Southern California IP Professors’ and UCLA Entertainment, Media, and Intellectual Property Colloquium Workshop, UCLA School of Law
Congratulations on the organizers for a successful event, and thanks to everyone who came out on a Sunday to attend.
Robert Merges, UC Berkeley, Justifying Intellectual Property (now available)
Quick tour of the book. Two ideas to emphasize: (1) propertizing labor, and (2) the importance of waiver. Looking for the best case to be made for the institution called IP. Midlevel principles: attempt to fit the foundations into something more policy-oriented but that would also be somewhat delinked from foundational principles so that we could agree on midlevel principles even if we didn’t agree that Locke and Kant were the guideposts. Part III: contemporary topics/applications of the first two parts. Much IP work designs its theory to justify a policy outcome or feed into a particular topic/issue, and he tried to avoid that—didn’t want to look for foundations for his ideas about pharmaceutical patents or digital copyright, but instead tried to build from the roots.
Foundations: utilitarianism: net welfare. The God that failed. Deontological/nonconsequentialist grounds—had a conversion experience. Efficiency is a policy tool, but he can’t justify the whole field on the empirical evidence, which seems to him undecided on whether IP is a net positive for social welfare. (This kind of sounds like he actually did design his theory to justify a policy outcome, albeit a really big picture one, that IP rights are justified.) Commitment to foundational pluralism: ultimate normative commitments will vary from person to person, so he designed his midlevel principles to accommodate those who have faith that IP rights are justified under utilitarian theories alone—room at the bottom: borrowed from Rawls, a shared vocabulary for shared public reason. (Is there room for people who have faith that IP rights are not justified under utilitarian theories?)
What do you call that shared policy space? Borrowed from Jules Coleman: midlevel principles.
Locke: applies more to IP than real property. Much of what’s out there is already owned in the real world, so the original acquisition/appropriation story is not necessarily relevant. As we sit here, though, thousands of people are coming up with potentially protectable stuff right now. So the story of how I come to own something that is as yet unowned repeats itself in IP constantly. Cartoon version of Locke is libertarian/Chicago school—or Nozick’s clever and facile idea that the problem with Locke’s theory is overclaiming: if I pour my can of tomato juice into the ocean, do I own the whole thing because I mixed what I own with nature? That’s a silly story. Locke’s actual treatment of appropriation is different. It’s not just any old owned thing that gets mixed, it’s labor. There’s always a natural/logical limit, proportional to the amount of effort put into, say, gathering acorns.
Kant: we had to invent the state to protect an aspect of autonomy: we have to protect certain kinds of claims to stuff to allow people to work out their selves in relationship to objects. Our relation to objects has something to do with self-development. Because we need this complex idea of possession we need law and thus the state. This is an interesting rationale for IP because it’s precisely the relationship between what you created and what people do or could do to it that create problems: someone reads Harry Potter and wants to create an altered version. Why should she have any say over that? Takes a complex notion of possession to justify that. Kant’s notion of relation between person to object has something to say about why we might want to regulate certain incursions far away to protect her autonomy. Kant also has certain limits, as important as with Locke.
Rawls asks how we create a fair society. We haven’t really taken on distributive justice in IP. Starting point: everyone ought to own more or less equally, absent a good justification. IP rewards creators and takes resources from users/consumers; there is an inherent inegalitarianism about this system, where very successful creators get rewarded quite a bit. Can be defended on the grounds that it helps the poorest members of society enough to justify the redistribution towards “creatives.” Not saying that all features of the current system meet that criterion, but it would be okay to say that focusing some resources on creatives is justified.
There is a kind of distributive justice discussion conventional in the IP world, justifying particular features of particular IP rights—should fair use apply to certain uses, or an improvement fall within the scope of the original patent. These discussions often have a distributive justice component, because we ask about fairness all the time. Tries to gather up these discussions and pull them together in a single approach. Fairness not at the level of a whole system, but at the level of the grant of the individual IP right. Stages at which fairness comes into play: (1) moment of the grant, where most discussion goes on—should we allow certain defenses or limits when the right comes into existence—e.g., the Girl Scouts singing songs at the campfire. (2) Deployment: the right’s been granted, and market conditions have changed/the world has grown up around the rights. Rights as deployed seem to have disproportionate effects compared to when they were granted. (3) Taxation: can defend a fairly elitist system in terms of grant of rights, but it’s absolutely true that society makes all kinds of contributions to every single IP protected work out there. There is therefore a social claim to some of the proceeds. That’s how you both protect an individual creator/property rights and also recognize society’s contributions. You could also limit her rights, but he doesn’t like that as much, and taxation can make sure that distributive justice comes in at a wider-focus angle.
