Thursday, June 09, 2016

TM roundtable, part 3

Session 2:  The Menu of Options for Design Protection:  Where Does/Should Trademark and Unfair Competition Law Fit?
This session will consider product-design trade dress protection in relation to other legal regimes that encompass design (design patents, copyright, and sui generis design laws).  To what extent is there overlap between these regimes?  Is such overlap desirable?  Costly?  Are channeling doctrines appropriate and/or available to minimize these costs?  What doctrines do these other regimes use to filter out designs that, for one reason or another, are inappropriate for protection?  What are the relative virtues of these doctrines, as compared with trademark law’s functionality doctrine?  Does the cheerleader case give us any insight on this question, with respect to copyright’s useful article doctrine?  What is the relationship between trademark functionality and trade dress functionality, both on the merits and in terms of preclusive effect?  What are the costs and benefits of a sui generis regime, either alongside or instead of trade dress protection?
Introduction:  Mark McKenna: Options: Utility patent, design patent, copyright (Varsity Brands may indicate more protection for design than others thought). Different defenses, duration, cost to acquire.  Different claiming mechanisms: words, pictures; ex post, ex ante. Design patents involve a specification prior to assertion against any particular defendant; stable understanding of what it is, as opposed to most trade dress cases which involve claiming in litigation, strategic.  [Burstein might say that continuance and splitting practices allow some manipulation of design patents along these lines.]  They don’t have to be alternatives; increasingly the case that people use many at the same time, often for the very same features.  Notice issues.  Using design patent to get secondary meaning, leveraging limited for potentially infinite protection. Competition policy can be undone by such leveraging.
Two views of how to think about overlaps: Ebbing and flowing over time in which predominates.  One is strong channeling sense that certain subject matter belongs in different buckets.  You don’t want one regime to undermine the placement of that stuff in the right bucket (or not being allowed in the bucket). Strong view of Sears/Compco—keep TM out of the way of patent.  This isn’t just about features that were, or are, or could have been protected by patent, but about keeping it out of patent space—keeping it away from patent law, not just from patents. That which is excluded from patent law is excluded for a reason; meant to put things into the public domain. Sears involves utility and design patent; Compco is just a design patent. This is about lack of protection for unpatented designs, not functionality. Courts then started to think of those cases in preemption terms, not policy terms.
Alternative view, pre-Traffix: overlap is itself not especially problematic; multiple systems exist but have different advantages and disadvantages. Lack of patent doesn’t tell you much/anything about TM. Overlap only becomes a problem in cases of real competitive need.  “Functionality” is defined as per Morton-Norwich.  A retail determination of need, rather than wholesale. This is a significant challenge b/c the issue is precisely what market you should be allowed to compete in. Need richer normative view of ok kinds of competition.
This view is much less concerned with overlap; patent protection is only one factor, often not dispositive—6th Cir. in Traffix itself reasoned that it wouldn’t cost much to use three springs or put it in a box. Not connected to semantic meaning of functionality—have a function—unless you assume that the only things that have competitive value are functional.  It’s really about competitive need; narrowing Sears/Compco to utilitarian features rather than the other aesthetic things they were dealing with.  Once you say this is just about competitive need, it’s easy to say that there are lots of things for which there is competitive need, so aesthetic functionality makes just as much sense as utilitarian functionality. But the terminology now seems totally divorced from original semantic meaning of terms.
Competitive need = aesthetic functionality becomes unconcerned w/channeling away from design patent. Away from categorical channeling of Sears/Compco.  Design patent drops out of the picture. 
Post-Traffix, strong Sears/Compco view is ascendant on the utility patent side: TM shouldn’t be in utility patent’s sandbox.  Both protection and nonprotection decisions are meaningful decisions about competition.  Useful features, not patentable features; patents are a cheat sheet but ultimately we’re concerned about utility.  No return to Sears/Compco on the aesthetic side; it’s all about competitive need. So there’s room for movement, especially w/heavier use of design patent.  Corresponded to otherwise broadening trends in TM law, so conflict is more significant than it’s ever been.
Why only patent law?  It shouldn’t be. We’ll see other thinking about TM/© interface per Dastar.  Why only restrict TM? That’s a useful and important Q, and other systems should consider their own boundaries. One obvious point about comfort w/utility patent: the short term has always been a prominent part of the discussion; TM by definition doesn’t. Infringement standards are also different—no all elements rule/less identity required to infringe. 
