Intellectual Property Institutions and the Panda's Thumb: Patents, Copyrights, and Trade Secrets in Economic Theory and History
PAUL A. DAVID
INTELLECTUAL PROPERTY ISSUES, ECONOMICS, AND HISTORY
The laws and administrative procedures concerned with intellectual property have once again emerged as a topic of widespread and intense discussion in this country and abroad. Many forces have converged to thrust the subject into the spotlight (see, e.g., Office of Technology Assessment, 1986; Benko, 1987; World Intellectual Property Organization, 1988; Rushing and Brown, 1990). Investment in R&D, for example, has become a central aspect of corporate and national strategies of global competition. The shortening of product life cycles, and the advance of techniques that make ''reverse engineering" and outright copying of novel products easier, have made it more difficult for firms to reap the benefits of innovation simply by guarding new technologies as trade secrets while quickly moving along their production learning curves to seize a cost advantage over potential imitators. Also, many awkward ambiguities and widening areas of legal dispute have been created by the application to new technological developments of laws pertaining to patents, copyrights, and trade secrets, particularly in regard to biotechnology and to computer and information technologies.
In addition to the forces being generated within the sphere of scientific and engineering research itself, national economic policy has contributed to the renewed interest in intellectual property. During the 1960s and 1970s, developing countries successfully resisted conforming to a regime of strong
international protection of intellectual property (see, e.g., Mody, 1990; Siebeck, 1990). However, during the 1980s, the U.S. government responded to the concerns of American producers—especially chemical, pharmaceutical, electronic, and information technology industries—by working vigorously to reverse the trend of the preceding two decades. Acting with some encouragement from other industrially advanced countries, the United States pursued a direct, unilateral course of action. It did not make any major effort to renegotiate agreements within the framework of the Paris Convention for the Protection of Industrial Property (patents and trademarks), the Berne Convention for the Protection of Literary and Artistic Works (copyrights), or other international conventions, nor did it offer some quid pro quo to developing nations that would agree to sign such conventions. Instead, by threatening within the context of bilateral trade negotiations to impose sanctions on developing and newly industrialized nations whose retaliatory leverage was quite limited, the United States achieved considerable success in convincing foreign governments to acquiesce to its position on the treatment of various forms of intellectual property. The pressures generated by the U.S. campaign, however, and the widening international markets for R&D-intensive goods and services have stirred a profound reconsideration of the merits and drawbacks of global "harmonization" of protections for intellectual property and of the desirability of achieving such uniformity at a strong, rather than weak, standard of enforcement.
Unlike the debates over intellectual property institutions in earlier eras, which had captured the attention of such great political and social philosophers as Thomas Jefferson, the current discussions reflect relatively slight interest in philosophical questions. Little attention is being paid to such issues as the "natural rights" of inventors and authors to the fruits of their creative efforts or the justice of claims advanced on behalf of all humanity to benefit from the collective, social processes through which new scientific and technological ideas arise (for exceptions see, e.g., Dworkin, 1981; Davis, 1989; Berg, 1991). Rather, in keeping with the more pervasively utilitarian spirit of the times, the statutes, legal rulings, administrative regulations, and other institutional arrangements affecting patents, copyrights, and trade secrets are widely regarded as public policy instruments that should be designed to enhance economic welfare by stimulating technological progress.
Even if the rhetoric of argument occasionally appeals to notions of justice and equity, modern economic analysis, and its characteristic preoccupation with questions of efficiency, now set the terms for policy discussions about the protection of intellectual property. On the one hand, economic analysis provides the most widely accepted, overarching interpretation and supporting rationale for public interventions aimed at channeling economic resources into invention and innovation. On the other hand, in continuation of a long tradition, economic analysis yields fundamental criti-
cisms of the systems that have been established to achieve that purpose by securing rights in intellectual property. Thus, it is instructive to begin by taking the economist's approach in discussing U.S. intellectual property legislation and national policies to enforce rights in such property internationally. At the very least, this approach provides a framework for identifying the major problems of allocative efficiency and the distributional issues that are at stake—from the viewpoint of society as a whole rather than from the perspective of the various private (and national) interests involved.
Economists as a body, however, have been unable to formulate much in the way of straightforward, practical advice to guide lawyers, jurists, and policymakers in these matters (see Priest, 1986). The fundamental cause of their inconclusiveness is not so much the tendency of economists to engage in theoretical speculations as it is their inability to achieve consensus on the answer to two difficult empirical questions. First, will faster growth in the stock of scientific and technological knowledge always be an unambiguously "good thing" for a particular industrial sector or national economy and, therefore, warrant the sacrifice of other, lesser societal goals? Second, how responsive is the supply of socially useful discoveries and inventions to the creation of greater private economic incentives? For policy analysts not to know the policy goal with any precision is a considerable handicap, just as it is for them to remain unsure about the incentives and constraints that would be required to achieve any particular goal, were one to be agreed on.
Unfortunately, however, the two questions cannot be answered any better by lawyers on the basis of their having delved more deeply into the details of existing or proposed intellectual property regimes. Nonetheless, those who from practice are most at ease applying the logic of microeconomic reasoning to intellectual property rights issues, must pay heed to the skepticism voiced by legal experts. They should take more pains not to allow familiar, simplifying abstractions to obscure a central fact about the nature of the world for which they would prescribe institutional reforms, namely, that the complex body of law, judicial interpretation, and administrative practice that one has to grapple with in this field was not created by some rational, consistent, social welfare-maximizing public agency. What one is faced with, instead, is a mixture of the intended and unintended consequences of an undirected historical process on which the varied interests of many parties, acting at different points (some widely separated in time and space), have left an enduring mark. So, it would be really quite remarkable if the evolution of legal institutions concerning patents, copyrights, and trade secrets had somehow resulted in a set of instruments optimally designed to serve either public policy purposes or the private economic interests of individuals and firms seeking such protections.
Agreement with the above does not deny the general notion of an evo-
lutionary drift toward social optimality in the effects of the law on resource allocation. Clever, modern Panglossians have come up with the proposition that the increasing likelihood that laws resulting in inefficient resource use will be exposed to economically motivated litigation, thereby creating "selective pressure" to remold property law in ways that tend to render it more efficient; that this pressure can work even if the outcome of the litigation is random; and that some beneficent "invisible hand" thus guides the evolution of legal institutions affecting economic performance. These ingenious but nonetheless dubious arguments are confined, even by their most ardent proponents, to the supposed workings of the common law system of judge-made law (see, e.g., Priest, 1977; Rubin, 1977; Goodman, 1978; Cooter and Kornhauser, 1980; and Cooter and Ulen, 1988, for discussions of deficiencies in the selective litigation thesis). The modern "law" of intellectual property, however, consists of statutory and administrative laws pertaining to patents and copyrights, even though the common law roots of the law of trade secrets create a complicating exception.
Thus, it is difficult to find even a speculative, theoretical justification for conceptualizing intellectual property statutes, and the administrative procedures they authorize, as institutional tools that were forged perfectly to "Promote the Progress of Science and Useful Arts." The latter is the specific purpose identified by the framers of the Constitution (Article 1, Section 8, Clause 8) when they granted to Congress the power of "securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." Nor should the prevailing statutes enacted under that authority be regarded as policy instruments designed to attain a social optimum defined more broadly in terms of economic welfare. Indeed, the first step toward understanding many of the policy dilemmas that arise today in regard to intellectual property would seem to be to acknowledge just this elementary point.
If intellectual property arrangements are to be viewed as utilitarian appendages of the body politic, it would be far more illuminating to recognize their essential nature as most closely akin to the "thumb" of the giant panda. The panda's thumb has been justly celebrated by Stephen Jay Gould (1980:Ch. ) as a striking example of evolutionary improvisation yielding an appendage that is inelegant yet serviceable. Although the panda can grasp and strip the leaves from the stalks of the bamboo plant, its thumb is not anatomically a finger at all, much less an opposable, manipulating digit. In actuality, it is a complex structure formed by the marked enlargement of a bone that otherwise would be a component of the animal's wrist—but for the effect of some genetic mutation—and the related extensive rearrangement of supporting musculature. It is, as Gould says, "a contraption, not a lovely contrivance," and one whose obvious mechanical limitations stem from its remote accidental origins.
Evolutionary processes in biology work largely with the materials that are readily available. So does institutional evolution, especially the processes of incremental change and adaptation in legal and other rule systems that give great weight to precedent (see, e.g., North, 1991). Accordingly, even though the legal provisions and administrative rules that make up the "patent system" and "copyright system" have changed considerably in form and function over their long history, they appear remarkably resistant to rapid and radical reform.
As the nature of technologies changes, however, it is increasingly evident that the familiar legal "contraptions" of patents and copyrights are rather ill-suited to some of the situations to which they are being applied (see, e.g., Office of Technology Assessment, 1986; World International Property Organization, 1989). They continue to be looked to as stimuli for the generation of useful innovations, but while enabling the private appropriation of economic benefits from new scientific and engineering knowledge, they have a variety of untoward side effects that may be distorting and even impeding the progress of technology. Moreover, the problems are not confined to those that might be solved by readjusting old and still serviceable legal tools or forging novel statutes to fit special technical circumstances. The process of more finely articulating and more vigorously enforcing private rights in intellectual property is certainly worth pursuing in some situations, but it cannot be looked to for optimal solutions to all of society's problems in designing institutional mechanisms affecting the production, distribution, and utilization of knowledge.
Identifying the limitations as well as the strengths of the private property approach is a central part of my task in introducing the subject of the global dimensions of intellectual property in science and technology. Setting out the basics of modern economic theory of intellectual property and reviewing the historical development of specific legal institutions that define and protect private rights in such property are also major aspects of my task. This assignment is a daunting one, for any of several reasons.
First, as noted, there is no settled body of economic theory on the subject that can be stated briefly without doing serious injustice to the sophisticated insights that have emerged over many decades of debate. Instead, the relevant economic literature is extensive, convoluted, and characterized by subtle points of inconclusive controversy concerning the appropriate course for public policy. Second, intellectual property law is an intricate, highly specialized area of legal scholarship and one to which I make no pretensions of expertise. Third, the historical development in Western societies of the patent system, the statutory protection of copyright, and the body of law governing trade secrets is a subject area that, unfortunately, has remained all too separated from economic and legal analyses of contemporary intellectual property issues. To link them satisfactorily would be no small undertaking.
Recognition of these difficulties should have been sufficient to dissuade me from accepting the assignment. In the event, as one can see, they were unavailing. Some considerable indulgence and forbearance on the part of the reader will therefore be required if my discussion oversimplifies complex matters of economic reasoning concerning intellectual property and the production and distribution of knowledge, points out only the most salient and early developments in the long history of these western European institutional arrangements, glosses over crucial distinctions and subtle points of modern law, and indulges in some provocative concluding comments on the current U.S.-led campaign for an international regime of uniformly strong intellectual property protection, as that appears from this economic historian's perspective.
KNOWLEDGE, PUBLIC POLICY ECONOMICS, AND INTELLECTUAL PROPERTY
The economist approaches the subject of protection of intellectual property rights, like many other issues, by trying to fit it into the generic formula for public policy decisions (see, e.g., Besen and Raskind, 1991:5). Somewhat loosely stated, the policy objective is to maximize the surplus of social benefits of the new information assets over the social costs of their production, that is, to maximize the "net social benefits." A further objective is to push the allocation of public and private resources in the direction of equalizing the social net rate of return on investments in knowledge and in other kinds of productive assets. This formulation gives rise to the following three classes of questions.
First, will the right amount of new information be created, and at the right times? The concern here is whether, in the absence of public intervention, private incentives would be sufficient to generate the optimal flow of additions to the stock of scientific and technological knowledge. Modern economic analysis recognizes that the peculiar, "public good" nature of information as a commodity creates serious resource allocation problems for competitive market systems. Further, it identifies the institution of private rights in intellectual property as one among a number of countervailing measures that the state may take to rectify the deficiencies of market competition.
Second, will the new information that is created be used productively, that is, in a way that yields the maximum flow of social benefits for the producers and consumers of goods and services? Unless intellectual creations are disseminated for others to enjoy as items of consumption or are used directly and indirectly in producing other goods and services, they cannot be expected to yield improvements in productivity and economic welfare. Consequently, a central set of issues for discussion among econo-
mists has been the likely effects that public policy measures meant to stimulate additions to knowledge would have on the diffusion of knowledge into commercial uses.
