Friday, February 28, 2014

Trademark Scholars Roundtable, part 2

Session 2: The Product Market Dimension

Robert Burrell: why do we treat territorial and product markets so differently? In the UK there was never any Q that nationwide protection was going to be the result of registration.  Always understood to be an advantage of registration, and an incentive to register. But you were initially confined very strictly to the goods named in the registration. It took a long time to chip away at that and have, for example, a defensive registration process for similar goods.

Part of this may have to do with the technology of registration, it was always understood you’d have to carve out some way of distinguishing what goods a registrant would get protection for and what goods they wouldn’t: never thought sensible to protect across all classes.

When this technology gets transported through the rest of the empire, it seems to have been accepted that you’d have the same nationwide scope, even though it’s far less obvious that Australia should have nationwide registration—there wasn’t really a single market between Sydney and Perth.  A nation-building exercise: the federal gov’t building a single market.  That’s one of the things the federal gov’t was empowered to do, bound up w/sense of nationhood beyond a simple free trade area.

Even in the UK, we quite soon had not quite a Dawn Donut rule but something similar, with a honest concurrent use defense, which provided both a defense and a way of getting onto the register. First person to register got nationwide protection; concurrent user was frozen into place.  We still see the effects of that today. A-mart, eventually sued by K-mart, which now only exists in Queensland.  Frozen there. We don’t rethink those in relation to confusion; just isolated doctrines. Nationwide protection was always illusory to a degree: courts push back on the idea of one registration being for the entire territory. But we’ve started from the opposite end with products, and think about defenses differently—broad confusion standards, but not much thought about the circumstances justifying exceptions. Why do we have honest concurrent use for remote user but not honest concurrent use for user on distinct enough product?

Bob Bone: Change in way of thinking about goodwill, no longer just product-specific. From brand to firm, and then goodwill becomes redefinable based on circumstances, not just protecting consumers.

Consumer protection orientation can be subtly changed.  How do we decide how closely related two products are?  It’s not intrinsic to the products—we ask about consumer perception. But in that case the factor is circular: the factor depends on likely confusion, but likely confusion is the question; we must be using some other principle to decide what risk level we’re tolerating. Consider a field of likely confusion—it gets higher in the center and weaker as we move out, where “out” is about different dimensions of difference such as similarity of marks, etc. Where do we draw the line?  There’s nothing intrinsically different about product and geographic boundaries, except normatively. So we may draw tighter limits on one than the other, but that’s because of reasons related to the type of dimension it is. Geography = sovereignty plays a role.

Move from the consumer perspective where we’re talking about likely confusion, to producer perspective. Different set of arguments about broader product markets. If brand is identity, then it’s odd to ask producer to change its name in order to move to a new market. Lots of problems with that position, but it introduces a new element.  So how do we want to structure product markets? How will these choices influence firm structures? Consumer perspective is all about confusion and whether it’s geographic or product differences is secondary, but not from producer perspective.

A note on dilution: used to think tarnishment was understandable and blurring not.  Still doesn’t think blurring makes sense normatively, or that its empirical assumptions are true. But at least it’s in TM language: selling power.

Tarnishment isn’t so simple: consumers don’t make a connection, so it’s not harming the reputation of the firm.  If you think of a strip club, you don’t think less of Tiffany’s the jeweler.  Under this theory (whose empirics, again, he doesn’t believe). Tarnishment is about product diversity.  When I buy Tiffany jewelry, Tiffany is part of what I’m buying.  Advertising: the advertiser can make the mark more/different.  When Tiffany is used on a strip club, the strip club can destroy that control, and so Tiffany will not be selling the same product “Tiffany jewelry.”  Prevents negative information from affecting the meaning of the mark. If we’d like to have regular jewelry and Tiffany jewelry, then barring tarnishment makes sense (if a bunch of Tiffany strip clubs would otherwise come into existence). If this is true, tarnishment isn’t a part of TM law; it’s doing something else.

Bently: on the assumption of nationwide scope—the mechanics of having a TM office would have contributed to that.

Red Bull decision envisions concurrent use in the field of dilution. Red Bull sued a Dutch company that had run a bar called Bulldog since 1975.  Moved from just serving drinks, for which it had a registration, to serving an energy drink.  Was this due cause?  Court seemed to think it could be: natural expansion of products from one market to another—whether it was natural to extend from bar into energy drinks. 

McKenna: Bone’s story makes tarnishment and blurring sound the same: both are about affecting meaning of the mark.

We do have a way of thinking about how owners move through geography, but not a good way of thinking about how they move through product market space. How do you think about superior rights, given that both producers may have started moving?

Bone: doesn’t know what it means for products to move closer together in the product space.  Is pancake flour closer to syrup than sugar? 

