Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.

The Free Republic Problem: Markets in Information

Miscellaneous Free Republic
Source: A Yale Law School Conference Paper
Published: Unknown Author: Yochai Benkler
Posted on 12/07/2000 02:52:16 PST by ICU812

The Free Republic Problem: Markets in Information Goods vs. The Marketplace of Ideas

1. The Free Republic Problem

Here's one for freedom of the press. The Washington Post and the L.A. Times are trying to get the government to prevent some readers from debating the political implications of their stories on a web-based forum. Well, not quite, but too close for comfort. The website is called Free Republic. It includes a forum where conservatives share news clippings and exchange opinions on line. Users who read articles they think deserve comment cut and paste them onto the forum. They then post a comment, and other users participate in a threaded discussion of the article.¹ In October 1998 the Washington Post and the L.A. Times decided that public discourse may well be a good thing, but not when someone else uses their stories to evoke it. So they brought a copyright action to prevent the users of Free Republic from posting the newspapers' stories to their political forum.²

The Free Republic suit crystallizes a pervasive tension between property in the information economy and the freedom to exchange ideas in the information society.³ This tension has two core components.

First, enforcing property rights in information goods requires government to prevent its citizens from using information that they want to use in ways that they want to use it.⁴ This is not a regrettable side effect of property rights. It is in the very nature of the institutional design of property rights in information. In order to make possible a market in information goods, government must delineate the boundaries of what is the "unit" of an "information good," and must commit to prevent, at the behest of the owner, anyone other than the owner from using that "unit" of information. This commitment to prevent some beneficial uses of the information is what makes possible a market in information goods, because permission to engage in the prohibited uses serves as the "goods" sold in that market.

Second, the choice presented by the Free Republic case involves not merely a local choice between two speakers, but a systemic choice between two types of speakers. Holding for the newspapers will benefit speakers like the newspapers-professional, commercial producers who seek to appropriate as much of the benefit that their production creates as possible. Conversely, it will make more difficult amateur discussions like those on the Free Republic forum, which require access to some existing information as an input into their own conversation but do not seek to appropriate the social benefits that their conversation creates. Though neither side would necessarily cease operation if it lost the suit, the dispute nonetheless requires a choice between two very different types of information production, with very different incentives, and perhaps most importantly from a normative perspective, very different types of content.

I suggest that (a) the choice represented by the Free Republic case is not only between having more or less information in the long run, but also about what kind of organizations will be producing information in the long run; (b) this choice about what type of organizations or individuals will be producing information is present whenever we shift the respective boundaries of the enclosed and public domains; (c) people and organizations who produce information for different motivations and with different organizational constraints are likely to produce different types of information content; and (d) therefore decisions about property rights in information must be accounted for in terms of their effects on the patterns of, strategies for, and ultimately the content of information production in our society.

To illustrate how enclosure generally has the normative implications I suggest it does, consider the following hypothetical built around the Free Republic case, but in the context of the newly enacted anticircumvention provision of the Digital Millennium Copyright Act ("DMCA").⁵ The Act prohibits circumvention of technological protection measures,⁶ whether or not the person circumventing is otherwise privileged to access the information, and it prohibits the manufacture, importation, or sale of circumvention devices or services. In other words, if a producer of information, like the Washington Post or the L.A. Times, distributes its product encoded so that it can be used only for authorized uses and upon payment, the Act would subject the Free Republicans to liability if they were to strip the news stories of their code in order to exchange the reports with each other and to discuss them. This liability is independent of the copyright liability for copying the stories themselves. Furthermore, anyone who produces, sells, or imports a device or service that would enable the Free Republicans to strip the code to get at the information would also be liable.

So now imagine the following. Presented with the unsavory prospect of enjoining politically-minded individuals from sharing and discussing news stories, the judge in the Free Republic case finds that the Free Republicans' use is privileged. Say the judge calls it a fair use-the use is noncommercial, the value of the work is primarily in its fact or opinion content, not its expressive form, and while the copying is of the entire story, it's economic impact is deminimis. (This is a hypothetical, after all.) The upshot of the case is that the Free Republicans are privileged to copy stories from the newspapers, and post them on their forum to instigate discussion.

Now say that the newspapers encoded the stories to limit access or reuse. The DMCA prohibits the act of circumvention per se, not the act of circumvention for the purpose of making an infringing use. Should the Free Republicans circumvent the code to reach or post the stories, under the DMCA they would be independently liable to the newspapers for that circumvention, even though they had a court decree saying that the underlying use motivating the circumvention is privileged. Note well here that it is law, not technology, that is preventing the Free Republicans from using the news stories. There may be off-the-shelf software available to circumvent the protection measure (say, available in Finland). By installing the code in their stories, however, the newspapers have invoked the legal prohibition on circumvention, and immunized their control over their stories from "fair use" or any other use privileged under the copyright law. The upshot is that the Free Republicans cannot use the information they need, and they cannot do so because of a legal prohibition that Congress enacted to implement its policy judgment that the benefits of this prohibition to the long-term production of information outweighed the costs in terms of access to existing information.

We could take the story one level further, retaining the underlying analytic structure. Imagine that the Librarian of Congress, pursuant to her administrative charge in the DMCA,⁷ exempts from coverage under the Act circumvention for the purposes of sharing news stories on web-based threaded discussions. She might, for example, think that such copying harms newspapers very little, but is crucially important to the viability of the discussion forum medium, which in turn is crucial to enable robust, decentralized political commentary and discourse. The DMCA requires the Librarian, upon making such a determination, to exempt the endangered uses from the ambit of the anticircumvention provision. But the Act bars such administrative determinations from being used as defenses in any action for violating the prohibition on manufacture, sale, or importation of circumvention devices.⁸

So now imagine the Free Republicans, with their court decree stating that what they are doing is a fair use, and with an administrative determination that they are free to circumvent the newspapers' code. They surf to the Washington Post's website, and they download a story. They try to copy it onto their website, but the code still prevents them from doing so. They hunt around in vain for some software or service to strip the offending code off the information that they are privileged to access. But they can find none. The devices they need to gain access to the information that they are privileged to access are outlawed. The Free Republicans have encountered an unusually crisp instance of a restraint on their freedom of the press. Not freedom of "The Press," the organized business of producing information about current events; but the freedom of the press of John Lilburne and Tom Paine; the freedom to use the devices necessary for effective communication of one's thoughts.

What this story helps us to do is to grasp quite intuitively the constraining role that law plays when it defines property rights in information. It should be fairly simple to see how the antidevice and anticircumvention provisions are laws intended to impede access to and use of information. They are, perhaps, enforced for very good policy reasons that have to do with securing long-term production incentives for information producers. Whether they are sound policy choices or not is not, however, the issue. The point is that the anticircumvention and antidevice provisions are congressional laws intended to and operating in the arena of access and use of information, and they are laws that operate in this arena by prohibiting people from accessing information or communicating it to others. As such these laws are no less a congressional action subject to first amendment scrutiny than a rule that seeks to enhance diversity in video programming by prohibiting cable operators from selecting all of the contents of programming carried over their cable system. In media regulation we have seen many congressional actions with parallel features to those of the anticircumvention provision that, upon reflection, courts have accepted as justified. The point is, though, that the courts uphold these media regulation laws only after subjecting them to exacting review. The anticircumvention provisions should no less be subject to that exacting review.

The point is not limited to the anticircumvention or antidevice provisions. For the story and the role of law in it can be rolled back up the story line. The judge's decision that the Free Republicans were making a fair use of the stories is what privileged their use in the first place. But an opposite decision would have prevented them from making that use, just as would the anticircumvention provision. Had the judge enjoined the Free Republicans' use, her decision would carry out another congressional law passed in furtherance of the public interest in information production, and implemented by preventing the Free Republicans from using the stories, in order to secure to the newspapers' investment incentives. This other congressional law is, of course, the Copyright Act of 1976. This institutional attribute is pervasive: property rights in information goods are laws that operate with the intent to regulate information production, and that operate through institutional mechanisms that consist in preventing many people from reading or using information that, but for the law, they could have read or used. This kind of law is the kind of law that in other areas, most prominently in media regulation, our constitutional framework has approached with an unusually high degree of caution. The same normative commitments that warrant caution in those cases ought to guide us in assessing rules that enclose information, even when these rules are formulated as "property rights".

3. The Free Republic Problem and the Distribution of Capacity to Make Meaning

Understanding enclosure of the public domain as raising "first amendment" type concerns helps to locate the normative parameters within which we ought to approach the assessment of property rights in information goods. It suggests that we should think (perhaps not single mindedly, but seriously) about how these rules will affect the quality of political discourse and the capacity of individuals to express themselves.

3.1. The Harper & Row Response

The standard response to the apparent conflict between property in information goods and what might be thought of as "first amendment values" is the one adopted by the Supreme Court in Harper & Row.⁹ By making information production into an economic activity, so this argument goes, we increase the production of information. More information is better; and it is "better" from the perspective of the values of robust political discourse and free expression. This quantitative version of the argument suffers from the conceptual difficulty that it conflates production of "information goods" with production of information that affects the quality of political discourse-let's call it "politically relevant information."

