The GSU decision — not an easy road for anyone

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)