McKenna: marketing literature has some answers there.

Bone: still normative.

McKenna: normative to say “too close,” but descriptive to say there are degrees of closeness.

Bone: in a sense blurring and tarnishment are similar. What gives the mark its attraction is making the product more appealing.  But you can also think of it as a particular type of selling device.  Psychological advertising: made a different product from the unbranded version. Can’t make that separation for tarnishment as he can for blurring.

Dinwoodie: in the product area we have a standard which allows us to meander, whereas in territory area we have a blanket rule: you have America.  Once you have a rule, you need exceptions.

McKenna: one other way to think about the difference—TM is build on connecting a mark to particular goods, which is the only way to assess distinctiveness.  We relax that at the infringement stage.  When courts say an inherently distinctive mark is strong, they apply that to other goods/services where it may not have the same meaning.  Geography isn’t as foundational to the TM system; more politically up for grabs.

Dinwoodie: that’s a common law way of thinking: registration lets you get whole classes.  You’re still measuring distinctiveness in theory against the set of goods/services, but practically you get a lot more.

Sheff: Metaphor of confusion manifold is elegant and interesting but probably not descriptively very helpful. If would be nice if each dimension was a vector that could be measured in terms of distance from origin (P’s product), but that’s not what’s going on. It’s not that product proximity is asking whether consumers are likely to be confused due to similarity b/t products, but that confusion is so empty itself—the factors are doing other things. 

See familiar mark in unfamiliar category: something goes on in consumer’s head. If consumer resolves discrepancy, the process of resolving that discrepancy feels good. If they can’t, the cognitive dissonance makes them feel bad. But what does that have to do with confusion? Probably not much. Whatever these dimensions are measuring, we hope they’re measuring effect on consumers, and Q is whether that effect has normative weight.  We might ask whether there’s negative feedback on the prior mark from discrepancy, though the marketing literature suggests there won’t be.

Dilution: who is allowed to participate in creation of mark’s meaning?  It can’t be the case that only the first user is entitled to participate in that.  Who else?  We have no global answer, but sui generis rules.  No sex/no drugs = tarnishment; what about user-generated content/consumer participation?  There we’re moving to allowing that, but what about uses by other firms?  We don’t have clear guidelines.

Burrell: discussing blurring is like atheists arguing the meaning of the Bible: none of us believes in it.  Tarnishment can be conceptualized as an alien invader, or as part of blurring, but you have to know what blurring is and we don’t.

RT: Bone’s description sounds like the province of patent; usually we expect competition in product markets to do the best job of provisioning them.

Borders/distances don’t actually shrink, absent invasion or devolution.  But what bars serve can change, perhaps in a different way—now that energy drinks are popular mixers with alcohol, competing in the market for bars may require selling them.  Compare Apple Computer’s move into music while music was becoming digital, something that wasn’t as salient, so it looked like Apple Computer was moving into music while really they were both moving. Possible tools for measuring closeness of product categories: Lakoff/psychological literature on how people group things categorically.  Legal precedent: rules of accession? Usually used for other purposes, e.g. people have argued with domain names that accession can provide rules of allocation to TM owners, but could it have relevance for product markets too?

McGeveran: some constraints are totally independent of confusion, like sovereignty—a true defense.  (Or at least it used to be.)

Bone: wants us to focus on harm, not on aspects like strength of the mark or similarity.  True that confusion is a grab bag of concepts, but using harm at the center of the infringement manifold has some beneficial effects. 

With dilution: he thinks it’s part of sellers deciding what they’re selling.  With tarnishment, the Q is whether they get to decide the quality or character of the good itself, or whether consumers will be allowed to participate—do we democratize what the goods are?  Or can the producer control? (But will they come when you do call for them?)  Of course the producer can’t control—the public does define this quality.  But it’s sort of like a market for lemons: the acts of particular private individuals can create an externality where people who like Tiffany jewelry may be saddened by the rise of the Tiffany strip club, if the subconscious effects story is true.  You buy this story or you don’t. But if you do, it seems that’s how you have to justify it. (OK, I’ve written a lot about how the empirics are wrong, and Bone does seem to agree, but it’s worth reiterating.)  True, as soon as you say “Tiffany jewelry” is a different product than “jewelry,” we are now fighting over the definition of the relevant market.

Burrell: even if you buy it, you still have to ask when tarnishment might be present.  EU: there might be tarnishment between whiskey and dog food, and he’s not sure that’s different from sex/drugs.

Litman: Does anyone buy this story as something that actually happens?

RT: No! Related point: if you say that making other people feel sad is an externality, you walk into an important issue in utilitarianism, which is whether we should ever count the sadness that some people feel while observing the behavior of others as a harm in a utilitarian calculus.  John Stuart Mill says no, and I’ve said the same. It turns out that if Tiffany stumbles, Cartier is willing to start providing the same benefits, and consumers even seem to enjoy the churn.

Grynberg: we don’t believe in dilution; we don’t think confusion is coherently configured. But we still have a dilution law and a confusion law, committing us to something. So maybe we should turn away from consumers and towards these other policies, like helping businesses.

McKenna: we should stop pretending that dilution is a thing. Attracted to something Tushnet said once: if we must, just say that famous mark owners own the mark, not connected to anything that a defendant’s use does to consumer perceptions, and have all the weight rest on showing fame and on rigorously enforced exceptions.

Dinwoodie: transaction/enforcement costs also have some bearing on that.

McGeveran: might have avoided some monkeying with the confusion rationale, if we no longer had very famous marks exerting a sort of gravitational pull on confusion doctrine.

Litman: Boy Scouts have that; are we happy with that?

McKenna: turns out that, though the new dilution statute has problems, its exceptions are much more robust than for confusion.  For enforcement costs: in order to allow people to protect incremental source meaning, we’ve created a series of rules that are expensive and difficult to apply, and the designs that make it through the hurdles are very few.  Dilution may be the same; there’s now a split about whether you have to show something more than association, and maybe why bother?

Bone: it’s possible that most claimants will give up if you have these hurdles. The Q that would arise: why just famous marks?  Why not give this right to more marks?  The common law asks what principles justify limiting this right to famous marks, and there is nothing but politics.  Courts would try to extend it. 

Heymann: Bently asked what has changed—and maybe it’s not distribution, but the way the consumer interacts with the brand as a personality. Dilution then can be a right of publicity for famous marks, like a right for celebrities.  Brand as persona drives the claims being pressed.

RT: Bone says it’s possible that most claimants will give up if you have these hurdles.  Though that still affects risk calculus in a different way than a small but absolute rule. He fears pressure to extend the right. But this wouldn’t be a common law right—that’s the point. Though I take the point that courts might try to stretch others’ rights to match the statutory right.  Not clear that we’d be worse off, especially if the standard required identical and not merely similar marks. Right of publicity analogy is great.  Notable that the right of publicity too has been under pressure to expand to noncelebrities in various ways.  One piece of evidence on statutory solutions, at least at the state level: NY did a much better job limiting the right than California did!

McKenna: the other issue w/the right of publicity is what counts as use. If you give courts any rein about what counts as use—“persona”—then they’ll run away with it. So any TM solution should require use of the identical mark.  (RT: maybe using the counterfeit standard?)

McGeveran: NY statute has a list, and has been effectively policed, unlike California’s standard.

Bently: we do have special rights for the Olympics, Red Cross, Red Crescent, and are extremely strong rights with no suitable First Amendment limitations/possible free expression defense under ECHR.  However, other institutions don’t seem to be clamoring for the same set of rights, so maybe they don’t inevitably expand.

Bone: all statutes are rent-seeking, so that won’t necessarily constrain it. Q is whether courts would use analogies to the statutes in other areas.  Heymann is right about branding of identity.  But what justifies protecting this identity? There may be a moral right, but that doesn’t work for a corporation.  If the reason is instrumental, then that instrumental rationale can be extended, and clever lawyers will try to do so. We may not be able to contain it by speaking to the identity of famous marks.

Dinwoodie: European standard uses “identity” but has been interpreted to mean “close to identity.”

Litman: points out that there are dozens of Tiffany’s strip clubs, as well as dozens of Tiffany’s restaurants (Posner’s example).

Mid-point Discussants: Jeremy Sheff

Looked at registrants and whether marks were in same category or not. Atlas has had as many as 127 prior registrations live at the time of applications in the same product class, and 268 live in other classes.  These numbers may be slightly off for technical reasons, but the basic point is there. Champion, Phoenix, Ace, ABC, Titan, Sentinel, Eclipse, Guardian, Star, Guardian, Eagle, Delta, National—similar story.

Takeaway: at least in registration context, TM system hasn’t seen any significant problem with concurrent use of standard character marks by different owners both in the same and different product classes.  This is particularly so with marks that might be arbitrary but connote “strength.”

Consumers are probably pretty good at sorting through different users’ applications of identical marks to different products, and PTO has long recognized this. 

McKenna: this is more evidence that the courts which conclude that registration is evidence of strength are clearly wrong.

Burrell: you’re just using classes, which are broad.

Sheff: yes, and classes have also changed in size and shape over the period studied.

