WIPIP, part 2: Copyright

Posted on February 9, 2015 by rtushnet
Session 1, Copyright
 
Jim Gibson [with Chris Cotropia]: Random sample of copyright cases filed 20080-2011, coded 957 cases for parties, claims, remedies requested, final adjudication. The filesharing cases and PRO cases (music) tell the expected IP story. Ps from a core © industry, revenue/incentive arguably dependent on ©, Ds engaging in wholesale copying/priacy. In contrast, the commonplace cases were surprising. Many Ps from non-core industries; many works from low authorship subject matter. Useful articles and non-copyright claims associated with © claims: TM, breach of contract.  So, how does this work with incentives? 
 
Ps were predominantly small firms, a bit over 64%, 72% of Ds were also small firms.  Individuals: 21% P, 13% D.  Industry of lead party: at the top: apparel/fashion/textiles, 13.568%, software 12.79%. By contrast, fine arts, performing arts, video games, public sector: very low.  Core IP industries, according to WIPO=advertising & marketing, commercial arts, Film and TV, fine arts, music, performing arts, publishing and software—1/2 of Ps are non-core.  Subject matter: literary work/software 13%, commercial art 12%–highest percentage.
 
Non-© claims: 61.62% had some other claim. TM: 35.23%, breach of contract, 20.10%, other, 48.04%.
 
How hard fought was the litigation? Number of times judge has to step in and settle a dispute b/t the parties is one measure.  Motions asking judges to make a decision: not much difference as b/t high and low authorship works; not much difference between P core industry and P non-core; not much difference (larger but not statistically significant) for intra-industry disputes versus extra-industry.
 
What else can we do going forward?  Possibilities: Was work revenue-generating? Is P’s business © driven?  Was there a preexisting relationship—complete strangers, bad breakup, make or buy decision that went bad? Mess with your competitor case? One dentist sued a rival dentist over similar ad copy.  What else to code re: authors as parties; our definitions included WFH.
 
So what if incentives aren’t playing out in the courtroom?  Resist extrapolation because of selection bias.  But there may be takeaways, such as whether we need a small claims court proceeding.  Limited reach of © litigation reform: will that get at problems we see in © generally? 
 
Policy consequences: higher threshold requirements?  Higher pleading standard? Shape of the shadow of the law: if non-core cases primarily inform doctrine, that has implications.
 
RT: Relationship between core and non-core and presence of other claims?  Do TM/© pairs have a particular pattern? [McKenna’s channeling? If you do see a group of hurt-your-competitor cases, we might want to be more robust about requiring people to pick a theory. There’s discussion of bleed between © and TM, and is this a source?] [may justify some disparate treatment of music—if cases are really siloed into types, then music-specific doctrine may not be as damaging as we sometimes think it is.]
 
Lunney: Are these leftover Dastar TM claims?
 
A: we coded for presence of unfair competition/§43—trade dress/a lot of useful article claims. But we should look deeper. There weren’t a lot of quiet title/ownership disputes. 
 
McGeveran: what about outcomes? How many settle, how many go all the way?
 
Patrick Goold, IP Law and the Bundle of Torts
 
Patent infringement as a tort?  Statutory, not common law, but analogous to common law torts in orthodoxy.  View that there is a unified, singular, discrete tort labeled © infringement, same as there is trespass or battery, and same for patent.  Is that really true?  His thesis: not really.
 
© is not a tort, but a set of torts. Reproduction is different from public display. Likewise, in patent, making, using, and selling are different torts. There is no such thing as the tort of real property.  There is a set of related torts: private nuisance, negligence, waste, trespass.  We should think about the work/invention the same way as real property: an object worthy of protection, and then a set of legal wrongs that can be done with respect to that work. Reproduction is different from public display as trespass differs from waste.
 
Bundle of rights = bundle of wrongs.  Infringe right to exclude = trespass. Right to enjoy = nuisance.  Physical integrity = damages, usually negligence.  Same with IP. Right of reproduction = wrong of reproduction.
 
Why does this matter?  Liability is one reason.  In other areas of tort, you see a spectrum of different liability regimes across the bundle.  Liability for trespass isn’t the same as for private nuisance.  Trespass is archetypal strict liability.  Need not prove harm or fault. Private nuisance—some say there needs to be an element of harm and element of fault. Negligence: fault and harm.
 
©’s unitary liability regime, most people would call it strict liability, but no advocacy for differentiation between the rights.  People have tended to say that there should be some liability regime that applies across the rights.  Same with patent—making, using, selling treated the same despite their different economic characteristics.  Reproduction and distribution in ©: economic theory says you should use strict liability when only the D can take care to prevent the tort. Use negligence in cases of bilateral care, when both can act.

Reproduction: the wrong is copying.  P could attach © notice to prevent accidental copying; TPMs could also prevent copying. This is an issue of bilateral care, so negligence rule might be appropriate to give author incentive to give notice.  Distribution: the wrong isn’t the embodiment of work in copy but distribution of the infringing copy thereafter. D can take care; author’s ability to prevent this is far reduced.
 
Another implication: IP over-inclusivity?  Focus on the idea that there is one singular wrong of copyright infringement tends to result in people thinking that there is a wrong of copying. It’s a short step to overinclusivity.  Copy-fetishism: Jessica Litman. The belief that every copy must be licensed or excused.  Link to idea of “wrong of copying.” Think less of “the tort of copyright infringement,” then we can move away from the fetish.
 
RT: [Do statutory limitations matter here in the conceptualizations of the right? Educational exemptions do target particular rights.  Compare fair use: reason to make it unitary. Relatedly, overlap in rights in digital age: Tony Reese.  And overlap with derivative works right and some other right, in almost any imaginable circumstance.  Also, possible comparison to move to unitary standard for online/intangible torts in §230 and European Directive, at least as to the gatekeepers whose conduct is generally thought important to target.  Maybe for gatekeepers there is a tort of causing harm online.]
 
A: first sale also comes to mind.  Those limitations define the scope of the right in question.  Distribution right has different scope from reproduction.  Liability regimes = different plane. Conduct that infringes the right—conditions of harm and fault that need to be addressed to determine whether there’s an infringement.  [I’m not sure I get this.  Which are exemptions? Can’t you characterize them either way?]
 
Fair use—maybe it should be unitary.  Common law = disaggregated bundle.  Civil/European law’s unified system might do better with a unitary idea. 
 
Q: overinclusivity might not be because the 76 Act focuses on copying. 6 exclusive rights, only one of which is reproduction; the others are overinclusive for other reasons.
 
A: true, even if you unbundled, each right would be overinclusive.  His argument is on top of that.  Copying covers things like distribution in our discourse, which is not logically clear.
 
Irina D. Manta (and Robert E. Wagner), IP Infringement as Vandalism
 
Rhetoric of theft: “you wouldn’t steal a handbag—downloading pirated films is stealing.”
 
Why might it be theft?  Both infringement and theft take things of value; consistency in enforcement. Anti: no complete deprivation of work; owner can continue selling copies of work; loss is difficult to calculate.
 
We argue: IP infringement is better characterized as vandalism or trespass than theft. If we were consistent on how we apply sanctions this would lead to a reduction of sanctions in the IP space.
 
Rhetoric of theft is old, and emphasizes gravity of the conduct. But it turns out to be difficult to define stealing: circular—taking something that’s not one’s own.  Doctrines like adverse possession, easements that create exceptions.  Justifications for theft label: incentive theory—author mixes labor with public domain, needs reward; stealing takes away reward.  Both stealing in property and IP infringement involve a form of free riding on the efforts of the owner.  That could lead to lost sales directly and indirectly for the IP owner.  That leads to rivalrousness: there are losses for owners/buyers if a good is devalued if too many people have it.
 
Problems w/theft label: Owner retains a copy!  Almost impossible to strip an IP good of all of its value.  Significant causation questions w/r/t harm.  What harm does an individual infringer do?  Lower risk of altercations in the IP context.  Last, IP law is much more disconnected from popular norms than property/theft law is.
 
Courts: not always clear but often refer to IP theft or piracy. US v. Dowling: SCt was unconvinced it was “theft,” but unsettled.
 
Our argument: vandalism or in some cases trespass. Like vandalism, there’s a destruction of some but not all value.  Owner can still/license good.  Vandalism like IP infringement can enhance value. (Banksy: society gets the kind of vandalism it deserves.)  Limitations: infringement doesn’t harm the original copy; generally no financial free-riding, though there are hedonic benefits; vandalism has to create damage or it doesn’t count as such—more like trespass (DMCA violations?).
 
Consequences: takes rhetorical punch from label; raises questions about possible punishment level. Sentence comparisons: theft. Punishments for © infringement are generally much harsher than for theft. Same thing for vandalism.  For same “value,” you’re better off being a thief or a vandal than an IP infringer. DMCA = 5 years in prison, $500,000 fine, while trespass leads to fairly small fines and in Texas (worst) up to 180 days in prison.

Generally treat IP infringers more harshly than thieves.  Especially puzzling given actus reus occurs more quickly in © and so one can accumulate more in a short timespan. Mens rea could be a few seconds, followed by realization of wrongdoing. Sentence disparity may not be principled but rather consequence of nature of federal law and political forces.
 
RT: suggest drawing connections between theft language and “broken windows” theories.  Implications for equality/IP enforcement in physical world is highest among immigrants.  However, broken windows theory provides one possible justification for high enforcement levels: this is more important than you think it is b/c broken windows create larger cracks in communities. This would also lead you to draw connections between actual penalties (rarely given) and low-level enforcement (omnipresent in certain communities, lifestyle offenses used to regulate life generally—connection w/Julie Cohen on surveillance state and IP maximalism, © owners’ hopes to create that kind of low-level enforcement online with copyright alerts).
 
A: lack of catching people isn’t as important—choice whether to enforce, not ability.  Fairness and justice is a concern. [I wasn’t really talking about the “need higher penalties because chance of being caught is lower” argument.  I was talking about the expressive and order-maintaining function allegedly served by policing against vandalism, and the effects of the same on surveilled populations as in Alice Goffman’s very interesting recent book.]
 
Q: Mindy Kaling has a great routine in response to the “infringement = theft” claim.  Framing issues: property or limited monopoly.
 
A: Even though we think © is about incentives, people in the world think it’s about natural rights. 
 
Q: other analogies—pollution, riparian rights? There’s something about vandalism that doesn’t get us that far away from theft. [I agree—there’s an implication of disrespect that I think is very strong, leading to a stronger moral rights conception if that’s your thing, and also interference with exclusivity is very different.]
 
