Sunday, August 09, 2009

IPSC: daily double--theory

Wendy Gordon & Boris Milman

Derivative Rights and the Rule of Law: Judge Posner and Copyright

Posner’s approach privileges property in a strong fashion. We in academia spend a lot of time talking about how bad the property model is, but it’s hard to pin down where this is happening because it’s usually so subtle; Posner is not subtle.

Gordon has used the Lockean proviso to justify limiting property rights, but at the core of Posner’s argument in Pickett and in Gracen, he relies on an upside-down proviso: privilege a prior property owner, and not the public domain, against second-comers. Is this peculiar, immoral approach justified by consequences?

Most of us think pure Kantianism is silly. Consequences should matter, but the bottom line: Posner in particular, and other courts like ERG, have created a double standard: a derivative work based on the public domain, like a Santa statue, we should be a little concerned that copyright might impede access to the original, but our originality standard should be only at most a little higher. If a work is based on a copyrighted work, like an inflatable costume based on an existing character, or a plate painted to recall The Wizard of Oz film, we should ratchet up the degree of originality required, in order to protect the copyright owner. But of course the copyright owner had the power to condition the licenses in such a way that it could get the copyright back; Posner rule is paternalistic towards big corporations, while the public domain—which needs help staying free—has no help.

Theory makes people overconfident—incredible stretches that create bad precedents, incompatible with empirics.

Jorie Gracen took over 8 hours just to plot the design elements: where Toto would be, where the sun would be, where the trees would be—no portion of the film looks like her plate. 40 hours painting; turned black and white reference stills into color. Under most standards, this is a copyrightable derivative work, created under license. Holmes: if a judge or lawyer should take hand at even the simplest sketch, he’d find out how hard it was. Roth Greeting Cards: 2 or 3 decisions was enough for copyrightability; here there are thousands, and still Posner rules that there is no copyright.

The upside down proviso: the proviso is that you shouldn’t give property if it would cause harm to use of commons. The upside down proviso is that you shouldn’t give property if it would cause harm to another property owner: “must not in any way affect the scope of any copyright protection in that preexisting material.” The Proviso is about protecting later generations’ ability to appropriate/create. The upside-down proviso is about protecting the people who got there first. This also drives Posner in Pickett, depriving creators of unauthorized derivative works of rights even in their own creative contributions.

Kozinski’s similar analysis in White: right of publicity should be preempted to the extent it creates an anticommons problem for copyright owners. Makes sense as a supremacy clause analysis. But where there’s no declaration of supremacy, it doesn’t make sense to prefer prior rightsowners to later ones. The upside-down proviso is paternalistic with respect to earlier owners—it applies even if they chose not to make their permission conditional, as in Gracen.

Matt Sag: Should the test for what counts as a derivative work be any different between works in which copyright has expired and works with three days left in their term?

A: Posner would say yes, because works based on the public domain need the incentive of copyright to be created, whereas the copyright owner can safeguard the value of the copyright. Gordon’s view: the question is whether you should tailor copyright standards to the individual situation. One answer is moral: for defending First Amendment interests, public interest in creating one’s own identity, those issues are nonfungible and need to be treated individually, for example in fair use or questions of essential facilities. When it comes to the economics, Jane Ginsburg’s approach is right: not whether an individual copyright furthers the public good but whether the system does. Gordon can’t imagine a system making gradations of originality working all that well.

Molly van Houwelling: Beanie Babie decisions—invokes a photo of a real pig. He seems to be protecting the public domain by a different doctrinal tool—the stuffed animals here don’t look like actual pigs, but if they did, that would make it harder to establish actual copying. A way of insulating public domain not through originality but through infringement.

A: Nimmer makes a similar point: that the defense of independent creation avoids some of the dangers Gordon worries about. Batlin makes a good point: the Customs people who blocked the second Uncle Sam bank blocked it because it looked too much like the first and they didn’t have the training or the time to find out that they were both based on a public domain work. So she doesn’t think that this is enough. But there are other ways of helping the public domain—thin copyright. Mark Lemley on the economics of improvement: a balance of power between all the provisions in copyright. If you have a high standard for getting copyright in a derivative work, then we should have a blocking patents type treatment for them; if the standard is low, then it’s not as reasonable to allow blocking.

Andrea Matwyshyn

Bourdieu, IP and Privacy

Can strong IP rights in databases and information privacy/security coexist? Lots of people are concerned with their privacy duties to consumers. The same data that exists in protectable database is the consumer data that’s being collected. If it’s lost, both the producer and the consumer lose. Consumer data can’t be framed as an alienable property right, because it doesn’t create enough incentives to protect it.

Framing data as a human right also doesn’t work, because human rights can’t be shared; qualitative mismatch. Looking for a hybrid. Consumers say they want privacy, but easily give away their information. Sometimes we suspend fundamental rights on a contractual basis—you can sign a nondisclosure agreement, for example. So maybe that’s where we should start: a consumer perspective.

Solution: Control over context, or selective embedding. Data has value as an intangible asset—copyright is about limiting use and context control. Feist was about the compilation, not the data. Trade secret: we talk about controlling disclosure in a way consistent with a possessory interest. Contract law/licensing grants some access and prohibits others. CFAA covers situations in which access has exceeded what’s authorized.

Case law: as data breaches become more prevalent, we see bifurcation between consumer and corporate interests. Losses suffered by banks when credit card info is lost are inherently connected to the losses suffered by computers.

Consumers maybe always wanted a limited licensing regime.

How do you calculate value of tidbits of information about consumers? Bordieu proposed that there are two types of capital: one economic and one social. Social capital is embodied information that hasn’t been put into the market yet. Apply this to data/privacy security. When I know my favorite color is red, that’s social capital that can be converted into market capital when I tell a market value. Value of this depends on how limited the knowledge of it is.

Social capital embodies “tells” regarding income, social class, and future consumption. Converts to economic value when embedded. The parallel is when an artist converts an idea in her head into a marketable commodity/painting. If you know a person’s favorite beer, age, and location, you can know many things about that person. Keeping the information about preferences and characteristics doesn’t generate economic capital, but it’s still a source of social capital. Embedded into the market, it’s converted into economic capital/value.

Consumers love free stuff, which is why they want to convert social to economic capital. They love having companies who know them and know what they like. But there’s also a creepiness trigger. Facebook keeps crossing the line. Maybe what’s happening is a perceived betrayal of the limited data license granted FB.

Beer example: beer seller knows preferences for purposes of selling beer. If habits sold to alcohol treatment company, which started marketing to her in her place of work, that would be a perceived transgression even if formally allowed by the license.

Realigning consumer expectations across fields of law: Congress tried to address the interconnection in DMCA §1201, allowing consumers to address privacy in the context of circumvention, but it hasn’t been explored in caselaw. Acknowledges, though, that the tension between proprietary control and consumer expectations is real.

Punchline: harmonizing IP with privacy/data security could recognize the idea of selective commercial embedding. Privacy policies have never been fully recognized as legal contracts. If we did this, we can describe them as a data license with an implied warranty of security and privacy. Presence of a data breach notification is further evidence of presumptive breach, which should be enforceable through ordinary remedies for contract rights. Not an exact science: all we can ask is reasonable care for the data.

Q: Privity required?

A: Privity is a slippery concept. Security auditing: done on behalf of one company for another. If it’s inadequate, who can sue? Done for the subjects of the data; are they third-party beneficiaries? We want there to be a connection between these parties, so this is an evolving area of the law.

Christopher Yoo

Reconceptualizing Personality Theory Through the German Aesthetic Concept of Spiel

Kant and Hegel’s personality theories are misread; people generalize from their theories about property theory, but Kant wrote about unauthorized copying of works specifically. In context of social movements of the time, Kant was German Idealism and later theorists were German Romanticism—abhorred Enlightenment excesses and didn’t like individuals pursuing their own conceptions of the good as collapsing into the French Revolution; rejected things we associate with author-centered theories. Why care? Because creativity doesn’t just self-actualize you after you create the object; the process of creating is what makes you a person.

Key concept: Spiel, play, associated with Schiller (playwright, responding to Kant).

Kant wrote two pieces on copyright. Divide with Hegel on alienability. Metaphysics of Morals: distinguish between right in thing and right in idea, you can alienate the copy of the book without alienating the right in the idea. But unauthorized publication is not a harm to the author, but only to a publisher; his justification is purely utilitarian. He also says you can freely copy artistic works because there is a nonmechanical agency by the second comer that is attaching their will to the work. There’s no protection for derivative works/translations at all, because other people develop their wills thereby. Yoo thinks this is inconsistent.