Commentator: Jonathan Barnett, USC
Suppose you do think welfarist analysis has something to say, what does this book have to say to you? Is law and economics analysis realistic in terms of how people actually think and behave? Is it complete? While we can explain surface structure of IP through efficiency, can we give an economic justification for IP? Is it operationally useful, in that even if we can make the models work we can’t ever get the data?
Book’s strong proposition: we can use nonconsequentialist reasoning to give a better justification for the field in general. Economic analysis doesn’t come up with clear answers all the time. Nonconsequentialism can help us decide close calls or even create burdens of proof.
So can nonconsequentialism give better answers or be less controversial? We seem to be at an impasse between law & econ scholars and others.
Barnett doesn’t see how Merges’s answers are more determinate; can’t be empirically tested. Also not less controversial, since redistribution (distribution) is controversial.
He also can argue that nonremovability from the public domain and proportionality (two of Merges’s core principles) can be reduced to efficiency, since efficiency recognizes transaction and other costs. So these are the outcomes we get anyway. Merges has another key principle: dignity. The core contribution of the book is that it looks at dignity from a distributional perspective. Distributional effects at the core of its IP analysis, and that’s not what efficiency analysis does.
Efficiency has no tools for analyzing how the pie is divided, and that makes it hard to explain what IP rights look like in the real world, e.g. provisions for various industries in the Copyright Act. If you do distributive analysis to produce determinate outcomes, you have to pick a privileged class. Merges suggests the pro author in copyright and the invidual inventor/small groups in patent (any group that doesn’t operate with large-scale advantages). The interests of that class should predominate over the short-term interests of users, amateurs, and extremely well-resourced technological intermediaries such as Google.
Distributional arguments usually focus on users and amateurs, but why should their interests prevail over those of the pro author/small firm? Book gives a nonconsequentialist case using real-world concepts for a robust IP system. That type of language has mostly been used to argue against IP. (Me: except before Congress.) That’s important because downgrades in IP has regressive distributional effects for individual authors and small firms, but to justify this we need to get back into messy empirics. Different types of IP systems amount to a social choice. Robust IP is a choice of atomistic production by autonomous IP providers. You may get big output levels through weak IP, but if you’re going to do that you contract the class of authors/individual inventors.
Merges: One chapter tries to address these issues. Likes Kant’s emphasis on autonomy. Why would we want to privilege smaller units of production? Kantian worldview and some Hegel/Adam Smith give us a good grounding for that. It’s hard to justify privileging that group on strictly utilitarian grounds, and yet he has a strong intuition that’s the right thing to do. Funny how far IP literature has gotten from the idea that a core purpose of the system is to allow professionals to make a decent living. Redistributive effect, but done to make jobs that pay a decent wage to people who are about high-quality content. Macaulay: tax on readers for the benefit of writers. IP literature has slanted towards critiquing this case, but many industries’ economic model is threatened and writers are group worth defending.
In the relevant book chapter Merges talks about the role of big aggregators/intermediaries. When he talks about the privileged class, he doesn’t think about them. Cartoon Network, Google, Microsoft don’t need a whole lot of special protection to make money. What role do they play in the ecosystem? Do they support the smaller units of production? Do they feed back into the viability of the atomistic creators? That’s a tougher question, both empirical and not. The idea that content platforms will pull all content into vertically integrated platforms is not quite right, but the question is one of optimal industry structure. Sometimes big guys can make little ones viable; to the extent that they do, we can’t be indifferent to what happens to them, and sometimes an attack on them can be an attack on smaller guys depending on the structure of the relationship.
Q: IP rights eventually expire—how does that fit?
A: He addresses this with respect to Locke’s sufficiency proviso. Spoilage is also important, as is Locke’s charity principle. Kant and Locke inherently limit property rights, and time limits on IP are a perfect example. Proportionality can connect with that—the difficulty of long time horizons in copyright may mean that fair use ought to expand over the life of a copyright. The further out you get, the less strong the original owner’s claim is and the greater third parties’ reliance interests are, especially with works that become canonical. That allows us to balance a fairly long time period of protection with others’ interests. Strong limits on strong rights.
IP is important to smaller units because it allows them to enter into contracts with bigger guys in less risky ways—they can disclose without betrayal. Allows them to substitute for being employees/units of a big company. Bigger guys also, when they sell the products of R&D, can bundle and sell in amalgamation. Hard for someone to come in and compete with the whole broader product offering. Even without IP protection, Microsoft would probably be ok given how hard it is to create Windows.
But if the big guys employ creative people, doesn’t that create a lot of jobs? Couldn’t we say that IP helps indirectly? Yes, we can say that: without copyright, a lot of the animators in the latest Pixar movie would be in trouble. But the theoretical point is that there may be other ways that Disney could figure out, technologically, politically, and through business models how to make money. The little shops are the ones who are exposed without IP. Comparative need. These are complicated ecosystems, so part of the big guy ecosystem is little guys and making distinctions can be somewhat artificial.
