Thursday, February 22, 2018

False endorsement: some bad questions to ask in a survey

Gibson v. BTS North, Inc., 2018 WL 888872, No. 16-24548 (S.D. Fla. Feb. 14, 2018)

Plaintiffs are professional models, actresses, and/or businesswoman who earn their livings by promoting her image, likeness, and/or identity to select clients, commercial brands, media, and entertainment outlets and carefully controlling their images. Defendants operate three “full friction and full nudity” gentlemen’s clubs. Each plaintiff had one or more of her images used on at least one of the Club’s Facebook pages without her consent.  Defendants’ corporate representative testified that they never asked the plaintiffs for their consent, and that he assumed J Dog Media, a separate company thad does the ads for the Clubs, created the graphic designs, backed up by a declaration from the owner of J Dog.

He stated that “in procuring the images at issue, these images were not altered in any way, and were located on other Facebook sites and pages, none of which were Facebook pages associated with, or describing by name, any Plaintiff, and all of which invited ‘sharing’ on the internet.” He said he chose the images “solely because the images depicted attractive females, and not on the basis of any person’s identity, celebrity, popularity, or any personal basis or characteristic associated with any particular Plaintiff.”

Plaintiffs alleged that the uses caused affiliation confusion, and submitted a survey with three representative images from defendants’ ads, two of which featured a named plaintiff. The survey included only “adult (21 and over) male Florida residents who had patronized a Bikini Bar/Gentleman’s Club/Strip Club in the past two years.” The expert testified that 94% of respondents thought the models in the advertisements had agreed to be in the advertisements; 91% of respondents thought the models agreed to promote the Club in the advertising; 9 out of 10 “thought the models represent the lifestyle to which the Club is oriented” [query: how could this be verified or falsified, ever?]; almost 4 out of 5 “thought the models enjoyed a lifestyle like that reflected in the advertising and might participate in some of the events described in the advertising” [mostly the same query]; 75% thought it was “extremely likely” the models were used “to make them think they represented the kind of women they would expect to see at the club,” and another 20% thought it was “somewhat likely” [ditto].

Lanham Act false advertising: Plaintiffs argued literal falsity in that the ads “necessarily conveyed or implied each Plaintiff’s association with[,] endorsement of, and support for the Clubs.” They didn’t show that the images were altered, and since they were actual images of the plaintiffs, they weren’t literally false. “[A]n unauthorized use of a photograph is not a literally false statement, even if the image was altered,” without more. The ads could still be misleading, as shown by evidence of deception.

Defendanted contest the validity of the survey based on the fact that the survey had a small sample size, included images of other women used by other clubs, and used only two of the 57 advertisements at issue in this case representing only two (or possibly three) of the thirteen plaintiffs. These “[c]hallenges to survey methodology go to the weight given the survey, not its admissibility” and a reasonable juror could find the survey unpersuasive, thus preventing summary judgment for plaintiffs.  Further, they didn’t explain how the alleged deception had a material effect on purchasing decisions.  “While the survey found ‘[j]ust over 90%’ of the participants were more likely to be interested in patronizing Defendants by seeing the ads with Plaintiffs than they were from seeing the same advertisements without Plaintiffs, viewed in the light most favorable to Defendants, this does not mean consumers would actually be more likely to frequent Defendants’ establishments based on seeing Plaintiffs in the advertisements.”  [Yeah, not necessarily the right control for materiality.]  Finally, plaintiffs didn’t show how they’d been or were likely to be injured—an alleged right to compensation for the fair market value of the use of their image had “nothing to do” with whether the use was misleading and wasn’t necessarily proximately caused by the misleadingness.

False endorsement: this is the same as TM infringement/false association. Again, the court denied summary judgment to plaintiffs.  Plaintiffs argued that their images were “by definition” inherently distinctive [this seems wrong under Wal-Mart; surely the plaintiffs’ images were either product design or close enough to get the benefit of the rule that in close cases courts should require secondary meaning], or had acquired distinctiveness.

Defendants argued that plaintiffs weren’t celebrities, which the court construed as an argument that their marks weren’t strong.  Interestingly, the court said that the standard secondary meaning factors “do not lend themselves as easily to Plaintiffs whose mark is their identity rather than their name,” which may be to say that identity isn’t really the proper subject matter of a “mark” [paging Mark McKenna], or in other words you can’t necessarily tell what in a plaintiff’s metaphorical “image” is performing the source identification function and what is just a non-trademark kind of distinctiveness (like my distinctive writing quirks).  The court didn’t require much definition from plaintiffs: “While Plaintiffs never explicitly state what exactly their brands represent, Plaintiffs’ declarations satisfy the first three factors [of length & nature of use, nature & extent of advertising and promotion, and & efforts to promote a conscious association between the mark and the business].”  But how can this be a coherent sentence?  What have they used for so long?  What is the mark with which they promote an association and what is the business, if not the “brand” itself?  Can a plaintiff be a mark for herself?

Regardless, plaintiffs provided no evidence of secondary meaning/public recognition, meaning that this factor was neutral in the light most favorable to defendants.

Similarity was “about as strong as can be.”  Similarity of goods/services, sales outlets, and customer base: plaintiffs’ lawsuit was premised on the idea that they’d never have endorsed these uses, leading service dissimilarity to weigh in defendants’ favor.  But plaintiffs argued that they vie for the same dollars as Defendants because “Defendants’ customers also fall within a subset of the demographic that is interested in beautiful women.” Plus the parties both use websites and social media such as Twitter and Facebook to reach their target audience. [That is, they exist in 2018.]  So that weighed in plaintiffs’ favor.

Intent: it was difficult to determine whether defendants intended to capitalize on plaintiffs’ goodwill.  If they didn’t even create the ads, they couldn’t have had bad intent, and J Dog’s owner said he chose the images because they were pretty, not on the basis of identity/popularity; this favored defendants.

Actual confusion: see above about the survey, which didn’t ask if respondents recognized the plaintiffs.

Statutory and common law misappropriation of likeness: Although “[r]epublication of facts already publicized elsewhere cannot provide a basis for an invasion of privacy claim,” the harm of misappropriation is different; plaintiffs secured summary judgment on these claims. 

Civil theft and conversion: Plaintiffs didn’t show the requisite intent to deprive them of the use of the images; at most there was a fact issue about who created the images. A reasonable finder of fact could determine that neither defendants nor J Dog intended to deprive plaintiffs of their images or any rights to them. [Also, depriving plaintiffs of the right to control all uses of their images is not the same thing as depriving them of their own possession of copies of the images, which is what conversion requires, especially to avoid copyright preemption. This should have been summary judgment for defendants, though it’s not clear they moved for that.]

Unjust enrichment: plaintiffs didn’t argue that the use of their images increased defendants’ customers or otherwise led to a gain, but that use of their images constituted a voluntary acceptance and retention of the benefit of using their photos without paying any fee.  The court wasn’t convinced that it would be inequitable to let them retain this benefit, especially that the misappropriation of likeness claims would provide compensation for the exact same harm.
Damages for misappropriation of likeness: Plaintiffs claimed they were entitled to the fair market value for the use of their images and reasonable royalties, though the court found their expert’s report problematic.  E.g., the expert valued one plaintiff’s “working day rate” at $5,000 without explaining what it meant. The court [wrongly I think] presumed that this was the fair market value for the use of each image; the same image was used three times [which would be one photo shoot, not three] but then the expert calculated actual damages as $30,000. Another plaintiff was credited with a “working day rate” of $100,000; for three different images and eight total uses, the expert calculated actual damages at $900,000. The expert didn’t sufficiently explain his calculations to avoid speculativeness.

Zazzle injunction reversed: irreparable harm is hard to show

Greg Young Pub’g, Inc. v. Zazzle, Inc., No. 16-cv-04587 (C.D. Cal. Feb. 8, 2018)

After a jury trial finding Zazzle liable for copyright infringement for producing goods with user-uploaded images, the court granted a preliminary injunction. Here, it vacates that injunction for failure to consider material facts and to provide a sufficient record on the multifactor test.

