Works in progress: Sean Pager

Sean Pager, Creeping Genericide: The Dilemma of Geographical/Process Certification Marks: I was late coming in because of panel transitions. Processes can move even when geographies don’t. Pager is interested in dual meaning – allowing California wine growers to identify method champagnoise but not to call their product Champagne.

Scott Boone: What’s the problem with tequila?

A: Consumers may not know the significance – tequila is made from a particular succulent cactus, the blue agave. Other agave liquors are not tequila.

Boone: Isn’t that analogous to not knowing who the source is but knowing that there is a source? So it’s not generic even if consumers don’t know production specifics, just know it’s special?

A: Begs the question of what consumers have to know. They know tequila isn’t bread, but if they don’t know that it’s limited to certain types of agave they might think that all agave liquor can be tequila.

David Welkowitz: If you don’t know what constitutes the status of being generic (given that the existence a group of products of different sources with the same certification mark makes it hard to tell what the unifying characteristic is other than the mark) it makes it hard to explain why you’d want to protect the identifier. Are you protecting the standard, or are you protecting the consumer’s understanding of whether there is a standard? The European system as described is about protecting the standard itself and thus the producers, not about protecting the consumer. That’s divorced from the common understanding of genericity, which depends on consumer protection.

A: Protecting the standard still has a consumer protection function, even if consumers don’t understand what they’re getting – they get the tequila to which they’re accustomed. Processes should presumptively matter – ingredients matter – more than geography matters. Processes and ingredients don’t survive and thrive at random. (Though arguably geographic reputation doesn’t either.)

Q: Can you really separate geography from process that way? The process and the grapes and the location all combine into champagne.

A: Wine is a special case because of the mystification of terroir; alcohol is also more generally regulated separately and more controlled than, for example, Roquefort cheese.

Q: Isn’t the real question for the consumer whether the tequila is good? As a factual matter, which actually changes the taste of the tequila, the place or the ingredient?

A: In general, he thinks process is more likely to affect quality than geography, though he agrees that it will vary by product. If tequila became generic, Mexican producers could still advertise “Made in Mexico” in huge letters – one question is whether a genericity finding would make it harder for producers of the original to convey origin information.

Q: We’ve all had good tequila and bad tequila – quality control is not necessarily the issue. Are we looking for some essence of tequila that makes it “really” tequila? If the process can be duplicated in another way, it’s hard to see the value of the certification mark.

A: If you can duplicate the process, he agrees you should be able to advertise your ability to produce tequila.

Q: European view: If I figure out the formula for Coca-Cola, can I market it as Coca-Cola? No, of course not. It’s easy to take products that are already generic in the US, but think about products that aren’t yet generic in the US – the reason for the European system is a different way of looking at the value of regions that produce high-quality products. Famous mark holders existed, but couldn’t be protected under TM because they were a group, not a single source. This is a different way of protecting against free riding on reputation – the logic is the same as TM but the mechanism is different. If you assume tequila is already generic, protection seems bizarre, but if you look at the people who made it famous, created something of value, and still make it now, exclusive rights make more sense. (Note: this European perspective seems plausible, until you get to the ban on “-style,” “-like,” “method,” etc., which is anticompetitive from an American point of view.)

A: It’s like dilution – you protect the mystique of Coca-Cola and the mystique of the traditional region.

Works in progress: Laura Heymann

Laura Heymann, The Trademark/Copyright Divide: Courts are starting to police the copyright/trademark boundary. Maybe we can move some things from the broad remedies of copyright law to trademark, where appropriate relief would be a disclaimer rather than an injunction.

Rogers v. Koons was brought as a copyright case, but what was motivating Rogers weren’t copyright interests in the market for the work, but for association and attribution rights. Another example: Jackson v. Warner Bros. where an Afrocentric artist sued over use of his paintings in the Whoopi Goldberg movie Made in America; he received letters expressing surprise that he authorized the use of his paintings in a movie of that caliber. His interest was more in reputation. Many copyright cases can be seen this way. (A problem I see is that TM law does acknowledge reputation as a reason to shut something up, not just apply a disclaimer – dilution. Also, I’m not sure you can fairly describe copyright as just interested in market harm.)

Creators in these cases aren’t concerned about losing a market they might otherwise want to exploit, but an integrity claim. At least they’re motived by an interest within the realm of TM rather than copyright. (Ah, but then we have to relax “use in commerce” to cover use in a film. As Goldman’s presentation points out, that might be a mistake.)

Courts should pay more attention when content owners try to use copyright to assert reputational or associational interests and should in fact use the doctrine of copyright misuse to force those claims into a TM regime. The second half of the proposal: that TM right should apply even after the copyright term ends. Source identification is important both to creators and to audience members. (Audience members? Did they really care in Made in America?) There’s no reason we need to abandon that importance when we move into the public domain.

Heymann is not proposing that any of these cases will result in a valid TM-based claim. The goal is to end up with fewer claims, and if the result is fewer claims overall that’s fine. But we need to render unto copyright what is copyright’s and unto trademark what is trademark’s. This is a more honest way of letting content owners make claims for what they really care about.

(What do we do with works that have no market, like unpublished letters? Do infringement claims have to be brought as TM claims for unpublished works? Or are privacy interests legitimate non-market interests within copyright under Heymann's scheme? If so, why not admit other non-market interests?)

Second-generation creators end up with what they want, which is the ability to reuse this material, as long as they have disclaimers. When we talk about copyright, we don’t focus enough on the economic interests at the heart of copyright law. (I just think this is wrong descriptively. There’s movement towards attribution in orphan works, as well as attribution requirements in educational fair use, and lots of popular conceptions of fair use refer to attribution – see also Creative Commons licenses, which require attribution.) There are two economic interests: the use of the work itself and the way it’s used, the packaging/the message it sends. But we conflate those two interests when we approach the copyright/TM divide. In other regime intersections, we’ve handled the overlap better – patent/TM, copyright/patent. In Sears, Kellogg, etc. we’ve been able to distinguish expired patents from unexpired TM rights.

As long as TM isn’t trying to do what patent does, create use rights in the article itself, it’s okay to use both, even though TM can be perpetual. Same thing with the patent/copyright interface in Mazer: the availability of patent has nothing to do with the availability of copyright. Yet we have a very different result in Dastar, where the Court cites the patent line of cases for the proposition that the right to copy passes to the public after the patent expires, then just adds in “the right to copy without attribution.”

The main issue that Dastar cared about was applying its rule to the public domain. Right now creators can use copyright as long as they can to achieve control, but Heymann’s proposal would preclude them from doing so.

Just because Romeo & Juliet is in the public domain doesn’t mean you can take a copy from Barnes & Noble without paying or perform it in traffic during rush hour; other legal regimes still apply. (Neither can you hit me with a copy of it, but what does that have to do with anything? Those are about the physical copy/performance, not the expressive work.)

Example: upstart airline starts using Rhapsody in Blue to identify itself, once the song goes into the public domain – it would cause confusion with United. But Dastar doesn’t recognize this. (I would distinguish using a work as a TM for something else as opposed to using a song as a TM for itself, which is what the bulk of Heymann’s proposal is about. Disclaimer remedies, as Heymann proposes, wouldn’t solve the problem with the Rhapsody in Blue example – and that’s because she’s proposing to solve a different problem.)

Eric Goldman: Could United really block another from using that song? Notice that you’ve shifted parties – it’s not Gershwin estate claiming TM rights that overlap with copyright, but someone who never had copyright rights in the work.

