Scholarly Communications @ Duke

More on GSU and the publisher response

By Kevin Smith, J.D. On May 22, 2012 · Leave a Comment · In Copyright in the Classroom, Copyright Issues and Legislation, Fair Use, Libraries

Now that I am back from vacation and have read the GSU ruling and some of the commentary more thoroughly, I wanted to add a few additional comments. In many cases these may be repetitive of things others have already said; I have not read all the commentary, of course, and some of this writing is for my own sake as much as yours. If it is redundant, I apologize.

I do hope most folks either have read or will read Brandon Butler’s “Issue Brief” for the Association of Research Libraries, which does a great job of analyzing the ruling.

The main thing that struck me as I read the fair use discussions for the individual excerpts that were in question in the case is that Judge Evans really did understand that fair use is a balancing test. Although some parts of her reasoning are more mechanical than I would like, she does not take a bright-line view of any of the factors. It seems worth a look, to me, at the way she relates the four factors in her discussion.

First, she finds that the first two factors — the purpose and character of the use and the nature of the original — always favor fair use in regard to the specific use and the specific works before her. This is true even though she declines to hold that the provision of short course readings is transformative, which is the key determinant in most fair use rulings over the past few decades. Instead, she sees this activity as at the heart of what fair use is intended to be, according to its own express terms. On the nature of the original, she holds that the works in question were all published and factual in nature.

This point about the second factor is likely to be the focus of some disagreement from the publishers. They wanted the court to hold that factor two did not favor fair use, apparently because of the effort involved in writing and publishing academic works. No one is more respectful of the labor of scholarship than I am, but hard work by itself does not get one a copyright or insulate one’s copyright from fair use. The Judge observed that educational works are exactly the works that fair use is focused on, with its favored purposes of research, scholarship and teaching. She also noted that permission fees make up no part of the incentive for academic authors to write their works (p. 81ff). This, to my mind, is the key reason for finding that factor two favors fair use in this case (although the Judge called it an “additional consideration”).

On the other side of the analysis, Judge Evans presumed, at least as an initial matter, that factor four, the impact of the use on the market for or value of the original, always favored the plaintiffs, which is to say it counted against fair use.

Which brings us to the third factor, the amount used. This was the “swing vote” in most of the individual analyses, and the Judge applied a clear but quite narrow rule. Ten percent of a work was acceptable as fair use if the book had fewer than 10 chapters, and a single chapter was considered fair use for books that contained 10 or more chapters. This rule, which was applied quite strictly to decide if a “decidedly small” portion of the work had been used (that was the standard Judge Evans applied for this particular type of use), had some odd results; as little as 3% of a work was considered too much for fair use in a situation where the work was long and had many chapters.

But notice something important. When the third factor disfavored fair use, the factor analysis stood at two in favor (one and two) and two against (three and four). In those cases the Judge went on to do additional analysis, asking several further questions. Was the heart of the work used? (she declined to assume that all e-reserve uses involved the heart of a work) Was there a “readily available” license for reproduction of the excerpt in digital form? Were licensing fees an important part of the value of the original work in question? (Judge Evans held that they usually were not, so this was an important question).

The results of this additional analysis were what finally determined whether or not a use was fair if the third factor threshold the Judge had set was exceeded. When the amount used was within her guideline of 10% or one chapter, the availability of a license for digital distribution did not sway the analysis against fair use. And when a license for a digital excerpt was not available, the amount used was less important (although the leeway here is small, for Judge Evans).

If this rule about amount seems unduly restrictive, libraries would do well to remember three other findings from the ruling that impact how we might view this guideline. First, she held that percentages should be calculated based on the total number of pages in the book, which makes for a much easier and quicker determination of amount than the method of counting the plaintiffs sought. Second, she held that individual chapters by different authors in an edited volume should by counted by the same method and not as if each were an individual work, which also simplified the process of decision-making. Third, she rejected the idea that a use that was fair use in one semester must be paid for in a subsequent semester.

All three publisher plaintiffs, along with the two organizations (the Copyright Clearance Center and the Association of American Publishers) that bankrolled them, have now issued initial responses to the verdict. Overall their objections are quite vague. Several of the responses refer to legal or factual “errors” in the ruling, but they do not specify what they are. The AAP does disagree with the Judge’s finding that the loss of permission revenues because of fair use imperils their business; they repeat this absurd claim even after the Judge pointed out, based on figures supplied by the plaintiffs themselves, that these publishers made less than one quarter of one percent of their 2009 revenues from academic book and journal permissions.

All of the publishers assert that there is error in the judge failing to consider what the AAP calls the “full context” of the activities at GSU. Sage and Oxford (who issued identical statements) suggest a “pattern and practice” of infringement , while Cambridge refers to “systematic and industrial-scale unauthorized reproduction.” It is hard to know what to make of these assertions, other than that they arise out of sheer frustration. Since the Judge has just found that only five of the excerpts before her were infringing, “systematic” and “pattern” seem like inappropriate words. The Judge had to decide the case based on the specific allegations and evidence before her, and the plaintiffs were the ones who produced those allegations. So if only five out of 99 (or 75, depending on where you start counting) were infringing, no pattern of systematic infringement has been proved.

The Cambridge statement has the strongest hint of an appeal, saying that they look forward to working with their “partners” to “resolve this issue.” It also has the chutzpah to evoke “our authors” as victims of the ruling. How many times can we repeat what publishers hate to admit in public, that the vast majority of academic authors are not paid for their scholarly writings (by publishers, at least) and do not consider permission fees when deciding whether, what or where to publish?

So where do we go from here? We need to remember that this is only a District Court ruling, which is not binding on any university other than GSU, that no remedy has yet been ordered, and that it is likely to be appealed. Nevertheless, it is a carefully reasoned ruling with lots of specifics for libraries and universities to consider. Some institutions may decide that the Judge has defined a safe harbor and that they want to anchor there; in that case it is worth noting that some of the analysis she applies would be difficult to replicate without detailed financial information from publishers. More likely, schools will look at this ruling and tinker with their own policies a bit. Given both its substance and the still-early stage of the proceedings (really, even after 4 years and 350 pages this is not the final word), this decision is not, in my opinion anyway, a cause for large-sale or precipitous changes at most institutions, a fact that even the publisher statements seem to admit.

The GSU decision — not an easy road for anyone

By Kevin Smith, J.D. On May 12, 2012 · 33 Comments · In Copyright in the Classroom, Copyright Issues and Legislation, Fair Use

Overall there is good news for libraries in the decision issued late yesterday in the Georgia State University e-reserves copyright case. Most of the extreme positions advocated by the plaintiff publishers were rejected, and Judge Evans found copyright infringement in only five excerpts from among the 99 specific reading that had been challenged in the case.

That means she found fair use, or, occasionally, some other justification, in 94 instances, or 95% of the time.

But that does not make this an easy decision for libraries to deal with. Indeed, it poses a difficult challenge for everyone involved, it seems. For the Judge, it was a monumental labor that took almost a year to complete. She wrote 350 pages, working through a raft of legal arguments first and then painstakingly applying them to each of the challenged readings. And for me, with a week’s vacation pending, I am trying to make sense of this tome before I leave, which is why I am writing this at four in the morning on a Saturday (please excuse typos!).

For the publishers who brought the suit – Oxford, Cambridge and Sage – there are some bitter disappointments. Judge Evans explicitly rejects the Guidelines for Classroom Copying as the standard to be applied. She also found the two major cases on which plaintiffs relied, American Geophysical Union v. Texaco and Princeton University Press v. Michigan Document Services, to be inappropriate analogies for the situation before her.

One holding that is certain to generate much discussion is Judge Evans’ rejection of the so-called “subsequent semester” rule, which had evolved from the Classroom Copying guidelines and led many institutions to assert that a liberal interpretation of fair use was permissible once but that permission had to be sought for subsequent uses of the same text. Judge Evans found this restrictive approach to be “an impractical, unnecessary limitation” (p. 71).

Perhaps most distressing for publishers is the Judge’s statement that permission fees are “not a significant percentage of Plaintiff’s overall revenues” and that their loss through the assertion of fair use does not threaten the publishers’ business. She calls this latter argument “glib” (p. 84).

In spite of all this, there are significant aspects of this ruling that will prove very difficult for libraries as well. Two aspects seem especially likely to stir up consternation.

First, the Judge applies a strict standard for the amount of a work that is permissible under fair use in this situation. The percentage she selects is 10%, or a single chapter. In putting this percentage into practice, she bases the calculation on the total page count of a book (this was an issue at trial, with publishers arguing that only the actual text of the work should be counted), and rejects any distinction for books that are edited, in which each chapter has a different author (p. 88). This is a less flexible standard than many libraries would like, I think, and it seems too rigid to be a good fit with the overall structure of fair use.

Second, the Judge bases many of her analyses of the fourth fair use factor on the percentage of the overall revenue that publishers realize for a particular title that comes from permission fees. She criticizes the GSU policy for not providing sufficient guidance for making a determination about this kind of market impact, but immediately acknowledges that the standard she is applying “would likely be futile for prospective determinations (in advance of litigation)” (p. 337). This is simply an unhelpful approach, since libraries and faculty members must make such prospective determinations without knowing all the information that publishers provided, under court order, to the judge. Recognizing this, Judge Evans says that “the only practical way to deal with factor four in advance likely is to assume that it strongly favors the plaintiff-publisher (if licensed digital excerpts are available) (p. 338).

Hard as it may be to be told to make this assumption, even though the judge has found it not to be the case in most of the instances before her, we should pay attention to her qualification of it. The permissions we should look for are those available for digital excerpts. In another place she asserts that she will consider the permissions market in the fourth factor analysis only when a license is “readily available” at a “reasonable” price for a “convenient” format (p. 89). She explains that the license must, to qualify as convenient, allow digital excerpts. This standard is actually hard for some publishers to meet in their current licensing postures. No infringements of Cambridge works were found, for example, probably in part because Cambridge allows only very limited licensing through the CCC. If publishers do not license in a way that facilitates reasonable educational use in the digital environment, the fourth factor will, she seems to be saying, cease to favor them.

Two more quick points, and then I will try to summarize the fair use analysis the Judge is recommending.

First, it is very important that the Judge finds that GSU’s copyright policy was a good faith effort to comply with the copyright law by interpreting fair use (p. 338). She criticizes the policy on two points – its failure to define the strict percentage limit she says is necessary and for not providing guidance about market harm, which she admits would be impossible. But she does not believe the policy was a sham or merely “cover” for doing whatever the faculty wanted, as publishers asserted. This would be a significant finding if money damages were at issue, due to the waiver of statutory damages in the law for educational institutions that make good faith fair use decisions.

The second point I think is interesting is the fact that Judge Evan considered hit counts when making determinations about some of the disputed excerpts. In some cases she cuts short the fair use analysis because hit counts suggest that no students actually read the material. In those cases she calls the copying “de minimis” and ends her analysis (see p. 96, i.e.). This means that it will be hard to dispute about specific readings without knowing whether or not they were actually read (we know, sadly, that many such assignments are not). Putting infringing material in a system, apparently, is not enough to establish liability; plaintiffs will have to show it has been used.

So here is the bottom line on the fair use analysis Judge Evans has outlined (pp. 87 – 89 of the opinion). The first factor – purpose and character of the use – heavily favors fair use because it is for non-profit teaching and research. She mentions the section 107 reference to “multiple copies for classroom use” and she does not employ a “transformational” approach. She finds that the second factor also favors fair use, since all of the works at issue were non-fiction and educational in nature. The third factor can go either way, depending on whether or not the excerpt is less than 10% or a single chapter. The fourth factor is where the difficulty lies; Judge Evans finds that it “heavily favors” the plaintiffs IF a license for the appropriate format is readily available at a reasonable price. There is lots of room to debate this part of the analysis, and lots of uncertainty, I think, about how it can be applied.

The Judge has deferred any action on crafting a remedy for the few infringements that she found. She has asked the plaintiff to suggest an injunction, which will be difficult to craft to fit this ruling and will have to look very different from the one they suggested 18 months ago. GSU will have a chance to object to any draft injunction, so the remedies phase of the case will go on for a while.

In general I expect librarians to be happy about the outcome of this case. It suggests that suing libraries is an unprofitable adventure, when 95% of the challenged uses were upheld. But there will also be a good deal of hand-wringing about the uncertainties that the Judge has left us with, the places where we need information we cannot reasonably obtain, and the mechanical application of a strict percentage. We will spend considerable time, I think, debating whether and how to implement Judge Evans’ rules into our own copyright policies. In the meanwhile, of course, the ruling is nearly certain to be appealed.

(Note that the opinion is not yet up on the free sites that I usually link to, but three different comments to my previous post have provided links to PDF copies)