I blogged earlier in the week about a letter that William Morris Endeavor Entertainment sent to its clients recommending that they opt out of the settlement. Or rather, I should say, I blogged about news stories on the letter, since I couldn’t get my hands on a copy. At the time, I though that William Morris’s stated reasons for recommending opt-out—that the settlement would “bind copyright owners in any book published prior to January 9, 2009 to its terms—seemed unconvincing. The Authors Guild issued a swift, strong reply and even held a conference call for its members to respond.

Well, William Morris has itself responded to the controversy with a new memo, this one apparently intended for public consumption. And this time, we’ve been able to track down a copy, which we’ve added to our collection at the Public Archive. My initial hunch, that the press summaries were misleadingly incomplete, has turned out to be correct. William Morris’s concerns are both much smaller and much larger than the news coverage (and the Authors Guild’s reply) would lead one to believe.

The critical point about the advice letter is that it recommended opting out, not objecting. William Morris has no big beef with the settlement itself. It just believes that its clients are better situated to take advantage of the Google Books program by opting out and striking individual deals with Google than by relying on the settlement’s rules. (In this sense, the Authors Guild may have overreacted, inadvertently transforming a small piece of advice to a few authors into a big dust-up, and making a calm view of the settlement seem like an angry one.)

The reasons for this confidence are interesting. First, William Morris thinks its clients have more negotiating power than the average author does:

We believe that our clients benefit from their stature, the negotiating power of the Agency, as well as long established precedent in our book contracts that other authors may not share.

Perhaps this is over-confidence, perhaps not. My best impression is that Google’s Partner Program is largely a take-it-or-leave-it deal. Google’s ethos is that if it’s not worth doing at scale, it’s not worth doing. Thousands of one-off deals would be unworkable for Google’s machine-heavy, people-light way of doing business.

William Morris’s attitude reflects also a calculated gamble that Google’s proffered terms are unlikely to change for the worse. Since Google seems strongly inclined to offer the same revenue splits and control to copyright owners not included in the settlement class at all (such as the authors of books not yet written), and Google seems inclined to offer the same terms to everyone, this strikes me as a fair assessment of the world. I don’t know that the upside potential really outweighs the foregone inclusion fees, but William Morris’s reasoning here is at least colorable.

The second interesting point William Morris makes is that the “terms of the Settlement are interminable.” (Having read them, I can say, no, it only seems like it.) In particular, William Morris is focused on the non-display uses—such as keeping the books in the search database, algorithmic research, and other large-scale automated processes. If you stay in the settlement and don’t Remove your book within twenty-seven months, you consent to all such uses forever.

Here, things get a little tricky. William Morris happens to believe (as I do) that the non-display uses are fair uses. The Authors Guild seized on this point, writing:

However, William Morris believes Google’s scanning is a fair use (an unusual position for those concerned with authors’ rights, and a decidedly outlier position for those in the copyright bar). What is more confusing is that William Morris encourages authors to opt out of the settlement while at the same time encouraging them to grant Google the right to use digital copies of their works for search purposes.

That’s a fair critique, and William Morris’s reply in its second memo is more of a point about long-term bets in copyright:

As of today, it appears to us that “non-display uses constitute fair use, but it is impossible for anyone to predict that such use will always be the case. And if such inclusion is not fair use, there may come a time when the author or their heirs might wish to remove their work from the Google database for any reason. The authors waive that right, forever, if they do not opt out of the Settlement.

Personally, I think that trying to reason forward about possible mutations in fair use law of this sort is an exercise in creative make-believe, and that there is no point in planing, one way or the other, for shifts of this form.

The third point William Morris makes is, to me, truly fascinating:

Few if any major publishers currently intend to make their “in print” works available for sale through the Settlement program. When it comes to “in print” works, the more restrictive instructions to Google from either the publisher or the author control. This means that if either the author or the publisher does not want an “in print” work sold through the program, Google will not have the right to sell it.

It appears that most major publishers will not allow their “out of print” books to be sold through the Settlement program either. In fact, we believe that most major publishers will take the position that none of their backlist is “commercially unavailable” as defined in the Settlement because the availability of on-demand and other electronic editions will constitute the works as being in print.

Read those two paragraphs again, slowly. If the analysis of post-settlement strategies is right, the implications are large. Let’s go sentence-by-sentence:

Few if any major publishers currently intend to make their “in print” works available for sale through the Settlement program.

If true, then most publishers plan to opt in to the settlement, collect their inclusion fees, and then promptly exclude their books from the Institutional Subscription and Consumer Purchases. That means that these two programs won’t be all that large, and the settlement is in fact not all that big a deal as to in-print works.

When it comes to “in print” works, the more restrictive instructions to Google from either the publisher or the author control.

This is one of the most interesting provisions in the settlement’s description of how copyright owner control will work. I’ve read it as being an author-protection provision; the author can veto the publisher’s attempt to sell a book, and thus possibly negotiate a better deal than the revenue splits specified in the settlement, even where the rights in the book haven’t reverted to the author. But …

This means that if either the author or the publisher does not want an “in print” work sold through the program, Google will not have the right to sell it.

… it’s also a publisher veto program. Publishers can veto all sales, and William Morris thinks that they will. This means that from a practical perspective, authors don’t have much control. Once the publishers say “no” across the board, it doesn’t matter what authors say. This would make many of the detailed provisions in the settlement and the Author-Publisher procedures largely moot. They’re beautifully engineered systems that won’t be used very often.

It appears that most major publishers will not allow their “out of print” books to be sold through the Settlement program either.

I have no idea what the source of this information is, either, or whether it’s true or not. But if it’s right, then the settlement’s two major programs will include few books that being actively managed by their copyright holders.

That means—again assuming that William Morris is right—that the settlement is really only about two classes of books. First, there are reverted books fully controlled by their authors—not a category to sneeze peanuts at, but not where the action in publishing is, either. And second, there are orphans.

I take two implications from this line of reasoning. First, the settlement is in a sense really about the orphans. We would expect almost all non-orphan works either not to be available through Google at all, or to be available through non-settlement terms. The settlement would give Google a license to the orphan catalog, and that’s the major work that it would do. In a sense, the settlement would be a three-way trade. Google gets the orphans and permission to index, authors get some money, and publishers get a renegotiated set of library agreements that put substantial limitations on what libraries can do with their digital copies.

Second, if we’re really being serious about protecting the interests of orphan owners, we can’t use opting in to the settlement as a measure of whether active copyright owners think the deal is reasonable. They could stay in the settlement but then immediately (and permanently) Remove or Exclude their books. That would indicate that they don’t really like the economic terms of the settlement, which would suggest that they haven’t fairly represented the interests of owners who may take a long time to show up.

This says to me that the settlement needs some kind of Removal/Exclusion threshold. If more than some specified fraction of registered (with the Registry) rightsholders choose to Remove or Exclude their books, then all unregistered rightsholders should be presumed to have done the same. This provision would be similar to the opt-out triggers some settlements contain, but based on the fact that the proposed settlement has both external (send a letter to the court) and internal (Remove or Exclude) opt-out provisions. Without this kind of trigger, the incentives of present class members and absent ones are misaligned.

Perhaps William Morris’s analysis of what publishers will do is wrong. It would certainly be easy for some major publishers to make statements about their intentions that would refute William Morris’s predictions. But if they’re right, the settlement looks very different—both a smaller and a bigger deal than it first appears.

I am interested that you say that the agreement only seems to be interminable. Who, on your reading, can terminate it, and how?

That was a joke about the length of the document.

I walked into that one.

Beware the Google Book Settlement agreement: if you spend enough time with it, it can damage your sense of humour!

There actually is a separate termination agreement between the parties to the settlement. Article XVI of the settlement agreement makes reference to it, but that separate agreement is a secret, and Article XVI says it will not even be available to the judge presiding over the settlement.

It would be really interesting to know what the criteria for the termination are. If DOJ decides to challenge the deal? If security standards don’t work and the corpus is hacked? If not enough books can be commercialized because too many ask for removal?

You seem to dismiss the significance of “reverted books fully controlled by their authors”. But the overwhelming majority of in-copyright books (including the overwhelming majority of “orphaned” works), are books for which the e-rights have reverted or for which no e-rights were ever asisgned to the publisher, and the e-rights to which in either case are fully author controlled. Yes, the settlement may be mainly about “orphaned” works (depending on how much effort is spent on trying to identify and locate rightsholders), but it is also mainly about works in which e-rights are held entirely by authors. Publishers don’t care about the rights to these books not because those rights are worthless, but because publishers don’t own those rights. The revenue being undercut by Google is the potential revenue to authors from self-publishing in e-book form, or licensing e-books through other channels.

Edward Hasbrouck: But the overwhelming majority of in-copyright books (including the overwhelming majority of “orphaned” works), are books for which the e-rights have reverted or for which no e-rights were ever asisgned to the publisher, and the e-rights to which in either case are fully author controlled

Thank you, this is helpful. (Similar points have also been made to me privately via email.) Of course, authors could also pull their works from the program en masse. With so many of them out there, though it seems unlikely, especially given that the default for out-of-print books is to allow display uses.

I’m not sure about the sense in which the Individual Purchase program and the Institutional Subscription could be said to “undercut” self-publishing or other licensing channels. Isn’t it more in the nature of a choice offered to authors, which they can accept or refuse? If anything, the Google option could pressure other potential publishers into offering authors even better deals, no?

“Of course, authors could also pull their works from the program en masse.”

That misses the reason why, as William Morris predicts (correctly, I suspect), “”Few if any major publishers currently intend to make their “in print” works available for sale through the Settlement program.”

Because of the different author-publisher revenue split under typical print royalty and subsidiary rights clauses of author-publisher contracts (85/15 or 90/10 vs. 50/50), publishers make more per printed book — even taking into account production and distribution costs — than per e-book, while authors make substantially more per e-book than per printed book — even taking into account the typically lower retasil price of an e-book.

Print publishers welcome incremental e-book revenue, but to the extent that e-books compete with printed books, publishers’ interest is in suppressing e-books to drive more profitable print book sales.

Authors’ interest is in driving as many potenetial print book sales to e-books as possible, both because they are more porfitable in the short term and because they build brand equity in the e-book, which the author can expect to control 100% once the book is out-of-print in hardcopy. (Costs of warehousing, etc. make it unprofitable to keep most books in print in hardcopy, while the aurthor can keep a self-published e-book in print for life at negligible cost.)

Moreover, to the extent that to the reader at least some books are fungible with others, and e-book sales of backlist books (most of them out of print in hardcopy, and thus 100% author-owned) compete with new book sales, publishers’ interest is in suppressing backlist e-books to drive new print book sales, while an author makes more from a self-published e-book of one of their older works than from a printed copy of their newest book.

There are fundamental structural reasons in typical contrcats why authors are on the side of readers — and both are oppoes by publishers — in seeking to make backlist e-books available.

One sees the same thing with musicians: Publishers’ Web sites, and music stores, push new titles, while musicians’ personal sites push sales of backlist downloads.