TMSR, part 2

Session 2:  Defining Markets: Doctrinal Settings for Market Analysis in Trademark Law

Trademark law clearly already undertakes market definition in the analysis of genericism and also for functionality. Are current approaches adequate? Are markets defined the same way, through the same inquiries, in both settings?   Should they be?   If so, what are the essential inquiries?  How does market definition operate within the factors test for likelihood of confusion – superficially, through the similarity-of-products or similarity-of-market-channels factors? Or more profoundly? Is market definition in fact ubiquitous in trademark law analysis because it is an inquiry implicitly undertaken whenever a mark owner defines the underlying goods or services with which the mark is used? In which other trademark doctrines—if any—is market definition doing substantial work, even if below the surface? Is there a coherent concept of market definition in operation across trademark law doctrines? If not, should such a concept be developed and imposed?

Introduction:   Jessica Litman [sadly missed, due to teaching]

Bob Bone [midtalk]: How careful we want to be about market definition may depend on error costs. If we’re worried about anticompetitive effects of product trade dress protection and we know other marks are available, we may tilt in favor of narrowing market definition.

If we cared more about commercial morality, market definition would be less important.  [Unless we thought that the TM owner’s market gave it preemptive rights in nearby markets.]

Discussant:      Bill McGeveran: in what sense is this even a market? Is that word even meaningful for some of the things that are packed into this prompt?  Two roofing contractors; one claimed a TM in orange and had a registration in orange, more or less. For what, though?  In one sense, clearly a defined market—two geog. competing providers of identical services. Scope? Hue?  Claimed orange on T-shirts that workers wore—but orange for safety equipment is pretty functional.  Is the market “the market for orange T-shirted workers”? Market definition is often not about defining markets.

Second, increasing detachment of modern brand-oriented law from TM is putting pressure on markets as a working concept. Increasing merchandising; detachment of online purchasing from presence; cross-licensing—all involve assertions of rights well outside the traditional “market.” One reason those doctrines have been so messy is that we’re using market-anchored concepts to talk about marks that are no longer anchored in markets.  This work is necessary, but if every doctrine instructs you to go back and consult the market, and you can’t identify the market, the doctrine is broken.

Dinwoodie: we’re more concerned as a group seemingly about defining the mark than defining the market. But if TM law is a law of competition, that matters; why be so comfortable w/allowing parties opportunistic definitions in litigation v. mark definition?

Silbey: Interaction b/t TM and other fields matters here—bankruptcy.

Lemley: Lynn LoPucki suggested that we could take franchising more seriously—if you hold yourself out as a McDonald’s, you’re a McDonald’s and your employees are McDonald’s employees.

McKenna: confusion isn’t found in nature; you have to know where to look to do a survey and once you ask a set of people you’ve defined a market.  Happens all the time if we’re trying to think about confusion. Happens implicitly every time an examining att’y decides that goods/services are remote enough to avoid confusion. Almost always done by uninformed, intuitive judgments.

RT: it seems as if one difference from antitrust is that antitrust takes markets as they are. Functionality in TM at least takes markets as they could be, if there were perfect information—we’re not going interfere w/ the potential, even if people don’t understand that this design produces some specific benefits.  Similarly infringement takes markets as they could be with the idea that producers can be in multiple markets and consumers can expect that cross-market presence—the lack of substitutability of salt for sugar doesn’t directly bear on whether a producer of one would produce (or license its marks for production of) the other.

Dinwoodie: bridging the gap; ITU/non use based systems are allocation of future markets—TM law in some ways constructs markets rather than reacting to them.

Grynberg: Market definition does little work b/c of approval confusion in cases like Diageo/red dripping wax for alcohol, or court just makes up its idea of the market.

Linford: path dependencies? Do you take Simon Tam at his word that he can’t get a record label to sign him without a ® for his band? 

McGeveran: we’re not defining a market any more, we’re defining a class of potential harm (poorly) and a class of potential perceivers of the mark.

Dinwoodie: vehicle for implementing levels of paternalism/competition/choice.

Austin: TM law assumes everyone wants to use a “virgin” mark and put it on everything, but doesn’t often ask whether that’s what producers or consumers want.  UK SCt: Starbucks decision gives an embarrassing attempt to segregate markets internationally. Eschewed consumer perceptions/desires and segregated markets along geopolitical borders, incoherently. Maybe there’s a degree of relaxation about defining markets b/c there’s so many ways to do it badly.