Laura Heymann: when a consumer sees “Virgin” on hammers, what do we assume the consumer thinks?  Maybe we want to ask what’s material to the consumer’s decisionmaking process.  What happens when the consumer experiences the same name in two different product spaces? Examine the assumptions behind our accounts of what consumers think and do.

Bone: we think consumers generalize some experiences.  May feel less positive about the entity to which the bad experience is linked—that’s intuitive. 

McKenna: Marketing research: when it happens, it’s because of sufficiently transferable skills in making and selling products that consumers believe the skills will transfer successfully—there, there might be some feedback. When the products are far enough apart, consumers won’t penalize the producer.  Happens when there’s fit.  Marketers consider extensions “far” when TM cases would consider them closely related—it’s not that it never happens, but it’s a lot more rare than TM doctrine things. 

There is a producer side story to be told: in those settings where what’s perceived to be a brand extension is a failure won’t have a feedback effect, it might nonetheless interfere with other extension plans. If Virgin the airline is perceived to be bad at making hammers, the airline may have trouble with other product extensions.  Normatively we can talk about that, but it’s at least plausible.

Bone: suppose I don’t have a prior positive experience with the senior mark, and I just experience the junior mark.  My first contact will be bad.

McKenna: would have to go back and look.

RT: Bone on the intuitive appeal of harm: Except the memory research and advertising research says that we’re very bad at resisting manipulation: ads make salty orange juice taste good; ads can make us think we saw characters at Disneyland owned by WB; we are very resistant to changing our already formed opinions, from brands to students (what happens when you tell teachers that their kids are particularly smart).  Because we get powerful messages from ads, our first contact with a brand is usually not consuming the product but getting the brand’s message. That’s why I don’t worry too much about the unusual person who has never seen an ad for the senior mark, encounters the junior mark, and then develops a bad opinion of the senior mark when he finally does encounter it.

McKenna: even really sophisticated consumers are really bad at resisting ads.

Much of the literature is about advising companies when it’s smart or not to engage in brand extensions.  It’s about measuring costs and benefits.  Marketing departments aren’t advising to expand into everything—there’s a lot about creating sub-brands and differentiating.

Dinwoodie: still suggests huge gap between legal and marketing assumptions.

McKenna: companies are in part interested in preserving space for future action.  Maybe they are thinking 10 years into the future.  (RT: There are companies thinking 10 years into the future?) Maybe in marketing departments.

McGeveran: execs who don’t want to be the one who lost space X if bosses want to move into it.

McKenna: concerns over free riding/leaking value might change incentives some, but hard to say it’s enough to cut back the incentive to advertise heavily. 

Litman: expects some overenforcement is that counsel believes the harm stories, however improbable.  Would be good for the world if the real empirics could be brought to bear on the problem. 

Bone: hard to falsify the proposition once committed, and people are risk averse so don’t want to change behavior.  Accountants claim they’re measuring something: stockholders think so too. Then we’ve got reinforcement of this thing that needs protecting or will be lost. 

Dinwoodie: there are also countervailing cost pressures; in house counsel might be taking some leads from business/marketing.

Heymann: Google sends letters saying “we see you, but we’re not (yet) asking you to stop.”  Could try to change the paradigm there.

Dinwoodie: if overenforcement is the issue, need something more systemic.

Heymann: but if clients are saying “do something!” lawyers need a legal pressure by which they can push back on that unless something actually should be done.  Much discussion of Jessica Silbey’s work.

McGeveran: Silbey emphasizes that it’s a personhood feeling.  Responsibility for integrity of the brand.

Heymann: sometimes they’d even hear from consumers: do something!

RT: but what is protection?  This is a question of scope: we don’t think that owning a building means you have to prevent people from building a similar building nearby or you’ll lose the value of the building, especially if the people in the new building are making a different thing than you are.  I don’t know if anything but persona/personhood concepts can be at work here.  (Jessica Silbey’s work again.)

Also worth mentioning: pressure from authorized distributors, not just consumers.  Costco v. Omega: Omega’s acts were all about protecting distributors. 

Dinwoodie: similar with keywords: if authorized distributors aren’t allowed to buy keyword ads by contract, there’s an incentive to pressure the TM owner to go after non-authorized distributors of legit goods for buying keywords.

RT: I noticed we didn’t spend as much time as we could have on product markets.  Why are they so hard to talk about?

Dinwoodie: You can get very sophisticated in defining product markets, in antitrust, but that’s not congruent with the rough justice that tends to prevail in TM.

McKenna: so many moving parts in the product market definition: are you interested in consumers? Managing business relations?  Ensuring future room for expansion? Or, with promotional goods, you just have a normative commitment to whose market this should be? We’ve all realized there isn’t one normative principle there on which we can all agree, making it hard to draw the line.

No comments: