Due to length limits in my blogging software, I’ve had to split this post into two parts. Part I covered the arguments of settlement supporters and opponents; this post covers the arguments made by the Department of Justice and the parties, along with a few brief comments of my own.
The United States of America
That concluded the presentations-in-opposition portion of the day. Next to speak was William Cavanaugh from the Department of Justice. (Judge Chin set no official time limits on him, or on the parties to the settlement; they all spoke, however, with a clear consciousness that they needed not to wear out the judge’s patience.) Compared with the Department of Justice’s filings, which have gone out of their way to recognize good elements in the settlement and to express regret about the issues it raises, his presentation in court was more forcefully negative. The Department of Justice opposes the settlement in its form.
He started by breaking the settlement into two familiar portions. One settles claims for past infringement based on snippet display: that’s appropriate. The other, “grafted on” would create the large commercial transaction. There’s no “nexus” to the fair use arguments over snippets, it’s not designed to remedy past harms, it produces results not achievable in the lawsuit itself. And that’s not appropriate for a class action settlement.
Look, for example, at the motion for approval. At pages 40–45, when it describes the lawsuit, it discusses the risks of losing the fair use fight over snippets, it discusses damages. There’s no mention of selling books or creating a subscription database. Had those issues really been in the case, Google would have had no colorable defense.
The class representatives had a duty to litigate those claims, or to settle them. They weren’t hired to serve as the class’s literary agent. The opt-out feature is unique to a class action, it comes up in no other realm or commercial context. It is a “misuse of Rule 23.” Such a deal may or may not be a good idea on its merits, but it’s beyond the appropriate use of Rule 23. (Hat tip to Amchem, which said the same over a global asbestos compensation program.)
In a lawsuit, the plaintiffs’ past claims could be extinguished by the judgment, but not as to their forward-looking claims. Thus, there’s a Rules Enabling Act problem. Here, Rule 23 is being used to modify substantive rights—and that’s why the “identical factual predicate” cases treat it as a separate issue from the adequacy of the settlement itself. But that’s at issue, too—the farther you go from the underlying issue, the harder it is to prove that the settlement is adequate, because it’s hard to identify the conflicts and what the class is giving up. (This was a dense, but very sharp, and very important part of Cavanaugh’s argument.)
And so we come to Firefighters. That case involved a court’s statutory authority under Title VII to enter and enforce a consent decree. Cavanaugh claimed it was distinguishable because:
- In each case, the forward-looking conduct is partly remedial to address past harm. Selling books doesn’t address past harm.
- The defendant’s activity was purely voluntary. In Firefighters itself, that activity was hiring a certain number of minority firefighters. It didn’t bring benefits to the defendant, which the current settlement does.
- When the defendant agrees to undertake remedial actions, it’s not protected by a release. Releases of future discrimination are simply not permitted. The Uhl case regularly cited by the plaintiffs involved a release of the same conduct at issue (trespassing to lay telecommunications cable) and it was necessary to settle the dispute.
Even under the Firefighters test, though, Cavanaugh explains, it’s clear the settlement doesn’t pass. Is this settlement within the general scope of the pleadings? No. Does it spring from the same underlying dispute? No. Is it consistent with the underlying law? No. That law here is copyright; this deal turns copyright on its head.
In policy terms, copyright does have utilitarian goals, but it advances those goals through exclusive rights. Institutionally, these issues are best left to Congress. When foreign sovereigns have issues, Congress can respond. And Congress is active in copyright; it’s amended the Copyright Act over sixty times. Digitization is a new field. If there’s to be a big shift and compulsory licenses, Congress should make that choice.
In defense of the settlement, Google says the copyright owner can say “no” at any time. But once you move out of what’s necessary to settle the case at all, no obligation should be put on the copyright owner. Certainly not a significant commercial burden.
At this point, Cavanaugh and Judge Chin engaged in a colloquy that I didn’t really follow. When they emerged from it, Cavanaugh talked a bit about the Registry. The parties say it wouldn’t work on an opt-in basis. Well, why not? Google’s competitors would be happy to participate and to fund it if it were giving Google access to works they didn’t have.
Cavanaugh also called attention to the Author-Publisher Procedures, which the DOJ has become increasingly interest in. It’s not simply an allocation of damages; it reformulates contract rights. Certainly, the class action wasn’t about the respective rights of authors and publishers. It’s unlikely that it could have been, given how individualized their contracts were. The Procedures may be efficient, but the judge should give serious attention to them.
On the antitrust side, he didn’t address much of substance. Instead, he told the court that the DOJ investigation was ongoing, in many markets, including search. (He didn’t have to say that last part; it may be significant, in a signaling sense, that he did.) The parties have addressed some issues, but not others. It remains concerned about the revenue split for non-commercially available books. These issues aren’t normally a concern in class actions, because they don’t normally create forward-looking agreements. The settlement creates a pricing floor to be used against competitors; it’s not necessary to achieve procompetitive goals, and may well be a per se violation. Unlike in BMI, there’s no consent decree and the rights aren’t being acquired through bilateral negotiations with rightsholders.
Overall, Cavanaugh’s presentation was professional and strikingly negative on the settlement. I was surprised at how few questions Judge Chin asked him.
Now, it was the parties’ turn to take the stage. Michael Boni did the principal lifting. He started off explaining that as a Philadelphian, he felt a bit like Rocky, getting battered for fifteen rounds; he hoped he could come back in the last one. (Perhaps not the right analogy to use in a New York courtroom when arguing before a long-time New York City kid.)
Boni started the counterattack by focusing on the scope of the lawsuit. It was clear, he said: the lawsuit was occasioned by Google’s plan to digitize the world’s books. It was giving digital copies to libraries—so the authors and publishers filed a lawsuit for injunctive relief. (But, I would note, they never pressed for a preliminary injunction to make Google stop during the lawsuit.)
Judge Chin asked—echoing many of the objectors—whether the lawsuit hadn’t started as a snippet claim. Boni replied that the plaintiffs had alleged much more. Judge Chin pushed on: If Google had engaged in wholesale display, it would have had no colorable defense, right? Boni took a quick step to the side and said that the plaintiffs were concerned with what they didn’t know about Google’s plans. They had no idea how much Google would push the envelope.
The “we didn’t know” argument stuck me as as a dangerous one for Boni. For one thing, Google’s lawyer, Daralyn Durie, would later complain that many of the objectors were driven only by “speculative” harms. For another, your own ignorance about what someone else is doing doesn’t actually mean they’re doing it—or give you a basis to sue them for it. And that, in turn, undercuts his “identical factual predicate” and “scope of the pleadings” arguments.
At any rate, Boni emphasized the generality—some would say “vagueness” of the plaintiffs’ pleadings. Under the Federal Rules, these general allegations are all that’s required.
At this point, Judge Chin shifted to the other half of the comparison: the releases in the settlement.Judge “Would it be against public policy to release a defendant from future claims of discrimination,” he asked. Boni waffled: “I don’t want to say yes or no.” Judge Chin pushed with the inevitable follow-up: “Is that different from here?” Boni said, as he had to, “Yes.” Here, there’s a release of known future claims, on a non-exclusive basis, with no transfer of an interest. But, Judge Chin asked, releases are usually based on past conduct. Why is this case different? Well, this isn’t the usual case, Boni responded, but it’s judged under the same, usual, Rule 23(b)(2) standard as any others.
Boni started reciting nine-point tests, but Judge Chin cut him off. Tell me why you believe this settlement meets the requirements. Boni-slash-Rocky stopped dodging and took up the real question head-on. The big issue here today is why not opt-in? “Here, we determined—”
Judge Chin cut in again, “I agree that if you did opt-in, you’d eliminate many of the objections. So why not?
“You’d have no settlement,” Boni replied. Judge Chin said something close to “I assume, and in fact after much of what’s been said here today, I surmise, that it’s about the orphan books, and that Google wants them in the settlement.”
Boni had a reply ready. “That’s a myth; it’s a political football.” In every class action, some class members don’t claim their works. But don’t they constitute the majority of the class, asked Judge Chin. There’s no deadline, Boni responded; they can turn their books off at any point. Judge Chin: What about those who don’t come forward? Boni: They’ll be looked for. Judge Chin: Aren’t the vast majority out of print? Boni: So far, some 620,000 out-of-print books have been claimed by 40,000 authors, through the notice program alone. We expect to find a lot of these people. We’ve had an 85% success rate; the UK licensing society has found over 90%. (These statistics are a powerful argument; some of the me-too objectors would probably have been better off looking for ways to call those numbers into question.)
