A Raw Deal for Libraries
by Karen Coyle
One of the most surprising, even shocking, features of the Google-AAP-Authors Guild Settlement is how hard it is on libraries. Given that Google Book Search could not have gotten off the ground without the cooperation of various university libraries, it is particularly disheartening that the proposed settlement treats them with such an iron fist at the same time as it expects them to foot much of the bill through subscriptions. It will be interesting to see how many libraries continue as partners, given Google’s bait-and-switch.
Take for example the digital copy that Google gives to a library in exchange for scanning its copy of a book. Previously, all library partners were given digital copies. According to the proposed settlement, however, only “fully participating libraries” will continue to receive copies from Google. These are libraries that (1) continue to provide in-copyright books for scanning and that (2) enter into a new agreement with the Book Rights Registry releasing them from liability for copyright infringement in relation to the Google scanning project. All other categories of libraries will no longer receive copies in exchange and, to make matters worse, they will have to destroy the digital copies of in-copyright books they already possess or otherwise expose themselves to the implied threat of a lawsuit from authors and publishers over copyright infringement.
Yet even these “fully participating” libraries are granted only a few permissible uses of their copies (e.g., services for the disabled, replacement copies, five-page access) while other uses that are arguably fair use (interlibrary loan, use in e-reserves and course management systems) are strictly forbidden. Compare this to the former agreements. In Google’s 2006 agreement with the University of California, available here, the university was allowed to use the digital copy to provide (unspecified) services to its library patrons (section 4.10). UC was also allowed to “add value” to the copy and possibly charge a fee for its use (see section 4.9, “Use of University Digital Copy”). How far we have fallen… Fully participating libraries must now give up such benefits and, if that wasn’t sacrifice enough, they must also guarantee the security of their digital copies as laid out in a 17-page “security standard,” under the threat of fees up to $7.5 million for security breaches.
Libraries have made huge investments in the books that Google is digitizing. Not only did they purchase, process, shelve and care for the books, over many years, but they continue to carry significant overhead costs for their continued use (including Google’s use!). Much of this investment has been made with taxpayer dollars. And yet libraries receive 0% in this proposed settlement while Google gets 37%. What kind of partnership is this? Taxpayers should be alarmed that their money has gone to provide a service that Google is exploiting on its own terms, in its own interests, with no monetary and little other return to the libraries.
By dangling the threat of future lawsuits, the settlement seems to be manipulating libraries into new legal agreements that 1) renege on the benefits of former agreements; 2) eliminate the digital copy for many libraries; 3) impose excessive restrictions on how remaining digital copies are exchanged; 4) impose harshly punitive security obligations; and 4) offer no revenue sharing or compensation for the libraries’ investments.
These new terms, we can presume, would not have been acceptable to libraries at the outset. Why should they be now?
Daniel Clancy Says:
I am Dan Clancy, the engineering manager for Google Book Search. I want to respond to some of the points made in this blog post since some of them are inaccurate and others misleading. Before providing a detailed response, I want to emphasize that Google is very supportive of public debate regarding the various terms of the agreement. However, it is important that such debate is factually accurate. We have offered to meet with the OCA to help clarify questions about the settlement, and we are still hopeful a meeting will happen. More information about the agreement is at http://books.google.com/googlebooks/agreement/ and, if you are an author, publisher or other potential member of the class, you should check out the Class website at http://books.google.com/booksrightsholders/. Now, on to a specific response…
Google spent a great deal of time with a number of our library partners to get their input throughout the settlement negotiations. The input of the libraries was extremely valuable, and the result is an agreement that offers a number of benefits to library partners, even compared to our current agreements with them. Our current agreements are actually quite simple: the libraries provide Google with books; we scan the books; we then can make use of these scans in our services; and many libraries also receive a copy of the material we scan with them. For in-copyright works, the libraries’ use of these works was limited by copyright law, as was Google’s.
The settlement agreement opens up new opportunities for reading as it provides explicit authorization that goes above and beyond what would be allowed under fair use. The biggest benefit of the agreement is the fact that the large majority of these books will be accessible in the U.S.. By default for out-of-print books, this includes free preview for up to 20% of a book, the ability of a consumer to purchase access to the book, access through an institutional subscription and the public access service. While there are many benefits to libraries, the core product offerings are the biggest. For institutions that choose to subscribe, their users will be able to access all of the books in the subscription at no cost to the individual user. For schools that do not have extensive libraries, this should prove very beneficial and even for schools with large libraries this extends the reach of these libraries. However, for schools that choose not to subscribe, their users will still be able to freely preview books and then can choose to the purchase the book, access the book through a local library which they can find through the Find It in a Library link or access the book through the access points at public and academic libraries free of charge. Currently, the books held in many of these libraries are only directly accessible to members of their community or through interlibrary loan.
These services are enabled by the efforts of our library partners who have preserved and maintained books over many years. As part of our acknowledgement of this critical preservation effort, Google offers library partners two options for their own access over and above the other benefits we are offering. For libraries that want access to the entire institutional subscription, Google will pay for a portion or potentially all of the cost of the institutional subscription based upon the number of books scanned from that library. For our partners where we are scanning a large portion of the library, the subsidy is such that these institutions will likely receive a free version of the institutional subscription. This means that for some universities, Google is absorbing the cost of digitizing their entire collection or a large portion of their collection, and in return their students, faculty, staff, visitors and other members of their community will be able to obtain broad electronic access to a large majority of these books as well as access to books scanned from other libraries. For other partners, this subsidy results in a significant reduction in their cost to obtain the subscription to all of the books we digitize from all partners.
Alternatively, we also are offering each partner an option called the “limited subscription” that will be free to them. A “limited subscription” provides members of the partner institution access to all of the books that we scanned with them that are included in the larger institutional subscription. Both of these offers extend as long as these books are in copyright and being offered as part of the program. Once the books are PD, we provide access for free. So, the simple story is: if Google scans a book from a library and is offering it for sale in an institutional subscription product, then that book will be made available to the students, faculty, staff and visitors for that institution at no charge.
In addition to the benefits described above, our library partners will also receive a copy of the digital files of the books we digitize. The settlement agreement includes explicit authorizations on how they can use these files as well as explicit restrictions. In some cases, these explicit authorizations will result in increased usage due to the certainty that they offer to our partners. While there may have been some uses that our library partners may have considered under our old agreement that are prohibited now, in most cases the need for these uses is replaced by the subscription service that Google is offering them at no expense. Furthermore, as noted above, many of the uses offered by the subscription service would not have been possible in the old regime due to copyright law. In addition, the scope of the copies that are returned to the libraries is greater. Under our current agreement, a library partner only receives copies of files from books we scan from that library. In the settlement agreement, for partners that go over certain thresholds of scanning, they will be able to receive copies of books that we scan from other libraries. This ensures that multiple parties will have copies of these files to preserve for posterity. As always, it is important to note that our library partnerships are non-exclusive. This agreement does not restrict a library in any way with respect to other digitization initiatives that they may consider.
While there are a number of other benefits to library partners, I’d like to respond specifically to some of the claims that are made in the post. For this portion, I will quote the statements from the original post and then respond.
“According to the proposed settlement, however, only “fully participating libraries” will continue to receive copies from Google.”
The settlement outlines different ways library partners can participate in the agreement. This is to accommodate the different requests from partners because the specific interests of our partners are often very different. Some of our partners are not as concerned about having a copy of a file returned to them, and as such the settlement agreement allows for a library to be a “cooperating library” if they are not interested in receiving a copy of the files. This is an option under the agreement – not a negative as the post implies.
“All other categories of libraries will no longer receive copies in exchange and, to make matters worse, they will have to destroy the digital copies of in-copyright books they already possess or otherwise expose themselves to the implied threat of a lawsuit from authors and publishers over copyright infringement.”
Libraries participation in the settlement is entirely voluntary. If a library does not want to have an agreement with the Registry, then they are under no obligation to do anything – including with in-copyright files. If a library partner does not wish to participate in the settlement, Google will follow through on all of our commitments to return books we have already scanned and will continue to scan public domain books with the partner. As for the “implied threat of lawsuit,” I’m not sure what you mean. Before the settlement the libraries were relying on their rights under the law to receive copies from us. After the settlement, the same is true unless the library chooses otherwise. There’s no implied threat of a lawsuit that comes from the settlement.
“they must also guarantee the security of their digital copies as laid out in a 17-page “security standard,” under the threat of fees up to $7.5 million for security breaches.”
This portion of the post is also misleading. The agreement includes a security specification that Google and the libraries must abide by. The details of this specification were agreed upon with a number of our library partners and are designed to be quite flexible and encompass a wide variety of good security practices. Each library will be able to define their own procedures in their security specification. In most cases, Google and our current partners already provide security measures that exceed the requirements of this specification. Under the current agreement, Google and the libraries face unlimited potential liability including potential statutory damages which bear little relation to any actual damage suffered. These sections of the agreement limit the damages available against a partner to reasonable amounts. The number cherry-picked for this post is the highest of those amounts and is only applicable if the library were to negligently or willfully breach their security specification AND do more than $7.5M of damage to rightsholders. Caps on damages never increase the damages payable.
I would also like to briefly respond to each of the points in the closing paragraph (original post in italics)
“By dangling the threat of future lawsuits, the settlement seems to be manipulating libraries into new legal agreements”
We don’t know of any threats of future litigation, and our existing partners have not been manipulated. As I said earlier, the agreement provides them with a choice that is theirs to make.
“that 1) renege on the benefits of former agreements;”
Google is living up to the commitments in all of our existing agreements for books already scanned and these agreements give both Google and the library the ability to choose which books to scan.
“2) eliminate the digital copy for many libraries;”
Any library that wishes to get a copy of the books we scan with them can do so and can also receive copies of books scanned from other partners that are held in their collection.
“3) impose excessive restrictions on how remaining digital copies are exchanged;”
As discussed above, the restrictions come with authorizations that we believe represent an improvement on the status quo in terms of libraries actual use of the files.
“4) impose harshly punitive security obligations;”
The security obligations are no greater than what the libraries already do, and the agreement limits liability as opposed to expanding it
“and 4) offer no revenue sharing or compensation for the libraries’ investments.”
It is true that, just like our original agreements, there is no direct revenue sharing in the agreement. Nor is there direct contribution by the libraries to Google’s costs. Instead, Google makes a large investment in the digitization of these files to provide many benefits to our library partners, not the least of which is furthering their missions of increasing access to the books including offering free access to the digitized versions to their entire campuses. Perhaps if libraries were for-profit corporations a different deal might have been desirable: one which put money into their pockets and did less public good. We don’t believe the OCA members would have liked that deal with the OCA and similarly we believe we have struck the right balance in this deal for our library partners.
Having said all of this, I would like to talk a little bit about why I personally am excited about the agreement. I have been involved with Google Book Search for close to four years now. I came to Google from NASA Ames Research Center in part because I was excited about the mission of the Google Book Search project and the impact that it could have on readers. I believe that there is a treasure trove of valuable information in the millions of books that have been preserved by libraries around the world and that it is important for this content to be discoverable and digitally available. I know the OCA members and leadership share this passion, and I have consistently applauded the OCA and others’ who are working to make this valuable content more accessible. While indexing and search is very beneficial and leads to increased discoverability of books, in the end, most users want to access the books once they discover them, and they want this access to be seamless. Today in Google Book Search, over 70% of the books we have scanned are in “snippet view.” While Google is confident we would have prevailed in the lawsuits, we still would have been left with the vast majority of these books inaccessible to users with no clear path to unlocking them. I personally am excited about this agreement because it unlocks an incredible number of books for readers to read and helps to realize the dream of increased access to information.
Karen Coyle Says:
Dan, I have to reply to your post, because I believe it does not reflect the reality of the library world. Or, said another way, you are looking at this from the point of view of a for-profit organization, and that’s not a view that includes libraries.
To begin with, you have ignored the question of “fair use.” This is a key aspect of the copyright law that allows the public to make use of copyrighted materials without the permission of the copyright holder. By placing the digital copies behind a subscription service and regulating access and use, fair use (that is, use that is uncontrolled) is not an option. You can be generous, you can be “fair,” but you cannot be the copyright law. This is key because it is the essence of the balance between the commercial interests of the authors and publishers, and the public’s interest in access to the intellectual output of our culture.
As non-profit, educational institutions, libraries enjoy broad fair use rights. They also make use of the “first sale” doctrine. Users can read entire works held by the library without any payment to the rights holder. There have been calls from the rights holders to eliminate this “privilege,” to require libraries to pay to lend books and other materials. Every move in that direction endangers the balance between the rights of the copyright holder and the rights of the public to freely (as in free speech) access information and culture. Without first sale, rights holders retain control over their published works in a way that could easily lead to discrimination and censorship. Open access to materials in libraries is the only defense we have against that. Among other things, this means that items in libraries cannot be withdrawn by the rights holders; no one can go back and revise history. It is this commitment to the public that makes libraries invaluable. With a system in place where everyone pays to view, and where rights holders can potentially recall their works, the rights of the public are no longer being met. Yet, I can easily imagine some cash-strapped communities deciding that they can eliminate their library and just provide access to the Google Book service. That hits me in the gut like a big gulp of 1984.
Next, you ignore section 108 of the copyright law. Section 108 allows libraries to make copies of items in their collections under certain circumstances. Primary ones are: 1) to replace a deteriorating item that is no longer available in the marketplace and 2) to serve disabled users. Both of these are listed as allowed under the agreement with AAP — that libraries can use their copies of the digitized items for these purposes. This bothers me because it puts into a contract something that should be left to copyright law. I’m sure that the library partners asked for this, but that doesn’t mean it’s a good idea. Libraries are already partners with the public, and copyright law governs that relationship. As libraries have moved from hard-copy materials that they purchase to contractual relations with information providers, they have been very careful to try to retain the rights of Title 17 (the copyright law). I see this contract as another erosion of those rights, which means an erosion of the rights of libraries to serve the public as copyright law intends.
You claim that the security requirements are not beyond what the libraries do today. You may have been talking to an extraordinary group of libraries, but I can tell you that most libraries cannot afford the expense of that kind of security. Yet by setting the bar high in your agreement, other library projects may be judged by that same measure.
As for A. Writer, the assumption that the copies made were an “illegal activity” is one that must be addressed. In making its agreement with libraries, Google clearly stated that it considered the digitizing of books for the purpose of searching to be fair use. It is not unlike the copying that Google does of Web resources in order to index those. The libraries themselves (and you can find this in their press releases and FAQs, or you can look at my blog post) determined, with the help of their legal counsel, that the copying was within fair use, as did some outside experts, such as Larry Lessig. In fact, some libraries have active digitization projects of their own going on. The difference, of course, is that Google is an obvious for-profit and would in some way be earning revenue from the digitized books, even if only for the searching. Had libraries been able to do the digitization themselves, and had they provided search capabilities to their collections, the rights holders may not have felt that they had a case. But you cannot deny that the libraries would not have participated if they had judged the copying to be illegal. And as for liability, as you know there is limited liability for state institutions (as compared to the private ones). So it turns out that the biggest liability in this project was the partnership with Google.
Your interest, of course, was in your particular project, but libraries in our country are an organic whole, with the actions of a few affecting all. We use the same standards, we share our resources through inter-library loan, we make broad agreements that benefit the entire community. This is totally unlike the for-profit world where each company looks after itself. We don’t do it that way in libraries, or at least we haven’t done so up until now. I consider the Google partnership with the libraries to be dangerous because it “commercializes” library materials. I know that libraries are impoverished, slow moving, while Google is rich and quick. I would love for libraries to be rich and quick. But in no way do I want them to take on the assumptions or point of view of a for-profit approach to information. Our society would lose so much if that were to happen.