Kant: we want people to feel free to make judgments of beauty. That’s a moment of self-actualization, exercise of freedom. Play: you can do anything you want. For this to be the case, you have to be entirely free from rules, because rules on what counts as beauty interfere with free play. Freedom has to be disinterested: free from hungers, preferences, drives. Can’t be an appetite/need to produce, nor can it be from a sense of moral obligation. He is reacting to the Enlightenment: this is an entirely subjective experience, but it may be universal—we should all be making judgments about the same thing, and because it’s free from preferences it can be universal. Baffling German statements! It’s passive—appreciation of beauty—not active/creative.

Another aspect of playfulness developed by Schiller. Antisubjectivist: response to Kant. He has a notion of dualism. The play drive brings it all together—rational and emotional/sensual/experiential sides. The play drive engages both other drives and unifies the psyche, making you complete. Engagement of personality: man is only fully human being when he plays. Play is also passive: contemplation of beauty. Contemplation of certain kinds of beauty is key: elitist conception of romantic author who experiences world and is able to rise above it and translate it for mere mortals. Popular novels, he thought, were not art. Art for art’s sake.

Hegel’s Philosophy of Right: Art receives less protection than literature. Some repetition/copying is inevitable and good. Teaching: you have to repeat ideas. Access to derivative works can help other people self-realize, so copyright has to balance. But allowing trivial changes to a work has utilitarian incentive consequences. Property is relational—what’s mine v. yours—but Hegel has very little to say about the scope of property.

Solution: Karl Marx. Has an aesthetic theory: creativity is a driving force of human beings. We need to create to bring the drive for unrest and the drive for unity together. You don’t need a romantic author’s intervention to do this—it’s a fundamental way to self-realize for everyone. Alienation of labor/capital: you don’t own your work. Writing for money is corruption. He has the notion of false consciousness, but art can break through that. They can get ahead of society and lead it.

New concept of personality theory: act of creativity is what matters. Creativity is a human need. Must be disinterested: creation for its own sake, not commercial.

Implications: give us an idea of the public domain. If we need room to create, and if we need access to preexisting works, now the public domain looks like a way to get there. The need for disinterestedness may shape the scope of the right—it’s a personal right. Do authors have a self-actualization right to have their works read, or at least readable? Or is creation alone enough.

Mark McKenna: Abraham Drassinower has a similar point about authorship—authorship entails allowing other people to be authors.

A: There’s a latent notion in a lot of this stuff that not everyone’s an author. Connected to whether we strive for an ideal or follow our own unique subjectivities.

Sprigman: Locke, as well as personality theorists, work in a way parasitic on utility theory. They have distinctive theories for why we have rights in the first place, but as soon as they start talking about allocating rights between people/distribution, they go to utility.

A: There are theorists who don’t believe that, and rely on self-actualization. There are flavors of Lockeanism and Kantianism that are consequentialists, but not all.

Sprigman: What’s a nonconsequentialist way of thinking about the proper distribution of self-actualization between you and me?

A: If it’s about self-realization, then you can define it so they don’t compete. I can create derivative works freely as long as I don’t disseminate them.

Gordon: But that might not be self-realization to some.

A: Then the question is the scope of the right to be read. Richard Fallon has a First Amendment paper on this dealing with two notions of autonomy: is it a notion that’s ascribed or a condition that’s achieved, and they’re fundamentally inconsistent. Same here.

Q: Any notion that the work we create has to be a public act? He always thought personality theory does a better job on why we protect diaries than incentive theory. Is there such a thing as a private work of art?

A: Maybe, but Kant and Hegel weren’t saying that.

Von Houwelling: Rights to create v. rights to artifact as reflecting on moral rights. We are nervous about creating rights that continue after they leave the authors’ hands. She imagines an intangible right to revise, not to withdraw existing copies but to redo; that would create less anxiety about restrictions on alienation of tangible things.

A: Goes to the idea of property as attaching your will to things. Dynamic quality of the relationship. Personality theory is usually seen as author-centered, but are there competing values that are nonconsequentialist? The way in which a work continues to embody your will makes it dynamic by nature.

