Broken glass: Lanham Act and defamation duel between glass repair companies

Diamond Triumph Auto Glass, Inc. v. Safelite Glass Corp., --- F. Supp. 2d ----, 2006 WL 2129498 (M.D. Pa.)

Diamond and Safelite are competing glass companies. They each repair and replace damaged vehicle glass. Safelite also runs a customer service program for various insurance companies, by which it refers insurance policyholders to Safelite or unaffiliated glass companies. Unaffiliated glass companies have to sign an agreement to be part of this insurance company network; Diamond was part for several years, but then left, partly because it believed Safelite was steering customers to Safelite shops over other network shops.

Diamond brought state and federal claims based on Safelite’s statements to policyholders who called the service program. Diamond argues that Safelite customer service representatives: 1) falsely conveyed in their greetings that they were representatives of the insurance companies; and 2) falsely warned policyholders that they could not guarantee Diamond's pricing or service after it left the network. Diamond also alleged that the CSRs falsely stated that Diamond was not a member of the network, though these statements were not provided by the scripts, and Diamond’s evidence was that they occurred on fewer than twenty occasions. Businesses harmed by isolated disparaging statements don’t have recourse to the Lanham Act and must rely on state law, so the court found that these statements weren’t “commercial advertising or promotion” given that Safelite handled millions of calls during this period and referred over 93,000 claims to Diamond. (As for the state-law claims, the court found that the CSRs could take advantage of the conditional privilege between insured and insurers, as long as Safelite acted without malice, but denied summary judgment on this issue – a victory for separate argument of state and federal claims!)

The court found no literal falsity. When greeting callers, the CSRs would state, "Thank you for calling the ABC Insurance Glass program. This is [name]. How may I help you?" Given Safelite’s agreements with insurers and insurers’ approval of the CSR scripts, the CSRs were affiliated with the insurance companies and authorized to represent themselves as insurance company representatives.

The warnings about Diamond's pricing and services after it left the network were also not literally false. Diamond argued that Safelite's warnings were literally false because it never charged more than the insurance companies could pay, it informed Safelite and the insurance companies that it agreed to their pricing, and it provided its own warranty in its invoices to all customers. The scripts, however, did not represent that Diamond's pricing would exceed the coverage, or that the workmanship would be of lower quality, or Diamond would not provide a warranty. After Diamond terminated its Network Agreement, it had no contractual arrangement with Safelite or the insurance companies and it was free to change its pricing, services, or warranty at any time without notice.

Diamond offered evidence of implicit falsity, which the court rejected. Specifically, Diamond’s expert testified that, based on her research into general consumer behavior, her review of a number of customer depositions, and her review of recorded conversations with CSRs, that the greeting led callers to believe the CSRs were insurance representatives and that the warning on non-network pricing and services caused “uncertainty,” which cased many callers to choose Safelite because it was guaranteed. Without ruling on admissibility, the court found that the expert report wasn’t enough evidence of falsity to survive summary judgment.

The CSRs were insurance representatives, given Safelite’s agreements; to the extent consumers believed the CSRs were insurance company employees rather than authorized third party administrators, there was no evidence this was material.

Independently, the court found sufficient basis for summary judgment in Diamond’s unclean hands, because it used the same type of script, telling customers they’d reached an insurer’s glass program rather than disclosing they’d reached a third party administrator, when it processed claims for two other insurance companies. This is a fairly rare application of unclean hands to false advertising, whose consumer protection rationale usually overwhelms any assessment of unclean hands; the judgment here is probably that Diamond’s acts didn’t really do harm either.

Diamond’s expert testified that the CSRs created “uncertainty” about non-network shops. But the Lanham Act doesn’t require that ads educate the consumer about all options, and using the calls to persuade people to use network companies was legitimate. Any doubt and confusion was simply the result of CSRs’ failure to disclose all positive information about Diamond. The Lanham Act doesn’t require Diamond’s competitors to do this. Diamond only found one customer who was deceived by the script into believing that Diamond’s prices would definitely be higher.

Diamond’s claims for tortious interference with prospective business relations failed to the extent they relied on truthful statements, but Diamond had other evidence, such as that Safelite serviced customers who had appointments with Diamond when those customers didn’t consent to Safelite’s service. Diamond lacked a sufficiently solid prospective relationship with customers who called and merely expressed a preference for Diamond, but it did have such a relationship when there were scheduled appointments – a pretty logical finding.

Safelite brought counterclaims against Diamond based on Diamond’s letters to Safelite’s insurer-clients complaining about Safelite’s alleged steering practices (false advertising), and on Diamond’s practice of giving gift cards to insurance agents who directed their policyholders to Diamond (violation of the Robinson-Patman Act, intentional interference with business relations, unfair competition, deceptive trade practices and commercial bribery, breach of the network agreement with Safelite).

Though the court’s description of the counterclaims mentions only the Lanham Act, it first analyzed Safelite’s “defamation” counterclaim. This probably occurred because the Third Circuit has some weird caselaw surrounding the defamation-Lanham Act intersection. Diamond argued that Safelite is a public figure and thus subject to the rigorous NYT v. Sullivan standard. The court disagreed, finding that merely running a nationwide call center doesn’t make a business into a public figure. Moreover (though this really shouldn’t fit into public figure analysis), Diamond’s allegedly defamatory statements weren’t part of a news broadcast, so they don’t get extra protection. The court denied summary judgment.

On the Lanham Act counterclaim, Diamond argued that its letters weren’t commercial advertising or promotion, because they weren’t sent to influence customers to purchase Diamond’s services. They don’t reference any effort to sell claims administration services. The court found a genuine issue of material fact, given that Diamond did seek to establish a call center and made sales presentations to insurance companies within a few months of sending them the letters at issue. (Comment: an unusual situation – because Safelite was both a competitor in glass installation and a participant in a different market segment, Diamond’s complaint about Safelite’s treatment of glass installers falls under the Lanham Act only because it also administered claims sometimes. If a pure glass installer had complained about Safelite, it seems that the Lanham Act would not have applied.)

Safelite might have difficulty proving particular lost sales or lost goodwill later on, but at this stage it didn’t have to do so. For the defamation claim, if Diamond acted with actual malice, Safelite need not prove loss of sales but may recover solely for damage to reputation. After Diamond sent its initial letters, several of the insurance companies responded to Diamond to rebut the claims. Even after receiving these letters, Diamond sent another round of letters including the same accusations. Thus, there’s a fact issue on actual malice.

Safelite also established a fact issue on whether it suffered losses from Diamond’s gift card program. Customers with no preference could be steered to Diamond by gift-card-incentivized agents, whereas the service agreements provided that Safelite would perform the work for customers with no preference.

Because Diamond took business from Safelite, the court refused to dismiss the deceptive practices/commercial bribery claims, though it did knock out the Robinson-Patman Act claims.

General commentary: The lawsuit doesn’t seem to have helped Diamond very much; this bitter and prolix battle seems unlikely to benefit anyone but the lawyers.

IP Scholars conference, property spirals

Sabrina Safrin, Chain Reaction: How Property Begets Property. How did we get from Chakrabarty’s propertization of modified genetic material to extensive patenting of unmodified genetic material? The creation of property rights for some generates demand for different types of property rights for others in related spheres. The paper looks at (1) patents on naturally occurring genetic material, (2) sui generis regimes for traditional knowledge, and (3) the patent paradox where the amount of patenting has risen dramatically while patents’ expected value has diminished.

Explanations: (1) People who see others getting property rights will seek rights for themselves regardless of whether it’s really good for them or for society overall, following the social signal that propertization is the right thing to do. (2) Assertions of property rights can also alter prior cooperative norms, as with tissue donation used for patenting instead of just research. Continuing to share under those circumstances makes the good citizen into a sucker. (3) Defensive property-seeking, as with defensive patenting to enable firms to fend off lawsuits and enter into patent pools.