Third, will the conditions under which new knowledge is created be such that the social costs entailed in its production are minimized? There are opportunity costs to devoting resources to the advancement of knowledge through scientific discovery, inventive activity, and the novel expression of ideas. The goal of public policy cannot be simply that of causing private agents or governmental agencies to conduct these pursuits on an ever-grander scale, without regard to whether they are conducted efficiently. Intellectual property institutions must be evaluated in terms of their implications for the social costs of producing new knowledge, as well as for the utilization of the existing stock of knowledge.
Information, Public Goods, and Competitive Market Failures
The argument most generally offered in support of public policy interventions to enforce patents, copyrights, and trade secrecy is that there is a "market failure." In the absence of governmental protection of private property rights, the argument goes, competitive markets would not give individuals and organizations sufficient incentives to induce the socially optimal amount of investment in public goods in the form of new scientific and technological knowledge. It does not necessarily follow, however, that the best remedy for market failure is to create valuable private rights in intellectual property. In fact, the problem is more complicated than even that of arranging for the right amount of some classic public good, such as national defense or lighthouses.
Knowledge may be viewed as a commodity, but it is not a commonplace commodity. It is highly differentiated and has no obvious natural units of measurement. It can have utility as a pure consumption good or as a capital good, and often as both. Knowledge is unusual in that as a pure capital good yielding a stream of material benefits when combined with other kinds of assets, it possesses an intrinsic value. Such is the case, for example, with information about the operation of a cost-saving manufacturing process or the design of a product with better quality attributes. Still more remarkable is information's extreme indivisibility and durability. Once a bit of knowledge has been obtained, there is no value to acquiring it a second time, or a third. There is no societal need to repeat the same discovery or invention because a piece of information can be used again and again without exhausting it. Karl Marx (1867-1894; 1970: Vol. 1, Ch. XXV:386), among others, was struck by the fact that scientific knowledge could be freely appropriated to productive processes, as are the physical forces found in nature:
Once discovered, the law of the deviation 0f the magnetic needle in the field of an electric current, or the law of the magnetization of iron, around which an electric current circulates, costs never a penny.
Related to this, and of even greater importance, knowledge differs from ordinary "private" commodities in being what economists refer to as a nonrival good; that is, it can be possessed and enjoyed jointly by as many as care to make use of it. This observation forms the point of departure for the classic analysis of the economics of R&D by Arrow (1962), but it is not a modern insight. Consider the following passage in a letter written in 1813 to Isaac McPherson, a Baltimore inventor, by Thomas Jefferson (reprinted in Koch and Peden, 1972:629-630):
If nature has made any one thing less susceptible than all others of exclusive property, it is the action of the thinking power called an idea, which an individual may exclusively possess as long as he keeps it to himself; but the moment it is divulged, it forces itself into the possession of every one, and the receiver cannot dispossess himself of it. Its peculiar character, too, is that no one possesses the less, because every other possesses the whole of it ... That ideas should freely spread from one to another over the globe, for the moral and mutual instruction of man, and improvement of his condition, seems to have been peculiarly and benevolently designed by nature, when she made them, like fire, expansible over all space, without lessening their density in any point, and like the air in which we breathe, move, and have our physical being, incapable of confinement or exclusive appropriation.
Jefferson grasped the essential point that the cost of transmitting useful knowledge in codified form is negligible compared with the cost of creating it, and that, but for society's need to encourage the pursuit of ideas, such information should be distributed freely. Indeed, on these grounds Jefferson proceeded immediately to reject the argument of the French philosophers that inventors and authors had a natural rights claim to property in their creations (reprinted in Koch and Peden, 1972:630):
Inventions then cannot, in nature, be a subject of property. Society may give an exclusive right to the profits arising from them, as an encouragement to men to pursue ideas which may produce utility, but this may or may not be done, according to the will and convenience of the society, without claim or complaint from anybody....
This does not mean that all types of knowledge can be transmitted at negligibly low marginal costs or that the private and social costs of filtering, interpreting, and utilizing information are insignificant. Recent discussions of the economics of R&D and technology transfers (see, e.g., Pavitt, 1987; Rosenberg, 1990; Arora, 1991) have recognized the importance of tacit components of technological knowledge and emphasized that the in-
formation contained in scientific papers, patents, blueprints, and other forms of codified knowledge often will not be sufficient to implement the technology in question; complementary know-how is required, and its acquisition is often costly.
Nelson (1990) makes the same point and goes on to associate codified knowledge with the "generic" parts of technological information—those that can be transferred readily and tend, thus, to move quickly into the public domain—and tacit knowledge with the "specific" bits of information that lend themselves better to being held privately. These particular identifications, however, do not seem either necessary or especially helpful. What is held secret and what becomes publicly disclosed are determined not so much by the inherent nature of the information as by the expected costs and rewards associated with each course of action for the agents involved (see Dasgupta and David, 1990). This much is obvious from considering the factors that enter into a firm's decision whether to file for a patent on a new process of manufacture or to protect it as a trade secret.
Nonrival possession, low marginal cost of reproduction and distribution (which makes it difficult to exclude others from access), and substantial fixed costs of original production—these are the three properties familiarly associated with the definition of a public good. When these characteristics are present, competitive markets—in which price tends to be driven down to the cost of supplying the marginal unit of the commodity—generally perform quite badly; competitive producers' revenues will not even cover their full costs of production, much less anything approaching the use-value of the goods to the public. Indeed, the attempt to make the beneficiaries pay for value received would so reduce demand as to result in an inefficiently low level of its consumption. In the literature of public finance economics, therefore, alternative allocative mechanisms are proposed as solutions to "the public goods problem." There are three principal alternatives. One is that society should give independent producers publicly financed subsidies and require that the goods be made available to the public freely or at a nominal charge. A second mechanism would have the state levy general taxes to finance its direct participation in production and distribution of the good, furnish and manage the requisite facilities, and contract when necessary with private agents to carry out the work. Here, again, the objective is to supply the good without having to charge prices for it. The third solution is to create a publicly regulated private monopoly authorized to charge consumers prices that will secure a "normal" rate of profit. This does not guarantee, however, that consumers will line up to purchase the goods and services in question. The legal right to exclude other producers from the market for a product does not, of itself, create a profitable monopoly of that line of business.
Although the nonexcludable and nonrivalrous nature of information quali-
fies it as a public good, information differs in two respects from the mass of conventional public goods, such as traffic lights, flood control systems, and airport beacons or radar landing beams. The first difference is that the attributes of the commodity—typically, the complete contents of the information itself—will not be known beforehand. Indeed, they are not automatically known to all the interested parties even when the new knowledge becomes available. This asymmetry in the distribution of information greatly complicates the process of arranging contracts for the production and use of new knowledge.
The second differentiating feature of knowledge is its cumulative and interactive nature. The stock of scientific and technological knowledge grows by increments, with each advance building on and sometimes altering the significance of previous findings in complicated and often unpredictable ways. As Thomas Jefferson remarked (reprinted in Koch and Peden, 1972:686),
The fact is, that one new idea leads to another, that to a third, and so on through a course of time until someone, with whom no one of these ideas was original, combines all together, and produces what is justly called an new invention.
On these same grounds, Michael Polanyi (1944:70-71), a British sociologist of science, maintained that patent law was essentially deficient because it sought to "parcel up a stream of creative thought" into a series of distinct claims each of which could constitute the basis of a separately owned monopoly, whereas "incremental progress interacts at every stage with the whole network of human knowledge and draws at every moment on the most varied and dispersed stimuli."
The same kind of creative recombinant process does not operate when one stockpiles weapons for defense or erects another set of airport landing lights. The light signal from an airport or lighthouse is a form of information, but it is the emission of the signal—rather than the bricks or metal or glass—that imparts the public goods character to those structures. This form of information, however, has no capacity for internal growth and elaboration. Unlike scientific and technological knowledge, light signals just do not evolve and acquire new utility through cumulation and interaction. As discussed below, legal and other institutional arrangements may be imposing high costs on research-intensive firms, and society more generally, by restricting access to some elements in those streams of creative thought and thereby making it less likely that the elements will be rapidly rearranged and recombined in new and fruitful ways.
Imperfect Institutional Solutions and Trade-offs
The importance of the foregoing differentiating features of knowledge notwithstanding, there is a striking correspondence between the three solutions for the standard public goods problem—subsidies, direct governmental production, and regulated monopoly—and the three main institutional arrangements that have been devised to deal with allocational problems in the production of knowledge and pure information goods (see Dasgupta and David, 1988). I refer to the latter arrangements as ''the three P's," because they can be described in highly idealized forms as patronage, procurement, and property, respectively.
Patronage stands for the system of awarding publicly financed prizes, research grants based on the submission of competitive proposals, and other subsidies to private individuals and organizations engaged in intellectual discovery and invention, in exchange for full public disclosure of their creative achievements. In Western democratic societies, patronage characterizes the pursuit of "open" scientific inquiry and the dominant institutional and social mode of organization associated with the conduct of academic science (see David, 1991).
Procurement is associated with government's contracting for intellectual work, the products of which it will control and devote to public purposes. Whether the information produced will be made available for public use is a secondary issue, although an important matter for public policy. "Sensitive" defense-related research is usually conducted under governmental auspices in secure, closed laboratories, but much publicly contracted R&D and the scientific work of governmentally managed laboratories and agricultural experiment stations are undertaken with the intention of disseminating the findings widely.
Property refers to society's granting private producers of new knowledge exclusive rights to the use of their creations, thereby forming conditions for the existence of markets in intellectual property and enabling the originators to collect fees for the use of their work by others. The specific legal contrivances of the patent, copyright, and somewhat more problematically, the trade secret fall within the property rubric.
The Intellectual Property System
Patents convey the most potent rights in the intellectual property system, for the patentee may exclude everyone else from making, selling, or using the subject matter of a valid patent throughout its term. Under the current U.S. Patent Act (35 U.S.C., Sec. 1-376), the usual term is 17 years, extendable by 5 years for pharmaceutical and medical device patents and by 14 years for design patents. The conditions that must be satisfied to secure
the award are also the most stringent (see, e.g., Chisum, 1989). In addition to being potentially useful to society, a patentable invention must pass three tests: originality (originating with the inventor in question), novelty (not having been invented independently by another), and nonobviousness (not already obvious to a person having ordinary skill in the pertinent "art").
Copyright, as defined under the terms of the U.S. Copyright Act (17 U.S.C., Sec. 101-801, 1982), subsists in an "original" work of authorship fixed in a "tangible medium of expression" from which it can be perceived. The boundary separating copyrights from patents is usually seen by intellectual property lawyers in the restriction of copyright to protecting its holder against the copying of the specific expression, but not against others' use of "the idea, procedure, process, system, method of operation, concept, principle, or discovery, regardless of the form in which it is described, explained, illustrated, or embodied in such work" (17 U.S.C., Sec. 102(b), 1982; quoted in Bender, 1986:920). Novelty, however, is not a requirement, and although statutory protection is provided against copying and other enumerated acts, independent origination is not precluded—more than one author can copyright identical works. Nor is unauthorized reproduction restrained, so long as it is not deemed to have significant adverse effects on the copyright holder's current or future economic interests—as when a work is copied for the purpose of scholarly study or quoted in part in other copyrighted material under the doctrine of "fair use."
Some further points of contrast between patents and copyrights derive from the absence of the novelty requirement. Works can be registered and deposited at the Copyright Office, but the owner's rights under the U.S. Copyright Act—except for the right to injunctions against infringers—exist independently of any formal registration, prior examination, or determination of the validity of the claim to originality. Thus, the scope of copyright protection ultimately must be defined through litigation. Counterbalancing the more restricted nature of the rights conveyed by copyright law, the term of protection provided is much longer than that for patents. In most circumstances, a copyright expires 50 years after the death of the creator, a convention that is now quite standard internationally.
Trade secrets are included here under the property rubric. Technological developments and the recent history of litigation have brought the trade secret closer to the patent and the copyright in its form (as law for the protection of valuable rights in information) and its ostensible social function (strengthening private incentives for R&D expenditures and abetting contractual arrangements for the limited sharing of technological information). However, regarding trade secrets as another type of intellectual property is somewhat problematic. Information that is kept secret can be a source of income, and as a valuable asset, it shares a quality that economists would automatically associate with other forms of tangible and intangible
protection as trade secrets. Under the American Law Institute's widely property. Nevertheless, all economically valuable secrets are not afforded cited (1939) definition (see Cheung, 1982:42; Bender, 1986:915),
A trade secret may consist of any formula, pattern device or compilation of information which is used in one's business, and which gives him an opportunity to obtain an advantage over competitors who do not know or use it.... It differs from other secret information in a business ... in that it is not simply information as to single or ephemeral events .... A trade secret is a process or device for continuous use in the operation of the business.
How is it possible for society to recognize and enforce a person's claim to hold property in something that is quite specific but must remain less than fully described and actually be hidden from public view? A further problem is that in the law of real and personal property, as well as in the areas of intellectual property law dealing with patents and copyrights, designating something as "property" has a particular meaning. It usually means that the possessor has the exclusive right to use or enjoy the thing, or to assign it to others for their exclusive use or enjoyment (see, e.g., Friedman et al., 1991:61-62). This special sense of the term is not satisfied in the case of a trade secret, because even when the possessor has taken measures to preserve its secrecy, the law provides no remedy if the information is disclosed by accident or uncovered through deliberate, socially conscionable ("fair") actions of others.
Unlike patent and copyright law, trade secret law (even when given statutory structure) is rooted in principles of common law, including theories of contract and tort, as well as property concepts (see Jager, 1991:49). Indeed, the general tendency is to de-emphasize rights to property in the information held secret and to protect its originators indirectly by enforcing relationships of confidentiality that have been established implicitly or through explicit contracts. Trade secrecy, thus, can be viewed as a means of increasing the security of the "default option." It offers a recourse that may be socially as well as privately valuable in circumstances in which patents and copyrights are unavailable, ineffectual, or unattractive means of appropriating the economic benefits deriving from the generation of new knowledge and its reduction to concrete practices. The relevant legacy from the common law of master-servant relations is the recognition of society's interest in the formation of relationships of trust between employers and their employees (and between principals and their agents). Because in many instances it is much more efficient, and in some circumstances absolutely essential, to give employees access to information that they could use to the disadvantage of their employer, trust in the confidentiality of such disclosures is desirable for all who would benefit from having the work done. Yet the original common law contexts, typically, were ones in which it was the master who had a valuable secret to safeguard. Given these historical
derivations, it is perhaps not so surprising today to hear employed inventors complaining that the law of trade secrets works to the benefit of their employers and leaves their interests largely unprotected.
Trade-offs in Organizing Knowledge Production
Although the attention of industry, the legal profession, and the wider community of researchers has been focused increasingly on the property mode of organizing knowledge production, the three modes must be kept in mind in formulating effective science and technology policies. Each of the three allocative mechanisms has been found useful in some fields and at some periods in the development of modern industrial societies; but the weight of reliance has shifted among them over time, and none has been accepted as clearly superior to the others in all contexts in which useful knowledge has been sought. Even the brief economic analysis that follows readily exposes some serious drawbacks to, as well as the principal advantages of, each of the three arrangements. (For a fuller presentation of the following analysis, see, for example, Wright, 1983; Dasgupta and David, 1988.)
Theoretically, at least, prizes and research grants can be established, or procurement contracts written for amounts that would award the producers of original intellectual works commensurately with the anticipated social use-value of their creations. As a practical matter, however, the patronage and procurement solutions are burdened by the fact that public authorities generally cannot set efficient terms for prizes in advance of their results. Moreover, in markets for which the expertise about the likely costs and benefits of particular research projects is unevenly distributed, contracting will entail high transaction costs even to arrive at rather imperfect agreements. The intellectual property solution avoids these drawbacks by letting the workings of the market determine the economic rewards after the fact. Thus, the avoidance of administrative arbitrariness in awarding prizes or granting subsidies for invention has been recognized as an advantage of the patent system by economists since Adam Smith (see Jurisprudence, A.ii:3133, cited in Smith, 1776:754, n. 69).
To secure the benefits of a rapidly accumulating stock of knowledge, it is desirable to promote speedy disclosure of new findings so that they may be disseminated, verified by replication, and put to use by others engaged in intellectual pursuits. Only in this way can the fullest scope be provided for the interactive process through which ideas proliferate and generate still more ideas. When directed toward that goal, the patronage and intellectual property systems must seek full and prompt disclosure. Each, therefore, is impelled to base the assignment of rewards in some way on the establishment of priority (see Dasgupta and David, 1987 and 1988). Patronage
achieves this by offering prizes for discoveries or inventions of a specified kind and by awarding research grants to those who have developed reputations for research success. Among academic scientists in the modern West, such reputations as a rule are based on validated claims to priority in discoveries and inventions deemed by expert peer groups to constitute useful "contributions" to knowledge. Priority is equally central to patent awards, although whether the touchstone is priority of invention or priority of registration and disclosure varies from one national system to another. Also, statutory copyright protections are traditionally accorded to the first author to disclose (by registration with the copyright-granting authority) a particular creative "expression."
A common implication of priority-based reward systems is that they give rise to competitions—among reputation-seeking scientists and patent-seeking inventors—that are characterized by a payoff structure in which the "winner takes all," or substantially all. (This pertains to the awarding of patent rights in cases involving rivalries among inventors, but it does not imply that the original recipient of a patent automatically captures all, or even the lion's share, of the economic gains deriving from a successful invention.) As a consequence, it is likely that from the viewpoint of society there will be too many contestants in the races for priority in discovery and invention. Those entering consider only what they individually stand to gain, and they do not take into account the effect of their participation on the expected outcomes for all the other competitors. The situation resembles the inefficiencies in resource allocations that arise when there is a "common pool problem" (see, for example, Dasgupta and Stiglitz, 1980; Wright, 1983, and references therein).
Further, in addition to the crowding of the field with contestants who might be more usefully engaged in other productive pursuits, there is a tendency for private rents to be dissipated in the scramble for the prize of priority and all that it would bring. The private value of arriving at a new finding a little sooner than the second-place contestant is likely to exceed greatly the benefit that society would derive from the slight advance in the date of discovery or invention. Such allocative inefficiencies, most probably, are more serious in regard to patents than to copyrights. Problems of "racing" do not arise in regard to trade secrets, but when the law of trade secrets is used to capture the value of new discoveries and inventions, inefficiencies in resource allocation will arise from sources other than the common pool problem, and this is likely to be even more severe, for reasons explained below.
In principle, the common pool problem of excess inventive effort could be avoided under a contract research system if the procurers were as fully informed as the researchers about the likely costs and potential social value of the findings. Under such conditions a contract could be drawn up that
would provide a single, successful research entity with an economic payoff that just matched the (certain) value of a patent monopoly that had been awarded for whatever duration was deemed socially optimal. As noted above, however, it is most probable that the writing of such a contract would be frustrated by the asymmetrical distribution of information between researchers and would-be customers for research results.
If the common pool problem creates a tendency toward excess investment of R&D funds in racing for patents, adjusting the terms of patent awards would seem to be a reasonably straightforward remedy. By shortening the life of the patent or narrowing the protection granted against infringements, the value of the anticipated prize could be lowered until private R&D investment was no longer socially excessive. Even with that problem fixed, however, the resulting allocation of resources would still be inefficient because the property solution, unlike patronage and procurement, inherently entails withholding access to the new knowledge itself or restricting the extent of its application by imposing license and royalty charges on the users. Three modes of use of the new knowledge may thus be curtailed: (1) it may have been the basis for the design of a new good or service that would enhance consumer satisfaction by increasing variety or offering superior quality; (2) it may have been the basis for a cost-saving production process; (3) it may have been an intermediate input in the production of further advances in knowledge.
By long-standing tradition, economists' critiques of the legal protection of intellectual property have focused primarily on the losses in economic welfare caused by the establishment of a temporary (but for copyrights, rather prolonged) monopoly of the application of information in modes 1 and 2 above (see, e.g., Plant, 1934, 1974). Exclusive possession of technological knowledge by a profit-seeking agent will restrict the extent to which that knowledge is applied for the production of commodities that embody the innovation or can be fashioned more cheaply by processes based on it. The more secure the possession is, the less the patent monopolist or copyright holder has to worry that charging a high royalty rate will induce others to seek to avoid paying it by investing in reverse engineering, "inventing around," or closely imitating his or her creation. With less risk of the entry of close substitutes to worry about, a royalty income can be extracted from the final customers by setting prices above the marginal costs of production and restricting output accordingly. The benefits of the new knowledge to society, and to consumers in particular, are thereby less than they would have been had the information been made available for exploitation on a competitive basis. Such lost benefits are referred to by economists as the "deadweight burden" of the patent monopoly.
More recently, however, growing attention to the allocation of public and private funds for R&D, and the concomitant recognition of the impor-
tance of scientific and technological advances as intermediate informational inputs into the R&D process itself, have added a new set of worries to the traditional concern with the deadweight burden of monopoly. These worries involve the adverse efficiency consequences of restrictions that intellectual property rights systems—unlike patronage and procurement arrangements—impose on access to information on research methods and results. Secrecy raises the costs to researchers and to society of the search for new knowledge. Because scientific and technological discoveries and inventions so often build on previous knowledge, many economically significant research developments (such as a commercially practical microprocessor chip) can be conceptualized as having entailed the successful solution of a large set of interrelated problems for which there exists a least-cost solution sequence. More realistically, there will be a number of such solution sequences that are substantially more efficient, in the sense of requiring less time or fewer resources, than others. Many firms with varying research capabilities can be engaged in trying to solve some or all of these problems. Because each step along the way represents a distinct "result," the economic benefits of which are likely to be worth appropriating by one intellectual property protection device or another, the payoff to priority tends to result in each firm's shrouding its efforts in every stage of the sequence under a cloak of secrecy. This is true even when the intention is ultimately to disclose a result, once it has been secured and found to be protectable under patent or copyright law.
Maintenance of secrecy by rival firms, however, makes virtually impossible the coordination of research activities required to achieve the optimal sequence. Some simulation studies by Folster (1985) suggest that the resulting losses due to wasteful duplication and delay can be very much larger than the excessive R&D expenditures attributable simply to the common pool problem. In other words, the lack of communication and coordination among the contestants in patent races would seem to be a more serious matter than the fact that too many contestants were induced to enter the race(s) to begin with. The greater the incentive is for firms to proceed with their R&D programs in complete secrecy, the more severe this source of inefficiency is likely to become. Here again, within the category of property devices for organizing the production of knowledge, patent and copyright protections possess comparative virtues (of intermediate-stage disclosure) that are lost to society when firms elect to rely on trade secrecy laws and attempt to appropriate the economic benefits of their inventions by embedding them in new goods that they can sell.
If trade secrecy per se has serious drawbacks and if, as has been pointed out, the adverse effects of the common pool problems associated with patenting and copyrighting can be mitigated by adjusting the terms of those property awards, does the foregoing analysis imply that there is not really
much of a socially useful role remaining for trade secrecy law? There has not been much opportunity for the development of a consensus on this point, since economic analysis of the legal protection of trade secrets is far less developed than that of the patent and copyright systems (see, as exceptions, Kitch, 1980; Cheung, 1982; Friedman et al., 1991). This may be due, in part, to the fact that patents and copyrights sprang from statutory enactments, which often ignite political debate and public discussion, whereas the protection afforded to possessors of trade secrets by the courts is rooted in the common law.
Optimizing Intellectual Property Protection: Issues of Length and Breadth
More than 30 years ago, Fritz Machlup (1958:80) remarked that although economic analysis did not yet provide a basis for choosing between "all or nothing" where intellectual property protection is concerned, "it does provide a sufficiently firm basis for decisions about 'a little more or a little less' of various ingredients of the patent system." For some lawyers and practical policymakers, this was perhaps too self-congratulatory an appraisal (see, e.g., Priest, 1986). Nevertheless, a number of economists subsequently followed Machlup's advice and examined the question of the optimal duration and scope, or length and breadth, of patent protection, taking as given the existence of the patent system. A similar approach has been taken in examining the economics of copying and the optimal level of copyright protection.
Length of Protection
Instead of accepting the historically given length of patent protection, Nordhaus (1969) and Scherer (1972) compared the size of the incentive effect on invention with the resulting inefficiency due to the deadweight burden of monopoly. By balancing the one against the other, they showed how the optimal patent length would change with market demand and technological factors. Three principal conclusions emerged with respect to cost-reducing process inventions. First, the optimal length of patent life is probably not uniform across industries and inventions because it is sensitive to (1) the price elasticity of demand in the end-product market and (2) the responsiveness of the costs characterizing the new production process to the amount of R&D resources devoted to its invention. Second, the more elastic the demand for the new product, the shorter is the optimal length of patent protection because higher prices will cause a proportionately larger reduction of the quantity demanded, and consequently a greater "deadweight burden." Third, the optimal length of patent protection will be shorter when
the technological opportunities being explored are such that greater production cost savings can be achieved with given levels of R&D expenditures. Nordhaus's (1969) work carried a fourth implication: The welfare losses resulting from setting the patent life at a suboptimal length are not very substantial except when major technological advances (cost reductions) are obtainable and when the market demand is very price elastic.
