Jason Mazzone, Brooklyn Law School, Who Should Regulate Fair Use?
Constitutional lawyers thinking about boundaries think about private/state action; separation of powers; division of authority between federal/state. Administrative agencies can develop expertise, fill in details of statutes, aid with enforcement, and provide specific rules to govern private conduct. So his paper is an effort to think about how agencies could play a helpful role in IP, particularly in fair use.
Fair use is a classic case where agencies could help. Standard ad law story: Congress crafted a deliberately broad and open-ended statute. The courts decide fact-specific disputes. Neither Congress nor courts allow us to predict in advance what use would be fair. Strict liability doesn’t tell you where the boundary is, only what happens once you’ve crossed. Result: copyright owners leverage vagueness and take advantage of fear of legal penalty. E.g., notices that claim that no portion of a book can be copied without consent. This chills a lot of valuable conduct. Another problem: contract law is being used to restrict rights that copyright would otherwise grant.
This is undesirable, and Congress and the courts haven’t helped prospective users abide by the law. When things go to private regulation, fair use erodes.
Administrative agency: Two models, though he’s still working through them. First, Congress would (1) make it unlawful to interfere with fair uses of copyrighted works, and subject offenders to civil penalties, a kind of consumer protection statute; (2) create an agency to enforce this statute, perhaps before an ALJ; (3) specify that federal fair use law, including agency regulations, preempt state contracts in case of conflict. The agency would define standards for fair use. Defendants could still assert fair use, but courts would defer to the agency’s regulations, which should be pretty specific, which would help make clear before getting to a judge whether any particular use is fair or not.
Second model: agency would have even more control over infringement determinations, including adjudicative authority. A copyright owner alleging infringement would be required to file a complaint with the agency; if no fair use defense were asserted, the case would proceed. If it were asserted, the agency would have authority to resolve the fair use claim; the case could go forward after, but again there’d be deference.
Agencies are good at adapting.
Rebecca Tushnet, Georgetown University Law Center, Economies of Desire: Fair Use and Marketplace Assumptions
I’m interested in the boundary between incentives and tastes. The incentive story, copyright’s standard justification: people only produce the stuff you want if they’re paid in money or other utility. Taste: people will consume the good at issue up until the cost is no longer worth it to them.
How do we get singers and painters and actors and writers? Creative production often functions like a taste, except instead of consuming creative works, people have a taste to produce them.
Won’t spend much time on the standard justification; just note that it’s everywhere in the literature, even among reform proposals. E.g., Judge Kozinski and Christopher Newman proposed eliminating fair use for derivative works, but requiring profits from derivative works to be shared. This has a baseline assumption that profits will routinely be available; noncommercial production doesn’t fit.
Creators’ accounts of production are about desire: addiction, obsession, excess. Arthur Miller said: “If somebody doesn’t create something, however small it may be, he gets sick.” Michael Chabon writes to entertain; his first story was a Sherlock Holmes pastiche, spurred on by the gaps in the original text that seemed to him an invitation to continue, as all enduring literature offers an invitation to continue. Steven Brust wrote a Firefly novel and released it under a CC license because, he said, “I couldn’t help myself.” Fans working in noncommercial spaces, often with limited cultural capital (though subcultural capital may be available to them), say the same things—writing is like a drug. Clay Shirky says that Perl is a viable programming language because millions of people woke up today loving Perl. People create ship designs for the game Pirates of the Burning Sea, researching in nautical museums, because they love to do it—some of them don’t even play the game; they just like to make ships.
Using the language of incentives is nonsensical with these types of creativity. You don’t need to give people incentives to speak; you need to threaten them to get them to shut up. The desire to create is self-renewing and self-reinforcing.
Implications: (1) Copyright law, even in its own economic terms, plays a minor role. Copyright isn’t the cost/benefit at issue. It’s the jobs the creators would be doing instead of writing/drawing/whatever: most artists pay for their time spent creating out of the opportunity cost.
(2) Incentive stories can mislead us about the value we want to protect. Under the First Amendment, we protect religious conviction not only, and not even primarily, because we worry about the chilling effects of persecution on religion. Devout believers have been willing to go to jail and even die for their causes; they’re hard to chill. We protect religious faith because it’s so important, and a core wrong of suppression is its disrespect of the believer.
(3) The standard move when people start talking about how creators gotta create is to punt to intermediaries/distributors, who are supposed to act like perfectly rational actors even if those crazy artists are unreachable through incentives. That’s not wrong, but it has implications. One of which is that intermediaries are essentially in the position of the proposer in the ultimatum game: Because creators really really want to create, intermediaries can offer a lower price, and creators will accept a lower price, than they would if incentives mattered more, just as proposers will offer a lower price and responders will accept a lower price in the ultimatum game when some reason is offered why the proposer has the role of proposer.
