offer lots of alternatives and that’s good for consumers.

Farley: how do cases stick (become the general rule) or not stick (disappear from general legal understanding of the rule)?

Lunney: the model isn’t clear on that.

Farley: seems like the theory is that defendants just drop off.

Lunney: hasn’t seen a well-funded D fight Boston Hockey in years.

Silbey: a lot of the good cases are cases where someone has engaged in expressive/nonintentional conduct and just decides to stand by their guns—e.g., Eat More Kale.  Cost/benefit calculations don’t come into it—these are not people who are thinking in rational actor terms.  To think about reform, have to think about these Ds too.  Classes of actors are much more coalesced in Veblen goods cases.  Boston Hockey is easy to adjust your model to if you’re in the sports apparel industry; you don’t have to like it but you can go about your business; you don’t care about consumer welfare writ large.  Other rules won’t stick because people aren’t actually looking at the case law—not worrying about trade dress, not worrying about expressive uses. Example: Bridgewater Candle—the copier was not thinking it was against the rules to copy the general business model and sales tactics.

Carroll: different information gaps/flows.  What info is the TM system putting out to participants about their ex ante decisions?  How do you get info to judges if you’re using the common law?  Perhaps some admin agency that can do factfinding.  Bring back the Office of Technology Assessment!  In TM, would that be in the FTC?  It wouldn’t necessarily need to participate in every case as consumer advocate but could make factual findings about general consumer understandings.

Lunney: note that once you have Boston Hockey, unsuccessful licensees become unsympathetic defendants.

Paper 7: Mary Wong, “New gTLDs” (with Jacqueline Lipton)

ICANN: we thought it was worrisome to have panelists all over the world making decisions, not just under the UDRP but the new Uniform Rapid Suspension System for slam-dunk cases and a totally new different procedure, Legal Rights Objection Process. We don’t know how it will work or whether it will govern future new gTLDs.  Whatever problems the UDRP has had, mostly the tweaks have been procedural, and the substance has been left alone.  Question: are courts a better place?  There are problems with legislation and treaties; so what are we going to do?

We found, looking through the research, is that UDRP studies were done pretty early.  Up until about 2003, there were tons of articles and hair-tearing.  That’s gone down a lot.  ICANN wants to review everything from UDRP to new procedures at once; she thinks that’s a monumentally difficult task especially given the political forces/inadequacies of consumer representation.

RT: why are we spending so much time and energy on new gTLDs?

Wong: It’s not really clear. The argument has changed over time.  The initial argument was worry over loss of control over .brand.

Grinvald: isn’t it a moneymaking endeavor?

Wong: that’s not the formal reason. Not just about competition and choice, but expanding the universe of available names; unpacks Google monopoly on search (but there’s an equally good argument that it increases Google’s monopoly).

Heymann: think about how this intersects with Qs of search. If the goal was to open the internet, could just create gTLDs that don’t communicate: .1234 for example.  Is it really contemplated that I should guess “ford.complaint” when I’m looking for a complaint site about Ford? 

McGeveran: who would do that?

Carroll: almost half the students use the location bar not Google.

McGeveran: but for search when you don’t know where you’re going?

Carroll: too early to tell.  Consumer attention remains the scarce resource online. ICANN’s original argument was tech infrastructure limited gTLDs; the geeks thought that was silly and they won.  From the brand owner’s perspective, new territory opens up in the race for attention.

Wong: they say they have to go deal with the new gold rush.  Now it’s clear that the gTLDs themselves aren’t a big problem—they’re so expensive that no squatter is going to register .louisvuitton.  Now they’re worried about the second-level domains: policing, enforcing, monitoring.  ICANN board was convinced that was a major issue.

Carroll/Ramsey agree: even using Google, people may use domain name to check relevance in the results. 

Carroll: new gTLDs could diminish Google’s power, but if they’re confusing then Google becomes more powerful.

Heymann: but if Wal-Mart gets walmart.complaint, then the signaling function of the domain goes away.