Implications: (1) We need to look beyond immediate consequences of granting property rights. Thousands of pages of briefs in Chakrabarty debated the issues, but not one mentioned the possibility of patenting unmodified genetic material. (2) This dynamic means less efficient and happy outcomes than Demsetzian theory suggests; the second wave of propertization has little efficiency justification and does not address tragedies of the commons – they’re reactive in nature: you put up a fence so I put up a fence. The overall scenario may be less efficient than the regime it replaces. Property rights are expensive to create and enforce.

For traditional knowledge, for example, the chain reaction thesis predicts that protection will not stay limited – it will return to protect Western knowledge; Portugal has already enacted laws to protect traditional Portuguese knowledge.

Sonia Katyal: If the world has already ratcheted up property rights, what alternatives do we have to reach the goal of distributive justice? Answer: Not really thinking about how to unring the bell, but to stop it from continuing to ring. Much care should be taken before we add new property rights or break down limits on existing rights. Maybe if we bring more into the public domain, generosity might beget generosity just as enclosure begets enclosure.

Comment: Is Chakrabarty what lets the genetic genie out of the bottle? If you can patent an extracted isolate from a natural, living body, the precedent is already there and has been since Pfizer in 1911.

Question: The future is inherently unforeseeable – you can’t see the consequences of refusing to adopt a property regime either. Answer: Fair point, but these changes have been short-sighted and have only assessed whether a change is good for a narrow band of society, without trying to predict what might happen overall.

Comment: Look at debate over sustainable development in environmental law – similar questions of intergenerational views of resources and how to get people to take the long view.

Question: Devil’s advocate: But isn’t property always better? Answer: For new innovations traditional witch doctors might come up with, property rights exist. But if it’s been around for hundreds of years, there’s no incentive benefit from propertization. Q: won’t the witch doctors transfer information to Western companies better if they have rights? A: The Western companies can pay for information transfer now.

Pam Samuelson: A sui generis database right didn’t succeed in the US in part because people feared chain reactions of property in data that would harm scientific information and news gathering. A: That may also have happened with surgical methods, which the AMA opposed patenting.

Unfortunately, because of the heightened airport security, I had to leave before the last presentation. The conference was a fabulous, intense experience, and I learned a lot. Kudos to the organizers.

IP Scholars conference, information overload externalities

Frank Pasquale, The Law and Economics of Information Overload Externalities. Pasquale wants to intervene in the Google Print debate as a justification for Google’s use. He recommends changes for fair use, specifically a factor-one privilege for categorizers and an expansion of misuse. He also wants to participate in debates over the nature of law and economics in IP. If information overloads are externalities, this has broader implications for IP, health care, and other fields.

Maybe we should consider different treatment of public and private actors – e.g., the NIH versus Google, with the former having a broader scope of fair use.

Buena Vista v. Video Pipeline is a major anti-snippeting case, giving Disney the right to license trailers only to people who won’t criticize Disney. Do the content owners get to control all convenient reference sources? Parts of Ty v. Publications Int’l suggest this is misuse. (Given my IP database, I realize that the snippeting debate concerns me greatly. Is a nonprofit a public actor in Pasquale’s sense?)

Bottom line: Google Library generates many positive externalities, especially in the developing world. Equality of acces, a right to know what’s out there, isn’t taken into account in the current fair use doctrine.

Trademark protections focus on reducing consumers’ search costs, and substantially similar mars can occlude the signal of mark owners. Categorization serves a similar function to TMs by generating metadata structures to help us navigate vast amounts of data. Both are probabilistic harms – one man’s trash is another man’s treasure; 60% of people might not be fooled, but if 40% are that’s infringement.

A privilege for categorizers: An initial problem in evaluation is the question of whether the “search results” are the accused work, or are particular entries the work? Cf. Hughes on microworks (see previous post).

We’ve got great scholarship on the positive externalities of categorizers. What does “cultural environmentalism” gain from adding in negative externalities? Even though we think a work may be valuable for any particular person, lack of coordination can render the overall landscape less valuable.

Mental harm is as real as physical harm, so data pollution is important. Why not Pigouvian taxation? Because of the probabilistic nature of the harm, First Amendment concerns, and lessons from environmentalism. But we can use externalities as a bridge concept to move between methodological individualism and overall social good. There are irreducibly social goods, and we need to understand how individualized transactions affect society as a whole while still being able to cash that out into the economic language of policy science.

What qualifies as a “categorizer”? Can Harry Potter fans create the “Harry Potter Search Engine”? Authors’ Guild’s professed nightmare is 10,000 wannabe Googles. Pasquale thinks the common law can work this out.

Why doesn’t Google settle? They’ve got billions of dollars. Amazon bargained, licensed, bullied a bit. That bargaining approach is the worst result, according to Pasquale. It would create insurmountable barriers to entry in the search engine market. You’d get all sorts of anti-criticism restrictions as in Ty and Buena Vista.

Question: What about a collective licensing solution? Answer: Deserves thought.

Trotter Hardy: There are always barriers to entry. So why is licensing a problem? People can borrow money to create a competing startup. Answer: Oblique answer with reference to net neutrality – why are people worried? It’s not that they think that they won’t be able to borrow money, but they fear being systematically disadvantaged, especially in a market that’s already nearly an oligopoly.

Mark Lemley: A universal library categorization might be effectively a natural monopoly. Duplicating it would be too expensive. Maybe if we don’t treat it as a private good, we’ll get entry that will be socially beneficial.

Search engines collapse cultural, economic, and social spaces, so you get different kinds of results when you put in “plastic christmas trees.” It would be good to have Christian search engines and liberal search engines etc.

Question: Should Google also archive paintings, sound recordings, photographs, newspapers, magazines, and everything else, providing snippets of them all without licensing? Answer: The problem is that licensing concentrates power too much; the problem you fear is one of concentration in Google’s hands, but there are other risks of concentration. There’s no difference between books and other types of texts, so go ahead with the snippets!

Question: Can Google create a market by giving royalties to authors then compete with Yahoo! on price? Answer: Maybe, but comprehensiveness also matters – so you’d want a metasearch engine that could look at both.

What about a government provision of this? Or of insurance against a major data breach? Some people are concerned about government control of information (or failure of government control, for that matter).

IP Scholars conference, trademark as producer protection

Mark McKenna, The Normative Foundations of Trademark Law. I found this paper extremely useful for my work on dilution, because McKenna tells the story of how trademark infringement became about consumer protection, rather than producer protection. Until the consumer search costs rationale became the dominant story, infringement was about stealing business – consumer deception alone was insufficient if it didn’t divert sales. Looking to consumer perceptions to define harm puts the scope of trademark law in the hands of marketers. We need to challenge trademark owners on the kinds of interests they are asserting, and how that translates to economic harm.

Traditionally trademark was a mediator between consumers and producers – the mark was there to get them together. The definition of goodwill has changed over time; it used to be the flow of customers trying to get back to the same person to sell them the same product as before.

Modern trademark law began with Schecter, who thought that the product was central to the image of the mark. Now, the products have faded to the background, protecting associations (Apple stands for innovative, cool, etc. in the computer space). All these associations are still in connection with products, but brand extensions are possible. Marketers think some extensions work and others don’t. Consumers have to think there’s a fit between two things, but they’re also the most dangerous because bad will from a brand extension can transfer back in to the original product category. Extensions don’t work when they’re unrelated – Apple Vacations – consumers wouldn’t care about them one way or another.

Paradoxically in dilution’s terms, then, brands are only endangered by near product relatives, not use of the mark in unrelated categories. The reason illegal art showing Mickey Mouse hanging by the neck is dangerous to the Disney brand is because it’s about the product. The things that are dangerous are captured by likely confusion – so this literature gives us another way to criticize dilution. To whatever extent dilution by unrelated products is real as a cognitive matter, consumers don’t pay attention to it when they’re making decisions.

Question: What about Virgin? Answer: It’s not a very good brand. If I started an unrelated Virgin company I’d do virtually no damage to it.

Marketers are very upfront about wanting artificial differentiation so they don’t have to compete on price. But all that advertising only takes, and makes an impression on consumers, when they can tie it to a particular product or service. Products never really recede to the background. Proof is that extensions that go too far fail, because the goodwill doesn’t transfer, and that the failure doesn’t harm the value of the core products.

Barton Beebe: The history piece of this is wonderful. Can you add a historiography, explaining Schecter’s and the student notes that came after him that are right now what federal courts rely on for the history of trademark? Also, disagrees with McKenna’s reading of the marketing literature – Beebe thinks lifestyle comes first, then you get the product. Response: it’s true that lifestyle matters, but you still can’t unplug it from the particular product. (See that recent book on selling revolution as consumer culture, where marketers tried to make every product hip and countercultural, but didn’t quite succeed.)

What’s the alternative to the consumer protection view? A return to unfair competition? Mark Lemley noted that some courts seem to think “unfair competition” is redundant. Answer: but not all free riding/competition was bad in the old view, only customer-stealing deceptive free riding.

IP Scholars conference, technology

Wendy Seltzer, The Fictional Physics of “Technological Protection.” Seltzer challenges the view of TPMs as just speed bumps keeping honest users honest. They create substantial difficulties with open source software.

Susan Crawford: What do you want to happen? Answer: Lawmakers should take note – rolling back the DMCA would allow a better equilibrium between TPMs and those who want to build interoperable systems.

Isn’t the content industry just going to say that’s exactly the kind of innovation we want to stop? The open source developer is just not a popular enough figure to change the debate, either in Hollywood or in Congress.

Andrew Chin, Software Licensing and Market Power in the Age of the Virtualized Computer. The third installment of Chin’s Microsoft trilogy. The uses of virtualized machines: they can run different OSes, including legacy applications; you can train people without worrying about them crashing the system; you can make full backup copies; you can monitor system activity at the meta level. Virtualization is the killer app for ever-faster processing power!

VMWare, analogous to Netscape, was the startup in the virtual machine space, dominating early on. Microsoft has a VM product and a plan to promote it as a service pack update. It’s cleverly designed to be Windows-dependent and work with Windows Server, full rollout scheduled for 2009. It will control what runs and how it runs, even if you have Linux or something else on top.

Given that virtualization allows servers to host both Windows and non-Windows VMs and consumers can switch from Windows to non-Windows applications, to blunt that threat, Microsoft needs to control virtualization on (almost all) servers that host Windows VMs.

Antitrust analysis will be very important. What kind of product is this? Is the combination a new product, an integrated product, or something else?

Question (if I understood it properly): but isn’t the beauty of virtualization that you can get anything on top of Windows, so it doesn’t matter what’s below? Answer: They may make it hard for you to run Windows as anything but the host. Question: But maybe Microsoft can still benefit from non-Windows hosts. Answer: But they are trying to dominate the space of machines that have Windows servers as the host, driving out other VM products.

IP Scholars conference, anonymous speech

Hmm, that post title might be misleading.

Thomas Cotter, Authorship, Audiences and Anonymous Speech. Positive and normative analyses of anonymous speech – how do audiences weigh such speech? The two recent prominent Supreme Court cases, McIntyre and McConnell, the campaign finance case, have very different approaches to the value of anonymity. If anonymity is content, requiring revelation of identity is content-based and probably illegitimate, but the Court doesn’t quite go that far in McIntyre. McConnell, by contrast with McIntyre, is fine with requiring revelation of campaign ad funding sources on the eve of a federal election because otherwise voters may be misled by anonymous/pseudonymous speech and because anonymous speech might further political corruption.

Because of the tension between these cases, it’s not clear what to do with online anonymity cases that come up. Any lessons from trademark law? In some ways, anonymous speech is like a generic product – we lack cues about quality that a name could provide. A person who’s known to be untrustworthy gets more benefit by speaking anonymously than by using her own name; anonymity can also avoid both deserved and undeserved sanctions for speaking out. Walt Whitman published anonymous reviews of his own poetry to drum up public interest; maybe McConnell is an example of this as well.

Rules for anonymous speech should avoid undue chill, even acknowledging that anonymous speech will therefore cause harm. A presumptive rule in favor of anonymity, perhaps, so that a defamation plaintiff will have to do more than file a complaint to get disclosure.

My comments: Cotter could consider PostSecret as form of anonymity with artistic benefits. There’s also relevant psychological literature: We forget the source of information more easily than we forget the information, thus causing us to revise the probability of its truth upwards over time. This makes anonymous speech more beneficial than it rationally should be and more beneficial than recipients sincerely believe it is – they say they discount its value, but later they don’t.

Comment: anonymity might not be “content,” but more like a loudspeaker/volume dial that affects transmission, so we could analyze restrictions as content-neutral. Response: Okay, but anonymity may be a choice that affects the message and is part of the artistic or political project of the speech. Pseudonyms like Publius can definitely be part of the message.

Michael Landau: When you don’t know the source of a quote – Chairman Mao or Abraham Lincoln – it affects your evaluation; that’s what makes “Who said this?” quizzes fun. Similarly, you may expect a different viewpoint from the Wall Street Journal than the Washington Post. The effect on evaluation is not just about volume. (Cf. research showing that papers with female names identified as authors are evaluated differently than papers with male names.)

IP Scholars conference, sixth parallel session

I missed the first presentation by Dan Burk.

William Gallagher, Strategic Intellectual Property Litigation: An Empirical Study of Enforcement of Intellectual Property Claims. Gallagher represented the defendant in Gov. Schwarzenegger’s suit over the bobblehead doll representing him; the case settled for the reasons such cases settle. He considered the result censorship, since the plaintiffs didn’t mind the bobblehead, only the fact that it was carrying an assault rifle. We don’t know much about plaintiffs’ decisions in IP, so he is studying patterns of private enforcement by interviewing California plaintiff-side lawyers who work on copyright, trademark, and right of publicity claims.

The lawyers say that both lawyers and clients prefer “aggressive” enforcement. They choose to enforce based on the perceived strength and sophistication of the target. They may even like a defendant who’s represented by a lawyer if the lawyer is a generalist who doesn’t know IP; the defendant’s lawyer may believe that statutory damages are available and so counsel the defendant when they’re not. Plaintiffs’ lawyers feel no ethical dilemmas in using aggressive enforcement techniques. Lawyers perceive an advantage for repeat players. They say they don’t overenforce rights, but they give examples that do seem that way. (Asserting copyright protection in cease and desist letters, so defendants won’t post them on the web ….)

Eric Goldman: What’s the definition of a plaintiff’s side lawyer in this context? Answer: Gallagher asked IP lawyers to identify other lawyers who “regularly” do plaintiff’s side work. He’s not sure it matters what the percentage is as long as the lawyers take the plaintiff’s side when they are representing plaintiffs.

Will really bad actors agree to talk to Gallagher? Answer: People like being told “So-and-so said you were a prominent practitioner,” so often they will meet or at least agree to meet. He gets over 90% who agree to meet; he’s still in the middle of interviews.

Compare these lawyers’ rhetoric to that of civil rights lawyers. Are there useful parallels? Compare to the letters available at chillingeffects.org.

Sometimes lawyers say they assert claims that, when push comes to shove, they can’t sustain. If you ask them directly about ethics, they claim they’re ethical, but is there a better way to get at these issues of ethics and even Rule 11? Plaintiffs’ lawyers aren’t afraid of defendants – 99.9% of defendants don’t have resources to resist, the law’s too ambiguous, it’s not going to be dismissed on the pleadings.

Mark McKenna: What about lawyers in the UK? Local counsel were much more reluctant to make the same assertions that US lawyers would routinely put forth, in his experience. (Mine is decidedly not that – the UK lawyers I dealt with over fan fiction were perfectly willing to make unsustainable claims.)

Peter Menell, Accessibility without Piracy: Reinvigorating Copyright’s Deposit and Catalog Functions in the Digital Age. Google Book Search is the new deposit and catalog for our time, furthering copyright’s access-promoting function. The promise is pantology: a systematic view of human knowledge. (You can read an entire book on this on Google, published in the 1840s, or order a copy from a Google partner.) Menell thinks that opt-out is a bad idea. When publishers publish works, they should not be able to leave the universal library.

Also, Google shouldn’t control the universal library either. It is a public service and serves a major public interest.

Question: The contract with Michigan puts lots of limits on Michigan’s use of its own copies. Is that problematic? Answer: Sure. Still, we’re paying for Google’s acts not out of tax dollars but out of Google’s success, which includes some rights for Google.

Sharon Sandeen: What about a digital copy deposit requirement for the Library of Congress copy? Answer: The deposit requirement has been moribund for a while, and Googe is revitalizing it.

Justin Hughes, Created Facts – Copyright and the Collapse of the Fact/Value Distinction.

This section of Hughes’s project fits with his article Size Matters (or Should) in Copyright Law. Ideas aren’t protected by copyright – that’s the negative standard. Originality is the positive standard. The third key component is that the merger doctrine is a negative standard that says where idea/fact and expression merge, the negative standard trumps the positive standard.

Consider the Flinstones: Fred is married to Wilma. Is that a “fact”? Barney isn’t married to Wilma. There is some truth condition to the former statement that isn’t true of the statement “Barney is married to Wilma.” We can know literary facts much better than we know nonliterary facts: Everyone knows that Batman fights crime in Gotham City, but we don’t all know who was the 22^nd president of the US (Grover Cleveland). So this gives us the Seinfeld quiz book case. There are lots of puzzles in this case, including that the author just took a tiny bit from each of 82 separate works, but she was nonetheless found liable for copying the expression of Seinfeld.

Three kinds of interesting cases where fact problems arise: (1) CDN v. Kapes, CCC v. Maclean Hunter – evaluative fact cases. In CCC, the court says the prices aren’t facts because they aren’t discovered but created, but the problem is that these numbers have a powerful effect on your life, much more than facts like how many trees grow in your back yard. So there’s a merger problem; the court solves that by saying that it’s not the sole expression of the idea of a car having value, but a particular expression. The court also says this is an opinion-infused idea rather than a building-block idea and can be protected. Even if you buy the distinction, though, this is a building-block idea when your car’s been totaled. CDN is even more disturbing.

(2) Practice Management, Delta Dental, Southco – nominative fact cases, in which creativity claimed in naming systems/taxonomies. These are not discovered but set up. This too is adopted into law as the only way to get reimbursement in Practice Management and Delta Dental. In Southco too, you can’t compete unless you can copy the numbers to identify what you’re talking about. Southco is clearly the most egregious case, because these people don’t need an incentive to give their parts separate numbers; otherwise they won’t be able to sort their parts. That’s a difference from Maclean Hunter, where the plaintiff was losing business because of defendant’s copying of the information produced at a cost.

(3) Veeck – law fact cases.

If the fact would not exist but for the incentive structure, perhaps we could protect it. Maybe the antitrust doctrine of essential facilities provides a way of resolving the problem through compulsory licensing.

IP Scholars conference, unpublished works

Elizabeth Townsend Gard, Unpublished Works in the Public Domain: A legal Assessment at Three. As a history graduate student working on a biography of a 20^th-century figure, Gard ran into numerous problems with the subject’s executor, and ended up going to law school to figure out what was going on. A newly discovered Beethoven manuscript is in the public domain, and that’s a big deal because of the change in the treatment of unpublished works established by the 1976 Copyright Act. The unpublished (unregistered) works of authors who died before 1936 are now in the public domain.

Publication, of course, is a mess under the 1909 Act. Microfilm is also a problem – if you microfilm papers for preservation, is it a limited or general publication? This is an actual problem with the Adams papers. They registered the papers under the 1909 Act; they do not seem to have renewed and don’t even seem to know about term extension, claiming that copyright expired in 2002 on the website.

Because of §104(a), we treat foreign works the same, so you can use a foreign unpublished work in the US under this rule too, but international issues will obviously loom large. Some countries protect longer; some countries protect forever. Orphan works problems abound in this field because archival data is limited.

Unauthorized publication: Henry Gates, Jr. buys a manuscript at an auction. He authenticates the text at great effort and publishes it in 2002 as The Bondwoman’s Narrative. Is the work under copyright until 2047 or did it expire in 2003? Gates didn’t distinguish between the preface and the main body of the text in his copyright claim. Did he get the copyright with the physical manuscript? Gard thinks the answer is no, because that old rule doesn’t apply to manuscripts. We might be sympathetic to him because of his effort, but the law isn’t on his side.

What’s the right response? People should use the works, even if there are scary copyright symbols on the works. Other scholars suggest claims for breach of warranty, unjust enrichment, and false advertising.

IP Scholars conference, fifth parallel session

I actually switched rooms this time, because Eric Goldman's piece is so relevant to my current work. So this is sort of a pastiche of panels.

Sonia Katyal, Property Outlaws II: Free(dom) Riding in the Age of Intellectual Property. Katyal considers IP, and property generally, as a delegation doctrine allowing individual owners to decide who can access property. IP law has more of a nondelegation doctrine than general property law, and Katyal discusses the types of “disobedient” users whose activities have generated this doctrine. There are acquisitive users (consumers, innovators) and expressive users (commentators, creators). Considering IP dissent as freedom riding preserves both informational and redistributive values.

The system can dynamically change to convert an outlaw into a law-abiding citizen – but Katyal’s not concerned so much about outlaws who are disobeying IP law as outlaws who are disobeying a particular IP owner. That is, just taking stuff isn’t enough to qualify you as a freedom rider. But there is a tension here, in that civil disobedience is expressive in being disobedience, so if you make protest legitimate some of its force may be lost.

Mark Lemley suggested that keeping it illegal but making it harder for IP owners to enforce the law might help resolve some of that tension.

Comment: consider the role of private forbearance (the flip side of what Lemley’s talking about) in shaping disobedience, just as prosecutorial discretion moderates civil disobedience in criminal law.

Tyler Ochoa, Who Owns an Avatar? Assessing Claims of Copyright Ownership in Virtual Worlds. The topic: law in massively multiplayer online roleplaying games that allow players to create characters and interact with the game environment and with other avatars. End-user agreements might control a fair amount of behavior, but might be preempted or otherwise affected by default ownership (e.g., for termination purposes). So Ochoa wants to ignore the EULA for now and start with default ownership rules.

The game programmer wants to say we own everything, because the avatar can’t do anything the code doesn’t allow. That’s overly simplistic, because there’s lots that isn’t dictated by the game provider. (1) Appearance – there’s lots of freedom in most games these days, such as City of Heroes. (2) Capabilities – somewhat more constrained. Are capabilities even copyrightable? They may define the character, as with Superman’s “faster than a speeding bullet.” (3) Behavior – though feedback is provided by the game environment, behavior can be relatively unconstrained. (4) Dialogue – created by the user.

What about fixation? Appearance and capabilities are fixed in data that resides on a server, and saved each time we log off. We’re used to thinking of works as fixed and finished things, but characters in games keep changing each time they’re played. This is a bit of a puzzle for copyright. Behavior/dialogue – what you think of that depends on what you think of the RAM copy doctrine, but with screen captures a lot of times dialogue can be captured and stored.

What originality is there and who contributes that originality? The game provider provides constraints, but within those constraints there is a certain freedom. Ownership depends on degrees of freedom provided. If all you can do is pick one of four characters, that’s not a copyrightable choice. But that’s not how it works today; you’re choosing from a range of options for each of 15 different variables. That eventually becomes an original selection and arrangement. All works are selection and arrangement in some sense. This is a crucial move: if you’re selecting from the game designer’s copyrighted individual components, like short stories, then the game designer has a lot to say about the rights in the resulting compilation/collective work. If you’re selecting from the crayons in the box, Crayola does not have rights in the resulting “compilation.” Ochoa thinks generally the attributes are uncopyrightable, more crayons than short stories, but things like the appearance of different types of shoulders one can choose are not quite the same as individual words from a dictionary.

Sometimes the character will be a derivative work. If you play a Star Wars game, they can give you a Wookie template and you can create your own distinct Wookie, but the Wookies are still Star Wars characters and so any Wookie will be a derivative work. (Hmm… I’m not sure I agree with that, but that also depends on whether you think that Superman is the sum of his capabilities, about which I’m also dubious.)

What about when you’re playing the game? The contribution of the game provider vastly overwhelms individual contributions. So a game played by lots of people consists of a collective work, to which users contribute material, just like the New York Times is made of works made for hire by reporters plus contributions by freelancers. The game provider thus has the right to reproduce your character inside the game. But if it wants to make a movie based on the game, it can’t include your character without your permission.

Question: Does this mean that characters can demand the destruction of imitative characters, or the forfeiture of their magic boots? Answer: The stock character doctrine would affect this, but if a person has really copied another’s character, then that may be infringement given how many alternative choices there are. If the character has value outside the virtual world, why let a copier profit from it?

Anupam Chander: You may be saying that it’s property, not that it’s copyrightable. We could approach it through a different lens. People are laboring in these virtual worlds. Copyright is expansive but maybe shouldn’t expand here. Answer: Let’s take law as given at first and ask what would happen under the law as currently written, and the originality standard is so low that approaching it from the lens of copyright could succeed.

My observation: The psychological and sociological literature on creativity might be of interest here. Some of that literature suggests that creativity is itself a function of constraint, that we do best when we have limits on some of our major parameters. Perhaps characters are freest when they are partially predefined: A game character who’s a troll who loves ballet and has a French accent, for example, might tickle our fancies more than a regular character in a novel.

Eddan Katz, Copyright Contraband. Katz is engaged in an attempt to bring moral coherence to the criminalization of copyright infringement and the attribution of moral culpability to seductive technology such as P2P and circumvention technology. Currently, we risk criminalizing ordinary technologies because their capabilities are more important to us than their actual uses, just as the TSA has banned a lot of ordinary materials because of the risk of bad content/uses.

In the P2P environment, copyright contraband appears via the NET Act. The relationship between the user and the technology means ordinary users will violate the law merely through joining the network because they’ll be sharing a folder of music files which will easily meet the statutory minimum amount. Intentionality is imputed to the technology by describing ordinary use as deserving criminal sanctions. The DOJ has only gone after warez traders and people with a much greater degree of social harm, but an internal memo from the IP task force report proudly states that P2P networks aren’t allowed on the DOJ network – internally, the technology itself is banned.

DRM preempts certain uses; moral culpability then attaches to decryption, a technological capability.

Inducement offers a useful test – the designer is morally culpable for the design of the technology.

Question: Is there really a problem of assessing mens rea (where we’re wrongly attributing intention to technology) or are we blaming the designers, as you suggest at the end? Answer: Sony and Grokster are both about capabilities, positive and negative uses. What he’s trying to show is that takes the potential harm and imputes intentionality to a device without seeing the larger context. We’re trying to force designers to foresee use and misuse, which requires conjecture about what they should have thought about.

Eric Goldman, Brand Spillovers. Offline use of the selling power of trademarks to promote other products is everywhere, mentioned in caselaw dicta, and assumed to be totally okay. Goldman is investigating those offline actual practices such as loss leaders, shelf space adjacency, and trademark-triggered merchandising (complements such as software that goes with hardware, couponing). Retailers are strictly liable if they sell infringing goods, and if they attach infringing labels on packages. But getting people to express an interest in a trademark and making money from that hasn’t triggered liability as far as he can find.

The online equivalent, however, has led to uncertain liability for search engines and others like pop-up advertisers taking advantage of a consumer’s expression of interest in a brand and creating physical or temporal adjacency. The cases are wrestling with whether there’s trademark use, confusion, etc.

What is the difference? An economic analysis: Maybe the retailers should be liable because TM owners create positive externalities for retailers, driving consumers to stores where they buy stuff from other people. But Goldman suggests there’s not much externality; the pricing mechanism can solve this because manufacturers and retailers are actually dealing with each other or with distributors in a chain. However, some beneficiaries of the redirected interest aren’t in that chain and thus the pricing mechanism can’t capture that.

Goldman’s other counterargument: If we’re trying to maximize social welfare, we should consider third-party welfare. Consumers benefit from these retailer practices. Retailers have the ability to reduce consumer search costs (by selling complements, etc.) though sometimes they deliberately inflate search costs (because the longer someone is in the store, the more she spends). Retailers have an incentive to make the shopping experience good enough that consumers find what they’re looking for, and that lowers transactions/search costs.

What we’re fighting about is who’s the best person to decide who makes matches between manufacturers and consumers: regulators, trademark owners, or retailers? As between those three, Goldman thinks retailers are the best. We can learn a lesson there for the online world.

My reactions: As a practical matter, it may be that online retailers live by the statute, and die by the statute: That is, online intermediaries may be “using” the trademark in advertising in ways offline retailers aren’t. Also, the costs of reorganizing stores alphabetically or isolating marks from other brands in the offline world are just greater, so it’s easier to imagine a non-spillover world online.

Bob Brauneis: Producers frequently offer to make the house brand for the retailer, allowing price discrimination, showing that it’s good for them too. Goldman: That isn’t a complete solution (especially when there are two or more national brands in that category).

Mark McKenna: There are spillovers to other brands: Pepsi along with Coke in the soda aisle. But no one goes to a grocery store with only one brand of everything. There are positive externalities in both directions for Coke and Pepsi; valuable brands may get as much as they give. Goldman: People do go to stores that only sell one brand of vacuum cleaner, though. (Trader Joe’s?) Online, the costs of selling only one brand are lower because it’s easier to go from store to store, which may justify a distinction.

Mark Lemley: TM owners are acting differently online just because it’s new. Comparison to offline uses are useful rhetorically, but the big point is that spillovers are not a big deal. Not all spillovers must be internalized; real economists realize this.

Sharon Sandeen: Retailer liability exists for shelf space sales/slotting fees, and that has prompted antitrust/unfair competition litigation. Response: Maybe Google will face that eventually, but right now there’s less market power. We tend to ignore the retailer’s role, but often the retailer can dictate terms, as with Wal-Mart – maybe we should pay more attention.

IP Scholars conference, fourth parallel session

Indirect Infringement Issues

Barak Orbach, Indirect Liability for Copyright Infringement in Multisided Markets. A multisided market is a business environment in which a market intermediary connects different group. Judges inaccurately characterize indirect infringers because they don’t deal with the structure of these markets. Dance hall operators are intermediaries connecting performers with audiences; video recorders involve TV networks connecting audiences, content owners, advertisers; file sharing application connect users and advertisers; flea markets connects buyers and sellers.

Important to distinguish between multisided markets and the classic supply chain of supplier to manufacturer to consumers, which doesn’t involve creation of direct connections between members of each group. Multisided markets create indirect network externalities based on the number of members of the other group: the value of credit cards to consumers is based on the number of places that accept it, not on the number of fellow credit card holders; vice versa on the value of credit cards to merchants.

Marketplace platforms connect buyers and sellers, as with a flea market, dance hall, or eBay. The problem is parties who engage in infringing transactions. Advertising platforms connect audiences and advertisers, and in most cases content owners: TV networks, newspapers, file sharing services, instant messaging. The problems are that a third party enables some connecting parties to circumvent the platform, or that the platform hosts infringing transactions.

Some harms come from piggybackers, for example by interfering with the audience’s payment (in attention) to advertisers, who pay content owners. Piggybacking may interfere with content transmission/storage (time-shifting) or enable conversion of perishable content to durable content (messing up the pricing system of content) or delete advertising.

Other harms come from freeloading platforms, which accommodate infringing transactions to enhance their own attractiveness and increase sales. Advertising platforms may enable content exchange without the presence of content owners at all. Some such platforms may be good – like email and IM services.

The taxonomy could allow courts to make inferences about intent, knowledge, facilitation, control, and financial interest, which helps us assess contributory and vicarious liability. It also shows that the controversial cases arise mostly in multisided markets of advertising platforms. Finally, it helps us to see that there is choice and flexibility in technological design.

Question: How would this lead to different results in cases? Answer: If Sony were decided today, commercial-skipping should be found infringing. The plaintiffs would have to be the platforms, not the content owners, and the focus should be on advertising avoidance and not content storage, which itself is not a bad thing.

What happens when there’s no business plan? Answer: All the bad applications planned to get revenue from advertisers. (Is this true? Freenet and Gnutella, etc.) In the case of anarchy, we may need different rules, but this analysis is helpful when you’re dealing with multisided markets.

Question: About Sony -- librarying is important; Texaco looked at the existence of archiving to find that Texaco researchers weren’t engaging in fair use. Answer: He doesn’t think it affects the economic analysis.

Molly Van Houweling, Safe Harbors. How do we deal with the combination of uncertainty, lack of legal sophistication, over-claiming, and risk aversion that characterizes copyright? Usually with insurance, but for some reason that doesn’t seem to work very well with copyright. Instead, we get constrained behavior. That can be worse than the behavior that might be allowed by rigid rules, even if they were generally pro-copyright owner. Her proposal: safe harbors that are more generous than what scaredy-cats are doing but are within the realm of the standard, so you could always go beyond the harbor and rely on the standard. It will entice scaredy-cats out of their little box, but also push some people who relied on the standard into the larger-but-not-as-large-as-the-law-allows box. Is the tradeoff worth it?

Van Howeling proposes horse trading: we could trade an attribution requirement for a limitation on damages, or bigger safe harbors for clearer pro-copyright owner rules.

Question: What about uncertainty from copyright holder’s point of view? Should we give them more certainty to justify production, such as “no fair use of works that sell over 50,000 units”? Answer: Possibly. She thinks the public choice problems are more serious on the users’ side than the owners’ side, though.

Sometimes there’s technical incompetence in crafting safe harbors, as with mandating the use of a particular technology like Macrovision. Perhaps some of this should be done by agencies that can change over time.

Question: What about change over time? Fair use is evolutionary, which has advantages when circumstances change. Answer: A hybrid solution should keep the possibility of evolution open, unlike the Classroom Guidelines where no one ever presses the limits of educational fair use.

Charles Adams, Applying General Tort Law to the Indirect Infringement of Patents, Copyrights and Trademarks.

Tort law could answer unsettled questions in IP law, such as what happens when the direct infringer has potential defenses. (The answer would then be there’s no indirect liability.) Interestingly, though the common law recognizes respondeat superior, there’s no general vicarious liability for the acts of independent contractors, whereas copyright recognizes such liability.

IP Scholars conference, third parallel session

Session 3, Copyright and Trademark Intersecting

Jennifer Rothman, Customary Intellectual Property. Rothman starts with TJ Hooper case from torts. Whose province is it to determine best practices, courts or industries?

Three problems with using custom in IP: (1) Customs develop in the context of avoiding litigation and preserving relations, as with the “clearance culture.” They’re not meant to apply to end-game litigation where relations are already disrupted. (2) Customs include suboptimal practices developed by interactions with non-repeat players. Large players have an easier time proving their version of custom in court. (3) There are just too few repeat players and ongoing relations in IP to justify resort to custom.

She’s not saying custom has no role. It may be evidence, as in trademark law where we assess likely confusion. Because of product placement, consumers are more likely to think a product that shows up is licensed – that’s a legitimate use of customary evidence. Likewise in contract interpretation where terms have specific meaning in an industry, custom has relevance.

Courts are already incorporating custom, so, if we’re going to use it as law, let’s at least say what we’re doing. Courts are now using custom implicitly rather than explicitly which makes it incoherent.

Principles for legitimate reliance on custom: (1) The party must have agreed to the custom or have opportunity to dissent. (2) The party must have known of custom. (3) Custom needs to be reasonable, so we assess public policy. (4) Customary agreements within industries such as gentleman’s agreement between publishers and libraries shouldn’t be applied to nonparties. (5) Custom should be applied equitably – right now it’s only used to reject defenses and never to accept them or expand fair use.

This analysis makes reliance on best practices statements problematic. They will likely be used as maxima rather than minima. Rothman wants to highlight the need for dissent. Customary law develops when people have the opportunity to consent. When they haven’t, we need to make clear that people dissent from “customary” practices.

Mark Lemley: Additional factor: Custom should apply if we can see the existence of a two-sided and not just a one-sided norm. If only copyright owners believe X and nonowners believe Y, neither norm should be adopted because neither is really a norm.

We can separate customary practices (behavior) and customary law (belief). For trade secrets, for example, we look at behavior rather than belief, since people may never have thought about the rules. We may want to distinguish between trademark-like beliefs that authorization is required and actual practices, and apply the practices instead. Rothman agrees that, though there is a feedback loop, it’s worthwhile to distinguish the two.

Matthew Sag: Technology changes customs – the shift from mix tapes to file-sharing. Do we allow the custom to port to the new technology? Response: This is where a legal vacuum may exist and get filled by customary practice, for example with photocopying technology where people decide that copying for scholarly purposes is fine and then publishers decide that there’s money there, and get courts to back them up.

Another example for the paper: NCAA would license marks to the broadcast network showing its games. The network didn’t care because it just wanted to broadcast the games. The licenses became useful against other parties when the NCAA used the licenses to show that it had rights against any use/display of its marks.

Greg Lastowka, Digital Attribution: Copyright and the Right to Credit. What are the ways we can recognize attribution interests in law? Copyright now is pretty limited, as is trademark after Dastar. Creative Commons and the orphan works proposal suggest that attribution is the way forward to support online creativity.

Advertising law is a possible source of useful principles. It’s hard to differentiate advertising content from other forms of copyrightable material, especially with things like product placement and viral marketing. Copyright law has generally ignored advertising, but it’s a big source of creative content and it’s supported, essentially, by attribution to the advertiser.

Proposal: Modify fair use to put attribution in as a fifth factor, borrowing language from orphan works. A shift from copyright as property to copyright as reputation-enhancing production is going to be long and complicated. His claim is not normative but descriptive: this change is happening and the law needs to react.

Comment: Didn’t Kelly take the link back to the original into account in finding fair use? You can use the positive economic effect on the original as a proxy for attribution, since attribution is what produces the positive effect.

Jennifer Rothman: What about subconscious copying? Wouldn’t this imply that such copiers should always be liable? And how far back to you have to go (as the concern was in Dastar)? Answer: We should fashion attribution rights along the lines of trademark. Credit for diffuse contributions, as with movie credits, is too broad – you should only require identification when there’s a substantial borrowing.

What about wrongful blame, such as copyright owners who object to being named? Courts can work out false attribution problems, such as a trademark lawsuit spawned by the fact that the user gave credit. Society can be harmed by the failure of attribution as an incentive to create (a copyright harm), and also by misattribution of a creation to the wrong person (a trademark harm).

James Gibson, Risk Aversion and Rights Accretion in Intellectual Property Law. This paper has strong resonances with Jennifer Rothman’s, above. The reach of IP entitlements naturally and organically expands over time with relatively little dependency on the positive law, because of risk aversion and licensing practices by ordinary market participants. Gray areas become less gray because licenses are common, and then there’s a whole new gray area. It doesn’t matter whether courts sanction the new regime, though that’s nice for copyright owners, but practitioners know the licensing culture and generally follow it.

Trademark’s feedback loop is a little less pronounced because it filters through consumer protection, but the dynamic is the same. Product placement, merchandising, and the like offer lots of benefits for the parties, so they agree to licensing deals. It’s not just risk aversion but cross-promotional opportunities. But it affects consumer perception because consumers absorb “persuasion knowledge” about marketing practices and then come to understand things like product placement as something authorized. It doesn’t sneak past consumers’ defenses any more, and by the same token they think that it is done with the trademark owner’s consent. Result: further expansion of entitlements. (Expressive defendants’ increased reliance on “use in commerce” and nominative fair use may then be the same phenomenon as the increased reliance on fair use in copyright.)