Discussant:  Lionel Bently: Utility patents are patents in the EU; there are also Community registered designs (25 years) and unregistered designs (3 years automatic from when made available).  National systems also have national registered designs, also 25 years; some have national unregistered design right, UK’s lasts 10 years for shape/configuration of product, not surface decoration. © is now required to be capable of being cumulative w/design rights.  TM protection for registered marks at Community and national level, as well as possible national unfair competition laws that extend to copying; specifically left open in ECJ TM judgment.
Functionality historically emerged as a way to encourage people to use utility patents to cover nonappearance aspects of design instead of design for appearance aspects. 1839: Britain first covers design, including shape/configuration of product. Patent system existed was expensive and involved required stages.  People tried to claim utilitarian aspects through designs.  Registrar thought they shouldn’t be used that way. New legislation in 1842 limited the register to ornamental design, and in 1843, to cope with demand, created new system for utility designs.  This was literally a channeling device.  In two stages, the legislation introduced functionality.  Finally turns up in 1919 as amendment to 1907 Act: features dictated by function shouldn’t be in the design system, designed to turf them back out to patent.  Very clearly a channeling device to get things into patent law, b/c of registered design system. 
Whether that continues into the European image, he’s not sure.
What about TM? For that, he doesn’t buy this channeling stuff. It’s TM policy: ordinary word marks provide information benefits when protected, and there are no harms b/c taking the words doesn’t do any harm.  Once you move to product features, there is depletion/harm that then has to be balanced.  In that case, you might not want the functionality test to be the same as between design and TM.
Design is about new appearances for products. People motivated to produce new technical effect aren’t motivated to produce new appearance.  This test accords well w/policy justifications for designs, though intent is a messy test.  IP systems apart from TM are largely designed to produce intentional behavior/investment.  So focusing on people’s intentions doesn’t seem necessarily inappropriate.  So I’d have a broad notion of design functionality, and an even broader notion of TM functionality.  I’ve always liked the standard that something is problematic as a TM where it confers any significant non-reputation-related advantage on the claimant. 
Dogan: I don’t think that’s very broad.
McKenna: Agree.
Dogan: as long as there are alternatives that work plausibly well in the marketplace, feature protection is allowed. Quite generous for protection, relative to what we have now for features that make product work.
Lemley: disagree as practical matter. Narrow limiting piece: essential to production of article.  Or affects cost/quality. I understand Bently’s test to be an effort to adapt “cost or quality” into the aesthetic framework.
McGeveran: a nonutilitarian gloss.
Dogan: relative to what?
Lemley: Significant is the question; but that’s implicitly there anywhere. You could make the argument that the Coke can is better than non-Coke b/c people like the color red; that’s not going to fly. The key is the non-reputation-related disadvantage, to distinguish b/t I want this b/c it looks nice and b/c I like the TM owner.
Dinwoodie: channeling part of Inwood is the first part.  Qualitex: strange “that is” language between Inwood and substantial non-reputation-related disadvantage; that can be very broad if you read it that way.
Dogan: if we think channeling has a value, how we think about design patents: Sympathetic to channeling in utility patents. Post-Traffix, courts have taken a pretty good approach; no evidence that there’ve been disastrous cases.  In design that’s harder, b/c if we accept design TMs at all, then there’s overlap by definition. McKenna came around to election; that’s the only real alternative if we want some separation/deference.  If we want design patent to incentivize design, then we have to have the same quid pro quo: after a time, it’s free to the public. That does lead her to election. If you take advantage of that system, we should take at face value your claim that your design has value, intrinsically, which then gets us to all these designs presumptively meeting Bently’s standard. 
Bone: can be useful to separate aesthetic value from generous view of secondary meaning. Source identification can be part of what it makes the thing aesthetically attractive. We’re not sure exactly what we’re channeling and where.  Sears/Compco dividing the world won’t work because there are other policies that will be stomped on. Trade secret: there are nonincentive reasons to have trade secret law. Policies that patent doesn’t take account of that we nonetheless want to operate in this domain, and that gets us to TM.  We may question whether TM policy is effected by giving protection in this regime, but it at least could do so.  Rule-like protection might be the best balance, but we have to argue that.
Kur: History of European design legislation—she was involved.  There was definitely no channeling purpose at all, at least for current design legislation.  It was only to incentivize product diversification, not aesthetics.  Aesthetic is the wrong term, except very broadly—capacity to communicate difference from what is already there.  [Making me even more confused about why Europeans insist TM has a product differentiation function.]  Unless feature is strictly necessary for function, protection is allowed.  Practice has developed differently.  Inner logic: there is no reason to exclude anything other than mandatory shapes.