A: Pollution is interesting, but not exactly the same.
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in copyright, http://schemas.google.com/blogger/2008/kind#post | Leave a comment

WIPIP Session 4, Cross-IP

Posted on February 6, 2015 by rtushnet
Mark Lemley (& Mark McKenna), Scope
 
Midnight in the Garden of Good & Evil copyright infringement case.  Court rejects invalidity claim: the photo has some creative elements.  So it has to go to a jury on infringement. But we don’t tell them to filter out the elements the photographer didn’t create—filtering is only in our instructions on actual copying; infringement/substantial similarity is a gestalt.  In TM: Reynolds Wrap v. Handi-Foil. Court says that the trade dress is legitimate but narrow: blue combined with light red combined with stripes.  But then the court says on infringement that juries could find infringing similarity. What’s striking about the similarities? They say non-stick and heavy-duty, food lifts off, square footage is the same and displayed in the same place, and Made in USA language.  There is certainly similarity, but entirely in things we’ve previously said are unprotectable under TM and shouldn’t be the basis for an infringement finding. And then there’s the iPad design patent case and Apple’s patent on the rectangle with slightly rounded corners.  DCt denied PI based on invalidity, Fed. Cir. says no, they’re almost certainly going to win; jury ultimately found noninfringement (not invalidity). But one reason is that, b/c of quirk of design patent law, they got to look at the prior art. Makes clear that they don’t own all rectangles w/rounded corners—Apple must own something smaller than that.
 
General problem: we chop up world into infringement, validity, defenses, decided at different times often by different decisionmakers. This creates creep in the rights, and the only way we know to fight it is to deny all rights. What IP law needs is an integrated scope doctrine: think about validity and infringement and defenses this way. One IP regime has a step towards that: utility patent has Markman hearing.  What that does right: asks a single Q, what is the scope of the patent, prior to a validity or infringement determination.  Markmanfocuses on words written by lawyers and not on actual invention; this is a mistake. Patent is also not immune from the problem—it says there’s no defense of practicing the prior art.  What the court means is probably that standards for invalidity and infringement differ and we don’t want you to smuggle invalidity evidence that doesn’t meet the standard into your infringement case.  But of course there should be a defense of practicing the prior art. Integrated scope proceeding would work; could lead a case to end b/c the P is claiming more than it has, but even if the case doesn’t end it’s an opportunity to articulate what’s protectable about the IP right.
 
Greg Vetter: Does trade secret also match as a proceeding where we do validity and infringement and defenses all together? What about unfair competition/misappropriation?
 
A: Trade secret has been separated from misappropriation and put into IP, and that’s led to a greater separation b/t validity and infringement; beneficial to treat it is IP in many ways but this is not a benefit. Courts if they thought you were a bad actor were willing to overlook the absence of a secret; we want to avoid that sort of prejudice in an integrated proceeding.
 
Q: Does estoppel help here?
 
A: every once in a while, but often they don’t use it.  Reynolds Wrap is an example. Some other doctrines otherwise hard to explain are haphazard efforts to manage this problem: “thin” copyright requiring virtual identity—makes it harder for overclaiming to occur.  Merger doctrine too.
 
Courts have a natural tendency to make boxes. That’s odd way to treat a common law doctrine. Courts are more comfortable if they feel they’re checking off what someone else decided; less comfortable deciding breadth.  But they should be.
 
Ramsey: is the law ok and judges doing it wrong? Or is the law wrong? 
 
A: we’re not arguing the doctrines are wrong. But once we find validity—the TM was almost functional but wasn’t fully functional so it passes—we ignore that at the infringement stage where the elements D copied are functional. So we need to cross the barriers.
 
Q: So should judges write a claim?  Should we get rid of juries? 
 
A: big difference b/t ex ante claim written by lawyer and ex post determination by judge. Much more comfortable with the latter as getting to right answer.  Circumstances exist in which words will help, especially if one has to instruct a jury.  Don’t find liability based on unprotectable similarities. Side by side comparison could do a lot more, though Egyptian Goddess sadly moves towards separation of validity and infringement for design patents.  It’s correct to say that judge is more likely to get right result than juries, but one implication of scope analysis is that some cases will fall out before reaching the jury once you take seriously what’s actually protected.
 
[NB: I’m not sure I agree w/the TM example. Arguably, if there is a secondary meaning in a trade dress—which might not really happen in these cases—then the fact that it’s mostly functional may put a duty on others to stay further away from the nonfunctional aspects than they otherwise would.]
 
Ari Waldman, Trust: The Distinction Between the Private and the Public in IP Law
 
Public/private distinction drawing is foundational Q of privacy law, and also for IP scholars. There are problems if either side gets too big.  In privacy/constitutional law, the public tends to crowd out the private.  But there’s also a problem in IP when we define the public too large in the context of minimal disclosures.  Public use bar—if you disclose/use/demonstrate invention you can’t get a patent.
 
Who wins/loses public use bar cases?  You can cluster winners and losers (only about 30 so far). Lone inventor versus large inventors have public use cases.  Lone inventors tend to lose public use bar cases; IBMs tend to win.  More research has to be done, but wants to think about possible reasons.
 
Rule: inventor must maintain control over invention during the use, it’s considered private, but if you relinquish control, it’s public.  Too often, that retention of control = assumption of risk doctrine that you run the risk someone will talk about it. Corporate inventors have extensive legal armies/cachet that allow them to force collaboration partners to sign confidentiality agreements.
 
Courts honor norms of big inventors, not young/small ones—the latter tend to ask friends and family. Norms of confidentiality exist but without the formality that exists in corporate settings. Courts don’t appreciate the norms in different social networks.
 
If you’re agnostic about privileging one type over another—not saying individual is better—it’s still not a good idea to privilege one set simply because the doctrine ignores uniqueness/variations of social norms from network to network.  Courts privilege formally negotiated agreements between collaborators, and arguably shouldn’t, controlling for other factors.
 
McGeveran: must first justify using inventor’s perspective.  Are individual’s perceptions a meaningful guide to what ought to be considered private? For privacy, maybe—goal is to protect reasonable expectations. For patent, maybe not, if our goal is to get inventions and enhance public knowledge.
 
A: some evidence that purposes of patent are also met by this type of analysis.  Experimentation: goal is to make the inventions the best possible.  A changed vision of how we control public use would enhance that.
 
RT: how much do these inventors know about the on sale bar? Is there any way that tweaking the rule could incentivize them?
 
A: Not sure they’re affected, but individuals generally reflect/respond to the law. 
 
Q: solo inventors are often norm-jumping from informal to formal.  Collaborative IBM types aren’t.  One message is that we can’t switch norms midway through the story. Can’t get a powerful exclusion right because you started in the world of your friends.  Inventor wants patent right—powerful market-based brutal and impersonal rights. Why not make them play the norm game from the outset in the impersonal mode?
 
A: but is that offensive to other patent law norms/frustrating other goals by making this contextual analysis?
 
Q: Paper apparently has a trade secret piece, which is stronger.  Your insight from privacy is about control.  When you choose to disclose to your inner circle, you haven’t given up on privacy—danah boyd, young people do care about privacy but manage it differently.  Trade secrecy is a good place for that.  [Analogy to naked licensing in TM might be helpful here—the doctrine there does recognize context.]
 
A: broader conversation about relational norms of trust—read his book when it comes out!
 
Jessica Silbey, IP and Constitutional Equality
 
Progress Clause: one part of project comes from qualitative/empirical data, about what people working in creative industries think progress is.  Instances of market failure.
 
Today’s cases: SCt.  What progress do they imagine?  If SCt is thinking of deeply rooted equality doctrines, that’s worth thinking about and responding to.
 
Two dominant strains of equality doctrine in two cases—both Justice Ginsburg, but can also be done with Aereo, Kirtsaeng, Petrella.  Eldred: Aristotelian equality promise of likes being treated alike. Classic neutrality—similarly situated classes treated the same; difference needs justification; deference goes up as the class gets less suspect. 7 members of Court said that Congress could extend 20 years to existing, not just future, copyrights.  What incentive rationale is there? What limit is there on that? The Court didn’t have factual explanations for this—a lot of hypothesizing, which is fine under rational basis, but what really might have been going on is an understanding of the value of formal neutrality in application—treating all copyright owners the same. Language: parity, alignment, even-handedness, existing and future copyrights “alike.” Harmony, sameness—these words show up again and again.  Personalization of the benefit for copyright owners is not about incentives but about the dignity of equal treatment. 
 
Rudimentary, and there are problems with the theory. Ignores the democratic flaws in copyright legislation; assumes all authors are similarly situated when they’re not.  Failed to consider those left out, while enacting a story about inclusiveness. A classic problem of formal neutrality. We talk about this a lot in equal protection law.  Justice Ginsburg knows that all too well.  Just application of neutrality depends on starting line being relatively equal for all those being compared.
 
Golan is actually about antisubordination. © restoration for foreign works in public domain because of failure to comply with formalities. Court saw 104A as reparations for unfair losses in previous years.  Reciprocation—foreign works put on equal footing w/US counterparts. Not about treating likes alike; it’s the other side of the coin: justifying antihierarchy approach where constitutional equality dehierarchizes unjust status differences.  Targeted special benefits are ok—affirmative action reverses unfair deprivation.  Language: foreign authors “subordinated” to domestic; 104A is a “remedy,” restoring authors to the position they would have had—that is language from discrimination cases.  Critiques Breyer as American exceptionalist, which also resonates.
 
Problem with that too.  Ginsburg says: Q is whether users must pay or limit their exploitation to fair use.  Rights must be obtained from marketplace, as they must be from US authors. This sounds like equal pay for equal work.  Contrary to equal pay laws, where there are no losers but discriminators and about whom we don’t care, 104A causes real harm to people who didn’t do anything wrong—now forced to pay. Leveling up ignores harm to stakeholders who don’t have political power—misses bigger picture about benefits for public domain, thinking © benefits only authors and not the public.
 
Petrella: also a direct response to the Lily Ledbetter case in which Ginsburg dissented.  Is this fundamental to the argument? Fortuitous?  Is it a complement to or displacement of a traditional property framework? What does it have to do with Progress at all? Blank check to Congress?
 
Vetter: would this mean that in AIA the fact that we left patents through March 2013 as first to invent and not first to file would be potentially unconstitutional?
 
A: not trying to determine constitutionality, but explaining cases through other frameworks, not necessarily as predictive or normative. From perspective of IP on the ground: SCt cases read strangely, at level of generality that feels unrooted.
 
RT:  Golan argument, you say that this ignores harm to the public.  But they were wrongdoers in this account, right?  Compare Mark Twain/piracy discourses—wrong even if legal.
 
Jonathan Mazur: Ginsburg seems to be imputing normative judgment to Congress—meant to treat likes alike. Does that make a difference v. constitution?
 
A: It’s not that someone made an equal protection argument.  Golanis right w/r/t deference.  But then why is the language there?  Equal protection jurisprudence is a leaky doctrine.  It changed constitutional law generally, became a fundamental value. Q is whether it does here.
 
Q: Why use a construct that is a red flag to colleagues?
 
A: That is exactly what happens in Aereo—Scalia calls the majority out. Kirtsaeng is a split, and looks more like a substantive equality case.  Might see more splits/disputes over proper framework. 
 
Q: Lexmark: treat them all the same?  [That’s super interesting, especially since you’re seeing resistance in the lower courts to applying Lexmark to §43(a)(1)(A) even though the opinion clearly instructs that should happen.]
 
Gerhardt: is equality the new “traditional contour”?
 
A: if so, we need to put pressure on how the lines are drawn, how the class is defined. Constitutional lawyers think a lot about that.  We need to do the same.
 
Gregory Mandel, The Plagiarism Fallacy in Intellectual Property Law
 
Public perception: IP law designed to prevent plagiarism.  Experiment—trying to figure out what American adults using Mechanical Turk thought about copying.  Copying someone’s creative product: 20% conditionally acceptable; 78% not acceptable.  Why/why not: 78%: ethical/moral. 6% mention any legal basis.  Response examples: “copying someone else’s work and taking credit for it is theft.”  “People should get credit where credit is due.”  Credit/misplaced attribution was the greatest concern.
 
Popular understandings of purpose of IP law.  Developed brief explanations of incentives, natural rights, expressive rights, and plagiarism; participants asked to rank by agreement w/them as justifications for IP protection. Plagiarism: 37% incentives and natural rights, 26% each, expressive rights 11%.
 
In-depth look towards IP in specific contexts: creative subject matter: book, music, painting, medicine, electronics, software. Type of copying: idea/expression (not full product)/copying creative product.  Perspective: what law is and what law should be.
 
Results: plagiarism fallacy appears to be rooted in and partly a cause of widespread perception that IP rights are too strong and too broad.  In identical scenarios, participants believed copying should be permitted to a greater extent than law allows. 
 
In each subject area, the copying in the idea condition would be permissible, complete copying not; expression (e.g., copying some qualities and chorus of song but not entire song; reverse engineering and duplicating patented chip).  Higher numbers = more in favor of copying—largely in favor of copying ideas. But in 4/6 partial copying scenarios, public believes copying should be allowed, and for music, complete copying is ok, and books and medical devices above 40%.  So preferences are highly contextual.
 
Tested mitigating factors: copying for educational purposes; noncommercial; permission; attribution. Results: baseline: permission made it nearly ok to copy.  Attribution, educational use, and noncommerciality all should reduce infringement liability.  All differences statistically significant.  Educational use generally does better than attribution.  The majority of people think that simply providing attribution should enable the free copying of intellectual works/inventions—not just downloading a song on the internet but other creative works too.
 
Exposing the plagiarism fallacy explains some puzzling behavior: YouTube videos that say “no copyright infringement intended” (or even “no copyright intended”).  Nonresponse to “infringement is theft” campaigns.  Many people may agree with “theft,” but have a different understanding of the meaning of “theft” than IP owners. People hear: don’t claim credit for someone else’s creative endeavors, and they think, “I can do that.”  People aren’t dismissive towards IP rights, but they understand them differently. 
 
Further findings: public is ignorant of IP law.  Multiple choice quiz—average of 4/10 right, 1.5 above chance.  Knowledge of IP law doesn’t affect opinions about what the law should be: if they did really well on the quiz, they still don’t change opinions about what IP should protect, suggesting that info campaigns are unlikely to change views.  The public views patents and © relatively similarly.  Answers across subject matters vary widely, but that variation seems subject matter dependent, not about artistic domain v inventive domain.  Variation isn’t across copyright/patent divide.
 
Demographic variation in preferences.  Older people, women, conservatives, and wealthier people tend to believe in stronger IP rights, and tend to report greater self-compliance w/IP rights than alternate groups.
 
Silbey: Findings on software: where we think rights should be weakest, public thinks rights should be strongest.  Does that have to do with industry status, distance from industry (people paint).
 
A: medical scenario involved vaccine, not familiar.  (Silbey says: that saves people, is understandable.)  Scenario was someone who writes a computer program copying functionality—reverse engineering.
 
RT: so the people who say they believe in strong IP rights, do they believe in the plagiarism norm very very strongly, or do they believe something else? How do they respond to things like educational/attributed use?
 
A: Only difference we really saw: People who believe in expressive basis favor weaker rights.  We are going to study differences in response to changed scenarios.
 
Sheff: most people on MTurk are consumers and not producers. If you think IP is distributive, it might not be persuasive to have respondents only from one side of the distributive problem. Could you manipulate respondents to be creators, like Sprigman, Fromer, and Buccafusco?
 
A: our hope is to run the same survey with creators.
 
Heymann: not surprising that people think IP is plagiarism because their first encounters in research papers involve attribution norms.
 
Q: relevance of TM examples?
 
A: we were surprised by plagiarism results, but TM is about attribution.
 
Lemley: if we think this is troubling, is it the law that should conform to people’s views or should people be educated about the law?
 
A: doesn’t take the position that belief should = law. We can have other objectives. Strong point: represents dominant view among users and some creators. Can’t expect IP system to function the way we want for incentives if there’s this widespread disconnect about the law.  Think about how we are going to get greater penetration among the public.  Look at creators.  [We could create an absolute educational exemption.  That might not have a huge effect on incentives but could really help.]
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in copyright, design patent, http://schemas.google.com/blogger/2008/kind#post, patent, trade secrets, trademark | Leave a comment

WIPIP Session 3: Copyright

Posted on February 6, 2015 by rtushnet
Session 3, Copyright
Abraham Bell (& Gideon Parchomovsky), Copyright Trust
 
Incentives to create: if we recognize too few owners, then we haven’t incentivized them enough.  If we have too many owners, we disturb the efficiency of use.  Many owners with power to license creates an effective commons, because the price of licensing would be driven down; play them off against each other. Many owners with vetoes = effective anticommons.
 
Many doctrines: work for hire, joint authorship, collective works, transfer, sole authorship, implied license.  Tasini: easy to end up outside the collective works privilege. Transfers themselves are imperfect/subject to termination.

Their idea: copyright trust, alternative form of ownership. Could be adopted by parties by contract or by court after the fact. Not all or nothing; greater v. lesser. Division between beneficial and trustee ownership, w/owner trustee and owner-beneficiary.  Beneficial shares may be unequal. Standard fiduciary duties. Mastermind will control the use of the work but not necessarily get all the profits.
 
Sources of inspiration: property, allowing division of equitable and legal ownership; corporations law—division of ownership and control.
 
Q: is this mandatory or voluntary?  If voluntary, they can contract into it and arguably does exist, e.g., for movies w/big stars; writers share income with publishers.  Fiduciary duty is not the same, but duty of good faith does exist.
 
A: Wouldn’t be mandatory, but there are situations where people didn’t think of it in advance—Aalmuhammed v. Lee, Garcia v. Google, Effects Assoc. v. Cohen.
 
Rosenblatt: relatively common/well known in entertainment industry: back-end deal; several unions provide for residuals. Not an ownership-type right but an income stream that comes as a result of success.  Devil’s in the details. One is individually negotiated, the other isn’t. Those are key details.  These cases happen because of non-union employment.
 
A: Mandatory isn’t the word to think of. Post hoc or ex ante.  We should deal w/ residuals—closest thing contractually that exists. 
 
Bob Brauneis: Now we have a test: author or not. You are adding a third category: author-trustee, author-beneficiary, and nonauthor.  How do you figure that out post hoc when the parties by definition haven’t come to an agreement? What is the test for the ownership-author versus the control-author?
 
A: Something very close to the Aalmuhammad court: among the people who otherwise look like authors, is there a mastermind?  If there is a clearly dominant person, that’s the owner-trustee, and everyone else who prima facie satisfies the criteria of authorship satisfies the criteria to be author-beneficiary.
 
Brauneis: so anyone who makes a creative contribution intended to be folded into a creative work is an author?
 
A: this is the problem with having a low threshold for © protection. Anybody can be an author.  Aalmuhammad is a drastic solution to the problem that everybody is an author; we are something in between.
 
Brauneis: I’m just curious about what that something is.  [Agreed: he seems to be going to “everybody is an author.”]
 
A: everyone who satisfies the statute on its own terms is an author-beneficiary. 
 
Q: if memory serves, Aalmuhammad got paid.  If I’m the trustee, how much money does he get?
 
A: practical problem of figuring out shares.
 
Q: The industry will indeed care about the answer.
 
A: Apportionment happens all the time.  © judges make value judgments all the time.  Thinking in advance is always the better solution, but if they haven’t, we should be able to land between zero and one.
 
Brauneis: you could stick with the even division but give one person control: percentage is separate from the issue of control.
 
Q: CARP requires 18 month proceedings with huge records. That could be very unhappy for district court judges.
 
Rebecca Curtin, The Transactional Origins of Author’s Copyright
 
Practice before the Statute of Anne might have affected that evolution in © history by which we started off with the Stationer’s right, protecting an infant industry à a right for authors.
 
Licensing Act expires in 1695: censorship in return for protection; Stationers try to get the quid pro quo up and running again, or if not that at least protect the industry. Doesn’t get political traction. Then stationers start arguing on behalf of author’s copyright. Rhetoric is there even if the rights are not.
 
Looking at what happened before to explain the Statute of Anne, and also potentially as a way to explain the renewal right.  Stationers’ petition, Feb. 26, 1706: “Discouragement of Persons from writing Matters, that might be of great Use to the Publick.” But there’s more emphasis on alienability and transferability in the rest of the petition.  Many learned men who used to dispose of their copies, and their families, are being hurt.  Looking for descendibility, devisability, divisibility.
 
1709 Petition: subtle changes; this one is ultimately successful. Protecting the ability of the author to sell to the stationers. 
 
Before 1709: do transactions give authors more rights than the legal default? There are examples.  Typical entry: name of printer/bookseller and title; author doesn’t have to be part of it.  But: a handful of alternative transactions.  E.g., additional note: copy never printed again w/out author’s consent. A reversion/right of first publication.  More complex contracts: similar to royalty streams. Milton’s contract for Paradise Lost called for a stream of contingent payments.  Another example: author paid in copies. If the book proves popular, he can sell those copies.  Author’s discount for copies, allowing author to get not just clawback rights but opportunity to become his own retailer.  These kinds of experiences allow people to be comfortable with the idea of authors’ rights, understanding that transferability would favor printers anyway because of the barriers to entry—little to fear from authors’ rights.
 
Jessica Silbey: tracing everyday practices through evidentiary fragments is great.  Are there other indices of bargaining power you are ascribing to authors? Who are they? Repeat players? Can you attach names?
 
Rosenblatt: reminds me of movie deals—clawbacks, pay or play, etc.  More like movies than the current book industries.
 
Ben Depoorter (& Alain van Hiel), The Dynamics of Copyright Enforcement
 
Enforcement against noncommercial online infringement.  Initial campaign; enforcement letters to campuses; settlement offers allegedly averaging $3000. 2008-2010—industry says it’s abandoning mass suits. But still pursuing high profile cases in the pipeline for statutory damages.
 
Empirical study: followup study of earlier enforcement studies.  Allows us to study enforcement more generally because of changes on axes of certainty and severity: low certainty but high severity (statutory damages); settlement letters medium certainty and severity; Copyright Alert System (CAS) lower severity but higher certainty of getting caught.
 
So, what was the effect on deterrence? What about norms?  Effect of CAS; influence of legal alternatives like Spotify.  Research methods: surveys and experiments; a lot of self-reported behavior and reactions to hypotheticals, which have downsides. But this is anonymous and there’s no obvious reason to lie one way or another.
 
Most students engage in some type of illegal downloading—music, movies.  Includes music swapping between devices.  41 of 349 claim never to have done that.  Legal subscriptions/purchases—correlation with that and illegal downloads. Illegal downloaders: 92% use legal alternatives too.  Future downloading: 29/100 say they won’t use illegal downloads in the future, about the same say it’s very likely—two distinct humps, not much in middle.
 
How have norms evolved?  Industry says it was about education, not deterrence.  Moral judgments about downloading music and movies—do you think immoral; does public think it; do peers disapprove? On a scale of 9, immorality of downloading music is at a bit over 4/9—not completely moral. Expectation of peer disapproval is almost nothing (2); expectation of public disapproval is a bit higher.
 
Change over time—4.5 in 2007, 4.2 in 2011, and down to 4.1 in 2014 (relatively small but downward trend). Moral judgment relating to one’s download history. Those who download a lot think it’s not immoral (3.9); never download think it is (6).  But this doesn’t tell you about causality.
 
How likely is it that file-sharers that download daily will face repercussions: 26.5% think it’s very unlikely, 60% total “unlikely.” Most think 1 in a million or 1 in ten thousand chance.  Not much difference in perceptions among downloaders and non-downloaders: both average a bit over 3/9 (unlikely that downloaders will face repercussions).  No different perception of enforcement likelihood. Deterrence is not playing much of a role.
 
Effect of receiving CAS letter: likelihood of perceiving greater chance of being caught was much greater, but very small N (7).
 
In 2014, less in the way of plans to download illegally than in 2011, but goes from around 4.5 to 4.27. Past behavior predicts future behavior.
 
Priming: unjust enforcement. What’s most unjust/disproportionate?  Statutory damages; private trolls; CAS and RIAA campaign are equivalently objectionable or okay.  Not really significant improvement.  Is there need for copyright reform/more balanced ©?  Statutory damages, but also CAS and settlement campaigns also trigger belief in need to reform. 
 
CAS isn’t a great deterrent for future infringement plans; a little better than the RIAA settlement campaign.  Backlash: do people intend to download more if you’re free from risk?  CAS = nearly 4 agreement; settlement campaign = almost exactly the same; statutory damages = more backlash by a bit.
 
Privacy concerns also may be important.  CAS is not “norm-superior.”
 
Heymann: You seem to include P2P and direct swapping.  Later questions talk about downloading. What are you trying to test?
 
A: we said any type of tech where you know you’re paying v. where you know you’re not paying. P2P isn’t as common in 2014 as streaming sites.
 
Q: did you go outside of © law?
 
A: no.  Students aren’t representative of society as a whole, but in this context they are the primary target audience of the enforcement anyway.
 
Q: what’s their appetite for risk in other situations, as opposed to this one?
 
A: we did ask for their views on others’ opinions/preferences, but that would be helpful too.
 
Elizabeth Townsend Gard (& Geena Yu), Is Fair Use Codable?
 
Thompson Reuters is licensing the Durationator for use by clients. Launches next week.  Every country in the world coded after this year.  What we’ve learned from coding the Durationator: how to think through law and communicate results in a simplified manner.
 
Course: research each case, group cases, create flowchart based on outcomes to see if they were similar in finding, and if dissimilar what mattered. Class split 50/50 on whether fair use could be coded, and pretty passionately so.  Wanted to create greater probability though no absolutes. Coding depends on the type of audience you’re trying to reach. Best practices approach might be more practical for local creators/artists, the community we want to help understand fair use.
 
Next steps: need an army for research—perhaps groups of law students throughout the country. Need many paths doing the same thing concurrently: the way every scholar did it; the best practices approach—put the same info in and put it different ways to see the answers you get.  Idea is to have experts put their system in and train it.
 
Q: qualitative coding?
 
A: we have to be exploring that. Not like Durationator.
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in copyright, http://schemas.google.com/blogger/2008/kind#post | Leave a comment

WIPIP Session 2: Trademark

Posted on February 6, 2015 by rtushnet
Jeremy Sheff, The Ragged Edge of the Lanham Act
 
Similar to Tushnet project; will focus on question about Lanham Act’s structuring of interface between PTO and federal courts.  Registrability v. enforceability: incontestability; §2(a) bars versus the common law—some of the bars have no analogue at common law.  Is a mark cancelled for scandalousness enforceable under §43(a)?  Renna v. County of Union said that unregistrable under §2(b) was unenforceable.  §2(e)(1) and (3): deceptive marks—if unregistrable, enforceable?  That’s not so much an issue of §43(a)(1)(A) versus (B)—use of deceptive TM might count as false advertising—so not even a right to use?
 
Judicial review and standing: direct appeal to CAFC, governed by APA standards of review; 21(b): de novo action in E.D. Va.; the director shall not be made a party to an inter partes proceeding under this subsection—leading to a motion to dismiss in Pro-Football v. Blackhorseon the theory that there was no standing. DCt ruled against that, but there are at least plausible arguments that Article III doesn’t map onto standing required to participate in inter partes proceeding.
 
B&B v. Hargis: preclusion and deference. Are the legal and factual issues in registration proceedings meaningfully distinct from issues in infringement litigation? Is the PTO entitled to deference on the legal and factual issues even if there’s no preclusion, and does that depend on distinction from issues that arise in subsequent infringement litigation?
 
What is registration for?  Procedural view: notice, evidentiary record for first in time rights system; inducement to register is to provide notice and avoid future disputes. Substantive view: those inducements have independent force and meaning: offense-side incontestability (inconsistent w/common law); nationwide priority; remedial advantages like criminal penalties and enhanced damages; evidentiary benefits like presumption of validity and notice.
 
Depending on whether you think it’s substantive or procedural, implications differ. Substantive: §2 has First Amendment problems. Standing is less controversial. PTO is probably entitled to deference. Possible preemption of state law or Congress should do it if it hasn’t already been done.  [RT: what should be preempted?] Implications of procedural view: divergence of §2 and common law protectability standards is probably unsupportable; standing in the absence of live infringement more problematic; less clear that judicial deference is warranted. [RT: not clear on that last point.  Can’t an admin agency get deference on factfinding even if there is generally judicial review?]
 
Upshot: really need to decide!
 
Lunney: Park N Fly: actual registration included design; could probably have proved secondary meaning.
 
A: but incontestability still is something you couldn’t get at common law and the opinion says it’s a carrot.
 
Welkowitz: what do you mean by deference?
 
A: maybe APA deference, to factual issues decided in registration proceedings.
 
Welkowitz: Judge said that it would be very hard to figure out how to give proper deference to the PTO in a jury trial.
 
A: there is a good argument that the question is substantively different between §2(d) and infringement.
 
Welkowitz: preclusion is an on-off switch, deference is not.
 
Ramsey: how robust is examination? That worries her about deference.
 
A: on a lot of issues, like §2(a) and 2(d), they tend to build more of a factual record. §2(d) seal of county gov’t, less contestable.  Maybe preclusion/deference only comes where detailed factfinding is necessary—which is where APA deference gets you.
 
Rebecca Tushnet, Registering Discontent
 
Sheff conferred a positive externality on me since our projects are so similar.  My starting point: Felix Cohen on transcendental nonsense and the functional approach—TM was one of the key examples of transcendental nonsense. Eleven years later he lost the battle in TM as the Lanham Act added a new set of legal fictions to the existing ones he criticized.  Are these concepts empirical?  If not, what should replace them?
 
Along with Sheff’s examples, I’m interested in the role of registration in the multifactor confusion test, as well as how to reconcile registration w/dilution, which was supposedly designed to prevent interference w/marks even on unrelated goods/services. 
 
What should be done? I agree, pick one. I’m attracted to substance: create a real divide between registered and unregistered marks, perhaps by double identity for registered marks and a robust harm requirement for unregistered marks.  Explicit balancing of non confusion based rationales as justifications of and limits on TM—since we’re so bad at the empirical tests anyway.
 
Rosenblatt: many inconsistencies w/current system and proceduralist view.
 
Sheff: agree—proceduralist turn would require changes.
 
RT: but also inconsistencies w/substantive view of registration—consider infringement test, which doesn’t consider the registered mark but only in context.
 
Bill McGeveran: if you’re attracted to a European model would there be use as a mark as a requirement?
 
RT: Maybe, maybe entailed by the very definition but if not should be done explicitly.  TM as a series of formal moves that, if made, entitle someone to a right.  If not made, don’t get caught in the cycle of “does the public perceive this as use as a mark.”
 
Ramsey: w/double identity you’d need a defense for comparative advertising/use as a mark.  Also, what about examining rigor if you’re going to go substantive route?
 
Sheff: question is ex ante costs of rigor v. ex post costs of resolving conflicts; reasonable people disagree.  Ask: is it a big deal if the PTO refuses your mark and you have to find another?  What gatekeeping function do you want?
 
RT: agree, rigor is an issue. One issue where the TM bar is in agreement is that there’s a ton of deadwood on the register.  Study: 2/3 of marks that applied for renewal couldn’t show use in one or all categories.  ¼ couldn’t show use in any.  This is a big deal; TM bar thought it was a slight problem but it’s not, especially in substantive areas.  Nonuse proceedings should be fast and cheap—Canadian model—and PTO should require more specimens as a matter of course.
 
Megan Carpenter, “Behind the Music”: Lanham Act 2(a): When scandalousness came in, Hays Code was in effect—people who committed immoral acts on screen had to be punished on screen.  Definition: shocking to sense of propriety, offensive to the conscience or moral feelings or calling out for condemnation.
 
Consumer protection is different: protecting morals rather than confusion.  Lessening in other IP regimes of regulating morality.  But political and practical realities exist.  Today’s legislative environment, unlikely that Congress will act.
 
