The GSU opinion is out, and it's good news for the University. Of 48 allegedly infringing excerpts that were up for reevaluation on remand, 41 were found to be fair use. After being chided by the 11th Circuit for her overly-deterministic analysis of the four statutory fair use factors, Judge Orinda Evans has dutifully given us 220 pages of less-mechanical-looking analysis.
(Though there was this little gem, which I swear is there just to tweak the noses of the appellate panel.)
The beauty of her original opinion’s mechanical analysis (and I’m not usually a fan of bright lines and thresholds!) was that it gave folks guidance about what would be clearly fair (or not) in the future when it comes to non-transformative use of scholarly books, using metrics (number of pages or chapters, etc.) that they could easily apply. Her new analysis, stripped of its bright lines and clear arithmetic, seems to amount to nothing more than her opinion about whether the use will substantially harm the market value (actual and potential) for the works used. How much harm is “substantial”? Well, in several places Judge Evans says the harm must be so extensive as to risk undermining the publishers’ entire motivation to publish the work. So, it would need to eat their entire profit margin (or enough that they decide it’s not worth the bother to publish). And in at least one place she seems to suggest that, because there is no marginal cost for a publisher to offer to license use of excerpts from a work, there is no real harm when GSU decides not to pay the license. This is heady stuff, and could offer a very wide berth for educational fair use of electronic excerpts.
What gives me pause is that the decision process here seems to involve knowledge of very specific market data about each individual work you want to use. Distributing excerpts of 25 pages of a 500 page book might be fair use if the book has made very little revenue from electronic licensing in recent years, but it might not be fair if licensing revenue was higher. That’s all fine and good for a judge looking at hundreds of pages of discovery, but what about a professor, ex ante? How can a prof learn this information? I don’t know; it’s closely-held business information that was only revealed in this case because it was litigated.
All that said, my prediction about this opinion has always been that, if it follows scrupulously the Circuit Court’s insistence on indeterminacy, then there will of course be very little guidance for future fair users on the face of the opinion. BUT, clever readers can start building their Excel spreadsheets, loading in the relevant data about each use, along with the outcomes, and looking for the (unstated) thresholds or patterns in the opinion. Start your engines!
A couple of last thoughts: It's fabulous that Judge Evans makes crystal clear that this case is only about non-transformative uses. Now, she (and the appellate court) used a definition of "non-transformative" that conflicts with all the case law I know of ("mirror-image copies" are non-transformative; tell that to HathiTrust and Dorling-Kindersley). But leaving that aside, this very tricky framework has virtually nothing to do with transformative uses. All the market hoo-ha that comes in under factor four would be irrelevant in such cases.
And it's also fabulous that Judge Evans is again awarding attorneys fees and costs to GSU. Now, presumably, those fees have climbed even higher. Perhaps now Oxford, Cambridge, and Sage (and their paymasters at the CCC and AAP) will decide it's time to stop paying GSU to kick their butts in court.
Finally, friends are telling me there are some editing errors in the opinion; namely, the summary of infringing uses at the end of the opinion includes works that were actually ruled to be used fairly in the body of the opinion. So, there may be even fewer infringing uses in the mix.
