Thursday, August 09, 2018

IPSC session 1

Standard disclaimers apply: These are my summaries, not the presentations themselves. As usual, I have to skip a lot of interesting presentations and I try to attend things I haven’t seen, no matter how good the ones I have already seen/read drafts of are.

Session 1: Trademarks

Jason George (and Lisa Larrimore Ouellette), Trademarks as Innovation Incentives

Not saying that the goal is/should be innovation. SCt has said multiple times that TMs aren’t about promoting innovation or discovery; that’s patent/©.  We’re saying that this has made us overlook the effects of TM on innovation, which could be good or bad.  Survey evidence: many innovative firms don’t use patents at all. Pharma/chem aside, firms view first mover advantage as more important in appropriating returns from investment.  Not concerned about lines separating doctrines. TM is important in maintaining/extending first mover advantage and thus creates incentives.  Purple pill for Nexium=competitive advantage in the marketplace; the Swiffer, which used lots of existing inventions but appropriated returns through branding.

Economists are now using TM as innovation indicator—registrations correlate w/R&D spending, stock market value, and self-reported innovation as well as patent counts. There are caveats. TM registrations don’t mean 1-1 new product; could be a product new to the firm or the marketplace rather than new to the world, but that’s also true of patentàinnovative product. Also industry-dependent, but in some industries TM better correlates w/innovation than patent, for example in info services.

This doesn’t mean that more TM means more innovation.  TMs have costs to innovation, as w/patents.  Barriers to entry; higher prices to consumers. Overincentive—TMs can endure forever.

How do first movers exploit TMs?  At least three mechanisms: staking out the most effective marks—Beebe & Fromer on the marketing literature on the limited supply of “good” marks.  Causing chosen marks to become synonymous with the innovation—arguably Lyft is a better name for ridesharing, but Uber owns the category mentally.  Using the first mover period to develop brand loyalty, as w/name brand drugs after patent expiration.  Can work together: cronut is an apt name for the hybrid pastry, and it’s a registered TM though it’s arguably generic (and the registrant may not have been the inventor [which seems to be a problem for the theory, as does the Swiffer example]) and developed brand loyalty. [Isn’t this an incentive to create a thing called brand loyalty or goodwill, not the thing that we have historically called innovation?]  Shredded Wheat case goes to the link—protecting a name synonymous w/product has a similar effect as patent protection post-patent expiration. 

Complementarity b/t patents and TM in industries w/low turnover and need for differentiation b/c advertising spills over to similar products, like the pharma industry. [Maybe you need TM differentiation to preserve market share in pharma precisely because pharma depends on patent-type innovation, which is generally free for others to copy once the patent expires?]  But TM provide distinct benefits over patents: Cost. You don’t need anything to get TM protection but bring product to market, and registration is cheaper/faster than patent examination. TM cover more subjects than patent—anything that can be sold in the market.  Quick innovation cycles are poor fits for patent as w/software, but first mover advantage can be significant. Less robust protection than patents, but weaker protection over a longer period can achieve the same incentive w/less deadweight loss.

Any full economic analysis of given TM doctrine should take good & bad into account when trying to measure its effects on a particular industry.  Where doctrines give wiggle room: genericide, secondary meaning, functionality. Unclear if change is needed. Added complexity from trying to factor in may o/w possible efficiency gain but that’s different from believing that these effects don’t exist.

[Why not call this theory a theory about protection against point of sale confusion or maybe just against counterfeiting?  Current formulation seems to be imprecise in just the way that distorts full economic analysis.  Does dilution protection support first movers’ incentives?  Especially given the evidence about how most (non-pharma) firms don’t make internal distinctions b/t IP doctrines (see also Jessica Silbey’s work), it’s hard for me to understand how one would distinguish between “existence of some basic TM right” and “this particular aspect of the doctrine” in performing the economic analysis of a doctrine’s incentive effects.]

Q: this is about what TMs do to compete/bring a product to market, not innovation. What is it about TM law that is stimulation innovative activities as opposed to bringing a product to market? Also: Don’t really get how registration is an innovation indicator. From economic perspective there’s such a low cost that it’s not a great signal. Innovation correlates w/ a lot of stuff like not being starved.

A: Rewards a number of different activities, but mechanisms affect first mover advantage in a lot of ways.  One could argue that first mover advantage is the key thing, and TM extends that.  Causes TM to have more direct effect on innovation.  Also true that correlation isn’t causation.  Don’t rely on those studies.