McKenna: maybe we’re using market in different ways compared to RT’s example (sugar/salt related?).  TM law says that if people think both would produce, then they’re in the same market. But we’d still look at the D’s consumers: if Morton’s Salt sued Morton’s Sugar, we’d still look at sugar consumers—if they overlap that’s fine, but we shouldn’t assume that.  That’s one way Ps manipulate their surveys: by baking that in to the structure of the survey [no pun intended, I assume]. If you find confusion among sugar consumers, you could find that Morton’s should own the rights in sugar, but you can’t assume you found in nature an existing set of consumers.

Dinwoodie: he thinks you’re allocating the sugar market to the salt producer, but it’s still two markets.

Mike Meurer: antitrust considers things like whether a producer might enter the market if there were a price rise by existing producers—could that be analogized to anything like TM and what the TM owner might like to do if things change?

Linford: Belmora as a case where US consumers may be tricked, though maybe not problematic if the chemical composition is the same; the party seeking to vindicate this harm may be the wrong party.  Is the Morton Salt company the right enforcer of rights against Morton Sugar?

Gangjee: Market highly relevant/picky definitions for registration.  Exhaustion/first sale: markets matter a lot. But by the time you get to infringement, we’re just interested in perceivers (per McGeveran).

Beebe: dilution signals a a fundamental issue about size of “market” and disappearance of market inquiry; dilution makes the mark everything. Expansive doctrines like IIC, approval, and post sale confusion tend to be market-limit-busting arguments.

Dogan: if we define markets broadly consumers come to expect more crossover/more easily confused. [I wonder about this.  How do we know how consumers learn the rules? In my class the other day we were talking about the public understanding that use of a hashtag on Twitter is not controlled by TM owners. Is it just a matter of its there-ness for us?]

Heymann: aesthetic functionality: what is the market in which a ring pop/baroque silverware producer is seeking to compete?  The court there wanted to know who they thought their competitors were.

Dinwoodie: are courts better on this for product design/functionality than for other issues?  It’s more important to define the market in such cases.

Dogan: TM owner meaning v. intrinsic appeal—very hard to segregate those things (which turn out to be important to resolve the issue of what the market is). No one would argue that Burberry plaid lacked substitutes before Burberry developed secondary meaning [though no one would dispute it had intrinsic appeal before that, I think, forcing us to figure out whether there was something nonsubstitutable about the intrinsic appeal].

Gordon: as a Veblenite, wants to reject the idea that consumers’ desire to purchase status should be part of the TM.  Burberry plaid purchase often has nothing to do w/actual appearance of Burberry, or with quality of product; they want to wear something that will signal status. That’s functionality to her.

Diamond: the survey expert is the person who does market definition in a TM case. We often make immediate assumptions and there might be different perspectives upon a deeper dive. 

Session 2,  Cont’d

Mark Lemley: point of TM law from producer perspective is product differentiation. That shouldn’t matter in a world in which I own the market for another reason; if I own the only patent, I shouldn’t need a brand, and using it is just preparation against impending expiration. TMs, to have value, should operate against a background of differentiated competitors. But how imperfect is the competition? A function of how broad the differentiation is and how broad IP rights are. And there is an antitrust/IP mismatch.  One implication: there may well be circumstances in which the brand defines its own market.  RT wouldn’t respond to a small, significant, nontransitory price increase of Diet Coke; if enough are like her, it’s its own market.

We reach beyond the market in lots of doctrines—post-sale confusion, dilution, etc.—look beyond expected consumers to a completely different group. The core may be about differentiated product competition, but we’re also starting to think of the brand itself as a market. Part of that is abandoning similarity of goods as a touchstone for confusion. Luxury goods, team logos, school colors: the thing I want is the identifier/participation in the group.  Has tried to resist this change, but it seems to be the way the law has gone.

If we took brand as market seriously for antitrust purposes—fine, you want to own the market, we will treat you as a monopolist for antitrust purposes—lawfully acquired monopolies are ok but get scrutiny of some of their practices.

Further emphasizes need for materiality/TM injury doctrine to measure when we should care. D’s benefit theory of remedies is completely circular once we’ve abandoned a market definition attempt; thus we should look for actual injury to TM owner, even injury that goes beyond lost sales.  Why do consumers care about this confusion? What is the connection b/t D’s benefit and P’s harm?