Wong: UDRP review will be relatively easy because it’s a contained universe. Big issue: should ICANN care?  Segregating some words into okay and not okay—dictating not just who should have the words but what should go into the sites.  ICANN has tried to avoid becoming a regulator of what speech is ok.

Ramsey: could .brand function as a certification?

Lastowka: how could it be different than cartier.com/reseller?

Silbey: might have different semiotic meaning.  Could also change depending on what happens.

Grinvald: need to think more about procedural justice: do panels have precedent, do they follow particular laws, do we need to worry about high complainant win rate?  Is this mirrored in other international initiatives—ACTA, TPP, shifting away from established judicial processes to administration.  Louis Vuitton’s GC: very excited about using the ITC to stop conduct they don’t like.  That is a deliberate, scary strategy; very confident about ability to use private forums to better advantage than public courts.  Cheaper, quicker, less chance for due process.  So ICANN developments reflect this larger movement.

Calboli: there are different approaches to free speech internationally that matter--.gay, .islam, .church.

Wong: Saudi Arabia has objected; objections can be not just to the gTLD but to the people running it.  This is an exception to territoriality, and the move to private systems can simultaneously contribute to fragmentation even as it internationalizes.

There will always be a demand for more, greater rights. There will never be enough. ICANN is the target now because TM owners don’t want to spend money on defensive registrations.

Lastowka: seems like a replay of original domain name system—fear of initial anarchy that has to be sorted out.  But does it make sense from a consumer confusion standpoint?  People can learn on the internet what domain names mean or don’t mean.  But what about the governance issues?  Are there cases about “are you .edu enough to get an .edu domain”?  Who can curate?

Wong: historical accidents.  ICANN doesn’t govern .edu; Educause does and regulates pretty strictly.  Others, like .aero, .mobi are not so successful, but over time some have successfully negotiated with ICANN to open the scope of their operations.  New gTLDs: answer depends on registry agreement with ICANN. We expect many are going to try to renegotiate the standard agreement; see a lot of variation depending on the nature of the gTLD.  Wong would rather have chaos than ICANN deciding what kinds of users are okay.  But .com/.org/.net have lost much of their signalling function.

Carroll: allow creation of company towns (unless ICANN puts rules in place).  (I don’t see how that’s different from giving me a gmail.com address.)

Wong: ICANN was never supposed to be a regulator, though, so it’s weird that it’s now making these decisions.

Ramsey: .xxx was accused of extortion—you could pay to prevent anyone else from registering a string.  Could change the fee for do-not-register.  ICANN should retain the ability to influence what the registrars are doing, otherwise they can run the company town and get rid of anyone they don’t like—getting rid of all the gay porn on .xxx.

Carroll: businesses could be lured in and then trapped.

RT: but they already are on Facebook, Amazon, Silbey adds eBay.

Carroll: this makes the potential worse.  (Hmm. Not sure I’m convinced.)

McGeveran: if the justification for new gTLDs is to open up new spaces, the fact that the hypothesized behavior is basically the same as what’s already happening on Facebook etc. is troubling.  Query: how important will it be for an entity to get entrée into the .islam space in order to be considered credible?

Farley: well, these companies are spending a lot of money to make it important.  Companies are looking around and thinking that they have to get in now if they’re going to get in at all, because the rules might change.

McGeveran: those are people with existing presences who want to maintain exclusivity; then there is the question of the tenants in those spaces; quite different sets of interests.

Wong: is it worse that there will now be book.amazon controlled by Amazon, which Amazon can allocate to sellers or to itself, or is there something different if Amazon owns .book?  Similarly, Google owning .google v. .search.

McGeveran: it’s all about whether the tenants control the landlord.

Ramsey: we want to know what laws apply, since this is operating in a space that is both private and international.

Discussion of expressive value of these designations: we just don’t know!  Regulating with an eye toward the future is hard.

Paper 8: Christine Haight Farley, “The Forgotten Trademark Treaty: The Pan-American Convention”

It’s a treaty in force, being applied in member states by US corporations, but we don’t know about it here. It’s not as if there are so many TM treaties with substantive rights that this one just got lost. It’s a 1929 treaty, and its first successful use was in 2000.

Why so ignored?  History/timing of this treaty and what’s happened in the US with treaty implementation. At the time, there was an understanding that a specific set of rights set out in a treaty would be self-executing. But the default has shifted against self-execution. 

Why you should care: because it is self-executing.  The SCt has said so (though possibly in dicta).  Art. 7 and 8 doesn’t really need legislative implementation; drafters’ intent seems to have been self-execution.  The Lanham Act used to name the Pan-American Convention but was deleted with housekeeping amendments (to genericize it to treaties). 

Rights are also very interesting.  There are sections on GIs and unfair competition, which are different from other int’l agreements.  But she’s focused on Art. 7 & 8 priority rights, which are radical.  Enough to be a mark, not well-known, and if someone in another country is aware of the mark they’re precluded from using the mark in a country in which you haven’t done anything or advertised.  One TTAB case used a constructive knowledge standard. It’s not a bad faith standard.  Pardo’s case: guy had an address 20 blocks from a Pardo’s restaurant in Lima, and that was enough (though in that case the script was also very similar).

Also the standard was not confusion but interference—like double identity.

Probably never anticipated Latin American businesses taking advantage of their rights in the US; rather the idea was to attract US businesses to Latin America. Over the years, US corporations have made successful claims.  History is more complicated than just US oppression, though.  GIs, and our agreement to them, are interesting example of that.  Complete tearing down of territoriality principles in the Americas.  The international registration, with a bureau in Havana, which is now defunct, was supposed to do that.  Art. 7 & 8 don’t require a registration instrument, but still attack territoriality/use.

Lunney: Tea Rose/Rexall case, before fed TM law, said that under the law of Kentucky if there was knowledge of Mass. that the Ky. user couldn’t be a good faith user.  Similarly looking to an outside jurisdiction to foreclose rights inside the territory.  Why isn’t that the same as the international approach here?  He doesn’t think that the Persons case follows inevitably from territoriality.

Silbey: but was there evidence of impending expansion? 

Farley: doesn’t think the case holds that if he had knowledge then he wouldn’t have been able to get a valid mark. 

Discussion of whether it would matter if there was secondary meaning in the jurisdiction of the junior user versus knowledge of secondary meaning in the distant jurisdiction.  I have always taught the Rexall/Tea Rose case as one about distance, with good faith not an element—if there’s no secondary meaning in your jurisdiction, then it’s perfectly fine to use a good idea for a mark you’ve seen elsewhere.  Lunney teaches that there is also a good faith/lack of knowledge requirement.

Farley: this convention is a change from “the sovereign decides how rights are acquired in its jurisdiction.”  Makes rights without use/consumer knowledge possible.

Heymann: Art. 7 & 8 also do say “upon compliance with requirements established by domestic legislation” and “in accordance with the legal procedure of the country.”  So how does that play out.

Farley: thinks this is pure procedure, but will investigate.

Ramsey: suggest what it would take to meet our obligations, if this is yet another treaty we’re not complying with, e.g., the Berne Convention.

Farley: will try to show difference in rights in careful comparison between this and Lanham Act.

Ramsey: can you protect GIs through TM under this?  We may have said we were already doing enough.

Farley: Brazil in particular had strong negotiators; they were ratifying earlier versions of this convention that the US didn’t ratify. US comes down with a memo that is not discussed, but is passed, and then the US joins. Trying to figure out what was going there.  But there was interest from Latin American countries, perhaps for creating a trading union—perhaps because of the ongoing Paris Convention.

Silbey: students often reject territoriality, perhaps because of the internet. The paper poses a challenge—maybe good faith doesn’t matter any more.  Maybe it’s not such a radical extraterritorial right if brands are likely global.  (Seems to me that we have to decide what we want to do for the vast majority of brands—the long tail, almost none of which are famous—versus the few that are at the “fat head.”)