When can teachers share copyrighted works (or excerpts therefrom) with students without payment or permission?
While other parts of the law come into play in narrow contexts, this is primarily a question about the scope of the doctrine of fair use. Educators have struggled with the dimensions of fair use for decades, and we are now at a pivotal point in that struggle. Friday’s decision from the 11th Circuit Court of Appeals in the Georgia State e-reserves case may be the last word on this issue for a long time. What does it say?
The battle for educational fair use has two fronts: the escape from the past, and the fight for the future. The GSU opinion secures victory on one front while leaving those of us focused on the second front free to fight another day. In that regard it is good news. However, while many are rightly celebrating this partial victory, it is crucial to recognize that the fight continues.
The first front is the struggle to shake off bad practices grounded in toxic and outdated visions of fair use. The primary sources of this decrepit fair use mode are the old 1976 Classroom Guidelines and the series of lawsuits and threats of lawsuits brought by publishers in the ensuing decades, aka the coursepack cases. Together, the Guidelines and the coursepack cases create a domain of educational fair use that is tightly constrained and highly deferential to the publishers’ profit motives. In the Guidelines-coursepack paradigm, a teacher can share materials with students only if the excerpt is so small, and the timing so tight, that it would be virtually impossible to pay a license fee to the relevant publisher. It is thanks to this paradigm that many teachers and librarians still think that they cannot use the same excerpt two semesters in a row (they can), or that no excerpt from a written work can exceed 1000 words (it can).
Friday’s decision from the 11th Circuit marks the end of the Guidelines-coursepack era.
The publishers argued vociferously that the court should adopt this mode of analysis, and, like the district court before it, the appellate court refused. Judge Tjoflat’s opinion says of the Guidelines (on p. 23, at the end of the lengthy fn. 12):
Whatever persuasive value the Classroom Guidelines may possess, we must keep in mind that they
(1) were drafted by partisan groups,
(2) “state the minimum and not the maximum standards of educational fair use under Section 107”, and
(3) adopt the type of “hard evidentiary presumption[s]” with regard which types of use may be fair that the Supreme Court has since repeatedly warned against.
Later in the opinion the court describes the Guidelines as suggesting, at best, “Congress’ tentative view of the permissible amount of educational copying in 1976,” before declining to apply those standards to the third factor inquiry on permissible amount. That is virtually the last we hear of the Guidelines. Because they describe an outdated safe harbor, the court treats them as essentially irrelevant to the contemporary question of the outer bounds of fair use.
The Coursepack cases come in for similarly cool treatment. The court claims that the cases “provide guidance,” but rejects the publishers’ argument that they are determinative. Indeed, the court goes on to find that GSU’s uses, unlike the uses of copy shops who were sued in the coursepack cases, are favored under the first factor because they are non-profit, educational uses. The court also expressly rejects the coursepack cases’ approach to the second and third fair use factors. Where the coursepack cases had held categorically that both factors disfavored the copy shops, Judge Tjoflat calls for a case-by-case evaluation, and describes several considerations that could lead to findings that favor GSU. Finally, on the fourth factor, although the court does not mention the coursepack cases, it diverges from their outdated insistence that market harm is the most important issue in the fair use inquiry. In other words, the coursepack cases were shut out on all four statutory factors. It is hard to see what remains of them as a relevant guide to educational users (as distinct from commercial copy shops).
All this is fantastic news, but where does it leave us?
We’ve shaken off the past, but the future of educational fair use is up in the air. The court’s opinion is deeply equivocal. As Nancy Sims pithily observed about the 11th Circuit’s opinion, “NOTHING HERE WILL HELP EVEN ONE TEACHER MAKE FAIR USE CALLS.”
The source of the opinion’s ambivalence is the equipoise between GSU’s favored status under the first fair use factor (the purpose and character of the use) and the publishers’ favored status under the fourth (the effect of the use on the market for the work). The court found that GSU’s uses are favored because they are educational and non-profit, but that the potential for cognizable harm to the publishers’ market is substantial because the uses are “non-transformative” (more on that term of art below). The relative weight of those two factors, as applied to the 50-odd particular uses before the court on remand, is impossible to judge at this stage, and even after Judge Evans makes her own judgments about how to weigh them in this case, the weight a court might give those factors in future cases of non-transformative use may be difficult or impossible to predict. One commentator has already reacted to this aspect of the GSU case with a plea for ‘bright lines.’
The deeper origin of this ambivalence and uncertainty, in case you haven’t picked up on it, is the court’s finding that the uses are non-transformative.
What does this mean? Transformative use is a powerful concept that courts and users have applied with increasing confidence over the last two decades; academics have shown that it is now the dominant consideration for courts deciding fair use cases. A transformative use is, very roughly speaking, a use for a new purpose distinct from the author’s original purpose in creating the work. Courts generally ask whether the new purpose adds value, and whether the new use is a “market substitution” for the original work. If a use is found to be transformative, the consequences for the user are dramatic. The question of market harm disappears completely, because, as the Supreme Court has said, transformative uses cannot effect markets in a way that is cognizable under the fourth factor. The only question remaining is whether the user took an amount that is justifiable in light of their transformative purpose.
