Nonrivalry and price discrimination in copyright economics.

• Author: Conley, John P.

INTRODUCTION I. A PRIMER ON PUBLIC GOOD ECONOMICS A. The Critique of Nonexcludability B. The Reconceptualization of Nonrivalry II. A FORMAL MODEL OF PURE PUBLIC GOODS A. Perfect Price Discrimination B. Imperfect Price Discrimination III. SPATIAL COMPETITION/IMPURE PUBLIC GOODS CONCLUSION INTRODUCTION

Legal scholarship on the economics of copyright has largely settled into a debate between two polar extremes. On one side are copyright "neoclassicists," who favor the expansion of copyright protection until it encompasses all of the present and future uses associated with a creative work, as well as reforms that facilitate price discrimination, on the grounds that innovation is best promoted if authors are able to appropriate as much of the value of their creation as possible. On the other side are copyright "minimalists," who favor limiting the number of uses contained within a copyright so that it provides only enough incentive for innovation and who are generally hostile toward reforms that facilitate price discrimination. (1)

Despite the differences in their conclusions, both sides generally frame the arguments in largely economic terms. (2) Indeed, both sides of the debate analyze copyright through the lens of public goods theory, which Paul Samuelson was among the first to analyze with mathematical rigor." (3) A core policy implication of public goods theory is that markets tend to produce too few public goods and underutilize those that are produced. (4) In the context of copyright, the economic analysis has focused almost entirely on the premise that not only are creative works nonrival in general, but also that any number of additional copies can be produced at zero marginal cost. In so doing, the current literature fails to capture the key economic features that give public goods their distinctive characteristics.

Samuelson imagined that there existed an exogenously fixed set of public goods and pointed out that markets would not produce public goods efficiently even if they were fully excludable. (5) He derived an optimality condition, now known as the "Samuelson condition," that requires that the level of a public good be chosen so that the sum of the marginal benefits derived by everyone who consumes the public good equals the marginal cost of production. Thus, satisfying the Samuelson condition in a market context requires determining the marginal benefit that each individual agent obtains from consuming the last increment of a public good. Since agents are charged in proportion to these benefits, they have an incentive to hide or underreport these benefits. Although economists have proposed a wide range of potential solutions to this problem, they have largely failed to produce a practical, incentive-compatible mechanism to induce consumers to reveal the intensity of their preferences. (6)

Reconceiving the application of public goods theory to copyright in this manner provides an important middle ground between the extreme positions that dominate the current debate. On the one hand, by underscoring that some degree of price discrimination is essential for the optimal provision of public goods, a more fundamental understanding of public good economics reveals that copyright minimalists' reflexive hostility toward price discrimination is misplaced. Indeed, it is quite likely that prohibiting or inhibiting authors' ability to price-discriminate will reduce economic welfare.

On the other hand, the more refined conception of public goods theory that we propose in this context challenges the neoclassicists' claim that efficient production of public goods depends on authors' ability to appropriate the entirety of the surplus created by their works. Instead, the Samuelson condition implies that when public goods are divisible (in that producers can vary the quantity of public good they produce), producers need only appropriate the marginal benefit (rather than the total benefit) resulting from further increases in the production of public goods. This means that some proportion of the available surplus remains with the consumer. (7) Our approach thus strikes a middle ground between the polar positions that dominate the debate and provides a basis for analyzing the optimal level of price discrimination. (8)

There are two additional complicating factors that make copyright goods a particularly difficult problem even within the context of public goods theory. First, the types of public goods that Samuelson had in mind when he developed his theory--such as roads and national defense--are divisible, in that one can provide one unit, two units, or any amount of these goods one chooses. Copyright goods, however, tend to be indivisible. That is, one can either write a book, compose a song, make a movie, or not do so. One cannot write half a book or two units of the same book. (9) Half a book (perhaps a book written with half the effort) would be a different copyright good, not half the quantity of the same good. Thus, copyright goods involve a fixed first-copy cost and then close-to-zero cost in allowing additional agents to enjoy subsequent copies, in contrast to the zero fixed costs and positive marginal cost typical of classical public goods.

Second, Samuelson followed classical general-equilibrium theory in assuming that the set of public goods is exogenously fixed. However, the major argument for providing copyright protection in the first place is that protection is necessary to provide incentives to produce new copyright goods. Thus, the set of copyright goods is endogenously determined in equilibrium. Economics has only a primitive understanding of how markets create new goods, but this is an issue that cannot be ignored in the context of copyright.

The interplay of these two factors greatly complicates the problem and makes it difficult to make definitive policy recommendations. Indeed, one reason for the persistence of the debate in law is that there is truth in both sides of the argument. Copyrighted works are often imperfect substitutes for one another. This creates demand interactions that change the analysis in important ways. Imperfect substitution adds a different dimension along which public goods can be underprovided: not only can there be too little of a particular good, but markets may also produce too few types of public goods. Perhaps even more surprising is the fact that if policy instruments are not correctly calibrated, it is also entirely possible that markets will provide too many works instead of too few. (10) Thus, conceiving of the copyright markets as endogenously determining the set of discrete public goods to provide through a process of monopolistic competition undercuts the seemingly intuitive policy inference that markets invariably tend to produce too few works. To the extent that the relevant equilibrium is one in which overproduction occurs, this insight in turn suggests that the tradeoff between incentives for efficient creation and efficient access to those works that are created--the tradeoff that frames most of the scholarship on the economics of copyright--may be fundamentally misplaced.

The purpose of this Article is to help disentangle these effects and clearly highlight the factors that should guide this policy debate. (11) The balance of the discussion is organized as follows: Part I briefly reviews the basic economics of pure public goods and summarizes the shortcomings of the way the debate is currently framed. (12) Part II lays out a basic model of pure public goods and analyzes the role that price discrimination plays in maximizing economic welfare. Part III extends the model by allowing for the possibility of competition between similar products that are imperfect substitutes for one another.

1. A PRIMER ON PUBLIC GOOD ECONOMICS

   Under the classic definition, pure public goods have two defining characteristics. (13) First, they are nonexcludable, which means that the good cannot be provided to one consumer without simultaneously providing it to others. That is, it is infeasible to prevent any agent from consuming whatever public goods are produced. Second, they are nonrival, which means that the consumption of the good by one consumer does not reduce the supply available for consumption by others.

   One oft-cited example of a public good is a lighthouse. Lighthouses are said to be nonexcludable, in that it is generally thought that lighthouse services cannot be provided to one ship without also providing them to others. The inability to internalize these positive externalities is said to lead to systematic underproduction of lighthouse services. Lighthouses are also said to be nonrival in that consumption of lighthouse services by one ship does not preclude other ships from consuming them as well. Another example is a fireworks display. It is difficult to allow one consumer to enjoy a fireworks display without simultaneously allowing others to do so. In addition, the enjoyment of the fireworks display by one person does not generally reduce the ability of others to enjoy it as well.

   As noted earlier, creative works are generally thought to be pure public goods. Absent some form of copyright protection, authors would be unable to prevent one purchaser from copying the work and sharing it with others. Moreover, the sharing of the work with one person does not in any way reduce the supply of the work available for sharing with others.

   1. The Critique of Nonexcludability

      Interestingly, Samuelson did not regard nonexcludability as an essential characteristic of pure public goods. (14) A moment's reflection on the classic examples discussed above reveals why. Consider the lighthouse. Metering access to a lighthouse is not impossible; it is simply very costly. For example, a lighthouse owner could engage a flotilla of ships to intercept other ships and ensure that only those that had paid for the lighthouse services were able to enjoy them. Alternatively...