Even w/o new legal framework, could get better, more consistent and defensible results w/marketplace context. Test: current context of the attitudes of the day, in the context of the relevant marketplace. In practice?  Empirical study: Most are rejected for containing a word listed as vulgar in the dictionary; but only 5.08% of rejections she examined considered market context.  But what is vulgarity?  Dictionary: lacking sophistication or good taste; explicit and offensive reference to sex or bodily functions; coarse and rude
 
Media like Urban Dictionary used about 25% of the time.  Mostly dictionary; next was other media (15%), a tiny bit of third party use, and 3.91% applicant’s own actions. The fact that it’s used in a context/market that accepts vulgarity, that’s used to support the rejection. TIT MITT for bras rejected because the goods would make the vulgar meaning clear. Google searches are common contextual evidence, but most often we see marketplace abstracted from that analysis.  Most are rejected for profanity and sexual reference, failing to account for narrow marketplace.
 
Adult-oriented goods/services, that also supports the scandalousness refusal. Where the goods themselves are not scandalous, like SHLONGWEAR for apparel, then the standard was the general public who’d be scandalized.  Catch-22.
 
Atypical to respond to Office Actions—usually abandoned; applicants are individuals and small businesses. Most common response when they do respond: context of mark.  Other common arguments: alternate meanings and third-party registrations. When context is argued, it doesn’t often succeed.  For every mark that’s been rejected, there’s one that passes through.  No predictability/consistency for Examining Atty or for the applicant.
 
By and large, these things are in use afterwards, though they can be hard to track down.  Used by applicant or otherwise in almost ½ the case; but that leads to a proliferation of uses.
 
RT: How many of these are really TM use and not ornamental? E.g., YOU CUM LIKE A GIRL.
 
A: A lot should be rejected for ornamentality, and that’s ok.
 
Rosenblatt: I always thought this was intuitive, and the examiner just justifies their personal intuition. If so, the examiner may go find something else even if you change the process.
 
Q: maybe, but I think “vulgar” is a much lower standard than scandalous. 
 
Farley: There is a lot of inconsistency, but there’s consistency on “shit.”  You’re telling a story of consistent difficulties; the dictionary might be more consistent.  Are you truly interested in consistency, or something else?
 
A: we don’t see all the things that pass through without challenge.  Examiners don’t have clear guidance.
 
Roger Schecter: merchants want to sell stuff. They won’t sell stuff consumers won’t buy, including because they’re offended.  If §2(a) is to have any coherence, then, the interests/reactions of nonconsumers must somehow come into account.  Maybe it’s hard for anyone to avoid anything in the internet age, but the likelihood that goods will be in a market channel where noncustomers will encounter them should be a relevant consideration.
 
A: Agree.  Similar to likely confusion factors.
 
Q: Adult stuff is approved (not always)—being for an adult product is necessary but not sufficient.
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in http://schemas.google.com/blogger/2008/kind#post, trademark | Leave a comment

WIPIP session 1: Trademark

Posted on February 6, 2015 by rtushnet
WIPIP Session 1: TM
Deborah Gerhardt (& Jon McClanahan), Colors
 
20 years of Qualitex: who is registering color? How do marketing folks view color? People remember color images more accurately/longer than B&W; color cues amplify legibility and familiarity; send signals about brand personality—can alter target audiences by changing color; prompts stronger emotional responses than words and images; people make up their minds in 90 second, and 62% or more of that is based on color.
 
What does that mean for TM?  Color can express a lot other than source. Personality, emotion:
[color emotion guide]
 

Various registrations for color—universities, Tiffany’s, Zagat’s cover.  Post-Qualitex: 143 Color only applications, 0.0%; 2071 for design and color, .06%; millions of registrations that are not color-based.  Number of colors claim: mostly just a single color, though a few claim up to ten.  Red/pink & blue are the most popular color combinations, followed by brown and white, then gray/silver & black, then violet/purple & green.  Color marks are registering at a much lower rate than design only, design & color, and neither.  38.1% color only marks go to publication, 23.0% to registration. Some design & color marks are really what we think of as color only, such as Tiffany’s blue for catalog covers—the design is the shape of the cover.  Design and color: 46.0% published, 59.7% registered; overall success rate for design only or “neither” is about 74%.
 
General trends: increase in applications, registrations; very few end up on supplemental register and that’s not increasing.  Marks with design and color (e.g., Target red target)—a lot more marks on the supplemental register, but publication and registration lines are really close together—fewer ITUs here.  Marks claiming color but not design, but there are so few (they are going to recode to move some design + color like the Tiffany one into the color category).  Color only=covers an entire object, or an entire piece of an object—the Louboutin sole is color only, whereas a stripe on a water bottle is not color only.
 
Marks claming one color by class of goods—neat slide with bubbles representing # of color marks by color.  Registered across all sorts of classes.
 
Glynn Lunney: shades?  Color depletion in Qualitex—how many colors does the PTO think there are?
 
A: people tend to register a particular Pantone color.  But how it’s coded: PTO codes 12 colors including clear, white, and black.  (Lunney suggests an inherent masculine bias to this determination.)
 
Lisa Ramsey: what evidence are examiners accepting as evidence of secondary meaning?  If women like purple, then why should any competitor be denied the color purple?
 
A: don’t know.  Red is protected for scaffolding, who knew?
 
Jeremy Sheff: a little dubious of the low numbers here.  It’s difficult to determine going by hand.  But just looking at applications w/colored drawing as part of the file, there’s over 180,000. That could be a word mark in a particular color, but that would still be relevant for color as stimulant for demand. Are you excluding that from your scope?
 
A: they aren’t necessarily claiming color.
 
Sheff: but it’s part of the brand.  May want to deal with that, unless you’re just interested in competitive issues in claming color per se.
 
A: Target has the red target and the outline registered separately, the second for broader protection.
 
Sheff: ultimately, are you issued in competitive issues/foreclosing color for others, or in the use for branding?
 
RT: Any ITUs? Examine Specsavers.  Registration/infringement.
 
A: Wants to understand what’s going on in the areas where similar colors are registered for the same class of goods.  Is there a color depletion issue? Maybe Jacobson was right that you didn’t need color only because color + design is available.  (That would have led to the same result in Qualitex, right?)   Vast majority of filers appear to be very experienced filers.
 
Lunney: are you going to break down reasons for rejection?
 
A: Definitely will look at abandonment data. 80% of color applications have Office Actions; 60% conversion rate, when it’s usually 30%–a lot of them are facing these barriers.
 
Q: year to year trends? Pantone color of the year—do you see it in the PTO?
 
A: another fascinating Q to look at.  Has also heard of different cultural preferences for color, but interestingly comparing color impressions 30 years ago to now there’s less of that—cross-cultural perception of yellow as happy.
 
Jeanne Fromer (& Barton Beebe), The Closing of the Linguistic Frontier in Trademark Law
[early in project: more coming soon]
 
Lisa Ramsey, Trademarking Everything?
 
Are we TMing too many things? How to convince brands to care about the issue/care about competition? Her focus: protectable subject matter.  Need legislative solution, likely international, though domestic reforms are also important.
 
Should we protect anything that’s capable of bearing meaning, as current doctrine allows: Taylor Swift’s “This Sick Beat” applications for everything from nail polish toT-shirts.  TM application for “I Can’t Breathe” for clothing.  Specimen carefully uses that term on the tag, but clearly interested in enforcing against people who are using it on the front of a shirt, despite PTO rule against registering ornamental uses.  People then do use the term on the tags, but threaten ornamental uses, often successfully.  Or Delicious for shoes, and not used ornamentally, but then they go after Victoria’s Secret which uses the word on a T-shirt, and VS can’t win on summary judgment.  VS too has used marks on the front of T-shirts, like “Very Sexy,” which made it hard for them to argue no use as mark. 
 
Etsy store for tiny gold tags on necklaces; taken down because someone else has a registration for Tiny Tags for tiny tags on necklaces (disclaims right to “Tags”).  This is on the Supplemental Register, but the recipient took it down.  Controls results for Google searches.  Facebook went to trial on “Timeline”—FB can at least afford to litigate.  Oprah couldn’t get summary judgment on “Own Your Power.”  Can we get brands to care about these abuses?  E.g., “How” registration leading to lawsuit against Chobani for “How Matters.”  These are descriptive uses.  Need to shift perception of what the point is: whether people should own words in goods/services.  “Candy Crush Saga,” etc.
 
Acquired distinctiveness is not a big barrier. Declaration from company about use of mark. Examiner might search 3d-party use.  But what about future users of descriptive terms/colors/designs.  Rigorous examination as in patent?  Maybe we shouldn’t have protection in the first place.  Likewise w/incontestability.  No independent examination.  Then you can’t challenge it based on lack of distinctiveness.  Can’t do lack of confusion/descriptive fair use on motion to dismiss/often even sj. 
 
Slogans too, highly distinctive.  Wells Fargo claims TM rights in “Envelope Free” for ATMs.  Apple trying to TM store designs.  Bellaband: court says they could prove protectable trade dress in website design.  Not like copyright where actual copying is required.
 
Shaming sanctions for TM bullies? Sometimes people withdraw applications or claims against competitors: Ice Bucket challenge, IPA lawsuit.  Also: require more in the way of combination; require visual perceptibility.  Look to patent law §101 for model for subject matter exclusions. Requires us to consider international reform. Other measures to balance competition/free speech—functionality isn’t doing the job.
 
David Welkowitz: how do you get around free riders? Even if brands get on board, they won’t want to give up their ability to get a TM.
 
A: That’s why we need legislative reform. No TM for use on the front of a T-shirt; no single colors.
 
Lunney: most of the expansion has come from the courts—symbol was added to the bill specifically to exclude trade dress, yet 40 years later Breyer turns around and says symbol means “anything at all.” May be hard to get legislative action. Problem is that you get sympathetic Ps and the court stretches the law, then generate a rule that applies even to ordinary cases. They don’t think of the case as a regulatory system but it does become such a system.  Lots of businesses in the real world have very small differences in their names; they can register a corporate name w/a small difference in the state corporate registry.
 
A: completely agree that courts are the problem. That’s why we need subject matter reform, because once you get a right it gets stretched.
 
U. Shen Goh, Branding Linguistics: What do Coca-Cola and Chinese Bakeries Have in Common?
 
Coca-Cola’s treatment in Canada v. US. 1975, started filing applications for Chinese language marks in Canada, but not US. Canada hasn’t been very friendly—decided that Pepsi-Cola was not confusing w/Coca-Cola.  Also Canada had no comparable American doctrine of foreign interests.  But in 1992, when renewal came due, they abandoned.  Even though Canada had just rejected the American doctrine of foreign equivalents—another case decided there was no confusion between 2 Chinese bakeries (Vancouver/Toronto)—Chinese word was the same but English translation was different; average Canadian consumer doesn’t understand Chinese and the audience is the relevant public at large.  So Coca-Cola abandoned.  Since then the case law changed and found that the average consumer is the target consumer/actual consumers of goods/services.  US does have a bit of shakiness in the doctrine, but does have foreign equivalents.
 
Currently we classify all foreign languages that don’t use alphabets or numbers as design marks in our databases. This is wrong.  Word marks receive more protection than design marks, and many marks not using Latin/Roman characters have multiple presentations/translations.  Simplified Chinese may not look the same to English-only speakers, but it’s similar to capital letters v. lower case in English.  Fluent = know it’s the same. Western databases should transcribe foreign language marks according to their official phonetic system.
 
Western system depends on letters—type in Coca-Cola, get matches. But if there’s no letter/alphabet, we traditionally treat it as design mark. Phonetic system: allows filing alphabetically and numerically. Improve efficiency of database; impossible to search foreign language marks alphabetically and you have to hire a firm to do a design search.  Would remove incongruity w/case law, which recognizes foreign language marks as word marks.
 
Solves inconsistency w/int’l trade: MFN/national treatment principles.  Mark in China is a word mark; would be a design mark in the US.
 
Judicial treatment: both Canada/US require that the foreign language mark be understood by average Canadian consumer/ordinary American purchaser. This evidence changes constantly depending on immigration, int’l trade, brand extension, and cross-branding. Not just a problem for large corporations.  We should assess foreign marks in the same manner as English language marks: assume a fluent consumer.  We treat “aubergine” as having a meaning; don’t question how high educational level of consumer is. Should always assess for distinctiveness. A trader should not be able to do in a foreign language what it couldn’t do in English.
 
Foreign marks shouldn’t be translated when assessing for confusion w/marks in the same language.  A fluent consumer doesn’t need to translate; the monolingual judge does and wrongly assumes that consumers do the same.  Do need translation for meaning, but that’s just practical.
 
Foreign language marks shouldn’t be translated when assessing for confusion w/English transliteration English language marks can’t expand into claims for homonyms or homophones. Traders who choose to spell an English mark are stuck with that. The same rationale should apply to a trader who decides to use a foreign language mark, and not use any/all of its possible transliterations. One Cantonese transliteration could be Cheung Kong, Chiong Kung, Chang Jiang, or more depending on one’s dialect.  If we do it for foreign language we should do it for English.
 
Foreign language marks shouldn’t be translated for assessing confusion w/ an English translation. English language marks can’t expand to claims for synonyms, etc.
 
Hana Bank: Korean language, want to transfer into English “Hana Bank.”  But there are multiple translations: Hana means one: One Bank, United Bank, Whole Bank: do they get to claim them all? Not fair to other traders.
 
Christine Haight Farley: you’re arguing not just for changed approaches but changed standards? It would be easier w/phonetic searching systems for parties to decide that something is too close in pronunciation, but you seem to advocate at the end for not having to use that.  Chinese companies: script versions of TMs, but Chinese consumers might prefer a different word—Michael Jordan becoming “flying man.”
 
A: you can use transliteration or translation. In China, many companies discover they’ve been usurped. Their Chinese TM may not be as popular as the TM that became colloquially popular. Coca-Cola ran a contest and picked the best—good brand management. 
 
Lemley: struck by statement about not extending protection to synonyms. Correct to say you wouldn’t get a registration, but as a practical matter you get control over them: both in registration where PTO will reject and in practice where likely confusion will control. Why treat foreign translations any differently? If people think of it the same way, why allow anyone else to overlap use?
 
A: There is a Coca-Cola case that allowed them to control Koke.  But that’s more for well-known/popular TMs. 
 
RT: disagree—Lollipops and Jellybeans for roller rinks held to be confusing, because not well known and inherently distinctive. But I love the rest of the proposals.
 
Abraham Bell (& Ted Sichelman & Gideon Parchomovsky), Trademarks as Club Goods
 
Why do people wear T-shirts advertising their affiliation w/brands? Not a new Q, but a new answer.  People are acquiring not any information about the actual physical product. Economic search cost explanation doesn’t give us a very good answer, b/c only the brand and not the product is of concern.  TM in theory incentivizes production of goods w/non-observable qualities, but the value of the brand is fully observable.
 
TMs as expressive goods? Express fealty to NY Yankees, Nike, etc.  Brands serve as self-expressive private goods and linguistic public goods: counsels in favor of weaker protection to foster expression.
 
We think TM has one function of creating clubs/signifying membership. Not simply self-expression but membership/identification w/ a network of others associating w/the mark. Our theory separates club aspects from expressive aspect of the mark.  Deven Desai talks about bits of this but does not delineate badges of loyalty from other expressive aspects.  TM is a way of telling you something about the individual bearing the TM: an entry ticket into a club.
 
Vital that TM owner is the one that owns the right to exclude.  Creating a membership group.  Club goods are intermediate b/t public and private. Semi-enclose things that would otherwise have risk of being public goods/private production of goods that could otherwise only be produced collectively.
 
Ted wears a Harvard T-shirt. Ted is getting branded with Harvard (like ISO standard). Harvard is getting branded w/Ted (other consumers know that Harvard is the kind of school that draws Ted, or at least Ted identifies w/). The value of the T-shirt goes up and down w/Harvard’s activities. Harvard’s control over production and distribution affects the value of the club membership.  So it makes sense for control over entry to be in Harvard’s hands.  If every single consumer controls the TM, it loses the ability to cabin this club.  [RT: But Harvard will sell anyone a T-shirt.  It does not police the boundaries of the club and it can’t police the used market nor should it be able to do so. Compare Republican and Democratic (or “Democrat”) Party.  They can’t protect the “club” aspect and seem to be robust.]
 
This gets us to exactly the opposite of the expressive theory. [But why would this lead to any right to control the image of the club?  Can’t we have the Coca-Cola Kid?] Not simply about T-shirts. Purchase and use of many products have a club goods aspect separate from self-expression. Drinking Coke, driving a Mercedes Benz, riding a Harley Davidson, using an iPhone, eating and McDonald’s.  Commercials for these brands often emphasize club aspects: e.g., Coca-Cola’s “I’d Like to Teach the World to Sing.” Create a membership group and police its boundaries.
 
What do we get this?  Facilitate dissemination of goods that can only come in these kinds of networks; large potential increases in social welfare. TM already does this but we haven’t been paying attention.  Best way of understanding dilution is being about this.  [Except then there shouldn’t be expressive exceptions.] Dilution = spreading club benefits outside the club.
 
Do we have some normative suggestions?  Exclusivity is important. Optimal spread of the good if costless is not infinite. Some point at which it’s too widely spread.  [How is “I’d like to buy the world a Coke” in any way limited?]
 
Lunney: complete misuse of the theory of club goods.  (1) getting funding to afford public good, (2) setting price that doesn’t create deadweight losses.  Treadmill in fitness center: could be public if there are few members; congestion = need to set price/fund creation. You haven’t shown me a congestion externality or using the congestion to set the price. This is just “if value, then right.”
 
A: the congestion is the exclusivity. If everyone does it. 
 
Lunney: McDonald’s would lose money if everyone ate there?
 
A: Yes.  Set price to control access. There’s a reason companies like Members Only name it Members Only. Even McD and Coke exclusivity is working to provide profit. [Citation needed.]
 
Gerhardt: Make assumption that TM owners create brand value and channel it down.
 
A: no we don’t.
 
Gerhardt: consumers invest in creating brand, should get return at least in form of being able to use for information.
 
A: the TM owner gets to police the boundaries even if it doesn’t create value.
 
Gerhardt: community rejected New Coke.  Good brands let brand story change in response.
 
A: but mark owners police who can enter the community.
 
Gerhardt: should we shut down Pinterest? TM owner doesn’t like being pinned by the kind of person who pins images of its products.
 
A: there are other interests—not saying shut down Pinterest. This is an important interest on the other side.
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in http://schemas.google.com/blogger/2008/kind#post, trademark | Leave a comment

IP question of the day, Harper Lee edition

Posted on February 4, 2015 by rtushnet

Amazon’s Write On is its project challenging Wattpad, trying to develop writers–including fan writers.  Participants aren’t compensated, but this is a commercial endeavor on Amazon’s part and presumably it will try to transition the successful ones to paid publishing.  So, here’s this morning’s email:

Imagine your favorite literary character’s past or future.
http://writeon.amazon.com/?ref_=ign_em_hd
Readers, rejoice! This summer, Harper Lee will publish “Go Set a Watchman,” which revisits characters from her classic novel “To Kill a Mockingbird,” 20 years later.
Inspired by this news, we have a special writing challenge:
In 500 words, write a story featuring your favorite literary character at an earlier or later point in their life.
Good luck, and have fun!

1. Assuming no authorization, is this an infringing commercial use of Harper Lee’s name?
2. Inducement?
3. More an advertising law question: Given that this is a new initiative by Amazon, who writes the emails to users, and who reviews them?  A new social media endeavor is often handed off to a specific person/group and not integrated with overall advertising, which is simple but can lead to gaps in review.

http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in http://schemas.google.com/blogger/2008/kind#post, right of publicity | Leave a comment

New article: images and the right of publicity

Posted on February 3, 2015 by rtushnet
Rebecca Tushnet, A Mask that Eats into the Face: Images and the Right of Publicity (38 Columbia J.L. & Arts, forthcoming 2015)
 
Abstract: In their eagerness to reward celebrities for the power of their “images,” and to prevent other people from exploiting those images, courts have allowed the right of publicity to distort the First Amendment. The power of the visual image has allowed courts to create an inconsistent, overly expansive regime that would be easily understood as constitutionally unacceptable were the same rules applied to written words as to drawings and video games. The intersection of a conceptually unbounded right with a category of objects that courts do not handle well has created deep inconsistencies and biases in the treatment of visual and audiovisual media, particularly comics and video games. These problems show up both in First Amendment defenses and in copyright preemption analysis. The possible arguments one might offer for treating images differently are insufficient to justify this disparity. The Article concludes that, absent the distortion produced by images, the right of publicity would properly be understood as sharply limited.
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in first amendment, http://schemas.google.com/blogger/2008/kind#post, my writings, preemption, right of publicity | Leave a comment

Reading list: Craswell on sports team nicknames

Posted on February 3, 2015 by rtushnet
Richard Craswell, When Nicknames Were Crowd-Sourced ~ OR ~ How to Change a Team’s Nickname
 
Craswell is an excellent writer with an engaging topic even for a non-sports fan like me. Treat yourself to this short monograph.  Excerpt:
 
Here is the Washington Times. . .:
“The Redskins are a private business enterprise, and the owner has the right to call his team whatever he likes.”
What few people realize is that this idea – the idea that nicknames should be controlled by a team’s owner – is a relatively recent invention. As we have seen, nicknames in the early days of spectator sports were almost never chosen by owners or by college officials. Instead, nicknames were chosen by the decentrralized process of the crowd. Anyone could come up with a new nickname, but the nicknames that survived were those that fans and journalists liked well enough to repeat. And in those days, whatever the crowd gaveth, the crowd could also taketh away.
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in reading list, trademark | Leave a comment

230 defeats false advertising claim against search engines

Posted on February 3, 2015 by rtushnet
Baldino’s Lock & Key Service, Inc. v. Google, Inc., No. 1:14-cv-00636 (E.D. Va. Jan. 27, 2015)
 
Defendants Yellowbook and Ziplocal provide online search engines for businesses, allowing search by type of service and geographic area.  Baldino’s is a Washington Metro area locksmith and security company, with the required licenses in Maryland and Virginia.  The internet resulted in a drastic decrease in revenue for Baldino’s, which it attributes to advertising of unlicensed and illegal locksmiths.  In 2014, for example, a Google directory search produced results for 1000 locksmiths in Virginia, but only 325 were listed as licensed; other defendants produced similar results.  Baldino’s sued for RICO violations and false advertising.
 
The court found §230 immunity.  The exception for violation of federal criminal law failed because Baldino’s failed to adequately plead RICO violations.  (I thought the exception just protected the feds, rather than preserving civil claims identical to potential criminal claims.) 
 
And the exception for IP didn’t apply because Baldino’s failed to adequately plead a violation of the Lanham Act.  (Apparently treating false advertising as an IP claim.)  Baldino’s failed to plead that the defendants made a false or misleading description of fact; the falsity came from the unlicensed and illegal locksmiths.  “To hold Defendants liable for misinformation appearing on their websites, which originated with third parties, is a drastic conclusion the Court declines to endorse. The Court believes the market incentive for Defendants to provide correct information to consumers is a better tool for accuracy than the Lanham Act.”
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in 230, google, http://schemas.google.com/blogger/2008/kind#post | Leave a comment

Pom squad at the DC Circuit

Posted on February 2, 2015 by rtushnet
POM Wonderful, LLC v. Federal Trade Comm’n, No. 13-1060 (D.C. Cir. Jan. 30, 2015)
 
POM ran ads from 2003 to 2010 touting medical studies that supposedly showed that daily consumption of POM products could “treat, prevent, or reduce the risk of various ailments, including heart disease, prostate cancer, and erectile dysfunction.”  Unfortunately, “[m]any of those ads mischaracterized the scientific evidence concerning the health benefits of POM’s products with regard to those diseases.”  The FTC ordered POM to stop; the D.C. Circuit rejected most of POM’s challenges, except to the portion of the remedial order requiring two randomized clinical trials before making similar health claims.
 
NB: With Tom Goldstein on the brief, the court of appeals may not be the last stop for this case.  I don’t think there’s a split, but a Court eager to expand protection for commercial speech might nonetheless be interested.  I suspect that if cert were granted, though, the Court would find—as it did in Nike v. Kasky—that the prospect of destroying a huge part of the regulatory state by making it much easier to engage in false or misleading advertising was too unappetizing for a majority.
 
The court here concluded that there was no basis to set aside the FTC’s finding of false and misleading statements.  However, it did reverse the FTC’s blanket requirement of at least two randomized controlled studies as a precondition to any disease-related claim—which is either a very interesting constitutionalization of a remedial standard, or maybe not that important; only time will tell.
 
The products at issue include both pomegranate juice and dietary supplements, POMx Pills and POMx Liquid, which contain pomegranate extract in concentrated form. By 2010, POM’s owners (the Resnicks), POM, and POM’s integrated marketing agency Roll had spent more than $35 million on pomegranate-related medical research, sponsoring more than one hundred studies at forty-four different institutions. The claims at issue here dealt with heart disease, prostate cancer, and erectile dysfunction. 
 
Take heart disease: One POM-sponsored study involved ten patients in the treatment group and nine in the control, and thus couldn’t be “at all conclusive” in the words of one of POM’s own experts. A larger study followed 73 patients and found no statistically significant difference in the tested measure or any other heart-related measure.  A still larger study followed 289 patients and again found no statistically significant difference; POM initially delayed publication but eventually allowed it. The final report identified some subgroups that appeared to benefit, but this was post hoc massaging of the data (a classic way to claim otherwise absent significance) and posed an increased risk of false positives.  Plus, even for the subgroups, the change was 4-9% decrease in arterial thickness, substantially below the 30% reported by the tiny study. 
 
Even though the bigger studies were done by 2006, “a consumer reading POM’s promotional materials after 2006 would not have known of those studies or that they cast doubt on [the] prior findings.” For example, in 2007, POM published a newsletter claiming that “NEW RESEARCH OFFERS FURTHER PROOF OF THE HEART-HEALTHY BENEFITS OF POM WONDERFUL JUICE.” The newsletter claimed a “30% DECREASE IN ARTERIAL PLAQUE.” Always, POM was silent about the contrary studies.  The evidence went on like this, but I won’t.  POM’s studies had grave limitations and/or found no statistically significant benefits.
 
The full Commission, by a vote of four out of five, found that 36 of POM’s ads made false or misleading claims, though injunctive relief would be justified even if based solely on the 19 ads the ALJ found to be false or misleading.  One commissioner concurred, saying that she, like the ALJ, would have found a smaller number of POM ads to be false or misleading, but agreed that POM and related parties had violated the FTC Act.  The injunctive order barred the POM entities from representing that any food, drug, or dietary supplement “is effective in the diagnosis, cure, mitigation, treatment, or prevention of any disease” unless the representation is non-misleading and supported by “competent and reliable scientific evidence that, when considered in light of the entire body of relevant and reliable scientific evidence, is sufficient to substantiate that the representation is true.” For purposes of that order, the FTC defined such evidence as at least two properly conducted, randomized and controlled human clinical trials (RCTs) that yielded statistically significant results. They’d have to be double-blinded unless the POM defendants could show that blinding couldn’t be effectively implemented given the nature of the intervention.
 
The rest of the order barred POM from misrepresenting the results of scientific studies in their ads and from making any claim about the “health benefits” of a food, drug, or dietary supplement unless the representation is non-misleading and supported by “competent and reliable scientific evidence.” But it didn’t require RCTs to support more general claims about health benefits.
 
Statutory claims first: “[t]he findings of the Commission as to the facts, if supported by evidence, shall be conclusive.” FTC Act § 5(c). This is basically the APA’s substantial evidence standard.  The FTC is often in a better position than courts to determine deceptiveness, since such a finding “rests so heavily on inference and pragmatic judgment.”  The FTC evaluates what claims are made in an ad; whether the claims are false, misleading, or unsubstantiated; and whether the claims were misleading—that last was not in dispute here.
 
Ads that merely convey efficacy have to be substantiated by a reasonable basis, determined by “the type of product,” “the type of claim,” “the benefit of a truthful claim,” “the ease of developing substantiation for the claim,” “the consequences of a false claim,” and “the amount of substantiation experts in the field would consider reasonable.” But for establishment claims, if the claim is specific, the “advertiser must possess the specific substantiation claimed.”  If the claims aren’t specific—claims like “medically proven” or images that “clearly suggest that the claim is based upon a foundation of scientific evidence”—the advertiser “must possess evidence sufficient to satisfy the relevant scientific community of the claim’s truth.”
 
There was no basis for setting aside the FTC’s “carefully considered findings of efficacy and establishment claims.”  The POM entities argued that the FTC interpreted POM’s claims too broadly by holding that if an ad truthfully references research connecting a food product to possible health benefits, it necessarily implies “the vastly broader claim that there is ‘clinical proof’ that the product treats, cures, or prevents a disease.” That’s not what the FTC did.  It clearly stated that not every reference to a test or study was necessarily an establishment claim. But the ads at issue here went beyond “merely describing specific research in sufficient detail to allow a consumer to judge its validity.” The ads referred to study results “in a way that suggests they are convincing evidence of efficacy,” drawing a “logical connection” between study results and disease effectiveness. “Moreover, they invoked medical symbols, referenced publication in medical journals, and described the substantial funds spent on medical research, fortifying the overall sense that the referenced clinical studies establish the claimed benefits.”  The court agreed with the FTC that when an ad represents that tens of millions of dollars have spent on research, that tends to reinforce the idea that the supporting research wasn’t just preliminary.
 
Nor was the FTC cherry-picking the record for aggressive ads.  There was no meaningful difference between earlier ads and more recent ads’ reliance on medical studies.  For example, in July 2010, less than three months before the FTC complaint, POM advertised that POMx was “backed by $34 million in medical research at the world’s leading universities” revealing “promising results for erectile, prostate and cardiovascular health.” It discussed the same old, tiny studies. The FTC concluded that “at least a significant minority of reasonable consumers” would construe the ad to claim that POM products could treat, prevent, or reduce the risk of erectile dysfunction, prostate cancer, and heart disease. The ad’s references to the described studies as “promising,” “initial” or “preliminary,” in context, were insufficient to neutralize the otherwise unequivocally positive claims of specific results. As the FTC held, the “use of one or two adjectives does not alter the net impression,” especially “when the chosen adjectives” (such as “promising”) “provide a positive spin on the studies rather than a substantive disclaimer.”
 
The FTC might’ve reached a different result if the ads had effective disclaimers, but they didn’t.  (The identified statement, “evidence in support of this claim is inconclusive,” is probably ineffective, by the way, but I can’t imagine that detains the DC Circuit much.)  Thus, the standard POM had to meet was “evidence sufficient to satisfy the relevant scientific community of the truth of their claims.”
 
The FTC then found POM’s claims deceptive due to inadequate substantiation.  The court of appeals held that this finding was supported by substantial evidence, mindful of the FTC’s special expertise in determining the necessary substantiation.  For both efficacy and non-specific establishment claims, the FTC found that experts in the relevant fields would require one or more RCTs to establish a causal relationship between a food and the treatment, prevention, or reduction of risk” of heart disease, prostate cancer, or erectile dysfunction.  The FTC emphasized a distinction between “generalized nutritional and health benefit claims” and “the specific disease treatment and prevention claims at issue in this case.”  Also, lesser substantiation might suffice for “claims that do not assert a causal relationship,” unlike POM’s ads, e.g., “POM Wonderful Pomegranate Juice . . . can help prevent premature aging, heart disease, stroke, Alzheimer’s, even cancer” and “Eight ounces a day is enough to keep your heart pumping.”
 
POM’s studies weren’t RCTs.  And their selective touting of ostensibly favorable studies constituted misleading omissions of material facts.  The FTC further found that POM was aware of the misrepresentations and the weaknesses in the studies.
 
POM challenged the FTC’s factual finding that experts in the relevant fields require RCTs to support claims about the disease-related benefits of POM’s products. That finding was supported by substantial record evidence. The FTC explained the need for a control group, random assignment of subjects, and double-blinding.  POM argued that some FTC experts admitted that RCTs aren’t always necessary to substantiate health benefit claims for foods and nutrients, but that contention took the statements out of context.  One acknowledged making recommendations about diet and exercise even without RCTs, but said that making a recommendation based on the “best available evidence” is “not the same as stating that a causal link has been established.”  Another acknowledged that “well-conducted, well-executed observational research is very important” for evaluating foods and nutrients, but he emphasized that a causal link between a food or nutrient and a reduction in disease risk “cannot be proven from an observational [i.e., non-RCT] study.” Yet POM did claim a causal link.  POM’s own experts offered other views about the need for RCT’s, but it’s not the court’s job to make its own appraisal of the testimony. 
 
POM argued that RCTs would be too onerous.  Practically, it’s difficult/impossible to “blind” a fruit.  But that doesn’t apply to the supplements—two of three of the challenged products—and several of the juice studies were double-blinded and placebo-controlled by using a beverage with “similar color and energy content” as pomegranate juice.  Also, the FTC required double-blinding only “when feasible,” acknowledging that, “in some instances . . . it may not be possible to conduct blinded clinical trials of food products.” Ethically, POM said it was “impossible to create a zero intake group for nutrients in an ethical manner—doctors cannot, for example, ethically deprive a control group of patients of all Vitamin C for a decade to determine whether Vitamin C helps prevent cancer.” But many ads made claims about the short-term benefits of consuming POM products.  Plus, there was “no reason to believe that it would be unethical to create a zero intake group for pomegranate juice.” RCTs could be costly, though the court commented that POM somehow sponsored dozens of studies, including RCTs.  But if RCTs were prohibitively costly, POM could choose to specify a lower level of substantiation with an effective disclaimer.  The need for RCTs was Pom’s own choice based on its ad claims.
 
Next, POM argued that the FTC’s substantiation standard amounted to a new legal rule, in violation of APA notice and comment requirements.  (Actually, the requirements for the FTC can be more onerous.) Nope: it “is well settled that an agency ‘is not precluded from announcing new principles in an adjudicative proceeding,’” and that “‘the choice between rulemaking and adjudication lies in the first instance within the agency’s discretion.’”  Moreover, the FTC’s decision wasn’t a major substantive legal addition to its substantiation standards.  It was consistent with FTC precedent for scientific establishment claims, and the FTC had required RCTs in other contexts. 
 
Defendant Matthew Tupper challenged the finding of individual liability as to him. He became POM’s COO in 2003 and served as its president 2005-2011. He argued that he shouldn’t be individually liable because Lynda Resnick, one of POM’s founders, had the final say on ads. That’s not the standard for individual liability.  Direct participation in the deceptive practices, or authority to control them, is the standard.  Tupper participated directly in meetings about advertising concepts and content, reviewed and edited ad copy, managed the day- to-day affairs of POM’s marketing team, and possessed hiring and firing authority over the head of POM’s marketing department. Even assuming that “authority to control” was a prerequisite for individual liability under the FTC Act, the court would therefore still affirm.  Nor did the FTC have to show knowledge of misleadingness—that’s only for equitable monetary relief, not injunctive relief.  When no restitution or monetary penalties are sought, the FTCA imposes strict liability. 
 
Tupper also argued that he voluntarily retired from POM, but that doesn’t mean injunctive relief was improper.  An injunction might be unnecessary if someone hasn’t shown a propensity to violate the law and nothing in the record suggested the possibility of further violations, but the FTC found that the POM entities, including Tupper, had a demonstrated propensity for misrepresentation and engaged in a deliberate, consistent course of conduct.  Plus, there was no assurance that Tupper wouldn’t return to POM or join another company that markets food products or dietary supplements.
 
Now, on to the First Amendment. Misleading advertising can be entirely banned.  POM said the court should review the FTC’s finding of misleadingness de novo, citing Bose Corp. v. Consumers Union of U.S., 466 U.S. 485 (1984).  DC Circuit precedent established that the factual finding of deceptiveness was reviewed under the ordinary (and deferential) substantial-evidence standard, even in the First Amendment context. Also, the court of appeals would reach the same conclusion even if it were to exercise de novo review, “at least with respect to the nineteen ads determined misleading by the administrative law judge and held by the Commission to form a sufficient basis for its liability determination and remedial order.”  (That could be a certworthiness problem, I think.)
 
Injunctive relief: Part III of the order barred  representations about a product’s general health benefits “unless the representation is non- misleading” and backed by “competent and reliable scientific evidence that is sufficient in quality and quantity” to “substantiate that the representation is true.” For that part, “competent and reliable evidence” meant studies that are “generally accepted in the profession to yield accurate and reliable results.”
 
Part I, however, governed claims about the treatment or prevention of “any disease.” The baseline that claims must be non-misleading and supported by “competent and reliable scientific evidence” was there, but for Part I purposes that last term was more narrowly defined as at least two RCTs yielding statistically significant results, double-blinded where feasible. The FTC clarified that this requirement applied only to unqualified representations, not effectively qualified disease claims.  But claims characterizing a study’s results as “preliminary” or “initial”—“even if describing a gold-standard RCT yielding results with an extremely high degree of statistical significance”—would fail to count as adequately qualified and thus would be prohibited. Instead, such an ad would need a disclaimer “unambiguously” saying that the evidence is “inconclusive” or that “additional research is necessary,” “even if the ad is substantiated by a well-designed RCT that experts uniformly consider to be conclusive, and regardless of the amount and quality of additional supporting evidence other than RCTs.”
 
The FTC agreed that the remedial order should be examined under Central Hudson. Obviously, the government’s interest in the accuracy of commercial information in the market is substantial. Central Hudson requires a restriction to directly advance the government’s interest and be no more extensive than necessary to serve that interest.  To the extent that there was a general RCT-substantiation requirement for disease claims—not requiring any particular number of RCTs—the order satisfied Central Hudson.
 
The RCT standard was the mirror of the FTC’s upheld liability finding: POM’s claims were misleading because they were unsubstantiated by RCTs.  Thus, a RCT requirement was “perfectly commensurate” with liability for past misleadingness.  Sure, POM only made claims about three specific diseases, but the broad scope of the order (covering any disease) was justified by POM’s “demonstrated propensity to make deceptive representations about the health benefits of their products, and also by the expert testimony supporting the necessity of RCTs to establish causation for disease-related claims generally.”
 
But requiring two RCTs was a trial too far.  (If you are from the FDA and you are reading this, you should be sweating.) While Central Hudson isn’t a least restrictive means standard, the FTC still had to show a reasonable fit between its means and its interest. The FTC failed to justify a categorical floor of two RCTs.  All else being equal, two would be better than one; but likewise three would be better than two.  And “[r]equiring additional RCTs without adequate justification exacts considerable costs, and not just in terms of the substantial resources often necessary.”  If two RCTs are required, “consumers may be denied useful, truthful information about products with a demonstrated capacity to treat or prevent serious disease.” That would be contrary to the objectives of the commercial speech doctrine.
 
Suppose, for example, there was a “large-scale, perfectly designed and conducted RCT” showing a significant reduction in risk for a particular disease, “demonstrated to a very high degree of statistical certainty … so much so that experts in the relevant field universally regard the study as conclusively establishing clinical proof of the supplement’s benefits for disease prevention. Perhaps, moreover, a wealth of medical research and evidence apart from RCTs—e.g., observational studies—reinforces the results of the blue-ribbon RCT.”  In that case, there’d be a substantial interest in informing consumers, without any misleading qualifiers that the evidence was inconclusive.  So, apparently the burden is on the FTC to show that this magical study does not exist. Query how much better off advertisers are if they have to figure out on their own which claims require replicated RCTs and which don’t.
 
Even the FDA has held that “[a] single large, well conducted and controlled clinical trial could provide sufficient evidence to establish a substance/disease relationship, provided that there is a supporting body of evidence from observational or mechanistic studies.” U.S. Food & Drug Admin., Guidance for Industry: Significant Scientific Agreement in the Review of Claims for Conventional Foods and Dietary Supplements 5 (Dec. 1999), 1999 WL 33935287 (withdrawn 2009).  The FTC’s two-RCT requirement “brooks no exception for those circumstances.”  But the FTC has elsewhere told industry that, “[i]n most situations, the quality of studies will be more important than quantity.” U.S. Fed. Trade Comm’n, Dietary Supplements: An Advertising Guide for Industry10 (Apr. 2001).
 
The FTC failed to show how such a rigid remedial rule had the requisite “reasonable fit” with the interest in preventing deceptive speech.  The argument that a two-RCT requirement was consistent with precedent didn’t make it fit this new set of facts.  True, the “FTC has usually required two well-controlled clinical tests” before certain “non-specific establishment claim[s] may be made.” But all the cited cases involved establishment claims about the comparative efficacy of over-the-counter analgesics, and reflected extensive consideration of the specific circumstances of such products, especially the subjective nature of pain sensitivity.  The FDA, as well as the medical/scientific community, required replication for clinical tests involving analgesic drugs. But that shows a history of requiring two RCTs only in “narrow circumstances based on particularized concerns.”  Consent orders have also varied on the quantum of evidence, depending on circumstance.
 
The FTC noted some expert testimony recognizing a need for independent replication, due to the possibility that a single RCT’s results may be due to chance or a unique sample.  But “insofar as the results of any particular RCT may be suspect due to deficiencies in the sample or trial, the baseline requirement for health-related claims independently bars any representations unless supported by ‘competent and reliable scientific evidence that . . . is sufficient to substantiate that the representation is true.’ That in turn requires that a study be “generally accepted in the profession to yield accurate and reliable results.”  Plus, the FTC’s experts themselves thought that one RCT could suffice for the prostate cancer and erectile dysfunction claims at issue.
 
Finally, the FTC appealed to POM’s demonstrated propensity to misrepresent the strength and outcomes of studies. But every party subject to a final FTC order has been found to have advertised unlawfully; the FTC didn’t explain how two RCTs were linked to the particular history of POM’s wrongdoing.  The rest of the order requiring sufficient scientific evidence could address that. Two RCTs might be justified in another case, but not here (even for heart disease, where the experts did testify that two would be required? Or can the FTC argue that later?).  So the order was modified to require at least one RCT before making disease claims.
http://tushnet.blogspot.com/feeds/posts/default?alt=rss
Posted in first amendment, ftc, http://schemas.google.com/blogger/2008/kind#post, remedies | Leave a comment