Christian Helmers (with Carsten Fink, Andrea Fosfuri, and Amanda Myers​), Submarine Trademarks

iPhone mark originally filed in Trinidad & Tobago on paper March 2006, by a company called Ocean Telecom that eventually merged w/Apple.  Filed USPTO by Ocean Telecom Sept. 2006 for extension from T&T.  Official announcement by Apple Jan. 2007.  Released June 2007; merger w/Ocean Telecom Sept. 2007. This reflects the submarine strategy: public disclosure is strategically delayed by filing in a foreign jurisdiction w/o online publication and sometimes combined w/use of shell companies. That can then be used to get US priority when filed w/in 6 months under the Paris Convention.

Small but increasing number of mostly tech firms. Used PTO TM dataset to ID TMs 2002-2016 claiming priority from one of 6 jurisdictions that are Paris Convention but don’t publish online register. Also extracted non-submarine filings from companies that use the submarine strategy.  187 unique company names doing submarine filings. Also distinguish filed in own name/filed by shell company; our prediction is the upward trend of filings will continue to increase.  Apple, Google, Zynga, Mattel, T-Mobile, Cisco, Amazon, LG, Bethesda Softworks, Intel, Beats Electronics, Facebook, Instagrams, Nest Labs, Tivo.  Amazon stands out w/shell company filing strategy, but Apple has done it more numerically so far.  Strong trend: class 41, 42 (computer & entertainment services), goods class 9 (computers): usually filed in products & services.  Mattel stands out here as toy company, but has used the strategy for online characters.

Used Google Trends: are submarine TMs associated w/fewer online searches prior to official product announcement/release?  Answer: yes. iPad Air, for example, had zero search intensity until official announcement; the USPTO filing date was after the official product announcement. Compared to regular TMs: the regular marks have much higher search intensity before the announcement date.

Thus submarine TMs do seem to be associated w/more secrecy than regular TMs. More likely to use when TM/product is more valuable; competition for products/names is intense; company pursues a global branding strategy. Various metrics, including: Where you tend to have a lot of refusals and oppositions, companies are more likely to go submarine.  Madrid filing/Madrid filing in China (indicating a global strategy) is positively associated w/submarines b/c they’re worried about squatting in China.

Daniel Hemel (and Lisa Larrimore Ouellette), Governing the Semantic Commons

Semantic space is elastic, but not infinitely or effortlessly so. English speaking consumers can be trained to remember marks like Krasny Oktyabr, but it’s a cognitive stretch. Example: Hagen Dasz.  Tragedy of the semantic commons include costs of proximity: costs arising b/c marks are too close semantically to one another include consumer confusion, reduced incentives to innovate/invest in quality.  Distance also leads to confusion: medication with name too long for expanded Twitter count makes it hard to identify the medication or to figure out what’s the same/different—leads to greater reliance on gatekeepers and greater start-up search costs.  Compare: last French/Italian wine you drank—what was the mark, not the varietal? For non-connoisseurs, it’s hard to remember and we become reliant on people who restrict market access.

Trouble w/TM is not just that semantic space is limited, it’s also that almost any change to TM law that reduces the costs of distance will increase the cost of proximity, and vice versa.  This tradeoff is almost inevitable. Consider geog. limits on TM protection, or higher showing of secondary meaning before recognizing a descriptive term as a mark: that increases ability to approach the other use, but also risks confusion for the people who do cross over geog. lines or who do think there’s a single source.

3 possible escapes from the dilemma: propertization. TMs are relatively weak property rights; assignment in gross doctrine limits alienability/divisibility, making semantic space heavily zoned—you can’t just do whatever you want w/your property. We could remove the zoning and allow property owners to allocate regardless of actual consumer protection, so the single owner could decide, or not, to allow one user in California and one in Iowa, or one user in one category of beer and another user in another category. Demsetzian logic: trust the mark owner.  Trouble is the Merrill/Smith argument about slicing property.  Other marks lose information value when you do this.  You can’t then look at any TM and understand what it represents, and the property owner doesn’t take that externality into account when making decisions for its own benefit.

Could use Pigouvian tax, or subsidize firms that expand the semantic commons.  Target tax benefit for advertising to those who call their product “Smirnoff” instead of “Natural Light.” 

Alternative ID systems: government generated, like SSNs, or bar codes (privately generated) so that you don’t have to remember what you drank before and can just scan the bar code.  There is need for coordination, whether by gov’t or by private systems.  Some health systems use 10-digit identifiers and others use 11-digit, and not all transfer the 10 to 11 in the same way, which is a problem of semantic space but not generally thought of as a TM problem. 

Q: is propertization sensible given the confusion results?

A: We know that land is finite and we’ve still given strong property rights.

No comments: