Much Ado About Shakespeares
Two major reasons for implementing copyright exist. The first is that of incentive; authors need incentive to create, and publishers need incentive to distribute. The second comes from the idea of natural right; authors have invested their labor, thought, and creativity into a product that others benefit from – so he should receive some benefit in return for providing benefits to others. These reasons support the idea of preserving copyrights over literal copies and works of substantial similarity, but less obviously so with regard to derivative works. Such derivative works can range from parodies and satires to serious sequels and vast improvements upon the original, but derivative works as a whole provides reason as to why we should lessen the incentive or curtail the rights of the authors of the original works from which they derive from. Different formulations of copyright laws with respect to derivative rights can provide for different rewards, and certain formulations might provide the proper rights to both authors (original and derivative) and optimal incentives for creations.
But to understand first why original authors deserve some protection from derivative works by other others, we should examine the first possible formulation of law – no protection from derivate works whatsoever. Only literal copies, and those with substantial similarity, are protected under the original author’s copyright. Derivative copies could be produced – as long as credit is given to the original author, and the derivative author disclaims that he is the original author. Thus, a kind of altered form of trademark law protects the original, established author and yet rewards derivative authors for the work they create according to its actual quality. After all, there were many sequels written to Don Quixote, and yet when author Miguel de Cervantes wrote his ‘true’ sequel, it was wildly successful – and much more than any sequel not written by him.
Here there are two objections. First, that determining what substantial similarity is would be ambiguous, and therefore would result in the usual costs of having courts decide. This problem, however, is inescapable from the idea that it is wrong to take someone else’s work and merely tweak a few words here and there in order to bypass his copyright. To do so would fail to reward the original author – for someone else could always take the fruits of his labor by simply taking what the original author had labored upon for so long and simply make changes that have no improvements. And to provide for such a protection would require courts to make a judgment – though it is a judgment that appears to be far easier to intuitively ascertain than, for example, the existing “fair use” criteria.
The second objection concerns the situation in which the original work was lackluster and unprofitable, but the derivative work substantially relied upon the original. Imagine if a producer twenty-five years ago took a look at The Simpsons’ pilot episode, noticed that it was of poor quality, but realizing the potential for the characters, hired a team of talented writers to create very successful episodes built upon these same characters. From an incentive-based view, creators like Matt Groening have arguably lost their incentive to develop these characters in the first place. From a rights-based view, Matt Groening would not be compensated for his labor in his creation, poorly executed it might be the first time. And if this producer chose not to give him any public credit whatsoever, then Groening would even lack the benefit of taking advantage of fame and reputation. Because many first works, including pilot episodes of very successful series, are of relatively poor quality and low profitability compared to a polished product years down the road, it makes sense to incentivize original works of poor quality, because its elements (such as characters) could prove profitable later.
But we should restrain ourselves from being blinded by success. What about those creations that could be vastly improved – and those that can improve them? Shakespeare’s works were sometimes based on prior works, and yet none of those original works have reached the acclaim and success that his derivative works have. Shakespeare saw potential in those works, a potential which he developed using his own talent and labor, resulting in classics that will continue to be praised until the end of Western civilization. Certainly, by expanding an original author’s right over derivative works of improvement, we might prevent classics from being created.
What about licensing as a solution? The derivative author licenses the copyright of some or all of the original work from the copyright holder. This practice might work for a conglomerate such as Disney purchasing the rights to Winnie the Pooh, but less so for the starving musician or nameless author who cannot put substantial money upfront (or even be able to contact or serious transact with the copyright holder in the first place) – in other words, the type of creator who appeals best to our sense of justice that copyright protects natural rights. Such an author might first produce the derivative work and show it to the original author in order to negotiate licensing fees; but the original author might, upon seeing how his original work could be improved, simply deny him the license altogether and implement those improvements himself, or worse, take it wholesale. Or, if the derivative author has no previous reputation whatsoever, his derivative work might not be read in the first place, resulting in a significant transaction cost problem. If the derivative author attempts to negotiate a percentage-based royalty without first showing his derivative work, then further transaction problems occur. The copyright holder might simply ignore the offer.
Such a blind negotiation would also result in significant transaction cost problems. When there is only one derivative author negotiating with the copyright holder, a classic bilateral monopoly problem occurs. But even if there are multiple derivative authors offering to negotiate with the copyright holder (who can of course license the same work to many derivative authors), the copyright holder’s inability to properly price discriminate using his monopoly will result in an inefficient outcome. Let’s use The Simpsons again as an example. Author A wants to write an entire Simpsons movie. Author B wants to use only Homer, one of the main characters, in a spin-off movie. Author C wants to use only Homer, but in a cameo in a movie. Author D wants to use only Homer, for a cameo’s length of time, but his use of Homer allows the plot to come together in a way that makes his movie wildly successful financially and/or critically. Author E wants to use Flanders, an occasional side-character, in a spin off movie. And etc.
Each of those derivative works uses the original work in a way where the value of the original work differs wildly from the other. Ideally, the copyright holder would like to know to what extent his copyrighted work will bring value to the derivative work, so he can adjust his licensing rates accordingly. But, as the comparison between author C and author D especially shows, sometimes the copyright holder needs to examine the derivative work to accurately determine its value. Author D might be willing to accept a ten percent royalties rate, whereas author C might only be willing to accept less than a one percent rate, and author A will be getting a huge bargain.
The same problem exists with mandatory licensing. A statutorily-imposed standard rate will not be able to produce an efficient outcome for the same reasons as above. Even if we were to use the courts to determine the “appropriate” rate, there are significant legal costs to be incurred in going to the court every time there was a derivative work (though arguably, this is already the case currently under the “fair use” doctrine), and the parties could not privately settle on a contractually-enforced royalties fee after the creation or publication of the derivative work (yet another bilateral monopoly problem).
Another solution comes from patent law, a solution that addresses the situation where one patent builds upon the other. Both patents are protected, but the ‘derivative’ patent holder (the holder of the patent that adjusts the earlier patented invention) must get permission from the ‘original’ patent holder. The problem here is that, in addition to the same bilateral monopoly problem in the patents context, copyrights under existing law can last over a hundred years. Whether an author writing a derivative work two years after the publication of an original work will have enough incentive from those two years of protection is dubious. And to have the length of copyright over the derivative work begin after the copyright term of the original work ends would have its own obvious problems.
Yet another category of solutions would be to give the copyright holder rights over derivative works for to an extent more limited than rights over literal (and substantially similar) copies. One crude way of doing this would be to say that, after the first one million of profits, an author loses his rights over derivative works. Yet it’s obvious that if the original work is of poor quality, it may never reach that amount, or reach it impractically slowly. Not to mention the problem that the original work might be worth more, or that inflation would mean that Congress would have to constantly update the law.
But what about limiting the number of years that an author has rights over derivative works? That is, J.K. Rowling has rights over her Harry Potter books for seventy plus lifetime years, but derivative authors can write books based off of Harry Potter and his world twenty years after publication. Such a law may require courts to determine which work is a “true original,” a work that creates the foundation for later derivative works, or else each consequent Harry Potter sequel would effectively “renew” Rowling’s right over derivative works. But this potential source of legal cost is vastly superior to that of, say, determining appropriate royalties percentages, because it has the virtue of being clear in an intuitive sense. Harry Potter and the Philosopher’s Stone is obviously the original, and Harry Potter and the Chamber of Secrets (the first sequel) obviously does not create an original foundation for later derivative works. It’s intuitive enough for poor, nameless derivative authors to reasonably be certain that Rowling will have no real legal claim against them, and that’s enough for them to begin laboring on their derivative work. This solution avoids transaction costs, minimizes legal costs more than most alternatives, and provides incentive for both original content creators and authors of derivative works. All that’s left would be to determine what length of time would be appropriate for ending the right over derivative works, an issue unavoidable by any scheme that seeks to provide incentive by monopolies limited by lengths of time.
One last solution (which happens to be not mutually exclusive with the above) would be to exempt from the derivative works right works that are of a medium or purpose different from the original. When a documentary about classic children’s literature quotes parts from the Harry Potter series, we know that Rowling’s financial success could not have been appropriated by such a film. Nor do we think of A. A. Milne’s Winnie the Pooh has having destroyed the financial success of the copyright holder of the stuffed animals his characters were based on.
One exception to using medium or purpose as an exemption would be with regards to advertising. When advertisers take a character and put words in his mouth to endorse a product, two things occur. First, the both the creation and author are misappropriated and a fraud is conducted on the public, as well the creation’s image and author’s reputation harmed or destroyed. This effect might be mitigated by requiring that derivative authors give credit to the original author and disclaim his permission or involvement in the derivative work’s creation (a good idea regardless because it enhances the original author’s natural rights without limiting the incentive for creating derivative works by derivative authors). The second is that the creation’s image is diluted. Imagine the backlash if Harry Potter was on every other billboard or television advertisement promoting the next product, no matter how quality the product is, or how fitting the product is with the character. Therefore, even though advertisements are of a medium or purpose different from the original work, we might want to preserve the original author’s rights over derivative works in this case. Even though a court might have to decide what is, or isn’t, an advertisement, such a line would be intuitive in the vast majority of cases, meaning that legal costs would be kept at a minimum.
And what about the status quo? Currently, the way that American courts determine the status of derivative works is by the “fair use” doctrine, which is theoretically an exception to copyright as a whole, and therefore there is no special treatment of derivative works. Furthermore, and more importantly, the four factors, as stated by Congress, are ambiguous – and the courts have used them in unpredictable ways. The most important factor is the impact on the market value of the copyrighted work, others being the “purpose and character of the use” (with special attention to whether the purpose was non-profit or educational) “nature of the copyrighted work,” and the proportionality of the original work compared to the infringing work. These factors are ambiguous when applied to direct, literal infringements already, but when applied to derivative works, prediction is impossible. With regard to the market value factor, firms already have difficulty determining or predicting effects in their markets to begin with, and courts are not expected to do much better. With regard to proportionality, it is never clear how much is too much. Courts have made ambiguous the factor regarding the purpose and character of the use, determining that all purposes have at least some non-profit motive. And so forth. It’s clear that the great ambiguity of the fair use doctrine and its unpredictable results, combined with the long length of copyright, means that a potential Shakespeare writing about 21st century life may give up on the idea of improving upon previous works altogether.
 “What’s So Fair About Fair Use” by Alex Kozinski and Christopher Newman, 46 J. COPYRIGHT SOC’Y U.S.A. 513, 520-22, 529 (1999)., found in Foundations of Intellectual Property by Robert P. Merges and Jane C. Ginsburg.
 See Anderson v. Stallone, where Stallone used Anderson’s script for a Rocky sequel without payment under existing law.
 There is a difficulty, however. Let’s say that Rowling’s sequel introduces a new character, or a new institution in the fictional world. Does the protection against derivative works begin with the original book, or with its sequel? If we decide that additions in made in sequels are protectable starting from publication of the sequel, what constitutes as a new addition? The courts might have some ambiguity in deciding this.
 Sony Corp v. Universal City Studios demonstrated that, under some circumstances, even wholesale copying was not copyright infringement. Given that the controversy went all the up to the Supreme Court, the case demonstrates that predicting the application of the Fair Use factors is ambiguous indeed.