This creates distributional concerns. Themes of exploiting labor turn up both in discussions of the mature copyright industries—music, television (writers), book publishing—and also in discussions of user-generated content. The incentive model, while it has no idea why people create, is happily indifferent to questions of justice as well once people have started producing: people who don’t need incentives wouldn’t participate if they didn’t get sufficient psychic benefits from producing. But failure to interrogate tastes should disturb us when it has significant distributional consequences. How did those tastes get formed? What are the alternatives? Analogy: women’s assumed natural proclivity for housekeeping and childrearing—we haven’t been thought to need incentives/payment, because we do it for free, even as economic structures and social conditions made alternatives hard to find.
(4) Implications for fair use. Joanna Russ, the sf writer, suggested that the “what if” of slash fan fiction was “what if I were free?” What would I do, what would I write? That’s a very different stance than the rhetoric of freedom in open source; I’ve always hated “free as in speech, not free as in beer” because of the unexamined assumptions there about what counts as free speech. What free is for creators from subordinated groups is actually a pretty fraught question, as John Stuart Mill eloquently noted in The Subjection of Women. Women as writers have rarely had the luxury of exclusive control to give away.
Fan fiction foregrounds desire. Desire is of course central to economic life: we have to want stuff to work for it. But desire is hard to contain, and stories about desire tend to make homo economicus uncomfortable. Stories about desire—stories told outside the money economy—are fundamentally based on the inexhaustibility of the imagination. This should lead us to resist any contraction of fair use for noncommercial works, even if licensing markets are said to be developing, and even if the nontraditional sexualities often expressed in fanworks become more accepted in mainstream media. The logic of scarcity that copyright imposes is inappropriate for these creative works.
Jane Winn, University of Washington Law School, Better Regulation for Consumers: Integrating ICT Standards and Consumer Protection (with Nicolas Jondet, University of Edinburgh)
Her focus: ecommerce. Commercial transactions tend to bump into IP.
What is better regulation? Deregulation is widely discredited in the US due to the collapse of capitalism as we know it; in an abstract way, better regulation requires defining the policy objective then achieving it in the most efficient, equitable way. So what are our policy objectives? Balancing interests of consumers, producers, other producers; maximizing gains from trade/return to investors; other complex stuff.
Once we’ve done that, how to match it up with regulation? Language of IP is dominated by ex post government intervention: rights defined and enforced in litigation. Better regulation would have full ex ante consideration.
IP hasn’t yet been integrated with consumer protection law, which hasn’t yet assimilated the effects of information economy. What is consumer protection law? Narrowly: economic, designed to correct market failures. Uncontroversial in the US. Broader: health and safety, social regulation. More controversial in the information economy—we don’t generally have bodily injury from the information economy. Most controversial: dignitary protection. Europe is more open to that.
Standards: the literature is mostly about product standards; what about information standards? Standards tend to increase competition, decrease costs, accelerate adoption of new technology. Costs: tends to reduce choice; problem of the long tail—lots of mainstream applications and fewer for small markets; can be profoundly anticompetitive; can increase switching costs. In information markets, network effects merely intensify these issues.
Common in the US to just dump a standard developed by an outside org into a statute. This is dumb. Problems of obsolescence, ownership of the standard. Technology-forcing standards are controversial. Environmental protection law has left a lot of scars in the US: perverse effects of trying to integrate regulations with standards. But our sense of how to draft statutes has moved on since the 1970s. Everyone understands that there’s a problem with standard development once the drafters know it will have regulatory force: raises the stakes, makes consensus more difficult; political compromise may be wasteful or infeasible.
Legislative solutions in the EU exist. Different countries have different approaches to harmonizing standards and regulation. France got out ahead and decided to have an agency regulate TPMs. Why? Because France is a coordinated market economy; US regulates indirectly by enforcing private entitlements, relying on market discipline, whereas France has a social/political consensus in favor of direct regulation to protect the public interest, which Europeans think can actually be defined despite public choice theory. Europeans see market failures much more readily than Americans do, and they think social concerns can trump economic ones.
Coordinated market economies have de jure standards bodies, which coordinate with regulators. Contrast to informal, voluntary, private-sector standards in the US. US information consortia are relatively non-transparent; they can develop standards relatively quickly, but outside America they’re perceived as illegitimate. In 2006, the French passed a copyright law; they knew they needed to respect TPMs but they were also concerned about interoperability and consumers’ rights. Created an agency, though it’s been paralyzed by controversy and bureaucratic sloth; its mandate is to guarantee things like fair use (how many private copies must be permitted by TPMs). Unfortunately, focus is ex post, not ex ante; no standards for developing technologies. Its goals are unclear, but the strength is that they’ve identified the problem: interoperability and enforceability of limitations on IP rights. The solution needs to be public law.
Opportunity even in a liberal market economy to redefine the scope of consumer protection to include dignitary interests.
Commentators: Steven Hetcher, Vanderbilt University Law School
On Mazzone: The argument is (1) fair use doesn’t work, so (2) let’s have an agency. Important to disaggregate standards from definitions. Problem is that fair use functions like reasonableness; it’s going to be pretty hard to define in more detail unless you want to reject US fair use. Also, if fair use is out of the case, substantial similarity is still an issue—how do we define that? It’s not that courts haven’t gotten around to specifying fair use or substantial similarity standards, it’s that they can’t do it.
Play: Blanche Survives Katrina in a FEMA Trailer Named Desire. Is it a fair use? Other cases won’t tell you because no fair use case is ever on all fours with others, especially if the key question is whether the accused work targets, at least in part, the original.
Possibly the EFF/Prince case requiring copyright owners to do some fair use analysis before sending takedowns will help develop the law further, though.
Agency model: once again, no one’s going to admit that their case is similar to previous ones, so they’ll keep refighting battles even in the agency.
On me: Hetcher is on board with the basic idea that creative production functions like a taste. Everyone’s on Facebook, contributing often copyrighted content. Criticism/question: okay, great, we’re making stuff. What about when we remix using other stuff? Lessig says “make amateur remix legal.” Tushnet says it’s fair use. In the world of fan fiction, the author did all the work—took the characters, but wrote each word. Now, you can copy a whole movie and just do a little change: CleanFlicks making movies family-friendly. Digital remix allows everyone to be a filmmaker. Where do you draw the line? (Note the privileging of the second creator’s use of words over the first’s use of images, and then the privileging of the image over the editing, even though it’s the editing that produces the effect of the image. As someone who’s written fan fiction and made fan videos, I can tell you which one I find harder, and it’s not the former. Broader question: Whose work counts?)
James Gibson, University of Richmond School of Law
On Winn: Captures and describes tech/legal conflicts, but also a depressing read, because the summary of the European efforts look enterprising and even radical in a good way, but they don’t seem to succeed. They managed to harmonize cellphone standards, but that entrenched a 20-year-old standard. Tried to standardize e-signatures, but no one seems to have paid much attention. Similarly with the French law on TPMs and interoperability—cumbersome in administrative terms, rather than mandating interoperability; the only people allowed to petition for interoperability information are other tech companies, who might not be the best stand-ins for consumer interests. The best of a bad lot is still not that good.
Possibilities: be less ambitious. Conflicts in standard-setting have various sources. Microsoft has leveraged network effects, but the cause of standard-setting problem there is different than when the standard comes from an ex ante international consortium, which is different than when the standard comes from the public sector. He’s not really convinced that TPM regulations have huge impacts on how much circumvention goes on, in practice. The government might well be right to back up TPMs at least sometimes, to prevent destructive competition between hackers and content owners. TPM-free music is now becoming available and the market mechanism may be working, even though consumer preferences about TPMs filter less well through the system than preferences about price.
On Mazzone: His paper asks whether fair use should be alienable. Are fair uses just tradeable/waivable default rules. If we think some fair uses should be waivable, then we should ask what counts as waiver; a lot of attention has been paid to shrinkwrap/clickwrap. But now there are these hortatory statements by content owners, saying no use without permission: is seeing them sufficient to remove our rights? There’s no contract, but is notice sufficient? How does it change our behavior when we’ve been given this notice, regardless of its legal effect?
Mazzone: Most law is administrative law; this state is an administrative state. Let’s talk about administrative solutions, even if it’s tough.
Winn: Political scientists in the US are in direct communication with European regulators; what they lack are lawyers/law professors in the US to talk to about alternative theories of regulation.
Mark McKenna: He still doesn’t know what fair use is for. It’s a catchall for a bunch of stuff not covered by copyright. First Amendment, market failure, both, others—all true at times. The agency would need to know what it wanted before it could implement that.
Analogy: UDRP—a cheaper resolution, though not an administrative agency.
For me: McKenna heard two different things: (1) people creating because they have to should be able to do so as a matter of fair use; (2) perhaps people who do this should be paid. These are very different questions. (I think (1) is easy, and (2) very hard. I don’t have a firm conclusion on (2), and in the end we lose so much by trying to assimilate everything to the market that I think it’s not a good idea, but I think it’s what we should be talking about, since someone is getting paid in this exchange—Google, the ISP, etc.)
Jessica Silbey: Creating creates us; it’s constitutive of our lives as people. So rather than asking whether people should get paid, we could also ask why have a market at all? If we are interested in incentives, ask “what would you make if you were free?” We assume markets matter more than they do, and the “intermediaries need incentives” answer is no longer persuasive given current technology.
Samuelson: She thinks fair use is more predictable than others do. If we move away from leading cases and towards policy-relevant clustering, we can make better predictions.
Litman: We have an agency trying to make prospective rules on fair use: that’s the interpretation of the 1201 rulemaking that Judge Kaplan of the SDNY has, as a way to preserve a fair use lifeworld for us. Figure out how your proposal diverges from that model and the degree to which it succeeds or fails in getting anyone any actual fair use rights.