Moderator: Kristen McCallion, Fish & Richardson
Gerard Lewis, Comcast
Courts have enshrined various business models—is that a good model going forward? Public performance: relatively new right. History of figuring out whether communication to the public was a public performance. Move of radio and then TV from foreground experience to background/pervasive. Change in technological limitations—innovation to improve the experience. Court initially said that cable TV wasn’t a public performance, just an enhancement of the transmission. Congress acted: framed exclusive right broadly, definition of public performance broadly. Exemption for homestyle receivers in public places was framed very narrowly. Made a number of assumptions about tech: tech available for private homes would remain distinct from that available to commercial establishments, an assumption no longer as valid now. Technical and detailed factors—number of receivers, length of wires, money taken in by establishment; contrast to broad definition of exclusive rights. Result: Fairness in Music Licensing Act, expanding the exemption, nearly 600 words. Next copyright act: don’t want to count bytes or other technical factors; that doesn’t seem to work.
Predictability and certainty are desirable, but worry about purpose built and specific provisions, such as AHRA model, where tech quickly becomes obsolete and doesn’t provide guidance.
Felix Wu, Cardozo Law School
Judge Chin in Aereo: Courts should resist the temptation to look under the hood. Aereo feels a lot like a cable system—functionally, is it the equivalent? A different case: ReDigi: tried to create a market for digital first sale. SDNY said there was no digital first sale, because there was reproduction and not distribution. If you didn’t look under the hood, the result might have been different. Not obviously a principle that favors owners/nonowners.
So what should we do going forward? Not clear that Congress has led/signalled at all. One key question to answer in the next formulation of exclusive rights: how much will we look under the hood and how much define in functional terms. (Why not go all the way to making available?)
We need to know the underlying goals of © law to have a functional definition. So his concept might be access. Reproduction might have been an attempt to mediate access to a work. Counting the number of books printed gave you some idea of how many would have access, though there’s not a pure 1-to-1 correspondence. But access is now potentially on both sides of the equation, not just incentives v. access. Consumptive/productive access. Is there a better way to split consumptive from productive uses? Though of course those categories are tied up.
Neil Netanel, UCLA Law School
Conducting empirical study of infringement/substantial similarity cases. Trying to figure out what the tests are, whether there’s any correlation with outcomes, types of work, identity of parties, etc. Paul Goldstein said the most important rule is that Disney never loses; wants to test that empirically (now that it has Lucasfilm, it at least inherited a loss or two!). Do courts filter out unprotectable elements, and how do they define those? For which issues do they rely on expert testimony? How often do defendants admit copying or access? When there’s an exact copy, a prima facie case has been made: defenses will either be ownership, noncopyrightability, or another affirmative defense. Most interesting/troublesome cases involve comprehensive nonliteral similarity (like the recent Disney/Snowman case, which may also provide a data point). Second Circuit and Ninth Circuit offer the major approaches. 9th Circuit’s extrinsic test seems to cover both “copying in fact” and “wrongful copying,” and then its intrinsic test seems similar to the 2d Circuit’s test. There are lots of variations on these themes even within the circuit. Students invariably ask whether the tests make a difference. We don’t yet know, and that’s why we need a study.
Lots of criticisms of these tests; Pam Samuelson & Mark Lemley prominent among them. Both argue for some version of inverting the stages, or bringing more objectivity from first stage to second. Initial results suggests it’s even more a mess than we thought. Initial survey was developed based in part on Second Circuit’s distinction between copying in fact and wrongful copying, and we asked if courts actually made that distinction, or applied the inverse ratio rule. We found out in the pilot that our RAs had no idea what we were talking about because even in the 2d Circuit courts often make no distinction whatsoever between copying in fact and wrongful copying; they just talk about “copying” or “unauthorized copying” w/o making it clear whether that meant license or authorized by law, or “protectable expression” without saying how much or what that meant. Redesign: ask RAs to identify statements about what would support a finding of infringement, and then determine which statements go to copying in fact/wrongful copying.
What should be the locus of reform? Preliminary experience suggests that some clear pronouncement of the test and the relevant factors is needed, whether from Congress or the SCt.
Christopher Sprigman, NYU Law School
To an antitrust lawyer, much about copyright seems odd, including the prima facie structure of a copyright case. On one end of spectrum, massive commercial infringement—copies of recent blockbuster. At heart of copyright law: consumptive, compete with originals in market, displace demand otherwise satisfied by legitimate work. At other end: parents post video of toddler dancing to a Prince song—not displacing, not a valid licensing market. Yet copyright law prescribes the same sort of prima facie case for either, despite the different effects on authors/markets. For an antitrust lawyer, this seems odd. Antitrust is based on the view that competition is efficient; copyright holds that unrestrained competition among copyists results in market failure. Sprigman has no quarrel with this more pessimistic account of competition in some ways, but different acts pose different risks, and the law should deal with it better.
Antitrust has the same problem and does deal with it better. Some conduct is highly likely to be harmful, like cartels. Exclusive deals may sometimes harm competition, sometimes be irrelevant, and sometimes help competition. Antitrust marks out a category for per se violation of the law: Ps need not prove market power or harm to competition; such harm is conclusively presumed. For all other conduct, antitrust uses a rule of reason. P is required to show market power and likely harm to competition. Makes sense: where we believe that conduct is very likely to do damage and very unlikely to produce benefit, we give P a cause of action that is essentially strict liability. For other conduct, we put P to the proof that the P ought to be in a position to produce. What’s in the category changes over time—minimum resale price maintenance has been moved into rule of reason.
Contrast with ©, where no P is required to show harm as an element of the prima facie case, even when the conduct seems unlikely to cause harm. Castle Rock: never showed that they wanted to enter the quiz book market, but court just hypothesized that they could and thus hypothesized harm.
Could we import this mechanism into the next great copyright act? Yes, we could have a category of per se copyright infringement—current liability structure should be for this category. Everything else should be rule of reason and P should be required to show harm as part of the prima facie case. He has suggestions, but they are a first cut—copyright litigation should be an engine of learning, which allows Ps to provide evidence about what causes harm (and what doesn’t), so the categories’ content can change over time.
(1) Only registered works should be in the per se category. Registration is a signal about potential harm. Registration should produce a bigger difference in treatment.
(2) In addition, use of work must also be consumptive/likely to displace demand. Thus it must be exact or near exact copy/performance.
Everything else should be rule of reason: derivative works, unregistered works, licensing markets not involving exact copies. Sprigman’s article on this proposal covers this in more detail.
Q: what about antitrust power of large aggregations of copyrights? ASCAP/BMI consent decrees ended up in pretty workable licensing scheme. Is there any point in looking at major film companies, etc. for something like that?
Sprigman: antitrust consent decrees are example of antitrust’s ability to learn. Price-fixing is possible, but blanket licenses also lower the cost of licensing, which requires ingathering copyrights. Blanket licenses are procompetitive, plus safeguards to prevent abuse. Safety valve: resort to court determination of rate; rules against exclusive licensing. SESAC is now under antitrust attack from private plaintiffs, and got past the motion to dismiss case because SESAC has struck exclusive deals for what Ps say are must-have categories of music and raised prices substantially. Conduct comes in various forms—antitrust can look at this as copyright hasn’t except a bit in fair use analysis. Need more because fair use is subject to a chilling effect.
Q: isn’t this too expensive to prove market harm, as in antitrust?
Sprigman: doesn’t think so, but we can learn over time. Could have various presumptions to help Ps. Antitrust has a sunny view of markets—tends to see them as resilient to anticompetitive conduct, while copyright doesn’t, and he thinks the empirics suggest copyright is wrong—markets adjust to piracy. May not adjust optimally, but we don’t actually know what the optimum is.
Q: could fair use get us to that world? How would a rule of reason reduce chilling effects more than fair use?
Sprigman: fair use has gotten more capacious relatively more recently. We’re seeing some market harm principles enter through the back door, but that’s not efficient. Channel litigation according to whether market harm can be shown.
Q: format shifting is becoming more relevant—backups, Kindle to hard drive, etc. Is there a way to bring this into formalization? (Private copying exceptions that plenty of other countries have, maybe?)
Wu: that’s an example of figuring out what we mean by access. Why would format shifting even be a problem? Because we’re focused on reproduction. If instead we focused on access, that would suggest that format shifting isn’t something we would care too much about.
Glynn Lunney: public performance right used to be against competitors; there used to be natural monopolies, with only one large theater in town, one or two TV stations. How should we structure public performance now that’s not true?
Lewis: transmit clause tries to deal with that, but doesn’t necessarily do it well. Physical natural monopolies are largely past (RT: um, at the content level maybe, not at the Comcast level!). Cloud services—can rent space on Amazon and be as much of a cloud provider as Netflix or Dropbox. Wu’s point about looking under the hood: we might not like what we see, or we might need to figure out what counts as a hood. Need to know the problem we’re trying to solve.
Q: of the various tests for nonliteral infringement, do any adequately address software?
Netanel: Typically in software cases the test is much more abstraction and filtration based. Merger, etc.
Sprigman: the way they vary it has to do with competition. Filtration is a decision about competition.