Right now we seem to reject the weight similarity of goods as a factor when it might help Ds and consider it very significant when it favors Ps; perhaps we should be more evenhanded. [I thought Beebe’s empirical survey suggested that similarity of goods could be dispositive when it favored Ds.]

Shari Diamond: Dilution and markets: fame is general consuming public. But when you get to tarnishment, what’s the relevant market? Can you be tarnished among nonconsumers? The statute is silent.

Courts spend a lot of time thinking about relevant population—it’s relatively easy for court to look at universe to decide whether the survey is helpful to the position the court has decided will prevail. Flexibility to declare the survey universe too broad or too narrow. For a prescription drug: is it physicians? Physicians who can prescribe the drug?  Users who might take the drug?  For toys: who buys the toys? But who sees it advertised and lobbies for it?

How much do surveys cost?  Online surveys have helped a bit for demographic diversity w/o too much costs. But it depends.  It can be $50,000-$100,000.  It depends on how hard it is to get a respondent who’s qualified. If you want a physician, you’ll pay a lot b/c the panels pay them a lot. If you have toothpaste, it’s much cheaper. Though that wouldn’t include expert time—just the survey itself.

Discussant:      Leah Chan Grinvald: real world consequences for how we operate for SMEs and others—uncertainty about how a court would treat a survey even if you have one. This really squishy notion of market definition hurts b/c they will narrow the market definition if they can or they may have a less visionary view of future market as they try to make it in their small market category. This has a feedback effect and contrasts w/ Amazon getting bigger and bigger.

McKenna: TM as information but we presume that TM owner has control over production of goods. As design and production separate (3D printers at home) it will be more an abstract notion of design rather than production quality. That’s another path to making the mark its own market.

Goldman: there are multiple market segments for any product.  Some may be confused and others may not be. 

Diamond: what is a judge supposed to do with that?

Goldman: may allow more defense arguments; when he sees 15% confusion, he thinks about the other consumers who aren’t confused.

Bone: the brand may not be a market in itself if it competes w/other marks that communicate the same kind of information. But at the end of the day, a brand is not a TM.

[much discussion about brands v. TMs, rights in gross v. rights appurtenant to particular products, rights that are appropriated from preexisting meaning (e.g. rights claimed in the Mona Lisa) versus supposedly self-generated value (versus strong sentiment that no one makes value on their own/Felix Cohen was right)]

McKenna: Any use of Coca-Cola [as a mark] would cause confusion under current doctrine. And that also affects what counts as “use as a mark.”  The only limits that allow other people to use Coca-Cola are nominative use, Rogers, etc. However: The doctrines we have created so that Coca-Cola can do that are not limited to Coca-Cola strength marks and that’s what creates havoc.

McGeveran: Basically any brand will get the right to put its mark on a T-shirt: undifferentiated licensing markets, in places like PacSun and Hot Topic, where NASA and movies and bands and all sorts of other things are simply displayed as possibilities.

Austin: classic view: seeing a Rubik’s cube on a T-shirt doesn’t mean you think it’s from there—we look at the label.

Lemley: but post-sale confusion gets rid of that. Proposal: when suing actual competitors, give latitude on mark similarity (Cyclone/Tornado for fencing). Maybe for brands, with this new vision, we’ve given them every right in the world except for the old carve-outs [first sale, Rogers], but that applies only to identical marks.  Schecter, but modified.  He wouldn’t allow rights in pure text, only the copying of the logo.  There are some cases where courts happily distinguish between using the word mark to talk about the TM owner, but don’t allow use of the logo. People might well distinguish in their minds b/t those two corresponding to TM v. brand.

Dinwoodie: Europeans are familiar w/ the difficulties that come from “same mark” standards.

Lemley: yeah, but it’s better than nothing.

Dogan: what do we do to emphasize social benefits of limits? Cheaper spare parts; price of T-shirts may go down. Numbers matter.

Linford: he’s not sure that maximizing price competition is always good. Consumers can lose, whether from Veblen satisfaction or otherwise.

Lemley: he agrees to an extent, but his reaction is yeah, that seems like a real social value and I don’t see why people would give it to the TM owner. If people putt better b/c they have a Nike putter, then we should let everyone call their putters Nike. Veblen goods are harder b/c barring the great unwashed is what gives value, but the power of a placebo should belong to the public.

Linford: you’re presuming it’s sticky—that if copying is allowed that value won’t be lost. But that’s an empirical question.