Transformative use protects for-profit publishers who reproduce entire creative works in new books sold on the commercial market, technology companies who digitize millions of books to power their search engines (for substantial profit), and appropriation artists who manipulate and alter other artists’ copyrighted works and sell the results for millions of dollars to wealthy patrons. With a transformative use story in hand, even billion dollar companies can disdain license demands; without one, however, the market reasserts itself and even core fair use beneficiaries (scholars, teachers, non-profit educational institutions) face the same uncertainty as GSU.
If GSU’s uses had been found transformative, the game would have been over long ago. As long as the professors’ choices about amount were justifiable in light of their pedagogical goals, the uses would be fair. Because they are not transformative, GSU’s laudable purposes and their reasonable choices about amount in light of their teaching goals must be balanced against the alleged harm from lost licensing and the alleged impropriety of the amount given the effect of the taking on the market.
The future of fair use in a non-transformative mode is necessarily uncertain, as it will track specific facts about the operation of markets for the works used, which, as Nancy Sims has observed, are not necessarily facts that individual teachers have at their fingertips. Uses permitted in this mode may well end up being relatively modest to nonexistent where licenses are available, depending on how Judge Evans decides to implement Tjoflat’s instruction to give “additional weight” to factor four. Faced with this uncertainty, or with an available license, teachers may reasonably fear that if they do not pay, they are at significant risk. Institutions who cannot afford to participate in the byzantine licensing schemes of the CCC and its publishing friends will be chilled from fair use activity. The rich will have rich learning experiences, and the poor will be impoverished.
So, the fight for the future of educational fair use, if it is to be much more than an unsteady allowance for uses where licenses are unavailable or unreasonably expensive, has got to be fought on the ground of transformativeness.
And, luckily, the court recognizes the continuing validity of transformative use beyond the very specific facts of this case. (Here educators should be thankful for the highly effective amicus briefs submitted by several higher education and library associations as well as from a laundry list of copyright scholars)(full disclosure: I helped with a couple of these briefs).
In footnote 21, on page 63, Judge Tjoflat expressly disclaims any intent to prejudge future cases dealing with works whose primary purpose or function is not educational. Indeed, he notes that the issue of transformativeness has been decided against GSU in this case as part of a finding of fact made by the trial court that, because it is not clearly erroneous, the appellate court will not revisit. GSU never really tried to litigate the transformative use question, so that finding is certainly not air tight. In any event, Judge Tjoflat acknowledges that future uses not at issue here, such as assignment of historical essays for purposes of comparative analysis, or of scholarship not written for educational purposes, may well constitute transformative use.
In fact, the universe of transformative educational uses that Judge Tjoflat and the 11th Circuit have not pre-judged is virtually limitless.
In a paper I’m working on right now, I argue for a wide variety of transformative educational uses, including:
Using popular culture as raw material for cultural analysis (motion pictures as evidence of the portrayal of women and minorities, e.g.)
Studying creative works as exemplars of success or failure in deploying techniques of persuasion, description, etc.
Using news reporting as a prompt for historical or political debate or reflection
Using fine art images to illustrate characteristic elements of a movement or era
These illustrations are not drawn entirely from my fertile imagination.
Many examples of common teaching practice that have a strong claim to transformativeness have been identified and promulgated in a series of best practices statements published by educators, scholars, and librarians over the last decade or so.
One such statement, the ARL Code of Best Practices in Fair Use for Academic and Research Libraries (which I helped to develop while I was working at ARL) seems to have anticipated several aspects of Friday’s GSU opinion.
For example, the ARL Code Principle 1 states that “It is fair use to make appropriately tailored course-related content available to enrolled students via digital networks.” The court recognizes that nonprofit educational uses are a favored purpose and remanded to the court to weigh this legitimate interest against the possible market harm to the rights holders (among other factors), i.e., to ensure that the use is appropriately tailored. Of course, for transformative uses, appropriate tailoring is easy to calculate by reference to the transformative purpose.
This is why the best practices also counsel that “Materials should be made available only when, and only to the extent that, there is a clear articulable nexus between the instructor’s pedagogical purpose and the kind and amount of content involved.” Judge Tjoflat agreed that purpose is crucial and upheld the district court’s holding that the question of whether the amount taken is appropriate is judged at least partially by reference to the pedagogical purpose of the use. In non-transformative cases, however, the market harm factor creates substantial uncertainty about how to know when uses are well-tailored.
Finally, in recognition that not every educational use is transformative, the best practices caution that “Closer scrutiny should be applied to uses of content created and marketed primarily for use courses such as the one at issue….” This is the same principle that Judge Tjoflat invokes in finding that, because GSU used the books at issue for substantially the same purpose as they were marketed for, the fourth factor should be given “additional weight.” This extra weight is a consequence of the non-transformative nature of the use - hence “closer scrutiny.” As the best practices recognize, materials not designed for educational use would be free of this constraint.
So, the GSU opinion from Friday describes one way forward, which the court below must follow, which is a careful and largely unpredictable balancing of the value of educational use against the value of a legitimate market harm to the publishers. This seems to be the new normal for non-transformative uses. However, the court also recognizes that there is another way open to transformative uses, and the difference in how fair use would apply in that context could not be more dramatic.