Thursday, August 11, 2011

IPSC session 2

Second Breakout Session

This was one of the sessions with an embarrassment of things I’m really interested in, so I hopped around. (Two simultaneous copyright theory panels, sigh!) Turns out there’s also a liveblog here from, I think, a DePaul student.

IP, Ethics & Morality

Zahr Said, University of Washington School of Law, Copyright Law and The Ethics of Non-Fiction

The law of deception v. the law of copyright, with different ways of handling accounts that purport to be nonfictional. What should be done about literary accounts that purport to be true but turn out to be false, or accounts that purport to be fiction but turn out to be true. If a memoir/nonfiction is falsified, it may trigger deception law but merit copyright protection. Copyright doesn’t attach to historical facts. Should the law deal with nonfiction writing as art, truth, or both? Copyright incentivizes fictionalization of historical accounts, but deception law penalizes fictionalization of nonfictional accounts.

Copyright seeks to protect the public domain and subsequent authors dealing with historical events. Another way to see this: copyright rewards embellishment. Case: Parker v. Dufresne out of Louisiana. Parker wrote 700-page manuscript about the murder of two women by a jealous boyfriend. In 2000, plaintiff approached defendant about collaborating on a docudrama, which didn’t happen, but later defendant published a 9-page short story “Based on a True Story.” Defendant said he was inspired by plaintiff’s manuscript. Court found no substantial similarity and said there was no copyright in facts. To Said, expression seemed to have been copied because of the details—custodian who found the body exclaimed “Sweet Jesus/Sweet Mother of Jesus!” Descriptions of body found in dumpster were quite similar, including sugar ants carrying crumbs up the leg of the corpse. Doesn’t this create an incentive to fictionalize?

On the falsity side, Augusten Burrough’s Running with Scissors triggered a defamation suit, which settled. A Million Little Pieces: memoir in which narrator overcame substance abuse issues, except that he made up a bunch of details and was sued for fraud by readers. Three Cups of Tea: man who promoted his work in Afghanistan turns out to have falsified a bunch of details about his success and has also been sued.

Tension: the more fictionalization you do, the stronger your copyright protection, but the greater the potential liability for deceit. Is this a real tension? Does it suggest we should play with the level of one or the other? What work does incentive theory do here? What happens when market conditions change so that there are big incentives to produce nonfiction? Frey tried to sell his memoir as a work of fiction, but Random House wanted it as a memoir and then didn’t fact check it. Should we think about copyright differently to rebalance the incentives?

My q: why aren’t you stuck with what you say? If you say it’s nonfiction, we take you at your word. (I also meant for copyright purposes.)

A: Who’s saying that, though? Related to the idea of authenticity in art. Off the mark to create liability for anything said in a work. She thinks this gets towards treating anything said in a book as if it were an ad. (I disagree, since the standard is not the commercial speech standard but the defamation standard. We had some back and forth on limiting doctrines like puffery and materiality.)

McGeveran: Thin copyright may not be all that meaningful—JK Rowling can’t stop people from writing wizard boarding school stories either. Both rules protect the same principle: some things in your work are free for other people to use.

A: doesn’t propose disturbing that. Scenes a faire does a lot of work before the court gets to “facts are not copyrightable.” People want nonfiction, so maybe you don’t need very much copyright protection. But then what’s the liability lever? (Well, yes: people want nonfiction that is nonfiction, rather than “nonfiction” that’s made up.)

Laura Heymann: in terms of framing, consider the harm alleged in these cases. One dividing line: Frey and Three Cups of Tea are memoirs, telling a story about the author himself; is there a difference between that and telling about other people? Other people can be harmed—defamation, false light—whereas lies about yourself might not generate actionable harm if we think about it. Is reliance on “true story” really all that important if you aren’t changing the way another person is represented to their detriment?

Copyright Theory

Edward Lee, Chicago-Kent College of Law, Copyright, Death, and Taxes

We’re on schedule for copyright revision in the US, and the UK just committed to revising its law. Berne/TRIPs are constraints on such revisions. Our options are somewhat circumscribed: we can’t impose formalities at least with respect to works of foreign origin, creating huge problems of notice externalities/orphan works. Three-step test for limitations/exceptions is another constraint.

Now consider taxes: US has complicated tax structure providing favorable treatment to capital gains, especially long-term capital gains. His proposal: use this approach to target income generated from copyrighted works. For example: if author elects a short term, tax could be 5% of ordinary rate, whereas if author elects full term of 95 years, standard rate would apply. Blair Witch Project earned nearly $140 million, which would face a nominal 35% tax rate. Electing a shorter copyright could reduce the tax to $2.5 million. (I like the idea, but doesn’t it require the corporate tax system to function first? Don’t the movie companies consistently report losses?)

Another alternative: tax credit. For example, $150 tax credit for free use of a copyrighted work in schools, when school requests to use it.

Hughes: Do you have any data on how profitable large copyright holders are? The year Titanic came out, the movie made so much money even Hollywood’s accountants couldn’t hide it.

A: doesn’t have data on that. Would be difficult, but most copyrighted works have commercial value only in the first few years. For profitability, you have to look at expenses. Tax incentives allow each copyright owner the opportunity to evaluate its own circumstances and choose the tax break.

Q: it presupposes that the copyright creator knows in advance what the profitability will be. (Well, the studios are not terrible about overall predictions and investing more every day.) JK Rowling might have opted for a short term and lost a lot of money thereby. Also, would this be a proscribed formality?

A: key is setting window to allow election of tax benefit. Might not have to elect the benefit in the first year. Might be able to choose it later. Not a proscribed formality because you never have to elect it at all.

Q: clever idea with a million implementation issues, and plenty of tax gaming. Tax expenditures: grand budget bargains are supposed to lower overall marginal rate by limiting tax expenditures, which makes your proposal inconsistent with the supposed emerging consensus. Why is this tax expenditure better than others?

A: It’s a difficult time, but we can look into the revenue possibilities of derivative works.

Q: doesn’t this incentivize firms to produce works that produce all their profit in the first few years? Is that what we want? The software industry doesn’t have to change any of its practices and gets a huge windfall, etc.

A: But then the software would quickly be in the public domain, available for further use.

Irina D. Manta, Brooklyn Law School, Cognitive Calibration for the Reasonably Biased Man: Objective Tests, Jury Effects, and the Copyright Dilemma

We like generally applicable laws that don’t take individual variations in personality/background into account, because it’s hard to measure individual variations and useful to presume that everyone knows the law (this is based in the reasonable person test’s background in negligence). Plus, just because someone is clumsy doesn’t make him any less troublesome to his neighbors than if he’s negligent, so it’s fair to make him conform to the standards of his neighbors. Thus, we expect reasonableness from everyone: reasonable behavior is behavior whose benefits exceed its costs. In some situations, difficulty calculating this means we look at the average person instead.

A number of critiques of this standard exist. People from different social/cultural background won’t always agree. Cognitive bias also complicates the model.

Dan Kahan et al.: SCt said no reasonable jury could find other than that the police officers in this high speed chase acted reasonably, but Kahan found significant variation in evaluations of reasonableness based on salient demographics. Summary judgment: judges tend to substitute their own judgment for that of the jury. Judge bias can have a powerful effect there. We may assume the biases will even out with a jury through various mechanisms, but there are many studies suggesting that groups have their own biases and suffer from polarization. Every level of review risks new levels of bias.

Copyright and reasonableness: we know that the parties are biased. Also, when people are faced with ambiguous statements, heuristic processing has strong effects, preventing true reasonableness. The tests we use in copyright might also invite cognitive bias: if we use Arnstein v. Porter (1) was there copying and (2) was it unlawful, then we might get hindsight bias: if something was taken, then it was probably copyrightable. Reverse halo effect: if a person is bad in one area, then they’re bad in all. If they copied, then they probably copied unlawfully. Confirmation bias: once jurors find some elements are similar in two works they might be primed to find more similarities. Current hypothesis: these biases tend to favor the plaintiff.

What is to be done? Get rid of jury/minimize its role? Judges have their own issues. Debiasing: mostly it doesn’t work, but there might be cautions that could work—“don’t play Monday morning quarterback.” Maybe offer a control song or other work to the relevant audience that could be used to compare to the competing works in a look and feel analysis. Some circuits may be becoming more aware of this. Maybe test audience perceptions via survey. And more general work on optimal jury composition. Looking for other solutions.

Viva Moffat: Number of copyright cases that go to juries is vanishingly small—so does debiasing juries matter a lot? Mostly it seems like summary judgment.

A: Just as important with judges as with juries, and wants to suggest things there too.

My q: why not eliminate substantial similarity because people just can’t do the inquiry properly, and have a reproduction right and a separate derivative works right where you’d ask questions that were different (about markets). That would encourage factfinders to think about different things than look and feel etc., avoiding the problems you identify with substantial similarity that are likely to be intractable. (NB: this is what I propose in my forthcoming Harvard Law Review article on copyright and images, which I hope to have up on SSRN soon.)

A: willing to explore it; copyright hasn’t been willing to do that.

Q: plaintiff frames the narrative in the complaint—that’s a powerful heuristic.

A: there are effects, but we may be able to reframe. Still, a good attorney knows how to exploit these biases; we want to make it harder for a party to buy a trial by having procedures that debias.

Q: why think bias is pro-plaintiff? If bias is random we might not care so much. Should we correct for bias by shifting the substantive law?

A: the biases she’s looked at suggest that the plaintiff benefits from being able to show similarity—once there is similarity, more similarities are likely to be found; once there is copying, people are likely to think the copying is significant. Plaintiff gets to frame the story, making it more difficult for defendant. Substantive law: we’d have to deal with other consequences. Maybe we’d want a high standard for similarity.

Eva E. Subotnik, St. John’s University School of Law & June M. Besek, Columbia Law School, Constitutional Obstacles?: Reconsidering Copyright Protection for Pre-1972 Sound Recordings

Thinking about potential constitutional challenges to federal law extension to such sound recordings. Takings is the obvious issue. Much precedent relates to real property. Most pressing to labels (no pun intended) is that their terms of protection would be truncated under a federal regime; state law remedies might also be truncated—punitive damages, secondary liability; fees for federal regulation; claims by users of sound recordings/reliance parties who might be making uses for scholarship/preservation; users/reliance parties who might become obligated to pay performance royalties under §114.

Term of years was most contentious at public meeting in June. One proposal is 50-year term to match other jurisdictions. This would mean that extremely valuable recordings from 20th century would immediately enter the public domain, rather than lasting to 2067. Another proposal: harmonize with 1976 Act as amended. Thus, essentially (and skipping over lots of detail), 95 years from publication, or life plus 70 or 120 years from creation. Terms would be quite different from common law as preempted by Copyright Act, mostly shortening terms.

Some of these interests are probably protectable private property, and probably there’s a public use: so is there a taking? Given the lack of physical invasion, we’d be looking at regulatory taking framework. IOLTA case: interest income earned by fund held in trust was the property of owner of funds, and thus transfer of interest was akin to a permanent physical taking and a per se taking. There is precedent that the taking should be assessed from the perspective of the full parcel, not clear how to deal with this for works that would immediately fall into the public domain. But there are other factors, and there is a difference between facial and as-applied challenges.

Their conclusion: courts should instead use Penn Central’s balancing test. Hard to speculate about economic impact; some recordings have little economic impact. Endeavors such as National Jukebox, through which Sony granted a free license to stream its old content—but now it’s cited as a possible untapped monetizing alternative. (I just love this: something they could only give away for free is now possibly monetizable. Heads they win, tails you lose.)

These recordings have also seen their protection shortened already (if you think that they had infinite terms), so that raises questions about investment-backed expectations. Courts might be troubled by extraordinarily early termination of state law rights. (What is extraordinarily early about 50 years?) Possible solution: minimum term of protection under federal law. Mechanism like filing notice of intent to enforce as under §104A might strike an acceptable balance for protecting works that copyright owners care most about while allowing preservation and access to other works. 1976 Act substituted federal causes of action for infringing unpublished works until then protected by state law; presumably legislation could be drafted the same way here, even though the rights and defenses might differ. Preservation of ownership identity has been a hallmark of disputes that span the 1976 Act’s effective date.

Q: curious about interaction between ownership of musical work & sound recordings. Pre-1971, record label tended to own both the music and the sound recording (as opposed to the later rise of the singer-songwriter with separated ownership). Is there a body of work with separate interests out there as a matter of economic reality?

A: there is reluctance to do preservation among librarians etc. who worry about separate state law rights. In applying for grants to do preservation, some experienced difficulty convincing funders to participate without the comfort of federal defenses.

Q: is Congress really interested in helping out archivists?

A: well, there’s some opportunity. If we can give comfort to Congress/courts down the road, hopefully that’ll help.

Bob Brauneis: would you have to do a balancing test for every separate work? Age may matter; unpublished works may have different/lesser investment-backed expectations than published ones. Hard on the industry.

A: that’s one of the complications—how do we identify likely claims? Hard to anticipate how claims would be framed. Vast amount of precedent discussing facial v. as applied challenges.

Brauneis: maybe there’s no facial claim here.

A: she thinks that’s reasonable. Takings may be non-generalizable. And yet people do make claims covering many interest-holders. Hard to know given the response to prior changes in copyright law, but these challenges generally weren’t brought in the past: is it a change in litigation culture? In takings doctrine? Based on participation in public meetings, takings is not a fringe part of the discussion.

No comments: