Thursday, February 26th, 2009 at 11:00 am
I’m going to respectfully disagree with Sajid when he says that the Google Book Search settlement is the best that we could have hoped for. He writes that “the job of the judiciary is not to set policy for the public good, but rather to resolve disagreements under the law.” I agree with this sentiment in principle, but I think that ship sailed several years ago when the judge gave the green light to a class action lawsuit in which the “plaintiff” is the class of every author in the country. When a decision affects the rights of every single author, it’s a policy decision regardless of which branch of government is making it. The subject is a little bit complicated, so I’m going to break my critique of the settlement into two posts. In today’s post, I’ll talk about Google’s legal arguments regarding fair use and how the settlement will affect future fair use claims. In a subsequent post I’ll talk more specifically about how the settlement will adversely affect the market for book search engines.
To understand what’s wrong with the settlement, I think it’s important to start with a quick review of how we got here. Earlier this decade, Google wanted to create a book search engine analogous to its search engine for the web. But they had a problem: books as old as 1923 are still in copyright, and for books that are out of print—the vast majority—there’s no efficient or reliable way to find and contact the relevant copyright holders.
Fortunately, Google had a lot of experience with building search engines and the associated legal issues. By the time Google Book Search (then called Google Print) was launched, there was both solid caselaw and widespread industry practice suggesting that making copies of a copyrighted work for use in a web search engine was legal under copyright’s fair use doctrine. Google reasoned that if the fair use doctrine legal for search engine software to make a copy of a website for indexing purposes, and if it’s legal for a search engine to display thumbnails of copyrighted images without permission, then it’s probably legal to perform the analogous operations for a copyrighted book. Based on this reading of the law, which I found plausible at the time, Google went forward with its project.
The important thing to recognize is that Google Book Search wouldn’t exist at all if copyright law didn’t provide Google with this opening. We don’t know how the courts would have ruled on Google’s argument, but if Google’s argument had been completely ridiculous, the publishers would have scored a quick and decisive victory. It’s precisely because the law in this area was so fuzzy that Google was able to move forward with its project and wrest a reasonably favorable settlement from the publishers and authors. If Google had tried to secure explicit permission from each author and publisher before including a given book in its index, it not only would have massively increased the amount of legwork that would be required, but it also would have excluded millions of “orphan works”—older works whose copyright holders cannot be identified—from being included in the index at all.
Unfortunately, due to a quirk in the way copyright law works, it’s unlikely that any other company will be able to follow in Google’s footsteps. The copyright statute says that in weighing whether a use is fair, courts should consider four factors. One of these factors is “the effect of the use upon the potential market for or value of the copyrighted work.” The litigants in the Google Book Search case fiercely disputed how this factor should be interpreted, with Google arguing that the court should consider only the effect on the market for the original work—the printed book—and publishers arguing that the courts should consider potential revenues for licensing the work for use in book search engines. But one thing that unquestionably helped Google was the fact that there was no existing market for “book search engine licensing rights,” so even on the publishers’ theory Google wasn’t depriving the publishers of revenue.
Now, there is a market for book search engine rights. And judges naturally look at the realities of the commercial marketplace when they’re weighing the fair use factors. So now that Google is paying publishers for the privilege of indexing their books, judges are far more likely to find that it’s not fair use to index a book without paying. I still don’t agree with this result, but it would be a far more plausible reading of the law after the settlement is approved than it was beforehand.
In a sense, then, what Google has done is burned the fair use bridge after they crossed it. As the first entrant into the book search engine market, Google was able to use its fair use arguments to gain leverage in its negotiations with publishers. But any potential Google competitors will be unable to use the same strategy because the existence of Google’s agreement with publishers weakens the fair use argument.
And that means that anyone wanting to compete with Google will be required to go hat in hand to each of the thousands of companies that publish books seeking permission to join the book search engine business. Publishers will have no incentive to give new entrants any better terms than they’ve already given Google, and they may push for even more favorable terms. And even if some large company (and it would have to be a large company—no startup can afford enough lawyers) manages to jump through all the hoops the publishers put up, the result will still be vastly inferior to Google’s search engine because it won’t include millions of orphan works.
In my next post, I’ll argue that this is not only a bad outcome as a matter of policy, but it’s an outcome that can only be achieved with the blessing of the courts. For that reason, the courts should think hard before giving the settlement their blessing.