Christopher Sprigman & Christopher Buccafusco

Valuing Intellectual Property: An Experiment

Experimental economics: underlying assumptions of IP’s theory of markets. Most IP rights are structured to protect property: right to exclude protected by injunction. But authors often aren’t most efficient exploiters, so IP relies on reallocations of rights. People will make wealth-maximizing decisions according to preferences. Unreflective application of Coase Theorem—rights will end up where they’re most valuable. Behavioral economics says no! People have heuristics/cognitive biases.

One such bias, endowment effects: initial allocation matters to valuation—classic experiments with mugs show willingness to pay and willingness to buy differ substantially. Is this true with IP? We decided to test it!

Haikus, short poetic form. We have students come in and they’re assigned roles, some as authors, others as owners, others as bidders. They write haikus. Treatment: participation in contest v. participation in lottery. Contest: Haiku will be judged out of 10 poems, $50 prize—ask authors how much they’d take to transfer their chance of winning the contest to a third party. Owners are people they lie to—tell them that this poem is yours; they’re given the same poems the authors have written, to control for quality. Ask them the same question: how much would they take to transfer their chance. Another treatment: tell them there’s a lottery with 10 other poems. Another phase: currently authors, owners and bidders only see one poem. Later: everyone will be able to see the other poems in the contest/lottery.

Hypothesis: authors will value poems more than owners, who value them more than bidders.

Results: yep. In the contest condition, large and significant gap between willingness to accept and willingness to pay—authors/owners want over twice the price bidders are willing to offer. The gap between authors and owners in valuation are, however, not significant, though authors want a little more than owners do. Bidders are acting rationally, bidding between zero and $5, and owners/authors are distributed pretty symmetrically around a mean of $25, though some want the whole $50.

As the authors/owners assess quality, they assess their valuation higher, because they think the chance of winning is higher: rational (assuming that quality valuation is rational).

Lottery condition: endowment effect is even stronger. Larger but not significant difference between authors and owners; ratio between 4-5 as between authors/owners v. bidders. Endowment effect studies involving lotteries usually have a WTP/WTA gap around 2x, but ours is 4-5x: a big difference.

Is this optimism bias? Creators/owners think they’re better than they are. That would seem to wash out in the lottery condition. Are people experiencing regret aversion? People don’t want to experience regret associated with creating and then selling the winning poem. (What if you were asked to donate it to a charity? Does it matter if you worry about an undeserving private party getting the benefits? Relates to gift theory as well as user-generated content/digital sharecropping issues. Or what if you have a chance to transfer it to someone who seems to have institutional competence—editing, promotion, etc.?) What about property rule v. liability rule? Some evidence suggests that liability rules eliminate the endowment effect. Other issues surround market experience.

Q: do you tell contest participants about criteria?

A: We tell them an English grad student will judge quality. We haven’t tried to specify.

McKenna: Optimism bias—not sure it washes out in lottery condition: people often play the lottery with a disproportionate sense they’re going to win. It might work differently, but it might still exist. Surprised there’s not a bigger author/owner gap, given investment between authors and owners.

Stefan Bechtold: Why not just a normal WTA/WTP question? Are you saying something about rights or just about contests/lotteries?

A: People criticize endowment effect studies for not asking people to put their money where their mouth is. We tell them we’ll complete all these transactions. There is a winner in each round, who gets paid. But they’re operating in a money market as far as they’re concerned. We also have a special challenge: this is nonrival property. Either you have a lottery ticket or I do. We have to induce value of nonrival property and then trade the income stream. If this is just a lottery study, this is a weird lottery—the results are distinct.

Bob Brauneis: The way you’re running this—all the poems will be entered into the lottery, only ownership is at issue. Control: entrance and exit—how much would you pay to enter into the lottery, v. how much would you take for avoiding the lottery. Regret wouldn’t be the same issue.

A: We tried to design this protocol to do it all and couldn’t do it understandably.

Q: what about a bargaining phase where bidders and owners get to talk? Initial bargaining positions—so what? Equilibrium includes bargaining.

A: They know the rules of the game, so this should reflect actual value.

Q: but you’re missing WTP information.

A: Q is how burdened by transaction costs these negotiations are; if the parties are really far apart, there are more transaction costs. Rationality affects the size of the transaction costs. And that also has implications for whether property rights are a good idea.

Q: Cell size/demographics?

A: Over 50 in each cell, which is statistically powerful.

No comments: