Panel I. Overview of Fair Use Since Campbell (Moderator, Professor Gomulkiewicz)
Barton Beebe, NYU
Updated dataset 1978-2014: 460 opinions. Available on his website. Significant limitations of case-counting: can’t substitute for qualitative analysis or other quantitative approaches looking at what judges actually do versus what they say they do. Also, can’t substitute for studying what authors and publishers do. But still provides some insights.
Findings: Nearly 70% are post-Campbell. 35% are from the Circuit. Fair use articles in law reviews outstrip fair use cases a lot, 3:1. Reversal and appeal rates are completely normal for civil litigation. Fair use win rates overall are almost 50%. Correlate most strongly with overall outcome: Factor one comes first, factor four slightly less. Two is close to irrelevant and three is a bit of sideshow. If Kienitz is any guide (almost an appallingly bad opinion), it looks like Chicago economists v. New York humanists or modernists.
Transformativeness: tried to show impact of Campbell on word count in fair use opinions as against other factors. Increased attention to the concept of transformativeness. 8 in 10 fair use opinions address transformativeness. What are the other 2 doing? Notwithstanding Kienitz, has transformativeness taken over? Do judges find fair use first, then transformativeness?
Quantitatively: nearly sufficient to trigger fair use finding, but not necessary. 79 find transformative use, 76 found fair use (outliers, Castle Rock; RDR v. Warners; Richards (someone wanted to make a new dictionary from another’s)); 111 fair use findings, 40 did not find transformativeness and 7 explicitly found nontransformativeness. One thing about reading the cases: a lot of indiscipline—a lot of district courts citing to Sony and Harper & Rowas if Campbell never happened—the subtleties of the presumptions have been lost on many district judges.
June Besek, Columbia: Changes to © law seem to fulfill the prophecy that the internet would change everything. Not anti-fair use—thinks a lot of the cases, including iParadigms(plagiarism), are rightly decided. But is disturbed by trajectory: that the first factor has become outcome-determinative—commerciality, creativity, size of taking all become irrelevant, and transformativeness means different market. Courts straining to find uses transformative because it makes the analysis flow so nicely. White v. West: briefs in a database isn’t transformative. We’ve gone from databases that are used as search tools to providing substantive content: White, TVEyes, thumbnail search engines next to framed inline link to the original. You’re providing content there. Cases disproportionately emphasize public benefit over private benefit to the creator. Why should the burden fall exclusively on authors and publishers to provide this benefit, rather than being funded by the public as a whole? We don’t allow people to reproduce textbooks at will. There is a public benefit in Google Book search, but keeping down the cost of patent filings? It’s not a question of whether the PTO will get things—conscientious lawyers will pay for the copies.
What does it mean to be transformative? Can it mean merely disseminating to a new market? Should Napster then have been transformative? The market would be very different now were that true, because record companies weren’t in that market—Napster claimed not to hurt CDs, but that wasn’t the relevant market. Digitizing millions of seldom-used books: true that brings them into circulation, but what if someone creates a specialized subset of books that are already word searchable—is it transformative to do that again, maybe to keep Google from over your shoulder. Maybe the algorithm is better/different/improves competition. Is that not transformative any more? That will mean lots of copies will be disseminated.
Is it transformative for Google to enlarge its database for books that are digitally published and not only available on shelves? It’s available for search somewhere else—will that weigh in favor of/against Google? Is it justifiable to prefer to have everything in one database? Slippery slope. What about public benefit in putting unpublished works up? Archives would love it because of their scattered userbase. Factor 2 considerations: are they simply overwhelmed? Maybe so for orphan works, but for non-orphans? That’s one of the next issues. Finally, courts are construing purpose very narrowly to find D’s purpose different from P’s purpose. Artificial construction of purpose: sometimes authors don’t really focus on purpose, or think about anything beyond immediate goals. Sometimes their purpose is “exploit whatever I can,” even if it doesn’t exist when they write.
Wish we could go to Congress, but unlikely they’ll act, and courts can’t refuse to decide cases in front of them. But some things considered as transformativeness now are well beyond what fair use doctrine should be addressing.
Rebecca Tushnet, Georgetown
My skin in the game: helped found the Organization for Transformative Works: a concept that has been taken up by many nonlegal fans to explain what it is they do when they make fanworks based on, but exploring and often critiquing, their favorite media properties. Both transformative content and transformative purpose elements; also transformative with respect to the creators and communities that form around them—we don’t often discuss the good that creating things does for creators as a factor in legal analysis, though I’d like it if we should—our submission in response to the PTO/NTIA Green Paper on remix spends a lot of time on that.
Most debates about the proper meaning of transformativeness are really about the larger shift towards more robust fair use. Critics of big transformativeness have an underlying dispute with fairness, not with transformativeness: they are uncomfortable with fair use findings in favor of exact copies, or sometimes in favor of inexact copies made with different but noncritical purposes. So, we hear less (but not nothing) these days about transformativeness’s potential interference with the derivative works right. If transformation means transformation in content, it can be hard to figure out the line between transformed content that is fair use and transformed content that is an infringing derivative work. Those criticisms don’t hit as hard with respect to purpose-transformativeness, where the defendant’s copying doesn’t necessarily implicate the derivative works right—the Google Books/Hathitrust litigation, for example.
This shift from the concern with interference with derivative works right to concern with interference with things like the existing library exceptions is an example of ideological drift—where legal arguments that used to be made by people on one side of a political debate shift to being made by those on the other side. There are clear political dimensions to fair use rhetoric, where many low-protectionists have shifted to performative claims about the robustness of fair use and high-protectionists now claim that fair use is uncertain and therefore not very important/potentially in conflict with the three-step test under Berne. Samuelson says that fair use is predictable, which has important ramifications for compliance w/the three step test: she is performing what she wants to be true (and I think is!).
Jeanne Fromer, NYU: Campbell was skeletal, but helped solve the circularity of the fourth factor. Courts are just now starting to work on the fourth factor. Campell suggests full bodied investigation of market effects; but that some effects don’t matter.
Full bodied: implication in Campbell that one ought to look at market benefits v. market harm. Strong value of transformative works to public + variety of market effects that may follow + recognition in third factor that there’s often a close connection between a work’s use and the need to call attention to the work in the defendant’s own work. Footnote: there can be market benefits, per Leval. Implicit: works can draw attention to, enhance, affirm original work’s role in marketplace. Fourth factor says “effect,” not “harm.” Finding new works on Google Book search = benefit. Court emphasized that in finding fair use.
Companies too are recognizing this more & more: Disney, once notorious for going after copiers—Air Pirates—now likes parodies of Frozen.
Circularity: absence of licensing fee: how do we weigh that? Looking at market benefit as well as market harm: consider effects that pass a certain degree of speculativeness.
Finally, market for criticism doesn’t matter. But why not? Hinted: Empirically unlikely, though that’s not entirely true. We should look at true potential effects and unlikelihood matters. Should also exclude effects unrelated to expressive value of the work: ideas, public benefit of work. Courts can focus more attention on that going forward.
Glynn Lunney, Tulane: Didn’t think Campbell would be a bigger case than Sony, but has seen change. Consistency with three-step approach: Look at WTO panel interpretation. Certain special cases. Does this mean rare/unusual/not run of the mill. Normal exploitation: same meaning: can’t be a large part of the market. Can’t unreasonably prejudice rights of author: interpreted as meaning too much. Reduces to a one-step test. The L&E doesn’t significantly reduce the actual or potential revenue to the rights holder.
Argument for fair use violating the three-step test: Difference is that the three-step test only considers the author; the fair use doctrine also considers benefits to public, and allows more loss to the rightsholder if the use is sufficiently transformative.
So will we see a WTO arbitration panel rejecting fair use? Very unlikely, and also unlikely that we would change the law any more than 110(5) was changed. ISDS could lead a country to challenge us. But this is really about weaponizing the three step test as an argument for narrowing fair use and beating foreign legislatures with a club when they consider adopting fair use (e.g. Australia).
Can we take this weapon away? One: is fair use a limitation or exception at all? Its characterization as affirmative defense is relatively recent. Well, is idea/expression a L&E? Is Feist? What about independent creation? Do each have to pass the 3-step test to be ok? I’ve pled all three as affirmative defenses, for strategic reasons. They may be implicit in the rights–§107 comes after §106, though. But §106 says subject to the rest of the statute! The only people who participate in WTO panels are the government, so they don’t give the full set of theories.
Reversal: if the three-part standard is inconsistent with fair use, it must change. Modify TRIPS to broaden the standard! We added in originality and idea/expression to TRIPs for software. If it turns out that there were a finding of violation, we should change the test.
Kate Spelman, K & L Gates
Took 150 years to get here. Fair use starts with a case about George Washington’s letters, reflecting the common law. Journal from when Campbell came out: clients were quite worried; this was a song with video of them saying “I don’t remember coming up with this song; it happened after a concert.” That had a stigma in those years: they were on drugs. (Cf. Blurred Lines.) There are still cases that smack of a morality play: that’s lots of what fair use was before Campbell.
Nobody called to ask about how to use fair use proactively before then. Usually fair use came up only after response to a C&D. Using fair use is now a normal part of economic transactions.
Q: will fair use be more of an international doctrine in 20 years?
RT: I feel bad about making predictions.
Besek: wouldn’t be surprised to see more fair dealing than there is now. Australian recommendations; Hargreaves Report in the UK recommended against fair use but more fair dealing exceptions. Unlikely to be consensus internationally: attitude that just b/c the US has it doesn’t mean it’s right.
Beebe: past—rhetorical mode of pessimism, trying to gain attention by claiming doom of fair use coming: Lessig’s “right to hire a lawyer” could become self-fulfilling once academics say there’s no such thing as fair use. It can create the doctrine. When we start becoming optimistic, that can also strongly influence the doctrine.
Q: One irony is that Campbell is a music case, and yet lots of music folks say there is no fair use in music. Are there non-parody music borrowing cases in fair use? If not, why not? Why so resistant?
RT: the rhetoric that there is no such thing as fair use for music is everywhere in the industry; judges fear they don’t understand music and so feel unqualified to judge what is transformative. Compare: photographers’ rhetoric, which is just as vigorous about lack of fair use, but judges are more confident they can interpret images.
Fromer: role of specific provisions for cover songs in deterring implication of other exceptions applying.
Lunney: fair use doesn’t necessarily add much to idea/expression because they think they’ve made substantial similarity perform that function. (Though fair use orients one differently to the role of tradition than idea/expression does.)
Q: rise of ancillary copyrights in Europe and potential levies—is there a global consequence for transformative fair use?
Spelman: Yes. The question is how even/fair are those consequences. The answer: not very. Technological consequences in Africa, cf. Arewa. Clients want global solutions. It’s so pleasing to get Amazon delivery, and we don’t think about how that happened. Google said that it would never be based in the UK because it doesn’t have fair use, but by tacit acceptance the UK does have fair use by using services provided from the US.
RT: I’m not sure it’s a consequence—these places don’t have fair use in the first place; copyright owners are asking for lots of stuff anyway. W/o fair use they’d still be demanding levies.
Q: what advice for the future/Restatement of Copyright?
Lunney: be cautious in all areas. 76 Act was supposed to be tech neutral language, without ever guessing that consumers would be copying in their own homes. Only address the problems you know are problems. Unintended consequences.
Besek: what is the mandate? Reflect and describe law, and/or make proposals for the future? There’s usually some of both in a Restatement.