Irreparable harm: Plaintiff alleged harm to its “competitive position” because Zazzle sold “unauthorized products,” and alleging “actual loss of market share,” as well as “harm by loss of reputation.” But there was no such evidence.  A conclusory declaration from plaintiff’s owner was insufficient.  Although they alleged direct competition, no evidence showed plaintiff’s products or those of licensees.  The fact of infringing sales and allegedly poor quality of images weren’t sufficient alone to show loss of reputation attributable to Zazzle without evidence about the quality of both parties’ images or a causal connection between the quality and a loss of reputation.

Irreparable harm from “lost control” of copyrights was equally insufficient. “After eBay, Plaintiffs cannot rely on the pure fact of infringement in order to establish irreparable harm.”  Grokster, involving similar arguments, had evidence of infringement on a huge scale, more than Grokster could ever possibly redress with damages, as well as evidence that plaintiffs’ copyrights were particularly vulnerable to repeat infringement. No such evidence existed here.

Plaintiff argued irreparable harm from the threat of future infringement and additional lawsuits, but the “mere likelihood of future infringement by a defendant does not by itself allow for an inference of irreparable harm.” There was no indication here that Zazzle would be unable to pay statutory damages, or that any infringements would require dozens or hundreds of future lawsuits. “The facts were uncontested that Zazzle promptly removed any infringements brought to its attention.”

Nor did plaintiff show lack of adequate remedies at law. The jury’s statutory damage award was more than adequate compensation: $351,100, compared to licensing revenues for the artworks at issue of approximately $21,489 and Zazzle’s sales of $5,622.  Nor was Zazzle likely to repeat its infringement in bad faith, given the court’s previous ruling that Zazzle wasn’t a wilful infringer and the de minimis infringing sales since the lawsuit was filed.  Only about $100 of infringing sales since the litigation began were to third parties; more than 75% of the sales were to plaintiff/its straw buyers.  “Considering that Zazzle reviews millions of images in a given year, Plaintiff has not been able to demonstrate that Zazzle is likely to allow infringement to continue in bad faith. And even if some infringement does continue, Plaintiff has not been able to show why the statutory damages award would not adequately compensate for that injury.”

The proposed injunction was also too broad. The court never decided whether Zazzle had a viable DMCA defense as to images only displayed on Zazzle’s website and never physically manufactured, because the plaintiffs withdrew its claims before trial. Thus, it was unclear whether the injunction covered mere display, or required Zazzle to take “reasonable” steps to address the display of images on its website as well as its manufacture of products.

Bye-bye injunction.

The other Redbox case: the color may fade, but not the mark

Redbox Automated Retail, LLC v. Xpress Retail LLC, 2018 WL 950098, No. 17 C 5596 (N.D. Ill. Feb. 20, 2018)

Redbox sued defendant DVDXpress alleging trademark infringement and false advertising; DVDXpress moved to dismiss and asserted counterclaims/affirmative defenses, some of which the court here rejects. The parties compete in the market for DVD rental services through automated vending machines called kiosks. Redbox has registered trademarks for its word mark and for the kiosks’ red color scheme.

Redbox alleged that DVDXpress recently began using kiosks that are entirely red in color, making them confusingly similar to Redbox’s kiosks. Redbox also alleged infringement from DVDXPress’s use the term “redbox” in the metadata for its website in order to capture searches for Redbox.  [Practice note: this use of metadata is probably useless and though it’s unlikely to be confusing in practice, the case law here is terrible. It is thus risky to use another’s trademark in metadata, though the court here does better than most (while conflating descriptive and nominative fair use).]  Finally, Redbox alleged that DVDXpress falsely advertises that customers can rent movies through DVDXpress twenty-eight days before the same movies become available through Redbox.
DVDXpress counterclaimed for tortious interference with its contractual relationship with Weis Markets, which signed a DVD Rental Kiosk Agreement with DVDXpress in July 2017 and accordingly asked Redbox to remove its kiosks, but four days after Redbox filed this suit, Redbox indicated that it wouldn’t remove its kiosks “on Weis’s requested schedule.” Redbox removed the last of its kiosks from Weis’s stores in late October 2017, after DVDXpress filed its counterclaim.

The counterclaim plausibly alleged that the continued presence of Redbox’s kiosks at Weis’s stores through late October caused Weis to breach the contract for exclusive kiosk provision or made it impossible for DVDXpress to perform by taking up the places for kiosks in the stores.  Delay, rather than total noncompliance, can be just as malicious as outright refusal to act.  Given that Redbox informed Weis that Redbox was seeking an injunction against DVDXpress and “would be concerned if more red DVDXpress kiosks began to appear in the marketplace,” it was plausible that the delay stemmed not from some independent logistical difficulty, but rather from Redbox’s illegitimate “desire to stop (or at least slow) the expansion of DVDXpress’s business.”

Acquiescence as an affirmative defense: According to DVDXpress, Redbox has been aware since 2002 of DVDXpress’s use of the color red on its kiosks, and in 2007, DVDXpress and Redbox were temporarily merged under then-parent company Coinstar, Inc. “One can reasonably infer that, while the two companies were merged, Redbox affirmatively approved DVDXpress’s use of the red color scheme on its kiosks. That plausible scenario is sufficient for the acquiescence defense to survive at the pleading stage.”

Fair use as an affirmative defense: “It is plausible that, when DVDXpress included the term ‘redbox’ in the metadata for its website, it did so ‘fairly and in good faith … to describe’ its own kiosks.”

Unclean hands: DVDXpress that Redbox deceptively advertised “no late fees” and Blu-Ray quality discs, and falsely advertised that certain DVDs would be available at Redbox kiosks on their release dates.  Unclean hands has to relate to the matter in which the claimant seeks relief: if the claimant’s right wasn’t the result of the alleged wrongdoing, then it’s unrelated to the litigation’s subject matter. Also, unclean hands “is disfavored where a plaintiff seeks to enforce laws, such as a false advertising prohibition, that protect the public,” especially where the defendant could challenge the plaintiff’s conduct through a counterclaim or separate suit. Here, the alleged misstatements had nothing to do with Redbox’s allegations; even Redbox’s own allegedly false ads about availability didn’t “help Redbox acquire the right (to enjoin DVDXpress’s assertions that its kiosks have certain movies twenty-eight days before Redbox) that it seeks to enforce here.”  A few courts have allowed the defense in similar circumstances, but the court here found that the “obvious hypocrisy in one company suing a competitor for doing the very same thing it did in the past (and got away with) … does not warrant denying an otherwise merited injunction or other relief, especially when the relief would end a deceptive practice that is harmful to the public.”

DVDXpress also asserted an abandonment defense, alleging only that “Redbox has discontinued its use of the color red on its kiosks in the ordinary course of business, at least, with regard to certain locations, namely at Walmart retail stores,” and that “Redbox has allowed a number of their kiosks to fade to a color that could better be described as pink.” That’s not enough to suggest discontinuance of red, and the court struck the defense.

Reading list: global fair use

The international copyright system requires all participants to recognise a freedom for fair quotation that covers much of the ground encompassed by the notion of ‘fair use.’ The obligation derives from Article 10(1) of the Berne Convention and thus applies to some 174 countries (and, because Article 10(1) must be complied with under TRIPS, is carried over into the WTO Agreement). In contrast to other limitations in Berne, Article 10(1) is not optional, it is mandatory. It creates an obligation, and thereby imposes a ceiling on the freedom of action of Members of the Union. The breadth of the obligatory exception is wide: as enacted in national law, it should not be limited by work, nor by type of act, nor by purpose. The exception should not be subjected to additional conditions beyond those recognised in Article 10: to do so is to breach the obligation. Subject to those conditions, the freedom the Article secures to users encompasses any and every act of quotation, the meaning of which reflects how the term is ordinarily used across all cultural forms. That includes free-standing uses, transformative uses and parodic uses. Its breadth reflects, but is not limited by, the desire to give effect to the fundamental freedom, freedom of expression. We have dubbed this ‘global, mandatory, fair use’, or GMFU, for short.
 Yet, despite this obligation, there has been a marginalisation of Article 10(1) and this paper explores how, and why, this distortion of international obligations has occurred. It suggests the dangers of pluralism in copyright when it comes to international obligations.