RT: Dogan says: Your claim by registering design is that your design has value: is that what you’re claiming? Why not: don’t copy my stuff?  An anticopying norm as described by Kur for Germany in the last session doesn’t have to have an incentive benefit, and it’s also consistent with her account of European design.  Maybe it’s just a private reward. Though I think that is  a bad idea, but coherent.
My preferred version of the aesthetic functionality test is a counterfactual: would the element have marketplace value if the plaintiff didn’t exist? Makes sense of most cases, though you then need another explanation for why Betty Boop case should come out as a D victory; also allows you to do defendant side functionality as in Louboutin, where there’s an obvious reason to have an all-red shoe even in a world without Louboutin: the counterfactual test allows you to remove the “reputation related advantage” and we’re trying to see whether there’s any advantage left after that’s done.
Dinwoodie: throw Dastar into functionality mix. 
McKenna: skeptical that courts are willing to pull the trigger on aesthetic functionality.  If we had a more limited design patent regime, there’d be less overlap, but now you can get a design patent on a logo on a screen.
McGeveran: Not sure there’s much disagreement in the room: not a pure channeling v. non-channeling divide.  Channeling is something courts use as a heuristic to reach policy-driven outcomes in a quicker way.  Bently’s history is one reason channeling occurs: institutional design/maneuvering around a bureaucracy.  Another is the heuristic.
Yen: What if some people might like Gs on their bags even if Gucci didn’t exist, but very few? How would that work?
RT: I think that’s incorporated in substantial non-reputation related dis/advantage.  Higher level of abstraction can be used for Louboutin: people predictably like single-color shoes.
McKenna: Justin Hughes argues that a preexisting aesthetic preference is required for aesthetic functionality—in many cases the P creates the demand/aesthetic preference.  [I think this is similar to but not as helpful as the counterfactual, b/c we can’t easily figure out what preferences were, and also b/c it’s possible that a few people would prefer, say, linked Gs on their luggage even before Gucci launched, b/c their intials are G.]
McGeveran: attractive is a lot harder to measure than useful in other ways. 
Dogan: most product designs are intended to be attractive in some ways. Gucci may be a good example of not being intended to be attractive: both to reflect source and to be attractive.  [This might be an example of distinguishing b/t the word/symbol mark and the design of a Gucci bag, which is probably, yes, designed to be attractive—but maybe product design generally just shouldn’t be protected.]
McGeveran: Green-gold pukey color in Qualitex likely wasn’t chosen for attractiveness, but rather for distinctiveness.  [Though might be different for clothing, where color plays a different role—it would make sense to have different amounts of functionality in different markets.]
Dogan: need to know how to measure value and in what market even under counterfactual test. Red sole w/contrasting upper might make high heeled shoes look sexier. We struggle w/this question of distinguishing visual features of product appearance when those should be protected v. free for all.
Jessica Silbey: election as a matter of client counseling.  It’s a common strategy to file for TM, design patent, utility patent together all the time. If you’re going to be penalized for not doing this when you start out by being deemed to have waived the protections you didn’t seek, that will encourage overprotection at the client level.
If we start from the idea that TM is not an anticopying regime, that concurrent uses are common—what would concurrent trade dress uses look like the way we have concurrent uses of protected TM words?  For aesthetic functionality, you should be able to copy aesthetic features like Burberry plaid, just as we imagine concurrent uses w/word marks, protected in separate markets, but it’s hard to do that with aesthetic functionality.  So if you start by saying TM isn’t anticopying, both product design protection and aesthetic functionality (nonetheless?- stop copying in many circumstances.
Alex Roberts: can we fold registration into this discussion?  Expectation is that litigation and registration scope would match up, though we know that’s not true in practice, especially w/r/t trade dress.
Kur: once the regime itself is flawed, the threshold doesn’t correspond to the contents of the regime, then we have a big problem even in the absence of overlaps. 
Max Planck center proposed to abolish our equivalent of aesthetic functionality; it doesn’t work.  We protect every little difference as design, so it became nonsensical. Need another reason to not protect shapes that provide inherent value, and we don’t have clear policies about why certain things shouldn’t be protected.  Worse, we make the assessment at the registration stage: whether it’s so appealing to the public that it shouldn’t be protected—they may be unable to do that b/c it may not be in use yet! And once the decision  has been made, it’s there forever b/c acquired distinctiveness can’t trump it.  Yet fashion is such that people may like some things now and not even care for it in 10 years. 

No comments: