I want to thank Richard Epstein for his thoughtful comments on my article, The Questionable Use of Custom in Intellectual Property, and the Virginia Law Review for asking me to reply to them. 1 In my underlying article I bring to light the tremendous impact that custom has on both de facto and de jure intellectual property (“IP”) law, and criticize the general preference of courts to incorporate such custom into the law. I set forth reasons why custom is of particularly limited value in the IP context. My position is not that custom has no relevance to an inquiry of what might be a fair or appropriate use of another’s IP. Instead, my position is more nuanced. Customs should be considered only for a normative proposition—such as what constitutes a fair use—when the specific custom was developed in a representative manner, is aspirational in nature (rather than simply a litigation-avoidance strategy), is applied to represented parties, and where an independent evaluation is made of what impact such a custom would have if broadly adopted.
In his response to my article, Professor Epstein reinforces his defense of the use of custom in the law and suggests that, at least in some instances, his position holds true in the context of IP. 2 Because Epstein focuses his response on copyright law and the copyright fair use defense, I will generally do the same in this reply. It is worth noting, however, that my article sweeps more broadly, considering customs involving trademarks, patents, and publicity rights.OPTIMALITY AND TRANSACTION COSTS
At its heart, our primary disagreement is not about the ideal scope of copyright’s fair use defense, but instead centers on what role private ordering should play in addressing that legal morass. Epstein unquestionably has more faith than I do in industries to develop optimal rules independently of judicial or governmental oversight. I am particularly skeptical that customs in the IP context will develop in an optimal fashion because of market inequalities, the complexity of the IP industries, and the dearth of ongoing relationships and repeat players. In addition, customs have developed in the shadow of the highly uncertain fair use rules and the public at large has rarely been involved in the development of these customs. As I elaborate in the underlying article, the fact that industries develop litigation-avoidance customs does not mean that they want those customs to govern when uses of copyrighted works are challenged in court. 3 Accordingly, the customs themselves are not intended to become legal rules, but instead to promote extrajudicial harmony and cost savings over litigating all uses.
My point is not about whether it is prudent for individual businesses to license works because of the unpredictability of the fair use defense. In the trenches, it may well make sense to license rather than to risk litigation, but that is a very different question than whether, when one decides to (or must) litigate a particular use, those customs should have bearing on the larger normative and doctrinal question of whether a fair use has occurred. I contend that courts should not use such litigation-avoidance customs to establish the scope of fair use. 4 The answer in my mind to the fair use mess is not to adopt highly restrictive customs and apply them. Rather, it is either to reform fair use or for the courts to engage more broadly in the four-factor fair use analysis.
Epstein suggests that he prefers using custom to determine fair use because he worries about the transaction costs of trying to distinguish between different types of uses of IP. 5 Such line-drawing problems abound in the law and to abdicate their determination to the industry is to entirely repudiate the fair use defense.CUSTOM AMONG STRANGERS:
THE EXAMPLE OF THE FILM INDUSTRY
Even though Epstein has supported the use of custom to resolve contract and tort disputes, he has generally not supported the use of custom to bind strangers. 6 Our respective definitions of stranger, however, are different. In response to my article, Epstein contends that the customs involved in licensing copyrighted works for use in set dressing for television and movies, in college course packets, and for scientific research are all customs that have developed voluntarily and among parties who have relationships with one another. As I document in my article and as I have witnessed firsthand from my time working in the film and television industry, this is simply not the case. There are rarely repeat players who are in ongoing relationships with each other. Users and owners of IP rarely have direct involvement with one another. In my article, I give the example of a documentary filmmaker who wants to include posters, postcards, and other images of Elvis Presley that appear in the bedroom of one of the documentary’s subjects. 7 I point out that the documentary filmmaker has no ongoing relationship with the Elvis estate, and that the Elvis estate is never likely to want to license any material from the filmmaker in the future. 8 The filmmaker and the Elvis estate may loosely speaking both be in the same industry—making and selling art—but they can hardly be said to have a relationship with one another.
Under Epstein’s views it seems that anyone involved in a particular industry is not a stranger to another person in the same industry even if they have no actual relationship with one another. Although I am uncomfortable generally with this approach, it is particularly problematic in the IP industries. There is no single industry involved and virtually every person in the world is a participant in the copyright markets. Almost everyone has created a copyrighted work, and certainly everyone has read, viewed, copied, displayed, or commented on such works. To conclude that we are not strangers to one another because we are all participants in the same copyright markets makes the stranger limitation virtually meaningless.
Perhaps Epstein would define the industries differently, grouping together only large film studios so that the custom would bind larger production companies, but not independent or documentary filmmakers. It is difficult for me to see, however, how BET, a cable television network, or HBO, the producers of the relevant television show, are less strangers to Ringgold, an independent visual artist, than a documentary filmmaker would be. 9 Ringgold and BET, or HBO for that matter, did not have an ongoing relationship and Ringgold’s primary business was not marketing her works to the film industry. Instead, the TV show’s set dresser independently and lawfully purchased a poster of Ringgold’s artwork from a third party for use in the background of a set used in a TV sitcom. This is not an instance of what Epstein terms “institutional arrangements to deal with mass transactions.” 10
Epstein notes that he might be more skeptical of the customs if there was a more “systematic industry dissatisfaction with the arrangement.” 11 In fact, there is much dissatisfaction, as I elaborate in my article. Both film studios and television networks, primarily in the context of litigation, have criticized the importation of litigation-avoidance customs to define the scope of fair use when a given use is actually contested. Moreover, the industry custom is hardly uniform among independent filmmakers and documentary filmmakers.REPRESENTATIVENESS AND VALUING DISSENT:
THE CLASSROOM GUIDELINES
One of the primary factors that I contend increases the value of a custom is whether it developed with the “input and participation of both IP owners and users and large and small players in the IP industries.” 12 When customs develop in a one-sided manner, I contend that they should generally be discounted. One of the primary examples I give of such a one-sided custom is the Classroom Guidelines, which were developed primarily by publishing companies to govern educational uses of copyrighted works. Epstein suggests that some universities and libraries were signatories to the Guidelines. 13 In fact, none were. 14 Nor is there any evidence that educators or students were involved in the development of the Guidelines. The American Association of University Professors and the American Association of Law Schools even went on record as opposing the Guidelines. 15
Despite Epstein’s suggestion that universities and libraries who did not agree to the Guidelines should be free to “litigate against the norms,” 16 the reality is that they cannot. Because courts have treated the Guidelines as binding customary agreements, the Guidelines have been incorporated into the law. No challenges to the Guidelines have been successful despite arguments that universities and others did not agree to their terms. My main criticism of the Classroom Guidelines is not that it was not a good idea to come up with some clearly delineated, private safe harbors for educational copying, but that such guidelines should not lead courts to bind parties who did not develop or agree to them. Even parties who agree to such guidelines should not be bound by their terms, if the understanding was that the guidelines would set a floor, not a ceiling, on fair use.CONCLUSION
Epstein and I both agree that the development of copyright customs is not surprising given the uncertainty about the scope of the fair use defense, but we part ways on whether those customs should be incorporated into the law. I continue to contend that they should not for the reasons I discuss more fully in The Questionable Use of Custom in Intellectual Property. Nevertheless, if the customs themselves restrict what gets made, as I claim they do, the question remains what can be done to relieve this pressure. Clarifying that customs will not be binding in the courts is one important step, but there must also be a movement by IP participants to explicitly and publicly dissent from the most limiting customs and to work toward fixing the fair use system.
Preferred citation: Jennifer E. Rothman, Why Custom Cannot Save Copyright's Fair Use Defense, 93 Va. L. Rev. In Brief 243 (2008), http://www.virginialawreview.org/inbrief/2008/02/18/rothman.pdf.
[Back to this issue]