Formal analyses of the economics of copying and copyright protection have tended to follow in Nordhaus's (1969) footsteps. Hirschliefer and Riley (1979), for example, evaluate the impact of increasing copyright protection by comparing the benefits from reducing losses due to underproduction of new works to the costs incurred in the form of losses due to underutilization of copyrighted material. The conclusions derived from this approach, however, are rather more ambiguous than for the analysis of patents. The reason is that the analysis turns on one's assumptions about the substitutability of demand between, and the comparative costs of, unauthorized copies and copies produced under copyright agreement.
More stringent copyright protection would decrease the social loss due to the underproduction of intellectual works, unless copyright monopolists raised the price of their products so much that consumers increased the demand for unauthorized copies. More stringent protection might also reduce underutilization losses if obtaining an unauthorized copy cost consumers more than they would be charged by a copyright holder who had a strict, enforceable monopoly (Novos and Waldman, 1984). In the same spirit, Johnson (1985) concludes that strengthening copyright protection could enhance social welfare even without stimulating the production of new works of authorship, so long as lax restraints on copying resulted in the demand for authorized copies ("originals") being reduced greatly in relation to total consumption of the work in question. These conclusions rest crucially on the supposition that the private cost to the consumer of obtaining a close substitute by copying an authorized "original" is greater than the copyright monopolist's marginal costs. However, as Liebowitz (1985) has pointed out, the latter assumption has in many situations been invalidated by advances in copying technologies. Indeed, he suggests that the complementarity in production between authorized originals and low-cost copies could, under some conditions, mean that Johnson's (1985) appraisal was vitiated. Instead, a more permissive law regarding copying might—by allowing utilization of highly efficient copying technology—actually increase the effective demand for originals as well.
Breadth of Protection
In another branch of the literature, the policy variable of interest has been the patent's optimal breadth, or scope, of protection. Continuing in
the analytic tradition of Nordhaus (1969), a number of more recent works, including some that treat the problem of optimal patent breadth and length simultaneously, focus on the balancing of incentive effects on invention against resulting inefficiencies due to underutilization.
Taking the flow rate of profit available to the patentee as a proxy for the breadth of a patent, Gilbert and Shapiro (1990) have derived conditions under which the optimal patent length is infinite: If total (consumer and producer) surplus declines at an increasing rate as patent breadth increases, it is optimal to grant an infinitely lived patent and adjust the breadth to ensure that the patentee makes at least zero net profits. As Gilbert and Shapiro would readily admit, however, the flow rate of profit available to the patentee is an unsatisfactory proxy for patent breadth because it fails to show how the given innovation is related to other innovations. They acknowledge that their static analysis ignores the fact that inventions build on each other and that a long patent grant may have "deleterious effects on the incentives of other firms to engage in related research, for fear that they will be at the mercy of the original patentee" (Gilbert and Shapiro, 1990:112). Klemperer (1990) considered the problem of optimal patent breadth in a static setting using a model of horizontal product differentiation, which ignores the possibility of vertical product improvement. The proxy for patent breadth in his model is the region of the product space covered by the patent grant. Correspondingly, two kinds of welfare losses have to be considered: (1) those caused by consumers' switching to less preferred varieties of the product that are unpatented and sold at competitive prices and (2) those caused by consumers' dropping out of the product class altogether.
Schmitz (1989) provides a more dynamic analysis of the trade-offs involved in broadening the breadth of patent protection. As a proxy variable for patent breadth, he proposes the anticipated probability of infringement. The economic importance of a particular line of research is represented by the absolute size of an increase in total surplus due to initial and secondary product development. When this parameter is increased, the relative contribution of those R&D efforts to the gain in total surplus remains unchanged, but the absolute change in the total surplus is increased. Consequently, the cost of invalidating subsequent innovations increases, and the optimal infringement probability—that is, the breadth of the patent coverage— should be reduced. The more important a particular line of research, the less subsequent developers should be constrained by the original patent claim, according to this analysis (Schmitz, 1989:4).
Kitch's (1977) "prospect theory" of patent protection and Beck's (1981, 1983) discussion of "unproductive competition" are two of the most significant early contributions to the analysis of patent breadth. An important feature of their work was the emphasis they placed on inefficiencies that
might be present in the R&D process itself. They identified the source of inefficiencies in the allocation of resources to R&D not simply with what Dasgupta and Stiglitz (1980) called the common pool problem, but also with ''unproductive competition for monopoly profit." Implicitly, however, the two problems are connected. Kitch and Beck's unproductive competition for patent monopoly encompassed wastage of resources on premature invention, duplicative R&D, unnecessary substitute inventions ("me-too" patents), and excessively rapid spending on research. These forms of inefficiency are clearly related to the previously discussed problem of lack of interfirm coordination or socially optimal scheduling of R&D projects, which stems from the competitive conduct of research under conditions of secrecy and from "patent racing."
As a solution to this problem, Kitch and Beck proposed to broaden the scope of patent protection to allow the rationalization of the entire development process for a given technological "prospect." Drawing an analogy to mineral resource development, Kitch and Beck have argued that by allowing the "competent" initial innovator to coordinate the subsequent development of "a technological prospect" through efficient bilateral monopoly contracts with other innovators, broader protection would eliminate duplicative effort, premature invention, and other forms of inefficiency in a competitive race for patent monopoly. Although they considered the possibility that the patent holder would contract with independent researchers, they implicitly assumed that the patent holder was a private firm, not a governmental procurement agency. Beck (1983:207) even proposed a competitive bidding scheme for future patents, which would be designed to "transfer the expected value of the patent owner's economic rent to the Patent Office, thus removing the economic incentive for unproductive competition."
Although this theory assumes that efficient bilateral monopoly contracts can be signed between the patent holder and independent innovators, the transaction costs in this process are likely to be nonnegligible. The entire argument projects a vision of organized and orderly development of technological prospects, the realization of which is problematic, to say the least. A major obstacle is the problem of asymmetric information and "thin markets" for specialized research capabilities, which are likely to make bargaining between the would-be "developer" and independent contractor-innovators very inefficient. In addition, although the "prospect theory" approach to patents addresses the issue of inefficiency in the development of an area of technology opened by a "breakthrough" patent, the proposed solution of broadening the scope of such patents would, except under the imagined auction scheme, further intensify the winner-takes-all nature of the payoff structure, thereby exacerbating the common pool problem.
