Google hopes to scan some of the largest library collections in the world and make them searchable online. This ambitious goal is intensely controversial. Some contend that “Google Book Search” could be a valuable research resource and boost book sales, while others believe it violates copyright laws. The Authors’ Guild and the Association of American Publishers, for example, have both sued Google for copyright infringement. Does Google Book Search constitute “fair use” of copyrighted materials? Will this new technology create winners and losers? Who will they be? This Joint Center conference investigates how the push to digitize printed information challenges copyright law and intellectual property rights, focusing specifically on the potential costs, benefits, and legal repercussions of the current controversy.
Registration and Lunch
Robert Hahn, AEI-Brookings Joint Center
Douglas Lichtman, University of Chicago
| || || |
Hal R. Varian, University of California, Berkeley
| || || |
Google hopes to scan some of the largest library collections in the world and make them searchable online. This ambitious goal is intensely controversial. Some contend that “Google Book Search” could be a valuable research resource and boost book sales, while others believe it violates copyright laws. The Authors Guild and the Association of American Publishers, for example, have both sued Google for copyright infringement. Does Google Book Search constitute “fair use” of copyrighted materials? Will this new technology create winners and losers? Who will they be? At a February 24 AEI-Brookings Joint Center conference, panelists investigated how the push to digitize printed information challenges copyright law and intellectual property rights, focusing specifically on the potential costs, benefits, and legal repercussions of the current controversy.
Robert W. Hahn
AEI-Brookings Joint Center
Google has created Google Book Search, an ambitious undertaking that scans and digitizes books from the world’s largest libraries, making books searchable online. If this effort succeeds, people anywhere in the world will be able to search the contents of a large number of books. Google plans to make 25 to 30 million titles available. Yahoo! and Microsoft have a similar project in the works.
Google Book Search has two key components: the Print Publisher program and the Print Library project. Under the Print Publisher program, publishers authorize Google to make the full text of a copyrighted book available within the Google search database. Users can view the entire page of a book where the search word or string appears, as well as a few pages before and after. This program is voluntary in the sense that publishers must “opt in” before Google will show a full page in response to a search query.
In contrast, under the Print Library project, Google does not seek express consent from the copyright holder, but shows only snippets of search results rather than full pages of books. Copyright holders must “opt out” if they do not want their works to be included in the Google Print Library project.
The Print Library project will scan books from Harvard, Stanford, Oxford, the University of Michigan, and the New York Public Library and index their contents. Users will be able to view the full text of out-of-copyright books, but only snippets of text containing their search query for books still under copyright. A snippet is a few sentences long.
Google Book Search could yield large benefits. The database would act as an online card catalog and book search engine, returning words or phrases where they appear in the text of millions of books. This enterprise, however, has raised concerns--particularly when a book is still under copyright and the copyright holder is easy to identify.
The project will create costs for some of the parties involved, most notably copyright holders. If more companies emulate Google, then rights holders must spend time informing each company of their wishes to be excluded. The Authors Guild also claims that authors are not being compensated for the work they put into creating the content that ultimately allows Google to collect revenue. Additionally, Google is not asking for permission from the copyright holders, while Yahoo! is, leading some to believe that the transactions costs of obtaining permission might not be so prohibitive after all.
A second concern is that Google Book Search might, under some circumstances, reduce the incentive to produce new creative works. A third concern is Google’s security for sensitive information, information that is someone else’s intellectual property. For example, should Google be held liable for a security breach? Should it be asked to demonstrate that it has taken appropriate measures to safeguard the property of the copyright holders? If so, what kind of measures should be required?
How will this program affect economic growth? Economists are typically most interested in how such rules will affect the overall size of the economic pie. While it is not the only question--in particular, we might also care about the distributional consequences associated with different rules or standards--it is an important consideration.
University of Chicago
The controversy can be described as follows: Google is attempting to build Google Book Search without asking permission for each book that it indexes, while copyright holders believe Google should seek approval before using their materials.
Five points are helpful in framing the issues:
1) Nearly everyone agrees that Google Book Search would have real value. If copyright law were to say that Google could proceed without asking permission, then Google would run the project faster, cheaper, and more comprehensively than if alternative means were utilized.
2) Authors need adequate incentives in the long run to create and disseminate work. Laws should not be myopic and should extend beyond the current technology.
3) There are real tensions. Security is an issue. What if the database gets hacked? Some categories of books, like dictionaries and thesauruses, will lose money. Google generates money from the Book Search project, and maybe some of that value ought to be shared with copyright holders. And in either case--whether or not Google must ask--technologies will be advanced. This creates tensions between helping Google and helping books and other technologies that have yet to be revealed.
4) We should recognize that Google is thinking and acting in its own self-interest, but also understand that in some cases, the copyright holder will be impossible to track down. In those situations we can be more sympathetic towards Google.
5) The public debate gets distracted by two flawed arguments. Opting out does not work easily with many providers of information. Google frames an opt-out argument as if Google is the only player. It will be difficult for copyright holders to opt out of all providers’ databases once more entities begin offering similar services.
Some scholars believe that a ruling against Google would invalidate its search engine. This argument is also flawed. There are many ways to distinguish the book search from the Internet search. Internet web pages are already available to be viewed in digital form, for free (in most cases) online, while books are not. Also, the Internet has a built-in digital infrastructure that allows designers to opt out electronically.
University of California, Berkeley
Of the books owned by libraries, about 15 percent are out-of-copyright, 20 percent are in-copyright and in-print, and the remaining 65 percent are in-copyright and out-of-print. Books in the last category are difficult to locate outside of these libraries.
Several factors determine if an action constitutes fair use:
1) Purpose and character, including whether commercial or noncommercial: Google is commercial, but it is selling ads based on the search query, not on the content of the books.
2) The amount and substantiality of the portion used in relation to the copyrighted work as a whole: Google shows the user a tiny amount of the copyrighted work.
3) The effect of the use upon the potential market for or value of the copyrighted work: the book search is not a substitute for the copyrighted work; instead, it may be a complement.
Many existing indices have been declared fair use. In Kelly vs. Arriba Soft, thumbnail images were found to be transformative and hence fair use. The information Google offers is transformed from its original state. Since book search results do not take the place of the original book, they will not cut into sales.
Courts have routinely dismissed charges anytime an index, compilation or catalog has been charged with violating fair use.
Is it still fair use when an entire work is copied to produce the snippets? The courts have determined many times that it is. In addition, people commonly compile lists, directories and catalogs (like TV Guide and Zagat), and these generally do not require permission.
An opt-in model would require Google to find every rights holder and negotiate with him, which is likely to create prohibitive transactions costs. The costs of the opt-out model, by comparison, are small. A well-functioning economic system requires that property rights be assigned in a way that minimizes transactions costs, which the current opt-out process accomplishes.
University of Chicago
If Google is the only firm that will ever make such a large number of copyrighted materials available, then opt out is cheaper. The problem is that the next company to engage in something similar will point to the Google precedent that copyright holders must opt out. It may soon become impossible for copyright holders to monitor the number of places from which they must opt out. Instead, we should build an infrastructure that at least allows authors to give or deny permission to use their works. If authors cannot be found, then fair use arguments are more salient.
The in-print/out-of-print distinction is not going to be an issue in the future, because nearly all new works are already in digitized form and print-on-demand options will be developed for older ones.
For fair use, the law specifies certain factors the court must consider, plus the judge must consider any other information she thinks is relevant. The law allows for other considerations, and we need to think about them.
Authors are not afraid of competition, but authors do not want Google to use the authors’ own work to compete against them.
University of California, Berkeley
Users who seek hard-to-find books that appear in libraries but cannot be found anywhere else will be made better off with the Google Book Search. A judge who is considering what is the right thing to do will recognize that this product will enhance welfare.
Google is not the only player in the market. An easy-to-use, widely accessible, generic rights clearinghouse would make a lot of these problems go away.
This conference summary was prepared by AEI-Brookings Joint Center research assistant Laura Goodman.