Boni finished by discussing substantive fairness to class members, including one of my favorite lines of the day: “To say it’s unfair to authors, is simply not fair!” Overall, the smooth, calm Boni was doing his best to soothe Judge Chin, but drew the sharpest, most involved questioning of the day.
Daralyn Durie was a bit punchier than Boni, but far less pugnacious than the hard-hitting Google brief filed last week. Her presentation was one of the day’s top few: confident, reassuring, and well-structured. Google chose its outside counsel well.
She led off by responding directly to Judge Chin’s earlier question about discrimination. No, a settlement couldn’t release future claims of discrimination, because discrimination is an evil. Dissemination of copyrighted works is good. “But isn’t copyright infringement an evil,” asked Judge Chin. Yes, said Durie, to the extent it doesn’t compensate the copyright owners, it harms their economic interests. What about those who want to sit on their rights, asked Judge Chin? Those who’ve put their works in the stream of commerce have intended them to be read, Durie replied.
It was a good set of answers, but notice the trick. By emphasizing compensation over control, Durie turns copyright as a property rule giving an owner the power to refuse uses into a liability rule providing only payment. (I’m sure her law firm partner Mark Lemley, a noted critic of the idea of intellectual property as “property,” would approve of that trend.) As one of my students pointed out, she was importing the theory of efficient breach into copyright law.
Turning to opt-out and opt-in, Durie said that opt-out was essential to fulfill that purpose. There’s no risk to the economic interests of the absent copyright owner. By default, only works that aren’t commercially available are included, which means they have no other channels of distribution, so there’s no current way for copyright owners to allow use or receive compensation. Money piles up for them, so they can only receive benefits.
It’s a clever argument, and persuasively presented, but also, I think, untrue. Google could get the commercially available status of a work wrong, and undercut a book actually being sold, say, as a download from the author’s web site. Or it could get the price wrong, and sell hundreds of 99-cent copies, satiating the demand for a book that could profitably have been sold to the same readers at $10 a copy. The result would be to dry up the potential market: again, economic harm to the author results. Some, indeed many, may well benefit from the settlement, but Durie’s logical proof that they can only benefit fails.
Opt-out, Durie maintained, is essential to the settlement, because there’s no other way to create this market. The current market is small. Google has scanned two million public-domain books, of which hundreds of thousands have never been accessed. Others are more popular. Ex ante, there’s no way to know which are which. (Know? No. Make reasonable guesses? There must be—or otherwise, Google’s algorithmic pricing couldn’t plausibly be fair to the copyright owners of out-of-print books.) And the cost of identifying the rightsholders is prohibitive, because no one has yet done it.
Judge Chin asked Durie whether this was an issue for Congress. She replied, quite sensibly, that Congress has the power to pass orphan works legislation—but that the parties have presented the court with a proposed Rule 23 settlement. If it passes muster under the Rule 23 evaluation, the court should approve it.
Durie continued by observing that the settlement will accumulate money for absent rightsholders. As soon as they show up, they can take control. As soon as their works become commercially available, the defaults switch and the settlement display uses stop.
Turning to the court’s power to approve the settlement, Durie said that the settlement clearly passes the Firefighters test. “Both sides say the answer is clear,” Judge Chin observed to laughter. As for being within the general scope, Durie pointed out that the plaintiffs’ complaints alleged the digitization of entire copies. If Google had been making whole books available, would Google have defended it, Judge Chin asked? As to selling whole books, no, Durie replied. Giving away the entire copy, perhaps: Google gave digital copies to libraries.
On the antitrust questions, Durie said the settlement is about consumer welfare. Without the settlement, there’s no way to access these works. There’s no danger of monopoly by Google; its current market share is 0%. The Open Book Alliance is the one trying to preclude competition by keeping Google out. The settlement, however, is nonexclusive; it doesn’t erect barriers, and even makes things easier for competitors.
Judge Chin asked whether Google would have a competitive advantage for unlocatable copyright owners. Durie replied that the relevant question is whether Google engaged in a wrongful exclusionary act, and then gave a rundown on the statistics that served as a nice rejoinder to some of the morning’s arguments. The 174 million “books” that Dan Clancy mentioned are just the ones they have metadata records identifying. But only about 42 million are in U.S. libraries, of which another 20% are in the public domain. 50% of the rest is in a language other than English, and once you compress the different editions of the same work, less than 10 million books are affected by the settlement. Only 5 million out-of-print books are implicated, and they expect the unclaimed percentage to shrink over time.
So: any exclusionary acts? No; the settlement creates no barriers to entry. Anyone can scan books; there’s no unfair advantage in search, which is open to all. The future is unknown; the objectors present only speculative concerns. But Google has disclaimed any antitrust immunity, so if Google were to become as dominant as Amazon (nice dig!), the antitrust impact could be reassessed.
Durie’s closing points were three. First, responding to Andrew DeVore’s concerns, she explained that the parties didn’t interpret the releases as broadly as he did—Google wasn’t, for example, seeking a release as to its AdWords programs. Second, on privacy, the Public Access service would be anonymous. And third, as to foreign copyright owners, this case involves U.S. copyright interests, uses in the U.S., so it’s appropriate for a U.S. court to apply U.S. copyright law.
Finally, Bruce Keller spoke for the publisher plaintiffs. Judge Chin wasn’t expecting him, which Keller clearly noticed, and therefore kept his remarks quite brief. He addressed four points.
First, going back to the scope-of-the-pleadings issue (by this point quite evidently the issue that most worried the parties to the settlement), he noted that the publishers had framed their lawsuit as one over the fact of the scanning. Google raised the snippet defense of fair use. But once the scans are made, the publishers in a lawsuit could demand return of the copies. Clearly we could also allow Google to keep the scans but set conditions on their use. (Judge Chin asked a question here that was little more than a paraphrase to make sure he understood Keller’s point.)
Second, even though Congress sets copyright policy, it’s also the courts’ obligation to rule within that framework, even on very difficult and enormous questions with policy implications, such as the legality of home taping as fair use in the Sony case. Judge Chin wryly commented, “Like in the Cablevision case_ which was also thorny, and there I didn’t do so well.” (The Second Circuit reversed Chin’s long and thoughtful copyright analysis in a shorter analysis that I think was more technically correct but also in some ways less satisfying.)
Third, Keller argued, the settlement doesn’t invert copyright law. It’s not different from a property owner’s right to exclude. Look at all the settled real property cases in which plaintiff property owners convey an easement (such as in Uhl with its easement for telecommunications cable-laying).
And finally, on fairness, this settlement has far more options than any other class action has offered to class members. In most class actions, which give simply one-time payment with a deadline, the redemption rate is low. Here, it’s not binary. Judge Chin: So the class members can go either way at any time? Keller: Yes. We kept the door open for absent owners to come forward at any time. It’s not perfect, but it’s fair, reasonable, and adequate.
As proof of that fact, look at the arguments made by opponents, which cancel each other out. Professor Samuelson wants greater access to works; Professor Nimmer thinks it provides too much. And Professor Courant is an academic author who likes the settlement. These different perspectives show how difficult the negotiations were. (One could also argue that these different perspectives could also suggest that the class is improper.) Again, the standard isn’t perfection, just that the settlement be fair, reasonable, and adequate.
Judge Chin announced that he would reserve decision, and that he had lots to think about. And with that, the hearing was over.
I was impressed with Judge Chin’s demeanor. He wore his intelligence and his authority lightly. His patience with the attorneys was, generally, inversely proportional to their self-importance. He didn’t ask all that many questions, but he regularly managed to put the attorneys on the spot with exactly the right question. On the whole, he stayed far away from the details and the specific tests; he seemed to be looking to garner parties’ clearest statement of their positions. He confronted them with difficult issues, but generally didn’t keep them pinned down once it was clear they couldn’t (or didn’t want to) address a question. His attitude was, by and large, pragmatic: he was looking for concrete problems and for specific solutions. I don’t have a much stronger read on how he is leaning in the case, other than to note that he is not signaling an urgent need to move quickly to a decision, and that he clearly recognizes which issues in the case are fundamental and which are tangential.
I expect a small flurry of motion practice in the next few weeks. Scott Gant signaled an intent to seek discovery on the notice; Amazon indicated a desire to brief its responses to the new cases cited by the parties last week. Based on his comments at the hearing, I expect Chin not to open up the case, procedurally in any significant aspects. I don’t expect discovery; I could see possibly some new briefing, but if so, under very strict page and scope limits.