Patent law generally doesn’t take account of licensing in most circumstances, so this doesn’t happen.

Are there any solutions? Less standards, more rules? That has problems of complexity and rights expansion of its own. Courts will have trouble too; risk-averse entities will comply with the law in the most rights-protective circuit. Perhaps a move to liability rules? That largely depends on your view of liability rules and hold-ups.

Follow patents’ lead: Courts shouldn’t simply ask whether there’s a licensing market but why there’s a licensing market. Courts already say that the fact that the defendant sought and was refused a license shouldn’t weigh against the defendant, and they should expand on that. Likewise, in trademark courts should impose a materiality requirement, even if consumers think there’s been a product placement.

Observation: Even among the relatively powerless community of independent filmmakers there’s strong support for a clearance norm, because they see themselves as creators.

Lydia Loren: There’s a collective action problem among licensees/potential licensees. Patent insurance/challenge pooling might be a good model. “Just say no to licensing” as a group goal.

There have been some moves in that direction, as with the filmmakers. But the insurance people are the ones who need convincing.

Mark Bartholomew & John Tehranian, The Secret Life of Legal Doctrine: The Divergent Evolution of Secondary Liability in Trademark and Copyright Law.

Two-part problem: the expansion in secondary liability theories for copyright, coupled with the lack of expansion in trademark secondary liability. Both IP regimes derive from the same origin, the common law of torts and agency. Courts deciding dual claims often refer to these common origins to justify their eventual holdings, even though the Supreme Court has said you can’t move from one to the other. The result is an unclear and unprincipled situation which allows for too much wiggle room and not enough clarity for market participants. The authors suggest that the myth of romantic authorship, combined with the panic over digital technology, produced an expansion of copyright liability that didn’t occur with the purely corporate property of trademark.

Comment: TM infringement may be more unpredictable than copyright infringement, given the digits of confusion, so the monitoring cost may be much higher for trademark and the harm allowed relatively law. Response: You could argue that copyright fair use is just as/more nebulous. In the Lockheed case, it wouldn’t have been that hard to tell NSI to maintain a list of major corporations – easier than maintaining a list of song titles as Napster was supposed to do.

Comment: Maybe TM owners don’t want expansion because if they had to police secondary liability or risk abandonment they’d have to invest inefficient resources in policing.

Comment: Our copyright cases seem to be more often about the technology of copying – the category of “printer” is now bigger than it was – but technology itself drives copyright cases in ways maybe it hasn’t driven trademark law. (Seems like Eric Goldman’s work on search engine liability applies here.)

Meanwhile, I've become a spokesperson for the Ab-roller

If you get the reference in the post title, you are a geek after my own heart.

Telebrands Corp. v. Federal Trade Commission, --- F.3d ----, 2006 WL 2243584 (4th Cir.)

Telebrands uses direct-response ads to promote its “compare and save” strategy – it finds popular items it can copy and sell cost-effectively. Some of Telebrands's more successful products include Ambervision Sunglasses, the Magic Hanger, the Safety Can Opener, the Audubon Singing Bird Clock, the Better Pasta Pot and the Roll-a-Hose Flat Hose.

Using this strategy, Telebrands introduced the Ab Force, an electronic muscle stimulation ("EMS") abdominal belt, in December 2001. EMS belts use electrical signals to cause the abdominal muscles to contract and release involuntarily. You can see the Ab Force in use in the first two pictures at this fairly creepy site.

Telebrands was copying other EMS belts, which were sold through infomercials as easy ways to lose weight, fat, and inches, and to gain abdominal muscle, without the need to exercise. Telebrands didn’t make any such express claims, but implicitly did so by encouraging comparison to those competing products. For example, Ab Force advertisements referenced "those fantastic electronic ab belt infomercials on TV." Other ads described abdominal belts as "the latest fitness craze to sweep the country," and a radio ad pointed out that the other belts "promis[e] to get our abs into great shape fast--without exercise." Naturally, the ads showed fit, well-muscled models using the Ab Force.

The FTC issued an administrative complaint alleging that Telebrands had made false and misleading claims in violation of the FTCA, specifically, unsubstantiated claims that the Ab Force caused loss of weight, inches or fat, caused well-defined abdominal muscles, and was an effective alternative to regular exercise.

An administrative law judge found that Telebrands had made the claims alleged and that the claims were material to consumers. Furthermore, the parties stipulated that Telebrands neither possessed nor relied on substantiation of the alleged claims, and that, in fact, the use of the Ab Force did not result in the claimed benefits.

The FTC imposed a “fencing-in” remedy barring Telebrands from making any representation, express or implied, about weight, muscle, health, safety, performance, or efficacy in relation to “Ab Force, any other EMS device, or any food, drug, dietary supplement, device, or any other product, service or program,” without substantiation. Telebrands sought to have the order narrowed on the ground that there was no reasonable relation between the fencing-in provision and Telebrand’s violation of the law. The seriousness and deliberateness of the violation, the ease with which a violative claim may be transferred to other products, and a respondent’s history of prior violations are the three factors the FTC considers for whether an order’s scope bears a reasonable relationship to the violation it’s supposed to remedy. Telebrands challenged the FTC’s findings that Telebrands intended to make false claims, that its implied advertising claims were transferable to other product advertising, and that its entry into three prior consent orders for unrelated alleged violations on other products constitutes a history of prior violations.

The court found the violation serious. This wasn’t just overselling a product that was essentially fit for the advertised purpose. Rather, without substantiation, Telebrands mounted an expensive, nationwide ad campaign claiming that the Ab Force could deliver results that it just couldn’t. And it was successful: Telebrands sold 747,000 units with gross sales over $19 million.

The court also found substantial evidence of deliberateness. Telebrands’s use of “compare and save,” given its extensive experience with direct-response ads in general and with the “compare and save” strategy in particular, prevents any suggestion that it acted unintentionally. It wanted to capitalize on the popularity of existing EMS belts, which its ads called “fantastic.” It knew the ads for those products, to which its own ads referred, made claims that the devices would cause people to lose weight and fat and develop muscle without exercise. The idea that Telebrands didn’t intend for consumers to believe that the Ab Force did the same things “strains credulity,” and its choice of visual images “calculatedly fostered” such beliefs. Its ads also called the Ab Force “just as powerful and effective” as the EMS belts sold in infomercials. Given that the ads lacked any explicitly identified purpose for using the Ab Force, consumers had to infer the benefit associated with that power and effectiveness.

The second factor is the transferability of the offending conduct to other products. As with the first factor, substantial evidence supported the FTC’s finding of transferability. The marketing strategy for the Ab Force can be applied to almost any kind of product or service. Telebrands seeks out heavily advertised products to imitate. “Compare and save” is one of Telebrands’s standard marketing tools, and unsubstantiated benefit claims can easily be evoked by reference to other products combined visual images.

The FTC concluded that the first two factors were enough to justify a broad fencing-in order, but still considered Telebrands’s entry into three prior consent orders, even though it had never admitted liability, as support under the third factor. Because the court agreed that the strength of the first two factors was enough, it didn’t reach the issue of whether this was legal error.

IP Scholars conference, second parallel session

Session 2, Reconceptualizing IP

Trotter Hardy, Copyright and Public Goods Revisited. We all know what public goods are. Ex ante, IP law overcomes the public goods problem. Ex post, we in the IP community stress continuing non-rivalrousness. The community exhibits ambivalence: ex ante, we tell authors we want them to create so we give them exclusive rights for practically forever. After it’s created, we say “you shouldn’t control this use, you’re blocking innovation, you’re locking up culture, you’re barring free speech.”

Even our language refers to works as public goods in the present tense. Once works are copyrighted, works are private goods, which was the point of giving property rights in them. As long as copyright hasn’t expired, it’s not a public good. Is this ambiguity in language because we believe works remain half-public? Copyright eliminates free riding, but there’s still non-rivalrous consumption: 100s of people can read the same novel. That view stems from looking at the wrong thing. Copyright doesn’t turn works into private goods, but creates rights, and the rights are private goods. Works can be non-rivalrous in consumption, but rights are rivalrous. If only one person can make copies, that person has something of value to sell. If lots of people can, the original author has a right worth less. It’s functionally equivalent to two people trying to eat the same apple.

The claim of non-rivalrous consumption is wrong. Others can’t reproduce works without “eating into” the original owner’s exercise of rights, even if they don’t interfere with others’ consumption. Current inquiry into damages and market harm thus emphasizes the wrong things – we shouldn’t look at damage to the market for sales of copies of works, but damages to market for the sale of rights – damage to the copyright. Thus, in Napster there were lots of surveys about Napster’s effects on CD sales, but that wasn’t the right question. Instead, we should ask whether the fair market value of the copyright has gone down.

Application to Campbell: The court looked at many markets, though it said there’s no market for parody. The Court would consider the market for satire, rap derivatives, possibly others. That leads to the question of what’s a relevant market. Hardy doesn’t think there are circularity problems but others do. A better way to do it is expert testimony on the value of the copyright, which is easy to do – it’s done in mergers and acquisitions, bankruptcy, etc. to assess the value of an IP portfolio. That would simplify the litigation and lead to more predictable outcomes.

Comment: Some copyrighted works are fashion goods, and their value might go up with the number of users, at least up to a certain point. Hardy isn’t sure that’s true of the right to reproduce copies as opposed to the value of the copies.

Comment: The right to copy isn’t physically rivalrous. The author can still copy without interference. The economic generating power of the rights is rivalrous. But isn’t that circular? The argument’s conclusions assume that the copyright should be generating revenue in these circumstances.

Comment: There are lots of ways you can use pieces of a work to damage its value – a scathing review quoting portions would affect the value of your ability to sell copies. Hardy doesn’t disagree and wouldn’t preclude that – he distinguishes between lowering value by competing and lowering value by pointing out the work is worth less in reality. Response: But that distinction works because courts are focusing on substitution in a particular market, which Hardy doesn’t like. Hardy says he’s not saying courts should ignore particular markets; the defendant can explain that the value of the work has gone down because of criticism, not substitution.

Question: Any implications for statutory damages? Hardy thinks there are none; it’s a way of showing actual damages.

Question: How do you value a copyright without valuing the works that can be produced? Answer: you can’t. Courts now have to look at all those markets themselves, parody, rap, etc., and Hardy wants to put that on the shoulders of the parties’ experts in the first instance.

Comment: The public is trying to maximize use value from the work; focus on value to copyright owner may not reflect total social value. Hardy thinks this will be reflected in the value of the copyright: if people want to hear the song, the market value will increase.

Ariel Katz, Substitution and Schumpeterian Effects in the Life Cycle of Copyrighted Works. A very interesting paper, formalizing ideas about copyright over time and applying them to compulsory licenses for musical covers, restrictive license terms and DRM, and the idea/expression dichotomy.

My comment: Katz’s concept of a prototype stage where the work deserves more protection and more of its value is in its ideas is an intriguing one. Two questions: (1) The value of prototypes may be not in their ideas but in their proponents – TV producers or tenured professors with proven track records, who you can bet on to do a good job. (2) Ideas may simply be in the air, as I believe attribution as a rule in copyright is (as several papers at this conference show) or as reality TV or the shift to season-long narrative arcs before that were. Both these things raise the risks of error if you give very strong protection for the prototype stage.

Katz’s response: Yes, his model could justify idea protection. Attribution norms solve many of these problems for academics, but not elsewhere, creating a gap in protection for early stages of a work. Nonetheless, there might be problems identifying truly victimized authors who were actually copied if we adopted greater protections for early-stage works.

Alissa Centivany, Copyright & DVR: Implications of Timeshifting, Commercial Skipping, and Networking. Decentralized DVRs don’t violate copyright law because they’re just timeshifting. Commercial skipping has been challenged as copyright infringement in the ReplayTV case, but not persuasively so. Currently, no DVR uses automatic commercial-skipping, only manual. By contrast, centralized network DVRs provided by satellite and cable companies probably need to be licensed. This is a tangle of copyright, FCC regulation, privacy, and DRM.

Critics of DVRs: Posner in Aimster suggested commercial-skipping creates an unauthorized derivative work that reduces the copyright owner’s income since free TV is financed by commercials. Others say similar things – eliminating commercials is an attempt to escape the true price of TV.

Is a program with commercials a protectable compilation? Or is a program one copyrighted work and each commercial separate? On compilations, the question is whether selection and arrangement constitutes an original work of authorship as a whole. But there’s no real indication that selection and arrangement occur on a conscious or sustained basis, or matter to people. An audiovisual work is protected to the extent the displayed images are related – but economic relatedness is insufficient, according to case law.

What exclusive rights are implicated by ad-skipping? Time-shifting of course makes a reproduction, but fast-forwarding never creates a fixed ad-free display. An unfixed derivative can’t be substantially similar to the original, otherwise thinking about altering a work would violate the statute. Also, bathroom breaks would infringe. I think the fact that the so-called derivative recreates a separate copyrighted work, the underlying program (which has a separate economic existence as sold on DVDs, on iTunes, and in syndication), also should affect the analysis. Suppose someone creates a new stereoscopic view of the Mona Lisa, and that stereoscopic view is a creative, copyrightable work. If I put the stereoscope image together and recreate the Mona Lisa, shouldn’t that be back in the public domain? I don’t think “others are free to copy the original, but they are not free to copy the copy” applies when my copying produces the original, just as if I copied a public domain work out of an otherwise copyrightable anthology – even if that decreases the demand for the anthology. The work I create is not “based on” the compilation, it’s “based on” a work to which the compilation copyright owner has no rights. If we take as given that copying the underlying work for time-shifting is fair use, then extracting it from a larger work should also be fair use.

Ad-skipping may affect the value of the work to advertisers, but product placement compensates.

There are also possible § 102(b) implications involving methods of operation.

Question: Can’t the copyright owner just argue infringement based on reproduction, and use commercial-skipping as the reason to distinguish Sony? Also, wouldn’t it be great if all the commercial content was bundled with the rest of the content, so copyright owners wouldn’t care if anyone watched on YouTube? Centivany agrees, that would be great.

Who is the copyright owner of this so-called compilation? The local TV station that puts in a bunch of the ads? If you start to think about the owner, that helps you see how it isn’t really a compilation.

Question: Will this create a wasteful arms race between advertisers and consumers trying to push and avoid ads? Centivany doesn’t think we should have only product placement, but it is a successful method, showing that we don’t need litigation or legislation. (I’m not sure we’re better off in a world where free TV only has shows that allow good product placement. I like Star Trek and there are people who like historical shows.)

Suggestion: Consider analogies to pop-up blockers and other ad suppressors, which might expand the doctrinal scope of the piece.

Julie Cohen’s work and Kelly v. Arriba Soft deal with attempts to control/manage post-purchase consumption, and might provide useful lines of analysis.

Sara Stadler, Copyright as Trade Regulation. Stadler’s focus is on works that are intended to exist in multiple copies, not works of fine art. She wants to use copyright as a regulation of unfair competition and avoid the incentive trap, where authors always demand more control to add creative incentives. Stadler would allow private distribution of copies, and only regulate public distribution. This solves the problem of personal copying. It also solves problems of DRM, which would only be justified to control public distribution. And it would take pressure off of fair use, as in the Google Book Search case, where the wholesale copying would no longer be an infringement.

Comment: Line between public and private copying is pretty vague. Response: The video distribution/hotel cases show that courts can work it out. The main problem is P2P, but everyone agrees that’s public distribution.

Consider other countries’ laws, which have exceptions for private copying.

Even if the private copying destroys the total value of the copyright, is that okay? Imagine a Ponzi scheme where everyone makes 5 copies for each friend and passes it on. Answer: Yes, her argument is a utilitarian one about the value of private copying even given harms to copyright owner. Also, in reality, the viral distribution isn’t going to reach everyone through that type of limited distribution.