Friday, February 16, 2018

WIPIP concurrent 3 (most of it)

Annemarie Bridy, Fearless Girl Meets Charging Bull: Copyright and the Regulation of Intertextuality

DiModica (who made Charging Bull) complained that Fearless Girl’s placement created an unauthorized derivative work and violated VARA by being a material alteration that prejudiced his honor—Charging Bull was no longer optimistic but transformed into a threat. Turning a semiotic dispute into a legal one.

Can © be leveraged to prevent confrontational dialogue w/preexisting works? What role does copyright play in regulating text to text and author to author interactions? And what can literary theory tell us?

Existing scholarship on role of romantic author: presupposes/valorizes creative genius and possessive individualism. But US © has important differences from European law. Bahktin/Gerard Genette: intertextuality views works relationally rather than as closed systems—dialogism within & between texts. Monologic discourse is poetic; dialogic discourse is associated w/the novel. Monologism privileges unitary voice of author as autonomous; dialogue has mix of voices that are layered in the work. Work is more open and porous. Genette’s hypertextuality: secondary author targets preexisting text & rewrites & completes it in a way original author might not have liked—hypertext as unfaithful continuation of its hypotext.

Dimoda embraces aesthetic monologism, which is encoded in author centric European © law. Strong moral rights, limited exceptions/limitations. US © is hospitable to intertextuality by design, even assumes it. Less author-centric and more focused on public interest in production of creative works. © protects second degree works such as compilations and derivative work, and also excepts a range of second degree uses, including fair use which is flexible/unfaithful/murderous (Genette’s term) and also compulsory license for musical work covers (less flexible but appreciates artistic impulse to remake/reengage the work). Very narrow moral rights that delayed US accession to Berne.

He can’t prevail: derivative work claim requires physical incorporation of elements of the protected work [but consider the case from yesterday finding that embedding was infringement]. Not every modification of a work creates a derivative work, Mass MOCA, and altering the viewer’s experience of a work doesn’t create a derivative work, Lewis Galoob v. Nintendo. Even if the placement did result in a derivative work, the semiotically confrontational nature of its relationship to Charging Bull would mean that it’s a fair use.

Rosenblatt: maybe if juxtaposition is fair use, the fact it’s juxtaposition forces us to go straight to fair use & concedes it’s w/in the scope of the right.

A: but physical incorporation is still necessary.

Ann Bartow, The Economics of Race and Gender in Art Law
Ownership disputes w/racial subtext. Control of Barnes Foundation stolen from Lincoln University, a HBU in Pa.; art now controlled by wealthy mostly white people in Philadelphia—thought the art should be in the hands of nice white guys. Georgia O’Keefe donated art to Fisk Univ., an HBU in Nashville; art now 50% owned and controlled by museum in Bentonville, Ark. founded by Walmart heiress. Leveraged highly contestable claims of financial mismanagement/lack of sophistication.

Georgia O’Keefe lost art theft based on statute of limitation grounds; in the vast majority of art theft cases, by contrast, courts found all kinds of ways around SoLs: possession as continuing crime; demand and refusal rules before SoL restarts; sales reboot the SoL; etc. Nude posing as a career strategy for O’Keefe—it was one of the only ways she could be accepted in the arts community. She didn’t feel it was voluntary. Hazards: sexual harassment, sexual assault, nonconsensual pornography, low pay and labor abuses, gawkers & creeps.

Another example: Barbara Kruger: appropriation art; Hoepker was the foreigner who didn’t comply with 104(a) so wasn’t able to enforce his restored (c). She might not have prevailed on fair use; museum used the art as an ad for the Kruger show, and put it on mugs and the like—she was exploited as she was being accused of exploitation.

Prince and Cariou’s Yes Rasta: both whites dealing in pictures of bodies of black men (and white women from soft-core porn for Prince).

More women than men attend & graduate from art schools/get art degrees. There’s no movement to get more men in art as women in STEM; but men are at the top. Every artist in the top 100 sales in 2013 were men. In 2014, no women in top 2104. Less than 4% of artists in Modern Art at the Met are women, but 76% of the nudes are female. Don’t do as well at art fairs; held only 24% of museum director positions and were underpaid compared to men. Art galleries control value of art by manipulating two signals: price and who buys. Too low and the perception is low quality and too high and the artists seem overhyped. Manipulate the secondary market. Also try to sell to tastemakers. Control value of art worldwide, b/c value is subjective and subject to signalling; white men decide what’s good and valuable. Museums can make or break artists with shows, deaccessioning, etc. Often in collusion with galleries, auction houses and wealthy collectors. China has wealthy collectors in old, European art which has also had an impact.

Antidiscrimination as a moral right. Attribution can facilitate discrimination, esp. for gender. Artworks can be racialized or gendered in ways that are perceptible and signaling, as w/software that can discern gender of text authors. Integrity: has multiple meanings, including the quality of being sound moral principle, uprightness, honesty and sincerity. Galleries, auction houses and museums as sites of resistance and reform.

Q: Internal splits in Fisk and Barnes led to majorities to surrender [Bartow notes that they required board takeovers] and there were reasons to see them not as central to the missions of the school. When the institution did try to monetize the collection, the AG limited Fisk’s ability to do so—it was accomplished in a number of different ways, including by using institutional power to threaten [Bartow notes threat to funding]. The capacity to divest the collection once the decision was made was also limited & cabined. [Sounds to me like the expropriation of black wealth or wealth foundation described in Richard Rosenstein’s and Ta-Nehisi Coates’ recent books.]

Rosenblatt: Compare appropriation art. Some are seen as legit and others aren’t. But when I dig in to these problems, they feel more like antitrust problems than IP problems. Why is IP the lever here?

A: there could be other things. And antitrust has a role to play around auctions. But moral rights is more empowering to individual artists. [CRT’s rights claims.]

Q: what if art is a male display behavior of taking attention-grabbing risks making a biological basis for male dominance in the field?

Chien: Uber pays men more than women per hour, but it seems to be b/c men drive faster and more dangerous routes. Correlated w/things that are correlated w/value. Are there ways to quantify what would be the legitimate basis for one thing to be displayed more? What are the right criteria?

Raymond Ku, The First Amendment Implications of Copyright’s Double Standards: It’s More than Just Entertainment
Touting the value of P2P for distributing creativity/access w/o deadweight loss. Reaction of former Register: what’s the big deal? It’s just entertainment.  Debate b/t Justices Stevens & Blackmun about ability to time shift popular entertainment: who cares about Honey Boo Boo?  Stevens struggled to find a political or educational use.  Sony makes more sense when you connect it to Bleistein: why are we judging the value of the programming that’s being watched? 

But we judge that all the time. We’ve been discussing how much creativity is sufficient for © protection and fair use.  When we do this, entertainment essentially always loses.  SAT/Seinfeld Aptitude Test case is the key example.  Interesting doctrine: Protection of creative facts—less need to copy them.  In Hoehling, by contrast, the movie studio didn’t need to justify its copying of Hoehling’s theory.  Case law suggests that for fair use, when we’re dealing with entertainment, for fair use no amount of added creativity is sufficient—Blurred Lines, 60 Years Later, The Harry Potter Lexicon. Even in Cariou v. Prince, court goes out of its way to say that its broad definition of transformativeness doesn’t cover entertainment.

We’re mixing up the idea of results and repetition. With factual works we do want people to reach the same results hopefully.  With fanciful works, the options are infinite and reaching the same result is not inevitable; repetition of results means repetition of expression. Independent contributions are not repetition. [  For factual and fanciful works, repetition is part of the process for learning for both; researching and performing respectively are similar in their contributions to understanding. Retyping Holmes’ common law is meaningless, but a musician that is able to replay or emulate Yo-Yo Ma or paint in Van Gogh’s style is doing something dramatically different, yet we say that copying in the latter circumstance is less justified. For both fanciful and factual, incremental contributions may not be recognized or appreciated as such; also, someone may say it better.

Factual: financial stakes are small in litigation, while for fanciful works the stakes can be significant.  For fanciful works in litigations, the work speaks for itself and everyone’s an expert; there are no standards/informal voluntary standards. Factual works litigation: there’s a practice of framing for the jury, governed by agreed on standards and guidelines (my article tells you the contribution it’s making; general, disciplinary, and institutional standards) whereas Fearless Girl and Charging Bull stand by themselves; courts may defer to professional/expert opinions.

RT: about results, what do you mean? I would have gone the other way, w/TS Eliot’s objective correlative—similar results can be achieved by multiple different expressions (compare romance novels or really any standardized genre, even the middle aged white male adultery novel).

A: When you’re trying to answer a Q, 2+2 has one answer but literary Qs have multiple answers. [Like Barton Beebe on progress & Jeanne Fromer on problem finding.]

Rosenblatt: is this a problem of fictional works versus a problem of ignoring SCt precedent in the form of Campbell’s dictate to consider different meaning or message, not just purpose? Are courts ignoring this b/c they disagree & think that entertainment is frivolous?

A: Easterbrook says he likes Cariou but it doesn’t explain difference b/t me creating a derivative work and me making a fair use, especially in entertainment where we often think it’s something the original author should control. 

WIPIP Concurrent 2

Nancy Kim, The License v. Sale Puzzle after Impressions v. Lexmark
Sales exhaust the patentee’s rights to any item regardless of what the contract says. But restrictions on licensees are different because a license doesn’t implicate the same concerns about restraints on alienation as a sale.  Even if the contract is enforceable, that doesn’t mean that the parties can’t contract about everything. They can’t recharacterize the world—you can call a contract w/a worker an independent contract, but if the facts establish an employment relationship you have an employee.  Same with license/sale.  “Authorized” sale is also important.

Contract can give rise for breach and infringement claims, but not for the same provision; not every breach exceeds the scope of the license and constitutes infringement. Principles: if the first transaction by the patentee is a sale, all subsequent transactions are sales. License agreement may or many not be followed by sale. If license allows licensee to sell patent items w/o restrictions, then the subsequent transactions to customers are sale.  If license imposes restrictive conditions on licensees’ ability to sell, they’re sales only if the licensee required with those conditions—e.g., limited to sales in North America, and then sales in France are unauthorized.  If licensee has obligation to impose restrictions on customers, then subsequent transactions are sales.  If mfgr didn’t have customer sign contract restricting use, mfgr might be in breach of contract, but it’s still a sale. 

With a standard form contract/adhesion contract if the patentee/© owner retains rights to reposess or control item, it should be considered a license, but otherwise a sale.  If negotiated or individual contract, distinction should depend on whether contract limits one of the patentee/© owner’s exclusive rights (that is, does it affect the scope of the license?).

Whether the purchaser/licensee infringed depends on whether the contractual provision was a covenant or condition, which means did it limit the scope of the license/implicate one of the exclusive rights of patentee/© owner, but the ©/patent cases differ.  Sale of © item may be treated as still having a license.  Example: a nonexclusive, nontransferable license to make 2 copies. This is not conditional language; transferability is not one of licensor’s exclusive rights as a © owner.  Also: Licensee agrees not to transfer, assign or sell—that’s still drafted as a covenant not a condition.  “provided that licensee may not transfer, assign, or sell … license is expressly conditioned upon compliance w/this provision”: if result would be forfeiture, courts in contract law generally are disinclined to find a condition.  How about: if licensee attempts to transfer, assign or sell this software, the license immediately terminates and is void: clearly a condition, and that’s what it should take.

Lemley: How confident are we that 9th Circuit’s condition/covenant distinction survives Lexmark?  Lexmark suggests © and patent are kin. 

A: she’s concerned about that.  Ct doesn’t directly address license v. sale b/c they say it’s a sale. Also different exclusive rights: with ©, you can sell the product and still not exhaust all the rights (e.g. public performance).

David Olson, Why Increasing IP Exhaustion Leads to Decreasing Ownership of Goods
Information costs for servitudes on chattel property are high; recording burden is high; may restrict resale/reuse of goods forcing inefficient manufacture of new goods instead.  With information goods/internet age cost of checking restrictions can be very low relative to value of item.  IOT goods that phone home/alert you as to any restrictions.  Cost of knowing them is just lower [depending on what you mean by “knowledge” given human practices & functional capacities].  Digital copy of book can also be created more cheaply; not much waste in idling/deleting my copy.  Allows for price discrimination, as does monitoring via IOT/collection of $ for downstream uses.

Still, most goods are sold free/unconstrained.  It’s often more profitable to simply make a first sale w/o monitoring or transaction costs.  E.g., cars.  Other times it can help buyers who can’t afford the total cost to give them only a partial right & have owner retain residual value—rental cars, tuxedos, etc. 

For the IOT, your refrigerator can be shut down if you violate a term.  HP is also licensing/renting printers on this basis—rent charged based on amount of use, and they send you the toner as you need it.  The relationship no longer looks like a sale; it is ongoing and involves patches, etc.  If your credit card expires, the printer stops working.  Pretty effective lockup through contract, and enforcement w/o any privity of contract necessary.  [Pretty good opportunities for fraud?]

Is this good/bad and how if at all should the law respond?  B/c price discrimination can be quite valuable to society, especially for cheap goods, the next move to preserve ownership would have to be inroads into contract law to stop these characterizations/prevent these restrictions being enforced via self-help.
Reasons to be cautious about making that move: now that exhaustion is so strong, we’ll see the same attempts through contract law; we’d be better off to allow experimentation and see how the market reacts. We’ve never seen mfgr limit all use/control all aspects; it’s not profitable and there’s bad PR from people not liking price discrim. Instead we’ll see monitoring and price discrim w/customization for users’ needs.  Consumer will not be able to see she’s being charged differently b/c the good will not be exactly the same for each one.  Like cars with different packages [this is a really disturbing example—we know that the poor/women/minorities pay more].  Can charge people willing to pay more, more, and charge low prices to those who can’t afford more b/c arbitrage will be impossible.  [When has this ever happened, compared to poor paying more?]

Better to use antitrust law than sale/license law to deal w/abuses such as market foreclosure.

RT: questions above.  Tell me why you’re hopeful.

A: discrimination law would still apply, but he was talking about standard packages; discrimination comes in at the overall price of the car.  People could still create data [unless contractually prohibited]; they might find that women are willing to pay more, which could be a problem.  Women sometimes are never offered a low enough price and they walk away; the transaction fails.  AI, which has the potential to lessen discrimination [do we know that?], means that if your goal is profit maximization you will try to figure out what people will pay and get past proxies like age and race.  [That does not seem to me to describe how the current algorithms work—they absolutely use your categories to be predictive about your characteristics, including WTP. Plus your other non-race and gender characteristics are tightly correlated w/them, so “non proxies” (e.g., where you live, whether you’ve gone to jail, how much money you make) are also proxies.  And then your reaction to the offers you get—which people in other groups won’t get—determines what happens next, so the discrimination just gets baked further in.]