Q: You get from Rawls: Much individual action is the result of pervasive social influence, so society has a legitimate claim on proceeds. But questioner’s read on Rawls is not about what society claims, but how people can come together in a society in which they can regard each other as political equals. First principle of justice in Rawls might therefore place significant restrictions on IP implementation—there are some basic principles that need to be guaranteed before you can start thinking about efficiency questions. One is right to freedom of expression/conscience, and current copyright system appears to implicate that. So do restrictions of justice precede IP/welfarist values?
A: Difficult question. There’s always a certain amount of bracketing necessary in this project, and he takes First Amendment limits on IP as fairly fixed by convention, though he knows that’s subject to critique. Though he thinks freedom of expression is part of first principles, a basic human right required by any just system, he chose to bracket the content/scope of freedom of expression and its limits on IP rights. Conventional answer: the Eldred majority says that question has been asked and answered via idea/expression, fair use, and other limits. Maybe there is a systemic effect (Lessig’s permission culture) that swamps the effect of the explicit limits, and Netanel has argued that this erodes democratic values. But he thinks the conventional tools are adequate to address the problem and takes them as given. Courts are pretty sensitive to the First Amendment dimensions of copyright problems and do a pretty good job. That’s not to say that, in the original position, people would adopt Eldred majority’s view. Rather, he thinks that just saying Rawls’s first principle would apply overlooks the importance of individual property rights as rights for each citizen. Most of us would not be comfortable with Rawls’s disregard of property rights today. Freedom of expression shouldn’t so dominate the right to individual property that individual property becomes a footnote, as it did in Rawls’s 1970s theory.
Q: You criticize utilitarianism for its computational complexity and thus its unworkability. Why isn’t this a problem for lots of nonconsequentialist theories, which often think that outcomes matter even if they’re not the only thing that matters? Also, what do you mean by offering a workable normative foundation for IP? He thinks that normative foundations are about right/wrong, not about how we come to know what’s right/wrong. There are many complicated questions we might not know how to answer, like exactly what makes democracy good/bad, and difficulty in figuring it out doesn’t mean it’s wrong to prefer democracy.
A: Not philosophically trained, so needs a delay. The more context knowledge you have, the more you problematize your field of study. Facile answer: know enough about welfare economics to have thoroughly confused myself over 25 years. I jump to this new raft and it looks clear. But 200 years of Kant commentary may leave me more confused than when I started. Also: when you look at Kant, he believes that accessing knowledge of right/wrong is not impossible and that there’s a universalist answer there if you seek it. Utilitarianism isn’t just complex, but also maybe ultimately unknowable. There’s a difference between difficult and unresolvable, and he thinks that utilitarianism tends to the latter; the equations just don’t resolve. Kant and Locke are at least possibly resolvable by contrast.
Wendy Gordon: Why IP? There’s lots of purposive labor in the world as to which rights could be given to redound to the benefit of the less well-off. Let’s say a geographic area has an unrecognized need for a particular product. Someone brings it in. Under a competitive system, people would rush in and the price advantage vanishes. Why shouldn’t the person who figured out this need get protection against these competitive harms?
A: He’s not sure that you couldn’t find some IP platform on which to defend some kind of supercompetitive right for that person. The hypothesized product isn’t novel, but if they came up with a catchy logo or distinctive set of practices with local flavor, maybe even distinctive appearance, we could turn it into IP.
Q: but if there’s no IP way to do it, then what?
A: maybe we should. There may be other classes of people worth privileging. Another way to think about GIs is to say that a well-known variety of tomato that hasn’t been tried in the Napa Valley before should get protection. (Protection for whom? Individual property isn’t GI status; to the contrary it’s shared with everyone who meets the criteria.) If you put a lot of work into it, and if someone else claiming that work seems unfair, we can justify misappropriation claims/property claims under some sort of restitution theory. (Maybe it’s just that I’m thinking about teaching property next semester, but I can’t help thinking of this poem.) We in IP tend to be hedgehogs in that we stick to our field. Any field, if you stay in it long enough, can be the grain of sand that reflects the whole world. There is a way to talk about deontological principles in a way that makes sense for my field.
My thoughts reading Merges’s excerpt: The frontier metaphor Merges uses for new creations might be better than the “public domain,” which as Julie Cohen has argued abstracts away from the landscape in which situated human beings find themselves. It’s easy to think of the public domain as always and everywhere reachable, but that’s not what the information landscape really looks like, as lived. I like the frontier in particular because (1) someone else was usually in what Westerners called the frontier and (2) the frontier is reached via the known; access to the known is a practical prerequisite.