Information goods are legally and technologically contingent upon those instances of the social process of making meaning that can, as a practical matter, be subject to exclusion by a defined owner, and hence can be the basis for a market. They are a medium-constrained category. Politically relevant information is the product of a culturally contingent process of exchanging symbols that causes us to hold one or another set of beliefs about how we should organize our lives as individuals or as a polity.¹⁰ It is a culture-constrained category. There is no necessary correlation between the cultural contingency that defines what information is politically relevant, and the technological-legal contingency that constrains what information can be demarcated and sold as goods. We therefore have no systematic reason to think that rules put in place to take advantage of technologically contingent tollbooth opportunities will encourage the production of information defined as "important" by its culturally contingent political valence.

The more sophisticated version of the Harper & Row argument¹¹ suggests that information goods produced in reliance on market incentives are qualitatively better as sources of "politically relevant information" than information produced by alternative sources because they are produced (a) free of the bonds of patronage and government, the two alternative sources of funding for information producers; and (b) in response to the interests and needs of as wide a social group as is the market, and hence contain what many people see as information relevant to them. It is to this sophisticated version of the Harper & Row defense of property in information that recognizing the choice represented by the Free Republic problem-the choice among different strategies used by information producers-is most relevant.

3.2. Effects of Enclosure on Information Production Strategies

I have elsewhere developed in more detail a model that outlines the effects of changes in property rights in information on the strategies used by information producers.¹² The core claim is this. Many people in society engage in information production. They produce and exchange symbols for a variety of reasons, and using a variety of strategies to appropriate the benefits of their production. While a more detailed typology is possible, it is important to understand three major divides in the extant distribution of strategies used by information producers in our economy. First, there is the division between commercial and noncommercial producers. Unlike in the production of wheat or automobiles, the noncommercial sector is of crucial importance in our information production system. Universities, government (through both direct investment in government labs or research institutes and indirect investment through NSF grants and the like), political and civic organizations, amateurs, are all integral and crucial components of how we produce information on both large and small scales. Second, there is the division within the commercial sector between those who appropriate the benefits of their investment by relying on property rights, and those who appropriate the benefits of their production by relying on early access to the information (like wire services) or by giving the information for free and appropriating the benefits through relationships created around the information (like lawyers who write for the National Law Journal). Third, and finally, there is the division, within those who sell "information goods" in reliance on rights, between those who produce information on a small scale, like individual authors, and those who integrate new production with ownership of large existing-information inventories, like Disney or Time-Warner.

The core point to understand about property rights in information is that they have different effects on these different strategies, such that changes in the institutional content of property rights can help some of these strategies at the expense of others. In particular, increases in the scope and reach of property rights benefit commercial producers who sell information goods, at the expense of noncommercial producers and producers who appropriate the benefits of their production by means other than sale of rights. The primary culprit forcing this tradeoff is the rather unique attribute of information, that it is both input and output of its own production process. Increases in appropriability therefore are generally thought to cause a simultaneous increase both in expected ability to appropriate the value of outputs (and hence increased incentives) and in input costs. By endogenizing the strategies used by information producers, my model underscores that while expansion of rights increases the input costs of all producers, and thereby increases the production costs of all, such expansion benefits, and hence increases production incentives of, only producers who sell information goods in reliance on rights. Expansion of rights increases the returns to commercial producers who sell permission to use, in part by raising the costs of other producers who do not sell rights-either because they are noncommercial, or because they appropriate the benefits of their information production by means other than sale of permission.

Moreover, increases in the scope and reach of property rights favor large-scale organizations that own information inventories over small-scale organizations (including individuals) that do not own such inventories. The mechanism for this effect is as follows. Input costs increase because of the need to buy more inputs-more uses of information-that previously were not subject to anyone's right-in other words, were available from the public domain at their marginal cost of zero. An organization that owns a large inventory of existing information can respond to the loss of public domain inputs by intensifying reutilization of its owned inventory, and will do so to the extent that its inventory provides even rough substitutes for information inputs otherwise available only by purchase from others. Since the marginal cost of reutilization is zero (information is nonrival), but the (supply) price at which information is available from the market must be positive, a firm will see lower costs to utilization of owned inventory even when the social cost of using intrafirm resources is identical to the social cost of utilizing market-purchased inputs. This is so because of the nonrivalry, independent of and cumulative to any transaction costs effects. Organizations and individuals that do not have such an inventory do not have the reutilization option, and are forced to cover from the market information inputs no longer available from the public domain. This increases their input costs more rapidly than the rise in input costs of large-scale vertically integrated organizations.

In other words, when Congress passes a statute like the Digital Millennium Copyright Act, the Sonny Bono Term Extension Act, or the proposed database protection legislation, it is making a choice among types of information producers. It is choosing to increase the costs of academic scholars whose libraries must choose between buying more publications and buying more access rights to a smaller number of publications, in order to increase Reed Elsevier's returns. It is choosing to increase the costs of amateurs, like children who would put together web-based projects about their favorite cartoon characters, for example, in order to increase the returns to Disney. It is choosing to raise the economic barriers facing participants in the Free Republic forum in order to increase the returns to the Washington Post.

Second, unless the scale economies of information goods production (large first copy, small marginal per-copy costs) change radically, commercial producers will systematically "aim for the middle," or for those products that will gain the most attention from the widest group of consumers. This is so for reasons familiar in the literature that has long criticized our polity's excessive reliance on commercial broadcasting.13 This effect will likely be somewhat mitigated in an environment that permits an infinite number of "channels," because it removes one scarcity that limits the ability of companies to focus on niche markets. Nonetheless, other scarcities, like human capital on the supply side and attention on the demand side, suggest that a commercial product will be produced only if it has an audience large enough, given the intensity of its members' preferences, to pay for the product's development and production costs. A noncommercial product may be developed simply as an expression of its maker's thoughts, but more importantly, in anticipation of capturing the attention of all those who would be interested enough to spend the time reading or listening, but not the money needed to buy. It is the diversity of a far-flung Speakers' Corner that is lost when we design our laws to benefit commercial at the expense of noncommercial speakers.Third, if in fact enclosure benefits large-scale organizations at the expense of small scale producers, then the Harper & Row argument must maintain that, from a democratic values perspective, information produced by such large-scale organizations is preferable to information produced by small-scale producers who do not own information inventories. There are two types of reasons to think the contrary.

First, what makes large-scale organizations more robust in a dynamic increasing-rights environment is their ability to access and reutilize large pools of pre-existing information that they own. This means that increasing rights give these organizations reasons to search for ways to reuse information they own because they own it, not because it is the most suitable information to make the best new product given available human capital. What one would predict, in effect, is increased output of the Lethal Weapon IV, Mickey Mouse band aids, and "best of" or blooper show variety. Organizations that do not own inventories see all available inputs of pre-existing information at their (above marginal cost) market price, and will not likely have a similar bias. Since producers will distort their production choices relative to consumer preferences based on the contents of their owned inventory, and since they will try to find the recombinations that will appeal to the largest segments possible, the probability that a consumer will actually be able to find his or her first-best preferred good is relatively low. Putative consumers faced with this behavior will therefore not invest in producing a first-best preference ordering about their interests in the first place. Instead, rational consumers will wait for producers to "invent" a menu and only then invest in deciding as a second-best solution what on the available menu is their preferred offering.¹⁴ This rational absence of consumer first-best preferences weakens the Harper & Row argument, that commercial producers effectively serve diversity because they well respond to consumer preferences.

Second, a democratic deficit emerges when a small number of organizations or individuals gain too large a share of the power to control the content of the information available to the polity. Again, this is an aspect commonly found in the literature criticizing commercial concentration in broadcasting,¹⁵ and is exemplified by the reported suppression of a news report about security failures in Disney World prepared for 20/20, the news magazine aired on Disney's subsidiary ABC.¹⁶ An information production system composed of many small-scale producers, commercial or noncommercial, will have less of a tendency to be dominated by a small set of political or other agendas than one populated by a small number of large-scale organizations, whether commercial or noncommercial, that centralize production of large portions of what occupies the attention of a society's constituents.

4. The Free Republic Problem

The Free Republic problem is a problem that all market-based democracies must face when they think of information policy. To make markets in information goods possible, governments must undertake to prevent many of their constituents from reading and speaking in ways that would be beneficial and feasible to these constituents, but that must be curtailed if the state is to serve the public interest by creating a market in information goods. Such a choice to sacrifice the individual interests in using information in order to further a social policy goal of making markets in information goods should raise serious normative concerns in liberal democracies. These concerns are compounded when the means of pursuing the social good of information markets appear to foster the development of large-scale commercial information producers who integrate new production with ownership of owned-information inventories, at the expense of noncommercial producers, both professional and amateur, and of small-scale independent commercial producers.

The practical consequence of recognizing the Free Republic problem is to recommend that we adopt a studied skepticism towards the adoption of new rights in information goods or other new restraints on users' capacity to access information. New rules that enclose the public domain must be put to a fairly rigorous test to show that they will not have the undesired effect on freedom of speech and diversity of discourse. Given the central role of large-scale organizations in our present information infrastructure, and the relatively expansive scope of rights already available to them, this is likely to prove a formidable task indeed.

1 See www.freerepublic.com, browse to Forum.

2 See Pam Mendels, Newspaper Suit Raises Fair Use Issues, NYT, Cybertimes, October 2, 1998.

3 In focusing on this tension, this paper follows Rosemary J. Coombe, Objects of Property, Subjects of Politics: Intellectual Property Laws and Democratic Dialogue, 69 Tex. L. Rev. 1853 (1991); Niva Elkin Koren, Cyberlaw and Social Change: A Democratic Approach To Copyright Law In Cyberspace, 14 Cardozo Arts & Ent. L.J. 215 (1996); Neil Netanel, Copyright and A Democratic Civil Society, 106 Yale L.J. 283 (1996). James Boyle, A Politics of Intellectual Property: Environmentalism for the Net? http://www.wcl.american.edu/pub/faculty/boyle/intprop.htm. A more complete statement of my argument here is Free As the Air to Common Use, First Amendment Constraints on Enclosure of the Public Domain, forthcoming 73 NYU. L. Rev. No. 2, May, 1999.

4 Treatments of this speech-restricting character of copyright and its potential conflict with the first amendment's speech clause include Diane Leenheer Zimmerman, Information as Speech, Information as Goods: Some Thoughts On Marketplaces and the Bill of Rights, 33 Wm & Mary L. Rev. 665 (1992); L. Ray Patterson & Stanley F. Birch, Jr., Copyright and Free Speech Rights, 4 J. Intell. Prop. L. 1 (1996); Mark A. Lemley & Eugene Volokh, Freedom of Speech and Injunctions in Intellectual Property Cases, 48 Duke L. J. 147.

5 Pub. Law 105-304 § 103, enacting new 17 U.S.C. § 1201 (1998).

6 The term refers to a variety of techniques that permit providers of information in digital form to use self-help to regulate access to their products. These measures can perform a range of functions, including gathering information about every use of a digitally-encoded piece of information as well as limiting or altogether preventing its use. See Mark Stefik, Shifting the Possible: How Trusted Systems and Digital Property Rights Challenge Us to Rethink Digital Publishing, 12 Berkeley Tech. L.J. 137, 138-40 (1997).

7 The Librarian must conduct a rulemaking on the record two years after enactment of the Act, and every three years thereafter, to determine whether enforcement of the provision as to any class of works will adversely effect the ability of users to make noninfringing uses of such works. Pub-L. 105-304 §103, enacting 47 U.S.C. §1201(a)(1) (B)-(D).

8 Id., enacting §1201(a)(1) (E).

9 Harper & Row, Publishers, Inc., v. Nation Enters., 471 U.S. 539 (1985).

10 By referring to organizing our lives as individuals, as well as a polity, I am purposefully adopting an expansive sense of the term "politically relevant information" to include in effect the practices of making meaning that structure our understanding of what options are open to us both as autonomous individuals and as self-governing citizens, and our valuation of these various options.

11 A stark statement of this defense is Paul Goldstein, Copyright's Highway, From Gutenberg to the Celestial Jukebox 232 (1994). Another, more nuanced version is Netanel, supra, at 342-62.

12 See Yochai Benkler, Intellectual Property and the Organization of Information Production (MS 1999) (available http://www.law.nyu.edu/benklery/Ipec.pdf).

13 See. e.g., C. Edwin Baker, Giving the Audience What It Wants, 58 Ohio St. L.J. 311 (1997); Jerome A. Barron, Access to the Press--A New First Amendment Right, 80 Harv. L. Rev. 1641 (1967); Owen M. Fiss, Why the State? 100 Harv. L. Rev. 781 (1987).

14 For a more detailed description of this problem as it arises in broadcasting see Yochai Benkler, Overcoming Agoraphobia: Building the Commons of the Digitally Networked Environment, 11 Harv. J.L. & Tech. 287, 365-68 (1998).

15 See sources cited supra, note 13; see also C. Edwin Baker, Private Power, the Press, and The Constitution, 10 Const. Comment. 421 (1993).

16 See Lawrie Mifflin, An ABC News Reporter Tests the Boundaries of Investigating Disney and Finds Them, N.Y. Times, Oct. 19, 1998, Section C., p. 8, col. 1.

1 Posted on 12/07/2000 02:52:16 PST by ICU812
[ Reply | Private Reply | Top | Last ]