A: That matters only if you assume that the copyright owner gives up both copyright and TM at the end of the term – otherwise we treat the copyright + TM owner worse than the TM owner. (But as Goldman pointed out, the Gershwin estate isn’t a trademark owner as we conventionally understand TM ownership. Treating the estate as a copyright-but-not-trademark owner is perfectly rational and nondiscriminatory.) Music may also pose special disclaimer problems. (Does that mean that the estates of composers can get injunctive relief, or no relief at all? Either way, problematic.) Disney can’t preclude all use after Mickey Mouse copyright expires, but it can stop someone from putting a big statue of Mickey in front of an amusement park.

Lemley: TM law can fit in the United case because of competition between the parties. If you believe in the “use as a mark” requirement, most cases won’t end up as TM cases because Jackson, for example, doesn’t involve a TM use. (1) Does that mean you want to broaden TM law to reach that, or (2) whether state law doctrines might come into play, the right of publicity/false light/invasion of privacy tort.

A: Right of publicity is an issue. She’s not looking to expand trademark law. The result may be that claims drop out together. There may be a cognizable TM claim in the Jackson case because people believed he gave permission for the display of the paintings in the movie, though he might lose.

Lemley: Aha, you’re back to expanding “use.” (Like “if value, then right,” perhaps we need to start talking about the “if confusion, then use” presumption in TM law. Or if that concedes too much, “if recognition, then use.”)

A: There could be enough of a use of a communicative product that would be enough – if you used it as a title (Lemley’s example). These claims become successful now because they’re allowed to be couched as copyright claims.

Here are my rather extensive notes on the article, which in many ways replicate my response to the talk: Patent and TM can accommodate one another because most of the time copying an unpatented article won’t be copying a distinctive and nonfunctional trade dress (e.g., Traffix, Wal-Mart). By contrast we can expect courts to routinely treat an author’s name as an indicator of source, even though it isn’t an indicator in the TM sense, especially if we call it a trademark interest.

Did Shakespeare authorize this edition of his plays? That’s a bizarre question from the perspective of authorship in the way that “Did Shakespeare write this play?” is not. In order to avoid massive liability for publishing public domain works, we’d need a bunch of new presumptions, probably drawn from nominative fair use, about the effects of labeling a work as being by a particular author. As with nominative fair use and “de facto confusion” v. “de jure confusion” in functionality, the presence of a particular form – perhaps “by Willa Cather, but this edition isn’t authorized by her estate” – would have to be treated as nonconfusing, without any submission of evidence by either side. But in that case, we aren’t really asking about consumer confusion at all, which is further reason that the game isn’t worth the candle.

Heymann suggests that disclaimers are the appropriate remedy rather than injunctive relief, but that’s not the TM law we have now. She’s really proposing a separate attribution regime. And I wouldn’t mind as much if she did do so, ensuring that disclaimers were the only available remedy. Even so, she acknowledges but does not address the empirical work on disclaimers suggesting they don’t work very well – consumers don’t read them, and don’t comprehend them when they do read them. If a disclaimer is legally sufficient nonetheless, that’s because it’s serving some function other than consumer protection and we’re back to moral right, which does fit within copyright.

Another way to state the problem: TM lacks a materiality requirement. It presumes, because of the connection to a producer’s rights, that authorization makes a difference to consumers, but that is not how the world of publishing works. Authenticity, on the other hand, makes a difference. (There's room, in other words, for both a pirate edition of The Hunt for Red October and an authorized, ghostwritten novel bearing the name Tom Clancy to be "Tom Clancy novels" -- but you will never convince me that the former is not a "Tom Clancy novel.") False advertising can take account of these concerns, much more sensitively than TM.

Heymann’s analogies, responding to objections that her scheme is too complex: orphan works also involve difficulties of attribution, where sometimes it will be hard to figure out who to credit. (1) That’s an attempt to make things better versus a baseline of non-use, with (2) limited remedies. Relatedly, (3) the practical reality is that no one’s going to sue over orphan works because there’s not enough money at stake, whereas there will often be enough reason to sue, or at least threaten suit, when a competing publisher of Beatrix Potter or other public domain work is trying to enter the market. The fact that rights thickets already exist is no reason to make them thicker and pricklier.

A response to some of my questions: Heymann suggests we just have a fundamental disagreement about what people care about when they look at communicative products, whether they care about identity of producers.

I totally agree they care about identity of the author, but they don’t connect that to whether the author authorized the particular copy at issue. What it means to be “by Charles Dickens” is not the same as what it means to be “made by the Coca-Cola Company.” This is what Dastar got right, especially when it left open the possibility of false advertising claims when authorial-origin claims were material and false.

Clearly there's a lot of great work right now about attribution and what it means for copyrighted works. I doubt Heymann and I will ever agree, but it's a good debate to have.

Works in progress: Eric Goldman

Frank Pasquale kindly summarized my presentation here.

Eric Goldman: Online Word of Mouth and Its Implications for Trademark Law: This comes from a synthesis of previous work asking what’s new about online trademark law. Exceptionalism: what’s unique, special or different about the internet? A single answer: The rise of online word of mouth. The universe of products, consumer opinions, and competitor claims is all available for you to see.

The internet enables consumer perspectives to reach beyond traditional physical limits. I’m happy to give opinions to anyone who’ll listen. Offline that’s limited by my social network. Online it’s not. Here is where intermediaries come in, mediating the development and dissemination of word of mouth, such as epinions.com, which had the business model of soliciting and packaging opinions. eBay feedback forums, bizrate/shopzilla’s merchant ratings, etc. These businesses propagate information throughout the network.

The effect has changed the paradigm for TM owners, who used to have a significant degree of control over the content shaping consumer brand perceptions, but online they’re losing that control. There’s increased competition for consumer attention online. The main way people find information is the first 20 spots on a search results page – that’s the ultimate channel for brand information in the marketplace – 10 results of editorial content (algorithmically generated) and 10 paid ads; almost no one goes beyond the first page. Even within this, the top result gets much more traffic than the 10^th. TM owners now have to compete with traditional players, like competitors, but also new people like their own affiliates, consumer opinions, resellers. It’s a zero-sum game, and one that a consumer could win. Consumer opinions now follow the TM owner in a much more direct fashion. (My search for “coca cola” pulled up killercoke.org as the fourth result, discussing Coke’s practices in Colombia.)

So now TM owners have to compete even to have their own say. Goldman thinks this is a good thing. Consumers derive value from peer opinions and value it more than other influences. TM owners should also be accountable for their actions; reputation following them is a good thing because it means accountability. That’s what we need for the market to work!

Problem: TM law can be used to excise online word of mouth – which is another difference from the offline world. If they can, TM owners will get rid of the bad stuff and leave the praise, creating a lopsided information economy. This doesn’t work as a practical matter, since uniformly positive reviews aren’t credible – we reduce the cognitive authority assigned to it. (My reading of the literature suggests people don’t reduce the influence sufficiently to make it worthless – and they certainly don’t backlash – so it’s still worth it for TM owners to try.)

Normatively: We should ensure TM law fosters online word of mouth.

Implications for TM law: Trademark “use in commerce” should not be just a jurisdictional requirement but should have substantive bite. The statutory definition is not particularly logical but none of the interpretations he’s seen are either. Once we acknowledge there’s this thing “use in commerce” we’re not really sure what it means, because it’s hard to divide the world between commercial and noncommercial. In this case, online word of mouth – consumers talking to each other – doesn’t fit neatly, because (1) it can have commercial effect and is part of the marketplace mechanism; (2) word of mouth producers may legitimately seek compensation for what they do, like having a server and hosting a site, just as journalists get compensated for their labor even if they’re talking about brands; (3) intermediaries make commercial referential uses, talking about the brand for their own profit.

Proposed solution: “Use in commerce” should mean that the commercial vendor is using the trademark as their own source identifier, as distinguished from others in the marketplace. If they’re doing anything else, it’s not a use in commerce.

Likely confusion: Initial interest confusion, a sledgehammer against things that include online word of mouth; hugely misguided. The problem is that courts are still struggling with difference between confusion about the product source and confusion about the content source. Usually IIC online is about content source, not product source. IIC has led us astray; we need to go back to product source confusion. Presumption: there should be no likelihood of product source confusion from online word of mouth.

Fair use: Goldman sees it as fairly narrow. When people use fair use to exculpate online word of mouth, it’s a bailout for something that wasn’t a use in commerce in the first place –nominative fair use is too narrow, and not recognized in some places.

Lemley: Beebe’s study is that empirically, only the court’s determination of intent matters. In the context of new uses, that’s enormously manipulable based on the court’s perception of free riding. Lemley worries about the ability of the confusion test to deal with this because we lack a normative baseline.

A: Agree: that’s why definition of use in commerce is our savior, because it takes care of this without fighting over intent.

Mark McKenna: Under current law, we define use as a TM as any use that indicate source, which includes sponsorship or affiliation, which means there’s really no requirement – the question is whether consumers think the entity talking about your product is licensed to do so. If TM owners can convince consumers this is true, you’re screwed.

A: Sure, but use in commerce can fix this.

McKenna: But if “use as a mark” includes “use to identify yourself” and consumers think that talking about a brand indicates that you are affiliated with the TM owner, then we’re back to the problem.

A: Nah, use in commerce can cut this off. (I think this is not entirely responsive. What does affiliation, sponsorship, etc. mean under this definition? Isn’t it gone or subsumed into source and therefore meaningless?)

Works in progress: Jacqueline Lipton

Jacqueline Lipton, Gripe Sites and Parodies in Cyberspace: Framing a Domain Name Policy to Facilitate Free Speech: This paper is part of a larger project on domain name policy. We ended up with the UDRP protecting TM owners and no one else. There are multiple kinds of domain name disputes, from celebrities to geographical indications to political fraud. ACPA does a really good job of what it was intended to do for the most part, as does the UDRP, but many disputes that arise don’t really fall within that sphere, and we need new ways to approach them.

Today’s topic: Parody sites and gripe sites, whether the registrant uses a TM as the domain name or in the domain name. Margareth Barrett has a similar paper; both take a more extreme view than many in the field would. Where the papers differ is Barrett’s belief that the courts can balance the TM factors; Lipton likes legislation and guidelines, even if that’s not all that realistic in the US.

Lipton uses a hypothetical, set forth fully in the paper, about law students who start a gripe site about a local restaurant after a bad meal there. Various tweaks – someone leaves a comment on the site recommending a rival steak house; the student rejects an offer of $500 but says “It would take a lot more than that”; etc. – raise interesting questions of commercial use, the applicability of ACPA, the UDRP, infringement, and dilution causes of action.

Courts and commentators have struggled with First Amendment implications. Lipton points out that the only real First Amendment protection in dilution is the “commercial speech” limitation – but many courts have found commercial speech more broadly than Kozinski did in Mattel. There’s nothing specific in the UDRP about protecting values other than TM values. Some arbitrators think free speech must be a legitimate use, but limit that to anything other than “trademark.com” – with no real explanation of why that should be. Is there something special about trademark.com as opposed to trademark.net or trademarksucks.com? Should we presume that’s reserved to a legitimate TM owner? Denicola thinks that’s right, Lipton is not so sure.

We should have a series of presumptions – if they have to be judicially constructed, that’s okay – that would be a sort of free speech zone. If someone other than the TM owner uses trademarksucks.com or a new critical TLD like .sux, we should presume they have the right to do so. Current law evolved without sufficient thought given to the critical and parodic sites.

Bruce Boyden: How significant is the problem? Since PETA, the gripe site people have been doing pretty well. Am I wrong about the trend?

A: There is a trend, but there are still cases, and arbitrations in particular are still finding against gripe and parody sites.

Q: Under recent UDRP decisions, is there more of a tendency to favor the TM owner against the gripe site compared to in recent federal cases, and if so why?

A: Her concern about the UDRP is that it has effectively global reach, and arbitrators defer to traditional TM law. Arbitrations are more deferential to TM owners, and the reasoning is murkier. There is more of a mess in the cheap, inexpensive system, which is probably only to be expected.

Mike Madison: Other areas outside of domain names where law has come up with a good guidelines or standards based way of accounting for First Amendment interests. We end up using case-by-case balancing in copyright, defamation, media regulation, and similar cases.

A: What we need is a TM-protective zone and anything outside it will be presumptively not TM infringement.

Eddan Katz: What about multilingual domain name issues? Coca Cola in Arabic or Chinese characters – should the TM owner have that right? Where are the boundaries?

A: That’s a great addition to the project. There’s a new .eu domain that needs policies worked out for it. The boundary has to exist to have clear rules and guidelines. The alternative is to have courts do it case by case, which has costs of its own. .com was intended to be a commercial space, but it became a default and now lots of people have .com who aren’t businesses (comment: like me), so Denicola’s proposed rule takes .com back to its foundations.

Q: What about political candidates doing strategic registration of opponents’ names?

A: They’re generally not TM cases, which is why TM doesn’t work. Has a piece on this in Wake Forest. She likes the California cyberfraud statute about misleading in an election campaign context.

Works in progress: Frank Pasquale

Frank Pasquale, Egalitarian Principles for Copyright: Economic analyses that are underrecognized in current IP scholarship. There may be implications for patent, though he focuses here on TM and copyright.

His project borrows insights from environmental economics: models of the commons, of scarcity, and of consumption. How IP can be used as a tool in a competition for position or for scarce resources.

Not all information is created equal. We should give good information, like categorization information (Google) special privileges in the copyright system. Some patents are bad, spam is bad, primarily position-enhancing information. Classic example of positional good: People in the front row at a football game who stand up to get a better view, forcing everyone behind them to stand up. Social harms: if you need a great suit to get a relative advantage, you spend money on a suit, but then everyone else needs to do so too and you end up with a social loss and no one better off. (The Star-Bellied Sneetches and the Plain-Bellied Sneetches.) Information can be this way, as with proprietary databases that help lawyers do forum-shopping, producing advantages against underfinanced opponents.

Second concern is not economic but philosophical: this amounts to an unfair commodification of advantage. No social good X should be distributed merely on the basis of characteristic Y – you shouldn’t be able to buy your way into the best college.

Replies to the economic critique: no quantification of positional effects. But there’s very little law resting on direct quantification of costs and benefits. We seek heuristics, rules of thumb, considerations.

There’s an inevitable mixture of intrinsic and positional goods. Parents will always buy tutoring for their kids if they can. But we can discourage this behavior.

How to disaggregate positional value and intrinsic value? Leveling up strategies – we want to make the good accessible to everyone. Leveling down – to no one, when we just want to render valueless the primarily position-enhancing information. The two bad things to avoid are wasteful competition (where lots of people are using the PPEI) and unfair competition of advantage (where inequality is persistent).

Poster child for PPEI: Ivywise Application Packaging, which produced Opal Mehta and got the author of same into Harvard.

The value of a positional advantage is inversely correlated to its pervasiveness, which is directly correlated to its equity. As the advantage becomes more widespread, it becomes less valuable, but may still incur the same cost/effort. So why not short-circuit the process? South Korea tried to ban private test prep, but it was overturned by the supreme court. This is an example of where you want more leveling up.

Leveling down: Mandatory disclosure of use and cost of services – for college applications, court filings, and the like.

Level ing up: Increasing access by decreasing copyright protections. Test prep questions and material shouldn’t be copyrightable. There will still be incentives to pay for the teaching, but spreading the questions will make help more accessible.

Another option: Encouraging price discrimination to help people who couldn’t otherwise afford it – but a very aggressive regulatory posture to make sure the price discrimination is reaching people. Encourage the socially valuable, education aspects of services like realistic college counseling, but share the competitive benefits.

At least with PPEI, the “more innovation” that we’d lose from decreasing protection isn’t that valuable.

Josh Sarnoff: John Adams has relevant discussion about leveling up (universal public education) and leveling down (prohibiting ownership of knowledge, though not copyright itself, and relying on reputational interests).

A: British Parliament once introduced a bill making it a crime for a book not to have an index!

Mark Lemley: Relatively paternalistic view of how people should be allowed to spend their money. This makes me nervous. At a minimum, you need to think about administrability of government involvement. Which trademarks and copyrights are to be extirpated?

Michael Carroll: It’s hard, but at least we can think about parts of the system where IP rights aren’t functioning as they’re supposed to.

A: Expanded fair use for test prep is interesting – some great legislative proposals about truth in testing, but copyright owners claimed that tests were trade secrets.

Carroll: Having represented ETS, we did use copyright to go after the test prep industry for copying tests.

Michael Madison: You could distinguish based on how concretely the good itself is defined. How do people use and value particular information?

Q: The law & econ question is whether we should encourage production of information that will divide us instead of unite us – this question comes up in contract law, whether information is something that will just allow one party to extract rent from the other. Do we let people keep information private, so they’re encouraged to invest? At what point does that become socially harmful?

Glynn Lunney: Obvious applications in trade secret law of how much you have to invest in protecting the secret. In terms of social value, luxury TM goods can be a form of negative externality: you buy the expensive car so you can show others what they don’t have. But we as a society seem to value that ability. We have manmade diamonds that are better quality than natural, but the natural diamonds command more in the market. If owning a diamond is just a signal of how much money someone has, why not just tax 75% of the price, since who cares whether the seller or the government has that money?

A: Why not just have people put a sign on themselves saying “I’m rich”? It only works because it’s indirect.

Sarnoff: Adams says the aristocracy is necessary to the government and the people, so he was not in favor of leveling property though he was in favor of leveling pure information.

A: Status signalling is an important topic and there’s good historical work on it.

Works in progress: Peter Winn

Peter Winn, Unauthorized Access, Computer Trespass and Privacy: Really interesting, though beyond my core field.

Winn is examining the CFAA and Stored Communications Act (federal), plus state and foreign unauthorized access statutes. What is the meaning of “unauthorized access”? The problem is one of overbreadth, as also occurs with concepts of trespass to chattels in cyberspace.

Two approaches: One way is to say that the legislatures were trying to address hackers, outsiders breaking into a computer, like burglary. The other way is to say that the statutes are intended to protect information in the computer system (which means that insiders can violate the law too, like a houseguest who pockets the silver). Many courts have interpreted the statutes the second way.

Very first case in which hacking was addressed, U.S. v. Morris: Morris was a grad student who created one of the first internet worms in the late 1980s. Morris was prosecuted under the 1986 version of the CFAA; Morris’s defense was that as a grad student at Cornell he had privileges to access the internet, and he designed his program to access the other computers in that system just as they were designed to be accessed – just as the code allowed him to do.

It’s a stupid argument at some level: that wasn’t the way they were designed to be accessed, because the programmers didn’t anticipate what he was doing. But from the machine’s viewpoint, there wasn’t any difference. The Second Circuit rejects Morris’s argument. You have to reference something other than the way that the code was designed. The Second Circuit was referencing norms – if Morris had called up the owners of the hardware and said what he planned to do, they probably would have told him “no.”

If we’re working with unwritten norms in the core cases, we get the expansionist cases – mostly civil cases – now companies sue former employees for taking customer lists with them. We’re no longer in the world of hacking, but courts are still applying intent standards and finding unauthorized access. Also with scraping a competitor’s website. Most interesting case: A law firm that issues an overbroad subpoena of an ISP where the opposing party has its email server, and gets the ISP to give them all the email without notice to the opposing party. After the firm was sanctioned, the employees whose emails were obtained filed suit against the person who issued the subpoena, and the 9^th Circuit rules they have a cause of action under the CFAA and under the Stored Communications Act.

When courts go beyond hacking, they’re using the idea of trespass in the internet context. Using it against employees subverting confidentiality, hacking, competitors all fits this model. But the employees whose email was searched have no possessory interest in the email stored at the ISP, nor do they have a leasehold, since the lease is their employers’. This is a case about rights in pure information. There’s a long line of 4^th Amendment cases recognizing standing for a constitutional trespass based on a purely privacy/information-based interest in not being surveilled, and the 9^th Circuit borrows from that line.

Why don’t we have a nightmare of overprotection in the digital world? Indeed, a lot of people who’ve written about this suggest that we’re shutting down the free flow of information. The solution: (1) We’re applying common law trespass to this idea of unauthorized access; let’s look at the common law, not trespass to chattels but trespass to real property and the time when trespass was a general cause of action for what we now call tort. The trespass to real property cases do not allow owners to dictate all terms. Particularly with property like inns that is private but open to the public, it is really very hard to bring a trespass case – a person who goes into the inn intending not to pay for supper is not a trespasser. A federal agent who goes on to private land where the owner is illegally selling liquor out the back is not a trespasser. If there’s a community practice of presumptive consent, then people can come hunt and fish on private land. Courts try to balance property and reasonable, appropriate uses of the property by the public. The common law wants to have it both ways. Posner decides a case about ABC, which promised not to do ambush journalism or hidden cameras and does just that to an opthalmologist who sues for trespass. Even in the face of fraud, there’s no trespass because fraud doesn’t go to the basic interest trespass was designed to protect.

(2) The need to reference something else: if you bring a trespass case, you’re referencing the system of property rights. Courts in expansionist cases are not simply allowing the business owner or the email senders to sue because they don’t like the defendant, but rather because they can find specific common-law torts and trespass is used to translate those into internet law. The proposal to shut down common-law rulemaking is to restrict the meaning of unauthorized access to code-based circumvention: regulation by nerds. Winn thinks instead that we should let the courts work it out. New Zealand has a very tight definition that doesn’t allow common-law rulemaking. But other Commonwealth/US courts are doing a good job of balancing the interests. If you ratchet it down to code-based circumvention, you’re just transferring the decision about what’s authorized to the nerds from the judges. And it means you can’t protect the privacy of third parties as with the email case, because only the owner has the incentives to create a code-based protection. (I’m not sure I buy this, since the owner has incentives to create code-based protections in order to make its hosting more attractive to consumers. And I’m far from certain that it’s a good idea to make competitors liable for scraping a site when it’s not trespass for them to acquire a printed catalog with the same information, even if they acquire that catalog by subterfuge. In other words, “balancing the interests” gives courts another way to go after what they think is unjustified free riding, which they are all too willing to do.)

Winn points out that there are alternate privacy protection regimes: you can put duties on information collectors to manage information responsibly. So we don’t need to do it this way. Right now, he says, third parties don’t have a cause of action against the privacy-violator along with the info-collector who allowed privacy to be violated, but I would find that a much better solution to the problem he cares about than relying on the flexibility of the trespass doctrine, which as he acknowledges is not traditionally about protecting third parties’ interests.

Works in progress: Scott Boone

Scott Boone, On Virtual Property and Copyright: His followup project to this, incidentally, is called “Han Shot First!” Which is great.

As we move from physical copies to digital copies, the personal property part of the balance between copyrights and rights in the copy has dropped out of the equation, as if all there is is the copyright: there aren’t copies any more. Boone is seeking a return of the property balance. Can the idea of “virtual property” bring personal property back into the equation?

By virtual property, he means property rights in virtual objects. There has been some previous work on virtual property, and more work on property concepts like trespass applied online.

Fairfield’s formulation of virtual property: it is rivalous, persistent over time, and interconnected (including across servers). Examples: email addresses/accounts, URLs, websites, chat rooms, virtual world objects.

How does this relate to copyright? Look at virtual world objects, in MMOGs where people interact in a persistent, 3D environment. World of Warcraft has 7 million subscribers, maybe 3-4 million unique players. The objects in these games, like weapons or houses, may or may not be subject to personal property rights. There’s a huge real-money market for transfers of these objects. But who really owns the rights? Game producers sometimes attempt to shut down outside markets. Korea and China have granted some sort of rights to players even in the presence of contractual agreements.

That transfer of a virtual object from one player to another doesn’t violate any copyright rights. There’s no copy made. (I really wonder about this. If RAM copies count, why doesn’t this?) It’s not a distribution – it’s either first sale or not a distribution in the first place.

There’s no inherent conflict between virtual property and copyright, at least no more conflict between personal property rights in a traditional copy and the copyright. There’s no experiential difference between virtual world objects and digital copies. Thus, if virtual world objects are rivalrous, why not treat them the same as physical property? Rivalrousness, or the lack thereof, exists at the level of code and code can create rivalrous virtual world objects – even if it is not likely with the current state of code. (So, technological measures inherent in a game can prevent the duplication that made copyright owners so leery of “digital first sale.” But I imagine that copyright owners would fear circumvention – not to mention the real objection, which is of course that they don’t like first sale in the first place, and if they can avoid injecting it in a new world order they will do so.)

Questions for discussion: At what point is a digital copy sufficiently rivalrous to merit property rights? Is a publicly ordered property system preferable to a privately ordered system of contracts and technology?

Q: Is there really no copy made in a transfer from one to another?

A: When Cartman sells a hammer to Stan, there’s no additional copy made, because Cartman ceases to have that hammer and Stan now does. The code that manifests the hammer exists on every player’s computer. The ability to use the hammer is the authorization that’s passed, which exists on the game company’s server. Unlike sending an mp3, no additional copy exists as the result of a transfer. Rather, just rights to use and control and exclude are transferred – built in at the code level.

Eric Goldman: What about the contract?

A: That’s the public ordering/private ordering choice. Contracts do cover all these instances, so the virtual property issue may never come up.

Eric: Authorization to create a derivative work – that’s a contract that shapes the scope of rights in a derivative work. We have a foundation that’s contract based and assets are created on that foundation. But the contract is still king.

A: It’s true that many game corporations have stepped in asserting their contracts. But we could say that game companies couldn’t put restraints on in-game alienation.

Eric: We already have an example of a legal system where the contract trumps (or channels) the rights in copyright – derivative works. What’s so special about virtual property?

A: The default rule is different, §109A. Digital property has the potential to be the same as physical, which makes the default rule of who has control that of first sale. That’s the difference – the non-contractual allocation is different than that for the right to control derivative works.

Beth Winston: Dan Burk has some articles on comparing seed and genetic use restrictions and the incorporation of the contract into the tech as compared to DRM.

A: The terminator gene is another place where we see the same questions arising.

Bruce Boyden: Within the game, the game designer has created some property-like aspects to objects within the game. But the game designer could change that.

A: And Second Life has!

Bruce: So it’s completely within the game designer’s choice – the question is who do you have property rights against. It seems unlikely that you have rights you can enforce against the game designer – and if you did, as the result of your participation, then Jack Balkin’s article on virtual liberty suggests there are free speech implications. (Boone and Boyden dispute whether the speech interest supports player property or not – Boyden says that player ownership would harm the game designer’s ability to speak, whereas Boone points out that player ownership would improve the player’s ability to speak. There is a tradeoff, but my reading of Balkin leans more toward Boone’s, since the player corresponds here to the citizen and the game designer to the state, whose “free speech” interests are at best fraught and instrumental rather than foundational and inherent.)

A: Boone hasn’t answered whether property rights added to contract and tech would be a good idea. If you start to go down the property route, you may start to dictate the technology, which could be a strike against virtual property.

Brett Frischmann: What’s the basic policy payoff? Should government intervene?

A: If digital copies are rivalrous, we should look at bringing personal property rights back into the discussion.

Brett: So you’re indirectly making the case to improve the infrastructure for DRM, since rendering digital objects rivalrous would allow this change.

A: I’m not advocating this change, but I want to consider it if the technology moves in that direction – property could moderate the effects of DRM.

(Seems to me that despite copyright owners’ fetish for control, game property is already at least as rivalrous as, say, music CDs. Those are subject to first sale even though they’re easy to copy and often copied – whereas game property, as far as I have heard, is only occasionally hacked.)

Works in Progress: Peter Yu and David Welkowitz on Human Rights and IP

Peter Yu, Reconceptualizing Intellectual Property Rights in the Human Rights Framework: Conflict or coexistence between human rights and IP? If there’s conflict, should human rights take precedence? Or are there enough safeguards in IP that there is no conflict? See, e.g., U.S. cases finding that copyright accommodates the First Amendment with built-in limits. Is IP a human right? This is connected to whether IP can be easily analogized to real property.

Lessons from basic human rights documents: (1) IP has always been controversial within international treaties. (2) There are multiple dynamics – the Cold War leads communist countries to reject protections for private property; hesitations about using UNESCO for copyright when there’s a plan to push for broader protections in a separate proceeding later (Berne). (3) Many of today’s issues have been explored in the 1950s, and we’re rejuvenating old, forgotten debates.

Moral interest: Protecting integrity or attribution. A lot of people who are pro-access to knowledge want to use human rights to preserve access, but a human rights regime would offer strong moral protection, working against their ultimate aims.

Material interest: Does right to property always mean right to private property? The treaties are not clear – property can exist as a right of a group. A lot of the communist countries wouldn’t agree to private property protections, so the treaties don’t specify. Other considerations: remuneration for scientists doing research, when many countries didn’t want to have to give them individual patent rights.

Treaties focus on “just remuneration for labor,” and also mention protection of the moral interest. They aren’t protecting full economic interests in IP. Property rules or liability rules would be acceptable under these treaties.

We could have a human-rights-based compulsory license, combined with just remuneration. But we assume that people have money to pay for a license. That might not be true. Freedom of the press belongs to those who own one? Also, when we emphasize compensation, we may overemphasize the material interest over the moral interest. Perhaps appropriate credit or appropriate disclaimers would be more appropriate.

Alternative: a “core minimum” approach. In human rights, there are resource constraints, and lots of countries can’t fulfill their obligations, but they may nonetheless not violate human rights treaties. The core minimum approach says that you have to devote the resources you have to your obligations. Drawbacks: if you focus on core and keep on expanding, no guarantee you can figure out where the boundaries are. We also don’t know what the priorities are: countries need a lot of different things, not just IP. Finally, core minimum means that we don’t look at the relationships between the various human rights. If you can apply core minimum to IP, why not to the right to health, education, or food? Why not use that to expand IP given that a core minimum protection for education etc. exists?

Progressive utilization approach: Need to keep expanding protection based on available resources. Yu proposes seeing it as a sustainable development regime: use IP to empower people and progressively expand rights. Three-step balancing: look within an IP provision first, then at general IP provisions, then at overall human rights obligations.

Challenges: (1) Human rights ratchet. Dangerous ideas about property will be imported into IP. You’re elevating the importance of IP to that of human rights, and that is dangerous. Yu responds: There’s always limited protection, and countries need not protect to the limit. Also, when we talk about protecting moral and material interests in IP, we are not talking about importing the whole US apparatus. Things like work for hire and rights to employee inventions are not human rights because they apply to corporations.

(2) Institutional capture. IP rights owners are more likely to have resources to manipulate the human rights community. But those corporations are already part of the discussion, using NGOs to influence process. Also, Yu thinks the human rights regime is robust and could fight against capture. And if we can get big players to think about food, access, and development, that’s a good thing. Language used in a human rights forum can spill over into IP, as after Doha when people are talking about rights to health and about access to knowledge.

(3) Cultural relativism. Are human rights just western imperialism? Yu doesn’t think so – others participated in forming treaties. Human rights aren’t fixed. Protection for indigenous groups and women has expanded. Also, the US is opposed to putting IP with human rights, as are other western countries. Is the US stupid? If not, maybe developing countries should persist in pushing for linkage.

Don Chisum: The French Revolution treated authors’ rights as basic rights, so there’s a history in the west as well.

David Welkowitz, Privatizing of Human Rights? The Application of the European Convention on Human Rights to Intellectual Property: A lot of attention given to human rights in the IP context has been about conflict – how can we limit IP by using human rights terms? Using public law to limit private law. From an outsider perspective, we may be entering a situation in which private actors with IP interests, including nontraditional IP holders, may be seeking human rights as a means to create and enforce IP rights.

Why the European Convention? It’s a good starting point because it’s fairly widely adhered to, 46 countries strong. It also has an accessible jurisprudence: judicial interpretations, cases, all the things US lawyers find familiar. Ironic, because most countries involved in the Convention are civil law countries. But case law is infiltrating into their system.

Article 8, privacy.
Article 10, freedom of expression: often invoked as counterweight to IP
Article 1 of the first protocol: Protection of property – every natural or legal person is entitled to the peaceful enjoyment of his possessions

Could argue that article 14’s antidiscrimination provisions also apply to IP, if for example trademark law discriminates against celebrities.

Von Hanover v. Germany, ECHR 2004 and Anheuser-Busch v. Portugal, 2006.

Von Hanover is the ruling family of Monaco. Princess Caroline is the object of a great deal of attention, frequently photographed in public and in private. This case involved a series of photos published primarily in Germany. Technically she sued under French law, but she sued in Germany. Germany’s Basic Law has a right to privacy. After not much success below, she reached the federal constitutional court. The German court found that she was a public figure and has to tolerate a certain amount of interference with privacy. There is a recognizable public interest in receiving information about her as a celebrity, even if she’s not a ruler and even if the public is mainly interested in entertainment. The court enjoined publication in France under France’s far more protective laws, but under German law she was entitled to have some pictures suppressed (pictures with her children, and pictures of her in a “secluded” place like a restaurant) but not pictures of her in public – even when a telephoto lens had been used to get the pictures. In the balancing the court did, it specifically said the German press was entitled to freedom to maneuver, exercise of editorial judgment.

In the ECHR, Caroline succeeded. The court balanced article 8 against 10. Court didn’t agree with the emphasis the German court placed on her status as public figure. She has no authority and is basically ceremonial, and the pictures weren’t taken as she carried out official duties. And the public has a minimal interest in entertainment value.

In awarding Caroline a victory, the ECHR said Germany has an affirmative obligation to provide her with a civil tort action to protect this right. Implications for IP: we’re used to a private enforcement regime, but this implies the state needs to back it.

Anheuser-Busch v. Portugal: Dispute over the rights to use Budweiser all over the world. The Czech company claims a protectable GI, and A-B claims trademark rights. A-B applied to register in Portugal in 1981, and the Czech company opposed. Negotiations ensued. After 8 years, A-B moves to cancel the Czech company’s GI rights, and succeeds. The TM office of Portugal is then directed to grant A-B a registration. By now it’s in the 1990s. The Czech company appeals based on a 1986 treaty between Portugal and Czechoslovakia protecting GIs; the appeal succeeds.

The ECHR didn’t uphold A-B’s claim, but the case is still of interest. The court agreed that IP rights are “possessions” under Article 1 of the first protocol. The rejection of the claim was narrow: since A-B had applied in 1981 and was immediately opposed, it was really one long application process, in which A-B had merely an expectancy that wasn’t sufficient for a possession. There was a dissent that would have allowed A-B’s claim even though the state has leeway to regulate under article 1 – a proportionality requirement. The dissent thought that invalidating the registration was disproportionate, given that the treaty came after the application was filed.

Welkowitz saw a lack of deference on the ECHR for the decisions of member states. In the IP context, lack of deference allows the court to create a common-law regime, ignoring a country’s own attempts to reach a balance.

Possibilities: A robust right of publicity? Based in dignity interests, perhaps. The alternative is to say that people have property rights in their images. Unauthorized use might be considered interference with possession, creating a private right to prevent such use. There is a contrary decision in the ECHR, in which a beer bottler used a portrait of someone’s grandfather on the beer and the court refused to recognize a right – based largely on the fact that the portrait had been given to a museum and displayed publicly. The court also said that there was no indignity, because the grandfather had been connected with the brewery. He’d been dead for 80 years, which creates further products for the claimant.

TM dilution: It’s possible to see TM dilution as a right under article 1, particularly uses that might make a mark generic – even in a movie or book.

On the other side: Rights of indigenous peoples. If traditional folklore is a property right, that might be a source of an article 1 right against letting outsiders get copyrights or patents on traditional material – individuals can have standing under the law.

Who’s likely to be advantaged? Celebrities are advantaged by being celebrities, famous for being famous rather than famous for a function. Complications for the EU: If the ECJ doesn’t agree with the ECHR, there’s no obvious mechanism for resolving the problem. Also could complicate international IP negotiations, if the rules are enforceable in court decisions and the US has to pay attention to the ECHR.

Will it end up cheapening human rights by injecting too much commercial value into the human rights discourse?

Works in progress: Michael Meurer

Michael Meurer, If You Can’t Tell the Boundaries It Ain’t Property: The project comes from a book by Meurer and James Bessen, Do Patents Work? The chapter is available at the conference website, and is worth reading if you’re interested in the subject.

There is a huge notice problem in patents, varying across types of technology. Claims can be submerged for long periods, amended during the application process without publication, and otherwise hidden. Even when the claims are public, it’s hard to figure out what’s covered, and most businesses don’t bother to try. As opposed to title insurance and E&O insurance for copyright infringement, both of which indicate the existence of mature and predictable markets, patent infringement insurance is essentially unavailable because it’s so unpredictable.

Much infringement is inadvertent. Defendants tend to be large and to invest more in R&D than plaintiffs – in fact, the hazard of patent litigation increases with more R&D. Pirates would do better by increasing R&D and hiding their tracks; this is further evidence that independent invention drives most patent lawsuits, as with the Blackberry. Parties are often in different industries and patent in different technology classes.

The harm of notice failures is greater than the harm of low-quality patents.

Structural features of the relevant industry to some extent drive patent prosecution. Maybe with more valuable technologies drive investment in patenting, but he thinks the causation flows the other way – the cost of defining clear rights changes the property people try to stake out.

Mark McKenna: What about protection by other means? Trade secret, copyright (for software), trademark.

A: That may affect the differences between industries. We may be giving more rights where they’re least needed.

Josh Sarnoff: Blackstone saw the difference between protecting real property, where possession helped give notice, and protecting ideas with embodiments. Abstract claims are used to control presently unanticipated potentials, and that’s just wrong.

A: Software is interesting as a general-purpose technology – along with business methods. There’s an abstractness to the invention which generates more abstract claim language.

Jay Kesan: Your analysis suggests that the only good patentee is one who invests a lot in R&D. You privilege technologies that require a lot of investment instead of the tiny inventor.

A: Independent inventors, nonpublic firms, and universities: We can estimate the patent premium for those different groups; for some of them (university profs) there is essentially no risk. The patent premium for small guys is smaller than the patent premiums for big firms, but the costs are also lower – there’s a net benefit for small guys, whereas there’s a net tax for big companies. Small guys will make most of their money by licensing. Still, improving notice quality would be good for small guys (by enabling licensing?).

Mark Lemley: What is to be done?

A: A rigorous indefiniteness doctrine. If we can’t make the notice function work and if patent isn’t really property, then we shouldn’t have strong injunctive remedies – and this could be industry-specific, depending on how easy to define technologies it is per industry. Even in pharma, the value of patents has declined over time, perhaps because of new problems with biotech.

My comment: I'm a little disturbed by the idea that the licensing thicket that exists in movies is a model for a well-functioning system. Right now, until insurers start accepting the documentary filmmakers' fair use principles, you can get insurance against copyright infringement if and only if you rigorously license or cleanse your film of obvious references to identifiable copyrighted works. This is hardly an ideal system.

The orphan works problem is a separate indication that notice is not exactly working perfectly in copyright. Moreover, the insurance system that works so well for Paramount works not at all for individual/amateur film productions, which now risk discovery by copyright owners if they're posted on YouTube.

What the E&O market shows is that the large-scale commercial film industry can produce works by editing and distorting their products more easily than a big tech firm can avoid patent infringement. But that's hardly a recommendation.

Works in progress: Mark Lemley

Note: I'm not a patent type. The shorthand flew fast in this discussion, and I couldn't necessarily unpack it. Lemley's paper is available at the conference website.

Mark Lemley, The (Unnoticed) Death of the Doctrine of Equivalents: Unrelated story: The CEO of Research in Motion, the company that just paid $612 million on a patent troll, just bought the Pittsburgh Penguins for less than a third of that.

Perceived abuse of the doctrine of equivalents was the troll phenomenon of the 1990s. All patent owners were asserting doctrine of equivalents, swallowing the rule of literal infringement. Courts then created a variety of rules to restrict the doctrine. Most significant and highest profile, prosecution history estoppel limits the doctrine of equivalents. The Supreme Court applies a “foreseeable bar” of surrender of rights under the doctrine.

Great test of how patent world reacts to exogenous legal shocks – 3 different rules for the doctrine of equivalents. Thus Lemley & colleague collected data for the 18 months before and after the 3 relevant opinions, which created a set of over 400 cases. Of course there are limits on using only written cases, but he doesn’t think they’re systematic problems in this data set.

(1) The changes in the rule in Festo had no statistically significant effect on virtually any set of outcomes, even in cases of prosecution history estoppel based on amendments and not arguments. (2) The reason is that by 1998 the doctrine of equivalents was already dead. Patent owners almost never win on equivalents.

Total win rate on doctrine of equivalents is 24%, small compared to theory of efficient litigation and small compared to results in patent cases, where patentees win more than half the time. Somewhat more likely to win in district court (31%), 17% in the Federal Circuit, but that’s an artifact of pre-Festo district courts – they learned from the Federal Circuit. Two-thirds of these wins are surviving summary judgment. Defendants win summary judgment 79% of the time. Less than 10% of cases actually involve final patentee wins.

Subset of cases involving prosecution history estoppel: plaintiffs win 14% pre-Festo, 31% mid-Festo, 33% after Festo. Not statistically significant, and radically counterintuitive. Couldn’t more people be raising bad prosecution history estoppel defenses because Festo encouraged them? No change in absolute number of cases filed. Also, most of these cases had been pending for a while when these cases happened. Finally, the share of cases talking about amendment prosecution history estoppel actually dropped.

Industry-specific differences show up very strongly in the number of cases. Mechanical inventions: 33% overall universe, 62% of the cases using the doctrine. Software: 8% overall, 22% equivalents. Electronics, 8%, 20%. Chemical and biological areas are underrepresented. The doctrine may simply work better in sets of claims that are harder to map to products. But there is no effect of industry on outcomes.

Also tested doctrinal formulations – it turns out accused infringers are happy with the “all elements” rule – patentees win 18% of the time when courts use that formulation, but more often with “function, way, result” tests.

Means plus function won’t help. The patentee won 5% of means plus function cases.

Why did the doctrine of equivalents die? Hypothesis: Markman. Once the Court decided that claim construction was a question of law, trial courts devoted a lot of time and energy to analyzing the literal language, and were usually in a position to grant sj one way or another on literal infringement. If they didn’t reject the doctrine of equivalents, they’d then have to hand the whole thing over to the jury to undo their hard work. Thus they were inclined to rule against the patentee. Unintended consequences of one area of law for another separate area.

Don Chisum: The doctrine may be limping, but it’s not dead. Patent owners and their contingent-fee lawyers look at a Blackberry case and think a 5% chance of winning looks pretty good.

A: Absolutely – it’s hyperbole. “Serious injury” would place at a lesser law review, but perhaps we should change the title now.

Michael Meurer: Empirical work on courts claimed that the doctrine was important to patentee-friendliness in the 1980s.

A: We don’t have the data, but we’re looking at the early 1990s. What we can definitely reject is the idea that the doctrine continues in strength through the late 1990s.

Jay Kesan: In claim construction, you’ve already interpreted the file wrapper and the rest of the history. So you’ve already decided the estoppel question, in a way.

A: But we see that patentees lose cases not just about estoppel, but about any issue of equivalents, including “all elements.” Your theory may be at work, but it can’t be the only reason.

Q: Did you look at argument-based estoppel cases?

A: Yes. The results are the same.

Q: Fed. Circuit developed the “clear and unmistakable” doctrine for limiting claims through prosecution history when assessing literal infringement – making literal infringement easier to find, so why do you need equivalents?

A: That’s true, you’re often not going to get to equivalents if you’ve found literal infringement. Some district courts will do both to insulate themselves from appeal, but they’re usually inclined to make the literal infringement result go in the same direction as the equivalents result.

Q: If patentees are doing so well on literal infringement, maybe only the really weak plaintiffs need to try equivalents.

A: That’s possible – if claim construction is really broad already, then equivalents is only invoked by losers. Our data gathering may be able to help resolve this.

Eric Goldman: What about the claims that did win under the doctrine? Were they a stretch, or a good use?

A: Hard to do that in a statistically rigorous way. The quality of the decisions is all over the map, from one sentence to extremely detailed – a plurality of the cases apply more than one test for equivalents. Without a baseline, it’s very hard to tell whether the cases are “good” or “bad.”

Dennis Crouch: The DOE is a settlement tool, providing fuzziness and allowing for compromise. Does that create a problem with the litigation dataset?

A: How would it systematically bias the dataset? Lemley has problems with the Priest-Klein hypothesis in patent law, but to the extent it applies it should push the win rate to 50%, making the results here interesting.

Works in progress: Dan Cahoy

Dan Cahoy, Confronting Myths and Myopia on the Road from Doha: His interest in the topic grew out of discussions about a potential bird flu pandemic, which could have created a massive need for Tamiflu, a patented product. If the patent had been overridden for public health reasons, which countries should have been allowed to do so and what should they have had to pay for it?

What is the purpose of compulsory licensing? Views have changed over the years. Traditionally it was considered a way to promote home industries, as with the requirement that patents be worked in the country. That’s disappeared as a rationale. The primary rationale now is to increase access to public health. Secondary category: remedial licensing to remedy antitrust violations.

Once access is our goal, the mechanism becomes of interest. Most research and commentary concerns what will trigger compulsory licenses, presuming that the compensation will be low. But compensation is an independent variable from what triggers the license, and looking at it might offer some solutions not otherwise apparent.

Myths of remuneration: (1) license must always be something less than market value; (2) a submarket rate is adequate or reasonable (there are problems determining what reasonable compensation is because a drug is usually just part of a company’s portfolio; what perspective are you going to take, the value of what the drug does for the user or the value of the investment to the shareholders?); (3) remedial antitrust licenses offer a model (punishment for badness isn’t the same thing as licensing for greater social good, and antitrust-imposed licenses are artificially low, often zero, because they’re designed to deter bad behavior rather than to affect levels of involvement in the industry).

His proposal: Move toward default regimes preserving innovation incentives but also access where necessary. Suggests default of market compensation, including in LDCs. Compulsory licensing is not a way to lower prices, but to solve emergencies and permit a relief valve. The model would not be that different from infringement. Under most cases, compulsory licenses wouldn’t be favored – only when a patent owner is charging a supercompetitive price.

If there were an industrial development rationale for licensing, that would change things.

How do we maximize access under these conditions? Under TRIPs, any country can do a compulsory license, and adequate remuneration isn’t defined, so it’s relatively arbitrary. We should try a hybrid: In a pandemic that impacts countries across the spectrum, industrialized countries should pay market rates, developing countries should pay development-indexed royalties, while LDCs should pay no royalty. If we really care about access, LDCs are the ones with no real ability to pay (and thus no potential gain to the patent holder). This would at least allow drug companies to do better predictions.

Works in progress: Mark McKenna

Mark McKenna, Trademarks and Contextual Meaning. Past project: consumers historically weren’t at the center of TM law, and we need a new way of attacking TM law other than saying that it’s gone way beyond consumer protection. It has always been producer-centered, with a thin candy coating of consumer protection to make it more palatable. There has been a substantial change in TM law’s focus, but not the one we’ve been thinking about.

Rather, the concept of the interest to be protected has changed. Trade diversion was the core of the old system; the new system is about protecting brand value. Traditionally, the mark’s purpose was matching, mediating a relationship between consumers and producers. It had no value in itself, only in preserving a flow of consumers finding the products they wanted. Schecter comes along and says that TMs have value in themselves, for reasons related to changing marketplaces. “Buick” means cars and related concepts, rather than simply linking car buyers to GM. Thus its ability to mean deserved protection from TM law.

The mark is now the center of a set of associations, with products receding to the background (if not gone). The mark has many meanings, but none necessarily has anything to do with the product. Thus it can attach to anything and can be harmed by distant uses.

Potential harm to the brand seems to be conceived in two ways, beyond trade diversion: (1) use by a third party not under the control of the mark owner that might create new negative associations that would impact the brand in its own markets – (a) could affect ability to recall the mark unaided, (b) could create new substantive associations; (2) use by a third party limits the scope of possible expansion of the brand, which these days is infinite.

In this new conception, all harm is dilution. Even if we’re saying consumers are confused about association in a distant market, the reason we care is that it might change the nature of what the mark means, which is dilution.

This conception ignores contextual elements, which the marketing literature recognizes. In marketing, context is everything. Most TMs have non-TM meaning. Apple isn’t just a computer (it’s a dessert topping and a floor wax!). Non-trademark meanigns compete if we are exposed to a term in a vacuum. At least Schecter was trying to focus on made-up words. Of course we’re not exposed to marks in a vacuum, which is why it’s not that hard to make the right associations and avoid the wrong ones.

When you start looking at context, whether an association will transfer from one product to another depends on fit between the products, the appropriateness of the association in the second context, whether cues are present to activate an association, and the dominance of the brand in its category (the effect of which is somewhat ambiguous). Dominance is a function of what brands you remember when cued with the category. Marketers want category dominance, but there’s some evidence that, regardless of dominance, use of the mark in other markets actually helps recall of the parent by activating its associations.

The tools of TM law map poorly onto these concepts. Strength measures exposure, not relative dominance. Distinctiveness has nothing to do with dominance; there are many dominant TMs that aren’t particularly distinctive in a TM sense – that’s because marketing is about brands, not about marks. Fame in the new statute isn’t defined with regard to market segment. We need some new tools.

Q: Marketing factors look like likelihood of confusion factors – “fit,” cues, etc. – are consumers going to make the association?

A: Except that in a lot of cases the legal analysis would find confusion where the marketing analysis would find no effect on the preexisting brand. Even if consumers think there’s a relation, they just don’t care – but the likelihood of confusion factors say that the plaintiff wins in that case.

Q: Migrating associations over time: brand managers think of brands as elastic and changeable. How does that affect the analysis? There are a wide range of brands – “Betty Crocker” doesn’t need much context to be identified.

A: Yes, there are a wide range, and a bunch of the dominant ones need context!

Q: Note that online, you may just get a contextless search using a TM term. That may affect the analysis.

A: Maybe. (But as questioner Eric Goldman has so persuasively written, lacking context for the mark doesn’t mean that we should infer confusion or dilution – other explanations are a lot more plausible when the term is used in guiding search.

Q: Is this about confusion or dilution?

A: Once you get to sponsorship/affiliation confusion, it’s only a baby step to dilution, so we should be concerned about those kinds too.

Lemley: Once you take the consumer protection out of the equation, why is it that we want to protect brands? There must be some subset of legitimate interests in brand value?

A: There’s a way to write consumer-centric TM law. But descriptively consumers haven’t been and aren’t central. McKenna doesn’t have a great answer yet – part of this is looking at the harm on its own terms. He’s not sure there’s something real left over.

Lemley: Yeah, but without the consumer, all we’ve got is “If value, then right.” (Which we know is bad.)

Q: What about free riding?

A: It’s not a harm to the market. It’s a potential harm to an individual market participant, but not a legitimate basis for law

My question: This objection to dilution makes a lot of sense. But it doesn’t deal with a separate big dilution problem: Insinkerator, Caterpillar, others who huff and puff when there are negative or humorous references to them. That relates to the fact that this presentation didn’t once use the word “tarnishment.” If you buy the associational analysis, are you committed to tarnishment?

A: Hoping no one would ask that. Harm to a brand can occur from acts that change the associations connected with the brand, so tarnishment is a potentially legitimate harm. There are reasons not to provide a legal remedy, such as the First Amendment, but at least it exists.

Q: That goes back to the question of why we care. Even if the brand owner is harmed, what justifies a cause of action? Maybe the answer in blurring is search costs.

A: He’s trying to call the search costs justification into question. The marketing literature says Buick for shoes won’t interfere with search for Buick for cars.

Irene Caboli: What other reason to protect TMs is there other than consumers? Property rights justification – people will make more of whatever they have rights to.

A: Sure, TM is industrial policy and we should speak in those terms.

Q: Is that compatible with the constitution?

A: Commerce clause these days allows a lot of breadth.

Gerard Magliocca: The standard being applied in many cases is not a reasonable consumer standard, but a dumb consumer (15% dumb) standard. You may have to deal with that to get where you want to go.

A: The marketing literature imagines consumers to be pretty dumb.

Q: Even if associations are nonharmful or beneficial, it involves risk. Why should mark owners have to bear the risk? Especially when noncommercial contexts increase risks to marks more these days, TM owners need some help from the commercial side.

A: Good question – relates to Mark Goldman’s point about mark elasticity over time. We may want to allow mark owners to control (to the extent possible).

My question: I can’t make the property rights justification work for brand value, which isn’t about maximizing the number of TMs created (the usual, not-very-good property justification) but about maximizing the associations in a person’s brain. What does it even mean to maximize brand value? How do we compare that to a situation in which there are a lot of brands, each with some value? It seems incoherent to me.

A: He’s trying to take the harms seriously and to talk about whether they’re real. Maybe they won’t justify legal protection, but we need to define them first, with less speculation.