legal theory: law and economics

Economic Analysis of Intellectual Property


The term "intellectual property" refers to a loose cluster of legal doctrines that regulate the uses of different sorts of information. The law of copyright protects "original forms of expression" -- novels, movies, musical compositions, computer software, etc. Patent law protects inventions -- machines, processes, (also) computer software, etc. Trademark law protects words and symbols that identify for consumers specific goods and services -- brands of cereal, clothing, automobiles, etc. Trade-secret law protects information that companies have tried but failed to conceal from their competitors -- soft-drink formulas, confidential marketing strategies, etc. The "right of publicity" protects celebrities' interests in their images and identities.

In recent years, a growing number of legal theorists have attempted either to make sense of this complex field or to propose ways in which it should be reformed. Most of these efforts have taken one of four forms:

The first and second of these approaches draw self-consciously upon well defined traditions in political and moral philosophy. Accordingly, they are discussed in a separate essay on "Philosophic Perspectives on Intellectual Property." The fourth of the approaches falls comfortably within the capacious family of arguments we have described as "economic analysis of law." Set forth below are brief discussions of its central claims and the problems its proponents have encountered. If you are curious concerning the origins of these various theories or the ways in which they have figured in recent judicial decisions, you may wish to consult the larger essay from which these materials have been adapted: "Theories of Intellectual Property." (###)

I. Varieties of Economic Analysis

The large majority of the writers who have attempted economic analyses of intellectual property have relied, explicitly or implicitly, on the "Kaldor-Hicks" criterion (also known as the "wealth-maximization" criterion or "potential pareto superiority") which counsels lawmakers to select a system of rules that maximizes aggregate welfare measured by consumers' ability and willingness to pay for goods and services. They disagree sharply, however, concerning the implications of that criterion in this field. Three quite different arguments -- commonly thought to be incompatible -- dominate the literature.

1. Incentive Theory. The first and most common of the three tacks argues that the optimal doctrine is the one that maximizes the difference between (a) the present discounted value to consumers of the intellectual products whose creation is induced by holding out to authors and inventors the carrot of monopoly power and (b) the aggregate losses generated by such a system of incentives (the consumer surplus sacrificed when authors and inventors price their creations above the marginal costs of producing them, the "administrative costs" of interpreting and enforcing intellectual-property rights, etc.) In rougher terms, incentive theory urges a lawmaker to establish or increase intellectual-property protection when doing so would help consumers by stimulating creativity more than it would hurt them by constricting their access to intellectual products or raising their taxes. Here are two illustrations of this general approach, both from the pens of William Landes and Richard Posner:

2. Optimizing Patterns of Productivity. Many years ago, Harold Demsetz argued that the copyright and patent systems play the important roles of letting potential producers of intellectual products know what consumers want and thus channelling productive efforts in directions most likely to enhance consumer welfare. In the past decade, a growing group of theorists have argued that recognition of this function justifies expanding the copyright and patent systems. In Paul Goldstein's words:

The logic of property rights dictates their extension into every corner in which people derive enjoyment and value from literary and artistic works. To stop short of these ends would deprive producers of the signals of consumer preference that trigger and direct their investments.

Won't adoption of this strategy impede public dissemination of intellectual products? Not at all, say the proponents of this approach. Sales and licenses will ensure that goods get into the hands of people who want them (and are able to pay for them). Only in the rare situations in which transaction costs would prevent such voluntary exchanges should intellectual-property owners be denied absolute control over the uses of their works -- either through an outright privilege (like the fair-use doctrine) or through a compulsory licensing system.

3. Reducing Rent-Dissipation. The final approach is related to but differentiable from the second. Its objective is to eliminate or reduce the tendency of intellectual-property rights -- and patent rights in particular -- to foster duplicative or uncoordinated inventive activity. Economic waste of this sort can occur at three stages in the inventive process. First, the pot of gold represented by a patent on a pioneering, commercially valuable invention may lure an inefficiently large number of persons and organizations into the race to be the first to reach the invention in question. Second, the race to develop a lucrative improvement on an existing technology may generate a similar scramble for similar reasons at the "secondary" level. Finally, firms may try to "invent around" technologies patented by their rivals -- i.e., to develop functionally equivalent but non-infringing technologies -- efforts that, although rational from the standpoint of the individual firm, represent a waste of society's resources. In recent years, several economists have devoted themselves to identifying possible reforms of intellectual property law (or of related doctrines, such as antitrust law) that would mitigate the dissipation of resources at these various sites.

II. Problems

Serious difficulties attend efforts to extract from any one of these approaches answers to concrete doctrinal problems. With respect to incentive theory, the primary problem is lack of the information necessary to apply the analytic. To what extent is the production of specific sorts of intellectual products dependent upon maintenance of copyright or patent protection? With respect to some fields, some commentators have answered: very little; other monetary or nonmonetary rewards (profits attributable to lead time, prestige, tenure, the love of art, etc.) would be sufficient to sustain current levels of production even in the absence of intellectual-property protection. Other commentators sharply disagree. The truth is that we don't have enough information to know who is right -- and are unlikely ever to acquire sufficient information. Even if we were able to surmount this enormous hurdle -- and concluded that society would be better off, on balance, by supplying authors and inventors some sort of special reward -- major sources of indeterminacy would remain. Is an intellectual-property system the best way of providing that reward or would it be better (as Steven Shavell and Tanguy van Ypersele have recently argued) for a government agency to estimate the social value of each innovation and pay the innovators that sum out of tax revenues? If the former, how far should creators' entitlements extend? Should they include the right to prepare "derivative works"? To block "experimental uses" of their technologies? To suppress their inventions? Determining which set of rights would be optimal well beyond our power.

Theorists who seek to optimize patterns of productivity confront less severe information problems. To be sure, they are obliged to make difficult judgments -- often with thin data -- on such questions as whether the failure of creators to license certain uses of their works results from the fact that such uses are worth less to consumers than preventing them is worth to creators (in which case, the absence of licenses is socially desirable) or from excessively high transaction costs (in which case, the creators should be compelled to grant licenses -- for free or for a governmentally determined fee). But inquiries of this sort are not as frighteningly complex as those that confront incentive theorists. However, scholars and lawmakers who take this road confront an additional problem: What is the set of productive activities the incentives for which we are trying to adjust? For the reasons sketched above, if we confine our attention to intellectual products, the optimal legal doctrine may be one that confers upon creators a very generous set of entitlements. Only thereby will potential producers be provided refined signals concerning how consumers wish to make use of which sorts of intellectual products. However, as Glynn Lunney has argued, if we expand our frame of reference, that solution proves highly problematic. In many fields unrelated to intellectual property, innovators are not empowered to collect the full social value of their innovations. The elementary schoolteacher who develops a new technique for teaching mathematics, the civil-rights activist who discovers a way to reduce racial tension, the physicist who finds a way to integrate our understandings of gravity and quantum mechanics -- all of these confer on society benefits that vastly exceed the innovators' incomes. Enlarging the entitlements of intellectual-property owners thus may refine the signals sent to the creators of different sorts of fiction, movies, and software concerning consumers' preferences, but would lead to even more serious overinvestment in intellectual products as opposed to such things as education, community activism, and primary research. Unfortunately, Lunney's proposed response to this problem -- reducing copyright protection until the creators of entertainment receive rewards no greater than the returns available to innovators in other fields -- would sacrifice most of the economic benefits highlighted by Demsetz and Goldstein. The optimal solution is thus far from clear.

Theorists bent on reducing rent dissipation have problems of their own. The most serious difficulty arises from the fact that reducing social waste at one stage of the inventive process commonly increases it at another. Thus, for example, Edmund Kitch has advocated granting to the developer of a pioneering invention an expansive set of entitlements, partly in order to enable him or her to coordinate research and development dedicated to improving the invention, thus reducing rent dissipation at the secondary level. However, as Robert Merges argues, granting generous patents on pioneering inventions will exacerbate rent dissipation at the primary level; an even greater (and more socially wasteful) number of persons or firms will now race to be the first to develop pioneering patents. Mark Grady and Jay Alexander have developed an ingenious theory for determining which of these dangers is more salient in particular cases. Primary inventions that have only modest social value but that "signal" a large potential for improvement are likely to draw potential improvers like flies; to cut down on the swarms, the developer of the primary invention should be granted a broad patent -- i.e., a patent of the sort commended by Kitch. Primary inventions with large social value but minimal "signalling" power should, instead, be given only narrow patents -- to reduce the risk of duplicative activity at the primary level. Finally (and most suprisingly), elegant inventions (i.e., socially valuable inventions so well conceived they cannot be improved upon) should be given no patents whatsoever, thereby discouraging rent dissipation at both levels. This typology, though intriguing, has many defects, both practical and theoretical. To begin with, applying it is likely to be harder than Grady and Alexander think; "how can we know when an invention signals the possibilities of improvements and when it [does] not"? Next, what are we to do with cases in which the invention at issue is of a type that both is highly socially valuable (thus creating a danger of waste at the primary level) and signals a large number of improvements (thus creating a danger of waste at the secondary level)? Finally, Robert Merges and Richard Nelson argue with considerable force that efforts, through broad patent grants, to mitigate rent dissipation at the secondary level may have serious economic side effects. Instead of enabling the original inventor to coordinate efficiently the exploitation of the technology, it may lead to "satisficing" behavior and an inefficiently narrow focus on improvements related to the primary inventor's principal line of business. In short, a combination of limited information and theoretical tensions render this third approach just as indeterminate as the other two.

Even if the difficulties specific to each of the three economic approaches could be resolved, an even more formidable problem would remain: there exists no general theory that integrates the three lines of inquiry. How should the law be adjusted in order simultaneously to (i) balance optimally incentives for creativity and monopoly losses; (ii) send potential producers of all kinds of goods accurate signals concerning what consumers want; and (iii) minimize rent dissipation? As Samuel Oddi points out, to date, no theorist has even attempted to answer this overarching question. Until that challenge is successfully met, the power of the utilitarian approach to provide guidance to lawmakers will be sharply limited.