Rosenblatt: rich could be allowed to make remix, poor not.  If what people want to do is on-label, then your logic makes sense. But if they want to do off-label stuff like remix, reverse engineering, follow-on innovation, doesn’t this bar that?

A: if we restrict producers to one price, the Q is do poorer consumers get to buy anything, or nothing?

Michael Burstein, Indiscrete Property
Information is often incompatible with the notion of thingness in ways that go beyond the mere costs of defining the thing: it is indiscrete and continuous, as opposed to discrete and separable as most tangible goods are. While a stock tip is relatively discrete, more often info conveys different value to different actors; info is more of a platform on which transactions occur.  Thus value depends less on definition than the structure of the underlying institution.  Should focus on rules that allow people to structure info flows for best use.

Case for info as discrete property, from two parallel movements suggesting strong IP to economize on use of info.  First, incentive/reward theory is unsatisfying; market failure resulting in underproduction doesn’t dictate IP as solution and other solutions could produce incentive w/less cost.  As empirical matter, it’s well understood that in many situations the IP incentive, or any incentive, may not be necessary for production of intellectual goods.  Next move: commercialization theory, where IP is about the subsequent use of creation/invention by giving incentive to solve market failure structurally similar to the earlier market failure theory, or by serving as beacon/signal/contracting device for people to come together and exchange info. New private law is aimed at reclaiming centrality of tort/contract/property in US law. Scholars offer info cost theory about economizing the cost of info specifying use rights.  Thingness of property matters b/c by defining a thing and marking it w/rough exclusion signals we create relatively clear signals about what you can and can’t do with it. Use rights in a car; you need no info about underlying owner of car to know not to steal it.  You just need to know it’s not your car.

Nonrivalrous info: inputs to info use are rivalrous, so we want to manage them. New public law theorists like Henry Smith argue here’s a benefit from reducing complexity of coordination, if the thing can be defined easily and the uses cannot.  In such cases, exclusion strategy is more economically justified than governance. Smith argues that patent is more thing-y than ©. This is not uncontroversial or unproblematic. Clarisa Long has made the exact opposite argument.

Problem: relies on thin concept of info such that all you need to worry about in defining a good is the cost of delineating the good. But other characteristics make it almost incoherent to talk about definition. Info is additive, communicative, and layered.  Additive: info bits build on one another so value is different depending on type or info.  “Data information knowledge wisdom” framework: data is info asset that’s product of observation, relatively discrete and objective; info is relevant or usable or significant—info uses data to answer a Q; knowledge is usually knowhow, turning info into instructions/decisionmaking aids, synthesizes across types of info.  Literature usually ignores wisdom and so will I.  As you move up, increasing correlation w/social meaning and value.  Info has value, but it exists in different forms w/different sources of meaning and value.

Communicative: JP Barlow: information is a verb, not a noun.  To have an effect in the world it must be transmitted & create meaning for recipient.  Does this not through ethereal transmission, but is contained in artifacts—physical manifestations that are vessels for the info. Some artifacts are more self-revealing of the info than others.  Sometimes info can be sticky/tacit (Eric von Hippel) and needs more than writing down. Info can be modulated self-consciously, as w/software modularity so you can hide certain elements of code & reveal others.

Layered: Heterogeneity—small molecule pharma—most direct way to communicate info is to show its structure, but you could also talk about the molecule/generate info about that.  You could talk about chemical characteristics, biological effects, efficacy & toxicity, all w/o revealing its structure.

Info exists in continuous indiscrete state in many cases.

In land, things are rough signals of value. But if info is continuous then thingness may send wrong signals.  Lose out on ability to communicate across users.  May be over and underinclusive.  Answer to Smith’s question about how you specify rights & duties up front may be “you don’t have to.”  Spectrum of discreteness may also help rationalize our intuitions, e.g. about patentable subject matter. If abstract idea is relatively indiscrete, and open platform approach is most efficient, that may help explain why IP doesn’t protect them. Likewise for doctrines about timing of patentability such as utility/reduction to practice—when the info becomes more discrete.

Rebecca Kai-Fang Lin, Artist’s Resale Right: Past, Present and Future
What makes art so special?  Work of art can be one of a kind; reproduction doesn’t create the same thing.  Low price when first sold.  Resale right was exception to exhaustion introduced in France.  Visual artists usually earn less than average income; this allows recognition and fair treatment and also helps them stay connected to the works.  Also can support cost of storage, conservation and restoration.  No evidence that resale right harms the art market.

Berne Convention introduced resale right in 1971 as reciprocal right based on country of artists.  Rightholder is author/authorized person.  Covers original works of art/original manuscripts of writers and composers.  Nature: inalienable right; applies in any resale.  Collecting societies should be able to collect the royalty and should have a right to gain information.  US, China, France and UK are the main fine art markets now, representing about ½ of global sales.

EU has harmonized: rightholder/authorized person though member states can limit who those can be; liable person is generally the seller but member states can diverge; the covered works are original works of graphic or plastic art; must be physical but copies can be included.  Nonassignable and inalienable right; applies in resales unless b/t private people or acting for nonprofit museum.  Sliding rate by sales price. Its share of the art market grew over the past years. 

France: liable person is acquirer & law officials jointly; applicable sales are those via a public auction or dealer.  Uniform 3% rate, and artists must publish in Official Journal if they want the right.  UK: seller and relevant person jointly & severally liable; copy of work is generally excluded; collecting society and right to information exists; sliding rate based on sale price. 55% of dealers not affected; average processing time of 10 min. 81% of artists use the income to pay for living expensives.

In US, only California—royalties of 5% of sale price, held to violate dormant commerce clause.  Japan lacks resale right.  Potential artist benefit would be big on market of EUR 2.19 billion; 280,000 visual artists could be covered; manga should also be included.  China has no resale right—proposal to cover limited to auction sales and limited to increase in value.  Int’l treaty may be a good idea.

RT: Collecting societies have a bad reputation for not distributing money fairly especially where the uses are varying and the market is thin (i.e. especially outside music).  If you require registration, what’s the point of a collecting society?

A: Tracking use is a big problem with other collecting societies but discrete sales of physical objects are more identifiable, and collecting societies can handle the administrative costs more easily than individual artists.

Q: Why would we want to benefit successful artists and not unsuccessful ones, when resale rights are only valuable to a very few?

A: mitigating inequality.

Chiang: What is the problem you’re trying to solve? It’s not free riding.

A: musicians can make lots of money from sales of copies, but visual artists can’t, especially for works sold early in their lifetimes.  Sellers benefit from the subsequent creations of the artists so artists should also benefit.  Unfair.

WIPIP panel one: TM tarnishment and (c)

[title fixed because I can't keep seasons straight]
Suneal Bedi, Bad Brands: Experimental Studies in Trademark Tarnishment
What is the reputation of a mark?  Working on PhD in marketing at Penn; marketing is the best discipline to answer this Q.  Brand associations and attitudes are the reputation of a mark. TM isn’t just source indicator; it invokes nodes of meaning.  Attitudes= liking; associations= other meaning, like “Mexican” for Chipotle. Thinks that tarnishing should have to affect preferences/attitudes—should move Chipotle down in the ranked preferences/affect market share. So why not look just at sales? The statute is focused on reputation, which is distinct from decreased sales though it should predict such decreases. There can be instances of reputation harm w/o harm to sales.  WIP looks at isolating decreased sales; have to use sophisticated quantitative methods. 

What predicts tarnishment? 1. Multiple exposures.  At minimal exposure, people don’t really care. It’s an S shaped curve: middle exposure is high effectiveness; after that there’s a saturation point, which is why it doesn’t matter for Pepsi to advertise at the Super Bowl.  Tarnishment is competitive advertisement harming the senior mark. But only awareness & internalization of associations can do harm.  A few exposures or one won’t affect reputation.  Empirical strategy: multiple exposures.

2. High cognitive load.  Elaboration likelihood model predicts attitude change in 2 different ways: (1) central (rational and thoughtful); (2) peripheral (positive or negative cues, simple inferences). People use peripheral processing in times of high cognitive load.  A double-take: your mind is somewhere else, but as soon as you look close you aren’t really confused. That’s an instance of high cognitive load, and that may predict tarnishment.

Studies include pre-tested brand associations and attitudes for several brands. Chick Fil A; fictional tarnishing brand, Chicks Fill A gentleman’s club.  People prefer Chick Fil A lower after exposure and find it to be less wholesome.  It makes it a little different but not a lot; not on tastiness or liking at all. What happens about multiple exposures?  Embedded banner ads in news articles; asked people to tell them about the news article—high cognitive load.  Zero ads v. 4 ads: Then there was an effect on rank, tastiness, and liking.

[This is all an argument about the nonrational, nonfactual, non-thought based nature of tarnishment which under current 1A doctrine should put it off limits for the law, just as in the tobacco cases.]

Felix Wu: what implications for tarnishment cases?

A: courts aren’t using the right empirical evidence, or requiring the right empirical evidence. Both under and overinclusive—actual harm is underinclusive, but 6th Circuit’s presumptive harm is overinclusive.  If it exists narrowly, Ps must introduce good empirical evidence.

Lunney: The Streisand effect—most such uses you’d never hear of except for the litigation.  Debbie Does Dallas, Pillsbury Doughboy. 

A: The distribution of awareness—how many people see it; also of the people seeing it, how many times they see it.  Both matter.  People who saw the movie saw her once in the costume, according to Lunney.  He says that it’s not therefore likely to be tarnishing.

Q: what about the ridicule factor? It doesn’t necessarily matter what the “tarnishing” is but just the challenge, mockery.

A: but how do we know it’s a mockery? Only if it’s associated with something perverse.  [Wacky Packages]

Lunney: is it ok to garrote the demand for something through criticism, a la Campbell v. Acuff Rose?

A: Wal-Qaeda: Jacoby’s survey says there was tarnishment; but overriding concerns like free speech matter.   Neutral on whether this is a good idea (thinks it’s kind of dumb).  He’s ok with tarnishment not being something the law regulates—the point of this was to discuss the standard of proving tarnishment now, which is bad.

RT: Charts show 8 ad spike/recovery in ratings. 

A: he thinks it’s a demand effect b/c respondents started to say “I think you were trying to get me to like Chick Fil A more” when he asked them what they thought the survey was trying to prove when there were 8 ads. [This seems worth further investigation]

McKenna: what makes us think that consumers’ reactions have anything to do w/the fact that it’s ads and not other things that might produce exactly the same effect. [I noted that Victor’s Secret image he used was from the Simpsons, not the actual Victor’s Secret shop.]  If there’s nothing unique about what the law targets, it’s cherrypicking from the same activities.

A: what’s unique is that the point of using it is to draw attention to the junior mark specifically people because people are aware of it.  Criticism = not using Wal-Qaeda to draw attention/coopt.

McKenna: now you’re just talking about free riding.

A: yep.  That’s the actual harm of dilution/tarnishment.

Charles Duan, Implications of Oracle v. Google for Technical Standard-Setting

Is it an infringement to implement someone else’s interface? Most standardss come from private standard setting organizations that invite participants from multiple groups who propose ideas and then the SSO chooses for standardization. Ideas contributed by 3d parties.  Patentable inventions?  SSOs have very comprehensive patent policies trying to prevent patent claiming. But fewer have copyright policies.  W3C’s policy changed in 2015, but Oracle v. Google was 2014—may have been a response to that decision. Confirms his belief that programmers didn’t think APIs and standards were ©able. Should courts accommodate industry expectations?  There’s value to that, though the law is not bound to follow experts in the field.

SSOs as a practical matter may want to start adopting © licensing policies.  Standard-setting organizations like IEEE might want to be more involved in these discussions in ©, though members may have conflicting incentives.

Q: what about ASTM—software standards varying from other types of standards?

A: distinctions—gov’t mandate in ASTM case. That was copying the standards text and not implementing the standard—as if ASTM were challenging buildings made using their standards.

Carys Craig, Substantial Transformative Taking: Holistic Comparison and the Non-Infringing 'New Work'
Bakhtin’s idea of dialogue: what an utterance chain looks like from a dialogic perspective and how that should influence ©.  Utterance as work: four connected things: what has been said before, the already spoken (relational meaning); distal not yet spoken (the still to be said;) the proximal not yet spoken (anticipated response); distal already-spoken (broader cultural meaning). So how can © protect the already spoken while permitting the dialogic utterance in the chain and leaving room for the not yet said?

What is a copy?  Analysis at the moment of prima facie infringement at which you determine something is substantially similar. Canadian case: Cinar Corp. v. Robinson: pitch and pinch: someone pitches a story about a film, which is then made into another film.  Claude Robinson had spent years building materials for Adventures of Robinson Curiousité; then he saw it on TV anyway w/parallel characters, settings, scenes throughout.  On appeal, Cinar argued that court was wrong to compare the works without filtering out the unprotectable elements.  Altai abstraction filtration comparison test: Canadian SCt said it wasn’t appropriate for this type of work which has to be compared holistically.  Shouldn’t dissect work, which eviscerates it—hard to get a sense of the authorial contribution.  Comparison, abstraction, filtration, then compare again if necessary.  Criticized for undue focus on protecting author over downstream creators, influenced by very sympathetic facts.

Maybe we need to look at whether the defendant’s work is in its whole something new—new expression. Holistic comparison can be used to shift to asking about the dialogic relation b/t the texts and to see the D as a possible creator too. Could consider transformativeness of the taking into our assessment of substantial similarity, rather than waiting for fair use and fair dealing.  By the time you’re searching for an affirmative defense, you’re already on the wrong side of copyright’s moral equation—seeking a reprieve by calling yourself an author too. We could do this at the infringement stage instead.  From Canadian perspective: transformative use isn’t actually available as a defense, but we could put it in infringement. 

Example: Larrikin Music Pub’g v. EMI Songs Australia (Down Under/Kookaburra song)—line taken from common folk song and reproduced in hit; court could find no way to avoid conclusion that substantial part of the song had been taken, and there was no defense. Holistic consideration could find a new work even w/some copying of the “golden nugget.”  Substantial transformative taking should be regarded as an original work—not derivative but responsive, new utterance in the chain in a discursive struggle w/P’s already spoken work.

A: not that surprised—old English fair use cases were about new works being created; not framed as a defense at all.

Q: what happens to the derivative work right?

A: Canada doesn’t have one and neither should you. But you should also resist the collapse b/t derivative and dialogic.  A derivative work has to be a substantial reproduction so we can use this framing to get over the derivative hurdle.

Rosenblatt: comprehensive nonliteral similarity v. fragmented literal similarity?  You maybe want to distinguish them.

A: yes.  Sampling cases—solving a problem that can’t otherwise can’t be solved.  Blurred Lines: they failed to filter out what should have been. But sampling, you can do the filtering and still look like you’re left with “golden nuggets” that are substantial, even though there’s a new work.

Q: so how do you distinguish b/t dialogue and “pinching”? [I would frame this perhaps as: was Bakhtin’s formulation even set up to contemplate “infringement” or other acts that might be considered illegitimate moves in a dialogic chain? What if we care about things other than dialogue—can incentives fit in?]

A: good Q. Wouldn’t necessarily get Cinar off the hook but would limit attempt to control downstream uses?

Rob Walker, The Great Disappearing Doctrine: Making Sense of Scenès à Faire
Doctrinal oddities: has a descriptive take but no solutions yet.  Scenes a faire is apparently a term from 19th c French dramaturgy, but he hasn’t been able to find  much/any use there, so if someone can figure out what the judge who first used the concept meant that would be great.  [Paging Francesca Coppa.]

Hard to distinguish what courts are doing/what is or isn’t a scene a faire, though we have some definitions, e.g. for software it’s expression dictated by peractical realities; incidents, characters or settings which are as a practical matter indispensable, or at least standard, in the treatment of a given topic—in some ways this is just an application of originality. 

What do we mean by indispensable? May be clear in literary work with traditional structure, like Chekov’s gun.  But what about avant-garde works or genre-busting works? Isn’t this idea of indispensability just a judgment about what a work needs to be a “good” version of itself, contra Bleistein?  Is Star Wars a sf film, a fantasy film, or a western?  What are the conventions that might be standard to it?  Is an earnest treatment of a topic using conventions in a traditional realist way unprotected, whereas a parody of the conventions could be protected, e.g., Blazing Saddles and Young Frankenstein and Spaceballs?  What’s the time at which we set a convention—at the time the P’s work was created, or the D’s?  Using the conventions of Elizabethan drama to write a new play would be strikingly unusual today, but not in Elizabethan times.

As courts get into the analysis, scenes a faire tends to disappear into the ether and is replaced by idea/expression and merger. Example: Fulks v. Knowles-Carter, in which avant garde filmmaker sued for infringement in Beyonce’s Lemonade video.  Court’s attempt to pinpoint scenes a faire collapses into idea/expression and merger. The exception: software cases where functional elements, programming conventions etc. matter.  But no guidance in literary/artistic context. 

Biggest problem: shifts focus away from proper analysis of similarities b/t the works and their substantiality; talks about the overall genre or milieu instead.  Underprotects realist works and over protects genre defying or defining works, creating a Bleistein discrimination problem.  Sherlock Holmes: when written, new/creative (maybe; he hasn’t done the research), but at a certain point this character solidifies into a set of tropes copied in many different versions over time—eviscerated?  [Public domain?]

Fix: return to focus on originality at the time the accusing work was created. In software, focus on functional/nonfunctional.

Q: relational context from Rosenblatt & Craig—Fulks submitted his video to Beyonce’s producers.  You lose the relational element.  Taylor Swift: the court says that players gonna play is too banal.  Making old lyrics generic.  Timing: can Swift make anything generic through success? Then we lose something about the contribution of the earlier work.

Madison: Judicial biography: the judge who used the term first was committed to balancing access & rights—a pragmatic intervention along the fair use continuum; another way to get at similar scope problems.

Q: why when we think about genre works it would result in underprotection? Why wouldn’t recognizing contribution of individual work result in right-sizing?  Originality might not give us right-sized results; shifts us to thinking about what the author contributed, not what the author took.

A: Will look into relational interactions.  May change his under/overprotection analysis.

Rosenblatt: every scene a faire was invented by someone but we often don’t know by whom.  Functionally an expiration doctrine?

A: likes that.  Certain things shouldn’t be protectable. If we start thinking about originality and take Feist seriously—the bar is low, but it exists—that’s the right conversation.

RT: Doesn’t originality assessment require an assessment of prior art? The reason we try to figure out what other works have done is to figure out what shouldn’t be attributed to the original work of the accusing author.  And one function of scenes a faire is that it deals with what the patent folks might call “Obvious to try”: someone was the first to use the missed phone call instead of the missed appointment at the top of the skyscraper but that doesn’t mean they should get the rights to that plot point; see also the Jurassic Park case—once you come up with the idea of a dinosaur island, certain tropes follow for anyone with minimal exposure to narrative structures.  That’s the function of scenes a faire that allows it to parse originality.

WIPIP: Opening plenary

WIPIP Plenary
[Standard disclaimer: these are just my notes; I miss stuff or it’s about patents and I don’t necessarily get it; I also have to pick and choose from many attractive panels and this year I’m only at the first day, which makes me sad.]

Colleen Chien, Innovation, Inequality, Innovators: 4 hypothesis: Two-pie hypothesis: distribution of patented innovation supports two stories, increasing abundance and increasing scarcity; innovation  can increase or decrease equality and IP can play a role (both and); quality of innovation depends on inclusion across the innovation pipeline (inclusive innovation); to measure inclusion we can look at inequality and small entities.

1894, Animal trap patent, William Hooker of Illinois—at the time, Midwest was 2d top region for patenting. Mechanical patent, individual innovator.  1976, different representative patent: Goretex, Delaware, chemical patent, innovator working with his father.  2015, Larry Ellison, California patent on database automation; 8 other inventors named, many South Asian immigrants.  California has been the top patent origin for a while now, mostly just the Bay Area.  Moved away from mechanica, instruments, chemistry to electrical engineering—now 50% of patents.  Individual inventors are a small percentage now.  CS grad students: Temporary visa holders are now over 60%; white Americans going down, Hispanics, African-Americans, Asian-Americans holding tiny.  Foreign company patenting going up.

Bigger pie: digital abundance, involving diversity and high skilled immigration, Silicon Valley leads the world and provides consumer surplus. But: less innovation in mfg; increasing domination by foreign innovators; domestic brain drain.  So what is the relationship between innovation, IP, and inequality?  Inequality of production and consumption can differ. 

Innovation drives sorting, segregation, agglomeration and concentration—best institutions get bigger market share, tournament rewards for individuals who manage to make it to Google/Stanford. Patenting and 1% inequality are correlated including geographically.  Rigged system (in left and right wings); privileged patenting and tax avoidance.  Patent and copyright rents (left wing critique); tech companies buying up academics and lobbyists and capturing all the gains (right wing critique)—IP lobbying top clients include Alphabet, Comcast, iHeart Media, Microsoft whereas in 1998 it was the RIAA, BSA, MPAA while only the BSA is left in the top ten.  Golden rule: he who makes the gold makes the rules.  Privileged patenting: later patent rates for those w/top 3d grade math test scores is highly associated with parental income.

Inequality in consumption: overinvestment in markets where rents are possible--$400 juicer in Silicon Valley while tropical diseases go understudied.  Rent-seeking prices, not just b/c of patents, but painkillers that once cost $138 now cost $2979.

Innovation and decreasing inequality in production: innovation can foster growth and social mobility. Consumption: price of electronic goods has gone down a lot, except for cable service; Larry Ellison’s surplus creation is at least partly shared.  Top 10% patenting entities took home 26% of patents in 1900s, now over 60%.  There are industry effects; 83% of electrical engineering patents are held by the top 1% of holders.  Except for chemicals, there’s increasing concentration in other fields as well.  No evidence that first to file disadvantages small inventors, but further disaggregation of the data (to screen out university entities) is needed.  Small entity share declines over the patent lifecycle—filing, actual issuance, first, second, and third maintenance fees.  Are they dropping out or becoming larger? Seems to be dropping out. Give us pause: we encourage filing at the front end, but people in small entities don’t seem to be getting as much value from them, so are we encouraging them to waste their money?

Michael Meurer: isn’t this IT and inequality as much as it is IP and inequality?  Increasingly hard time showing up in your data for startups; perhaps it’s increased investment required in sufficient scale of IT that is a big part of that.

A: There are significant industry effects, true. But tech is not destiny.  Automation in textiles: low skilled textile workers gained, but high skilled artisans lost their jobs.  There are other non-IT specific mechanisms such as capture of institutions; education system; pharma lobbying to set the rules.

Michael Madison: Local/regional aspects—Cleveland, different than/overlapping with Silicon Valley. Encourage you to push on that part of the argument.  Industrial, economic, social history differs in different places.

RT: Having just finished Ta-Nehisi Coates’ recent book, I wonder whether this isn’t all the result of the specific history of suppression of African-Americans over centuries.  The IP professoriate too tracks the results of this suppression.  IP and IT may demonstrate the results but are they places to look for solutions?

