Google Glasses: came with restrictions on use/resale. Digital cameras have done this for years with ToS for the software embedded in the cameras. You couldn’t have a use restriction on an analog camera, but the restrictions seem to run with software embedded goods. Odd to have such a difference v. chattels in general.
Why are servitudes disfavored? Chattels are smaller, mobile, hard to distinguish/fungible. We can glean how these differences, which are differences on balance rather than differences in kind from land, change information costs. (1) Absolute cost of determining restrictions on mug v. determining restrictions on use of land—you have to figure out if it’s the same mug because placement isn’t certain. (2) Relative cost of determining restrictions, given lower cost of many chattels. (3) Aggregate cost: we buy and use many more chattels than pieces of land, and figuring it all out could take more bandwidth than we have. Could have rules distinguishing cars from other chattels, since we deal with fewer cars than chattels generally. (But that piles on quickly.)
What should this say about software-embedded goods? We should be suspicious especially as we move to the internet of things about unexpected use restrictions.
Creepy and Orwellian world if all the things you own keep track of where they are, which is what you’d need to change the difficulty of identifying each object and making sure it’s the same.
Commentator: Miriam Bitton
Paper mentions decreasing information costs: chattel registries for cars, registration for insurance purposes. (Nature of obligation: ownership interest is the usual thing a registry has and the real property regime is pretty hostile to use restrictions even when recorded. And try to register a use restriction when you register the title of your car; it won’t go well!) Other legal systems don’t necessarily use the same land registry, which means that the information cost theory has to be more attentive to differences (history, path dependency?).
My reaction: note the literature on failures of notice in consumer protection/nudging contexts—may further justify restrictions on what kinds of limits can be written into the ToS.
Paul Heald: common-law cases finding nonpossessory security interests in personal property are fraudulent. The language of these cases is vitriolic. It’s fraud for the potential that the person granting the interest will sell off the article, even if the person does nothing more.
Q: in practice, registries are replete with bad recordkeeping, fraud, etc.; acknowledge that it’s a way to handle the information problem but it doesn’t lower cost substantially especially as contextual uses of objects change.
Q: another thing to look at is costs to judicial system: were the servitudes really imposed or not. That may offer a difference between built in and contractual restrictions.
Interesting discussion; I had to leave to ensure I was on time for the next panel!
Wendy Gordon, Dissemination Must Serve Authors: How the US Supreme Court Erred
Commentator: Rebecca Tushnet
In Eldred and then Golan, the Supreme Court accepted the proposition that copyright expansion retrospectively extending the term and clawing back certain works from the public domain could be justified not as incentivizing authors, but rather as incentivizing certain distributors to invest in distributing the newly repropertized works. Professor Gordon suggests that this conflicts with a proper understanding of copyright’s author focus, not to mention with Feist’s insistence that creativity and its support is the only appropriate justification for copyright rights.
If copyright is for authors, what led the Court and various commentators into error? Gordon offers three kinds of explanations grounded in economics, history (which I would call politics), and law. Properly understood, the assumptions of these arguments don’t justify a stand-alone embrace of disseminator interests.
Economics starts with the idea that copyright makes dissemination easier and more profitable, and you don’t get “progress” without dissemination. I’ll return to this in a bit because I think it’s the most important part of the argument, but let me quickly sketch her other explanations. The second explanation, history, is that publishers and other distributors profit from copyright and therefore engage in extensive and successful interest group politics to ensure that the law favors them. The third explanation is legal; Gordon argues that it’s easy to mistake the form of copyright—distribution rights and a special focus on publication, both of which make dissemination look independently important—with copyright’s substance, which remains author-focused. Both publication rules, pre-1978, and the distribution right exist only in furtherance of the first, economic objective: publication and distribution are key points at which legal support may seem necessary to keep too much value from leaking out of the chain from author to authorized publisher; without a distribution right, a reproduction right would be toothless since third parties could make the reproductions and then disappear.
Since the second, historical/political explanation for publishers’ rights can have limited moral or conceptual force, and the third legal explanation dovetails neatly with the first, economic explanation, I want to return to the economic argument. Gordon argues that authors of copyrightable works face a version of Arrow’s information paradox. Unlike inventors who might be able to keep processes and machines secret and still derive value from them, authors generally can’t use their ideas and expression without disclosing them. Because disclosure is the only option, legal protections that substitute for other barriers to copying encourage publishers to take the risk of paying authors for permission. Gordon is skeptical that as much encouragement is needed as it once was, but the point is that the protections for publishers are dependent on the idea that we want authors to get paid. Dissemination in itself is not valuable—disseminating books and disseminating widgets both need to get done for the economy to work, but that doesn’t mean that special legal protections for book disseminators are justified, and indeed they have no special claim compared to the claims of widget distributors. Investment in advertising, selecting which products to offer, helping consumers choose among their options, and distribution infrastructure are required in all markets, and Gordon argues that special subsidies for these common costs of doing business need justification beyond the idea of incentives for creation.
Gordon engages in particular with Professor Jonathan Barnett’s work, also presented here, that publishers make a unique and costly contribution by evaluating and choosing which works to publish, and by using losers to subsidize winners, “increasing the chances that the next, latent bestseller will get the exposure it needs to take off.” If copiers can cherry-pick winners, this strategy doesn’t work. As Gordon points out, this argument is not expression-specific; indeed, it can be recognized in INS v. AP and Doris Silk v. Cheney Bros. Gordon’s conceptual point is that Barnett’s argument is at base still all about authors: all the arguments about finding the next great American novel et cetera hinge on the role of the individual creator awaiting discovery and subsidy.
She also has a related point about competition, prefigured above: all businesses must search and sort. And all businesses disclose, in some ways, the results of their searching and sorting, at the very least through price signals—any business that has discovered a profitable niche can be seen and competition is likely to enter to drive down prices, unless the law prevents that. Non-authorial contributions of distributors may well be extremely capital intensive. But Gordon points out that this proves little without comparison to other industries, and other ways that capital needs might be met.
One fruitful question for further discussion, it seems to me, is whether competition is really the default, though, or whether we in fact live in a world with barriers to entry that make competition a phenomenon on the margins. Consider Amazon’s price comparison app, which—quite relevant to Gordon’s argument—allowed consumers to compare prices on books, lawn mowers, or anything else in a physical store also sold through Amazon. Retailers complained that this harmed them by forcing them to bear inventory and other costs that Amazon avoided, allowing Amazon to compete on price only by free riding on these other stores. Is this any different from the free riding publishers fear? And if it’s not, does that prove that copyright is special pleading or that retailers need more protection from internet free riders?
Yesterday, Barnett emphasized that are other possible mechanisms to capture revenue, like lead time, but from an efficiency perspective we shouldn’t just be trying to get a reasonable level of output; we should get an optimal level of output, and for that, we don’t know whether lead time advantage is sufficient. Without copyright, intermediaries have to choose mechanisms like lead time. As I understand Gordon’s point, it’s that if you’re serious about optimality, that requires an assessment of the alternatives for capital investment, and if copyright provides a method of capturing value that is unavailable to investors in other industries, then you’ll see overinvestment in copyright works/overproduction compared to the optimal result.
Gordon also makes a side point worth further discussion: the “invest in the next big hit” justification needs some more refinement given that big hits generally make almost all their money in a few months or a year. Why copyright should persist longer than that to subsidize publishers is less clear, certainly on a blanket basis not requiring any renewal and allowing expansive rights over derivative works.
Gordon concludes that “Only a comparative institutional analysis can show whether disseminator industries need help that is more or different than other industries need, and whether, if such help is needed, copyright and its roughly 95 years of lead-time-advantage is really an appropriate tool.” Again, I’d ask whether this comparison is as dangerous to copyright-for-publishers as she suggests: it’s not clear to me that there any low-IP mass industries left; our examples of low-IP phenomena like tattoos and stand-up comedy are centered around craft production—perhaps ironically, where the physical performance of a particular authorial individual is important to the otuput. But these are the questions we need to be asking: do disseminators of expressive works need anything special that copyright provides, compared to disseminators of other things? Are there non-copyright ways—grounded perhaps in competition policy—of providing what they might need? What are the costs of favoring disseminators, especially in cases where there doesn’t seem to be a need to pay authors who are creating things anyway?
Gordon’s own comments: paper has various things to say, including explaining what common-law copyright was for (like trade secret) and why it didn’t need to persist after the 1976 Act. Distribution right is a kind of specialized secondary liability. Dissemination is a legitimate part of copyright policy, but not disseminators. (Sounds to me like the standard line about antitrust: the law protects competition, not competitors.)
Publishers count, according to Golan, because they disseminate; she fears the next step that their interests justify anything expansive. Holding makes disseminators’ self-interest hide under cloak of progress. Gordon believes dissemination must aid authors to be constitutionally relevant.
Adam Mossoff argues that even if professors don’t need copyright royalties, our publishers need them. Her question is: does this mean that giving copyright to professors serves publisher interests per se? The answer is clearly no. If copyright is needed for scholarly publishers, and if professors need journals to gain reputation/flourish, then copyright is only conditionally justified: it still has to be tied to authors. Thought experiment: what if journals perished? If there were other ways to get the job done, sorting for quality, we shouldn’t care if journals died. If there aren’t other efficient means, copyright should care: but that shows we only care about the incentives of authors.
Balganesh: you’re arguing from purpose and structure. What about copyright-like structures to further disseminators’ interests, such as the Broadcast Treaty, with rights independent of right in the content of the broadcast.
A: Jeanne Fromer explains Golan as the SCt’s desire to avoid the question of whether treaties can bypass constitutional commands about progress. This reminds her of that. SCt is always using copyright this way—it’s a garbage can. Given the presence of the IP clause, there’s a tension—she thinks it would be improper to make IP-like rights without limited times/creatorship or inventorship.
Q: how do you deal with Lady Gaga, who released Born this Way on video free to view before she released the album—does she care about copyright?
A: sounds like advertising.
Q: celebrities may be able to ignore copyright/use network effects. If this happens a lot, why worry about intermediaries?
A: But advertising is consistent with the copyright incentives for authors model. She does think that incentives are overstated as justification for authors, given the nonmonetary incentives.
Justin Hughes: Congress should be held to furthering progress by means of authorial rights—do you mean that, or do you mean “when it grants exclusive rights, they have to be to authors”? What happens when you subsidize authors with NEH grants—would that be ok to further progress?
A: yes, of course—she’ll tweak the statement.
Q: what about the public? Can we consider the benefits of preservation of works to the public?
A: what she fears isn’t so much that Congress will enact things that aid dissemination, but that without the limitation she recommends that Congress can do whatever it wants at the behest of disseminators, without considering public interest. If we could get over the epistemic problems of proof, and courts could see that parts of the argument were selfish and parts in aid of distribution to the public, she’d be more sanguine.
Q: how does the progress clause do this work? This seems like a rational basis problem. Congress says it’s trying to help authors, even if academics disagree/don’t need copyright.
A: she’s undermining the stance she took in Fair Use as Market Failure, where she proposed that any social benefit mattered. She’s not sure how to square that with the idea that disseminator interests of publishers shouldn’t matter, if dissemination interests of copiers do matter (even if they’re just making pure copies).
Q: Arrow’s information paradox doesn’t engage directly with copyright; why not point to Landes & Posner’s discussion instead?
A: They’re brilliant, but they’re completely unpersuasive arguing that centralization is good for progress.
Ann Bartow: are there differences between Wikipedia and Encyclopedia Britannica as disseminators; the latter edit/quality control.
A: to the extent they’re editing, they’re doing copyright work, even though editors politely don’t claim copyright. Lots of intermediaries do creative things—colorization of film, even. Treat them as authors when they’re doing that sort of work.