A: Inequality in patenting is one way of tracking inclusion; it’s one way of rigorously tracking the pipeline.  There are some “valleys of death” in the pipeline; IP policy may not be the place for a fix, but it is one way to measure what we are losing.  Institional history of patent system in specific: patenting was initially limited to citizens, no slaves or nonwhite foreigners; married women’s patents were also questionable until the late 1880s.  Article on the invention of the slave: cotton gin etc. had many contributions from slaves but masters took the credit.

Mark Lemley: This maps to all US history: all the trends you describe look like trends in the economy generally/patents aren’t actually different than the economy but move in lockstep with it, e.g. Gini coefficient.

A: Yes, as an empiricist wanted to see what was going on w/innovation system and its self-image as egalitarian and meritocratic.

Saurabh Vishnubhakat, Rethinking Patent Law's Validity Power
Power used to reside ex ante in the PTO, ex post primarily in the courts; ex post power has been progressively reallocated from courts to PTO: Bayh-Dole Act, 1980, American Inventors Protection Act, 1999, America Invents Act, 2011—largely a story about expertise. Neglected story: a desire for political input. A certain industry/tech, business methods, has been singled out for special treatment, which isn’t just an expertise story.

How it’s been used in practice is part of the politicization of the patent process, particularly the validity power: Adjudicatory process has been used by PTO to do both things Congress intended and things it didn’t/stakeholders were worried about.  PTAB panels have been packed w/admin judges who are picked by those on high. Judicial oversight: resisting review of case selection (PTO uses nonreviewable decisions as shield from Fed Cir review); evading review of statutory boundaries; evading review of adjudicatory duties.

Panel stacking: 3 APJs per panel usually, but can be expanded, which is not random. Supposed to include field experts.  Members are chosen by agency leadership and can include leadership.  One senior and one junior person at least—to ensure that new people are acculturated.  The effect is to control outcomes, especially on rehearing: rehearing may be granted with a different, larger panel.  Colloquoy b/t judge and PTO counsel: “the Director is trying to ensure that her policy position is being enforced by the panels.” Consistency & uniformity are desirable, but there are rule of law concerns.  Target v. Destination Maternity: original panel was 3 APJs; were about to decide an issue in a way leadership would have disagreed w; before decision expanded sua sponte to 5 including 2 lead APJs, but still got 3-2 decision holding firm; panel was expanded again, to 7 with 2 more senior APJs, and finally got the “right” outcome, 4-3 the other way.

Judicial oversight: more subtle, more defensible but still ultimately problematic. Past: a finding that no substantial Q of patentability existed was nonreviewable.  Should a decision to proceed also be nonreviewable, meaning no interlocutory review?  Preliminary decision to proceed would merge into final merits decision, if any. Cuozzo majority disagreed; best argument was the canon against surplusage, given what the APA says.  But then what does the statute mean for a decision to proceed in order to be “final and nonappealable”?  The high standard for presumption of judicial review to be overturned from the APA was not met, in SV’s opinion.  Cuozzo was a strong victory for the PTO, allowing additional claims to unreviewable discretion. Why is this so potent a vehicle for policy autonomy?  The validity power is exercised separately—power to screen cases for likely merit, and separate power to adjudicate cases actually chosen. Decision to screen is directed to Director by statute, which is subdelegated to the PTAB which has power to adjudicate, but there’s no statutory command and it may be problematic to have the same panel do both. Screening is unreviewable, and adjudication power is reviewable.  That reviewability is a key factor in the argument that the scheme is constitutional and appropriate. Natural agency incentive: make things look more like screening than adjudication, and it’s not that hard to do b/c screening std is whether there’s likely success on the merits.

Stacking has some justification in uniformity; SV thinks that there’s even an argument for Chevron deference when the Director is sufficiently clear.

Costs: injury to stable property rights; injury to credible commitments of Congress and agency; self-reinforcing injury to oversight—can you trust the Directors who come after not to exercise their political power against us?

Solutions: make it more judicial.  Have some decisions be precedential if you like them; make decisions precedential and train APJs to follow them.  Done for ex ante clarity.  Also: separate screening from adjudicating, raising the cost for the agency to evade review.

Betsy Rosenblatt, Copyright’s One-Way Appropriation Ratchet
Rolling Stone review: Beck incorporates a “cross pollination” of styles which shows him to be one of the most innovative and forward-looking artists—he used hip hop, soul, etc.
Courts sent a very different message from Rolling Stone when hip hop artists sample—start with scare quotes around “rap music” and intoned “thou shall not steal.”  Similar message to NWA from 6th Circuit: get a license or do not sample; this doesn’t stifle creativity at all.
In 2016, the 9th Circuit disagreed w/6th Circuit for de minimis copying; the infringer in the 6th was the NWA, whereas in the 9th the noninfringing defendant was Madonna, and the plaintiff was a group of African-American musicians, even though Madonna adopted the cachet of a subculture not her own.
If you make slides for class of music cases, though there are exceptions like He’s So Fine, mostly white artists appropriate with impunity and are often lauded, but minorities who appropriate find legal challenge and often moral condemnation.  This is a social as well as legal product, reflecting societal and judicial bias but it’s also systematically baked in to the discourse of value in © law and practice.

© presents as facially neutral but inherently/discursively assigns value to some expression over others. Doctrine combine with judicial attitudes and risk imbalance to reflect and reinforce racial bias.

Responses are often to give more ownership to oppressed peoples, and there is sense to this, especially if we’re interested in profit maximization.  Merges/Hughes proposal.  If one’s goal is to create a few especially wealthy members of minority groups, expanding substantial similarity is a tool, but that is unlikely to promote equality broadly and ignores the wider effects on creativity by favoring artificial concept of originality.  That proposal also doesn’t help those who don’t have resources already or who want to talk back to inequality.  Don’t double down on a flawed system; adjustment has to come by making it thinner in a way that recognizes the dialogic nature of expression.

The traditional is reduced to the primitive; the doctrine freezes works and attributes ownership to anyone who fixed them.  Devalues improvisation and creates a “right” version of the work. Derivers from “unauthored” works get to own the whole thing; this lets them create the narratives/colonize the work of minority cultures, taking from but not giving back—Deep Forest can stop others from similarly using the work of Solomon Islanders. Makes it seem that they have more value than those they’ve built upon—assigns value to things that might just be practical limits on ©’s grant of exclusivity.

Have to use dominant works to critique dominant narratives, but cramming this into fair use distorts the activity—makes Alice Randall call The Wind Done Gone a parody.  Makes the use for critique more risky.  Semiotic disobedience would be even harder w/thicker ©.  People who don’t trust the system are more likely to withdraw/refrain under conditions of uncertainty.

Minorities are less likely to register, to own (as opposed to transferring—Spike Lee in Malcom X; George Clinton’s works are owned by an unrelated party).

Takeaway: more of the same isn’t better. Not interested in tearing down copyright.  More recognition that all communication is in some way derivative.  A deriver should be understood to own only what’s contributed to the “prior art”—allow more derivation and appropriation by all. 

Vishnubhakat: you seem to posit crowding out of © space; doesn’t that mean that if you allowed closer imitations the “more” innovative minority voices would be crowded out?

A: Doesn’t posit crowding out; doesn’t mind appropriation as long as the appropriators don’t get as much out of it as they do now.  Doesn’t see it as a zero sum game: there’s an infinite capacity for creative expression.

Lunney: Note that the most wealthy African-Americans come disproportionately from the © space compared to whites; should that change the analysis?

A: No: many of them are actually from negative spaces like fashion, cuisine, talk shows, the famously constrained hip hop. Also making a few people rich is not a social justice strategy.

Q: One way downward ratchet?  Some examples of cases are minority v. minority. Should we relax the standard for fixation? 

A: gut instinct is that relaxing fixation would benefit improvisers in particular, which might be good. But it has all sorts of practical difficulties.  A lot of these doctrines exist for good, practical facially neutral reasons. We should think about mitigating their disparate impact.  Termination of transfers also has its place.