There are circumstances, nonetheless, in which faster technological advance and consequent welfare improvement might be obtained by using a
monopoly to "internalize" the process that generates innovations. David and Olsen (1989, 1991) point out that when private production of a durable good is subject to learning by doing that generates process improvements that cannot be protected as trade secrets, an industry composed of competitive suppliers may perform suboptimally from a social welfare standpoint. If the industry in question is supplying a new technology embodied in a machine or other producer good whose adoption depended on progressive reduction of its supply price, the diffusion into use and incremental improvement of the new technology would be affected adversely. The firms would move along their learning curves more slowly than is socially optimal, because they are not able to capture the benefits of future cost reductions that are a by-product of gaining more production experience. The price of the industry's product, therefore, also would fall more slowly. By correcting the externality, the grant of a monopoly franchise for production of the new good could lead to a second-best welfare optimum even when there was no prospect of future inventions being induced by the promise of patent rights.
In some respects, the above considerations resemble the concern for the efficiency of the technology development process that motivated the analyses by Kitch (1977) and Beck (1981, 1983), but this aspect of similarity should not be overstated. The monopoly franchise envisaged by David and Olsen's analysis would be designed solely to optimize the rate of incremental improvements and the resulting diffusion of the technology into use, without regard for the possibility of inducing some future breakthrough that would introduce yet another new technology. The situation is one in which patent protection would be granted not for invention but for the introduction and adaptation of a basic invention that had already been developed elsewhere. (As discussed below, these were precisely the historical circumstances in which patent grants were first used.)
To analyze further this facet of the patent system, David and Olsen (1991) develop a formal model of interdependent diffusion and learning and show that creation of a monopoly franchise may lead to an overall gain in social welfare, depending on the exact form of the learning function and on conditions governing the demand for the new product. Other things being equal, stronger learning effects at low levels of production experience tend to strengthen the case for granting a patent to the local "innovator" (not necessarily the inventor) of a technology that has yet to be brought into use. Because a learning monopolist will want to produce at a higher level than a competitively organized industry would, there will be a gain in social benefits from accelerated diffusion of the new product and reduction of its production costs. On the other hand, the learning monopolist will want to stop producing when the new good is less extensively diffused than it would be under conditions of competitive supply, which would entail some wel-
fare loss. David and Olsen show that the point at which it is privately optimal for the monopolist to cease producing will always occur prior to the end of the time for which the exclusive patent right has been granted. Indeed, the best strategy for the monopolist may be to shut down production long before the patent/franchise expires and competitors are free to make use of the knowledge gained through production experience. Permitting this lull, in a sense, is just the social cost of using monopoly to correct the problems caused by the externalities in learning by doing.
Because the expiration date of the franchise enters into the monopolist's determination of the date on which to suspend production (and hence affects the extent of "cumulative underproduction" by the monopolist), there will exist some optimal finite length for a patent franchise that is not intended to stimulate future patent filings. Under the conditions assumed in the David and Olsen (1991) model, the optimum franchise duration is shorter when the interest rate is lower, when initial cost reductions due to learning are less drastic, and when the distribution of willingness to pay among the potential purchasers of the new commodity is more skewed toward high values. Yet even when the optimal franchise duration is brief, the associated improvement of welfare over the competitive supply alternative can be quite large.
Although temporary monopolies may provide a simple way to fix the knowledge spillover ("learning externalities") problems that cause competitive markets to generate too slow a pace of technological advance, it will not be a "first best" remedy. Indeed, monopoly cannot be looked to as the most efficient market structure from the standpoint of stimulating product innovations. Arrow (1962) pointed out long ago that an entrenched patent monopolist would have weaker incentives than a would-be entrant to engage in an R&D program that would yield substitutes, even superior substitutes, for goods that already were profit-generating items in the product line. More recently, Merges and Nelson (1990:5-6) persuasively formulated an analogous case, although in broad terms and without invoking results from any formally specified model, for restricting the breadth of patent protection. They argue that even though competitive investments in R&D can result in inefficiencies, technological development tends to proceed "much more vigorously and creatively under a regime where there are many rivalrous sources of invention, than in a setting where one or a few organizations control developments."
Although a satisfactory mathematical characterization of the process of cumulative and interactive product innovation has yet to be developed, recent models have focused on the potentially adverse impact that granting broader patents may have on the pace of technological advance. This work raises important trade-off issues that have been neglected for too long by the theoretical economics literature. The work of Scotchmer and Green
(1990) and Scotchmer (1991) examines the implications of the point that although broader protection provides stronger incentives for R&D aimed at achieving breakthrough inventions, it may seriously weaken the incentives for second-generation innovators to elaborate and improve on the work of the pioneers. Scotchmer and Green implicitly dismiss as infeasible the sort of integration and internalization of the whole line of development envisaged by the prospect theory of Kitch (1977) and Beck (1981, 1983). In their analysis, affording a broad scope of protection to the first patentee puts followers at a disadvantage in negotiating the terms of licenses for technological elements complementary to those they themselves will seek to patent or imposes on them the added costs of trying to "invent around" the blocking, first-generation patent. Many of the cumulative welfare gains attributable to breakthrough technological advances, however, derive precisely from the latter category of incremental, follow-on inventions.
Achieving socially optimal patent breadth is thus a matter of striking the best balance between the net rewards offered to inventors in the first and the second-generation categories. One implication that would seem to follow is that as basic scientific advances reduce the costs of successfully inventing around breakthrough patents in a particular technological area, the breadth of patent protection awarded the pioneers could be increased without diminishing the net incentives that would exist for derivative, second-generation R&D projects.
Essentially the same considerations that arise from recognizing that new scientific and technological knowledge (and intellectual products more generally) spurs the further production of knowledge also arise, in principle, in regard to the protection of copyrights. This point forms a central feature of the analysis by Landes and Posner (1989:335), that is, "too much protection can raise the costs of creation for subsequent authors to the point where those authors cannot cover them even though they have complete copyright protection for their own originality [of expression]." The net effect of increased copyright protection on the supply of (equivalent) works thus depends on the balance between the encouraging incentive effects (for authors and publishers combined) and the discouraging effects of "driving up the cost of expression." Under these conditions, Landes and Posner (1989:344) find that the more the cost of expression rises with increases in the level of protection, the lower is the optimal degree of copyright protection. A social welfare rationale is thereby suggested for leniency in infringement proceedings, that is, permitting more extensive use of copyrighted material to create new derivative works and maintaining broader protection against literal copying.
INTELLECTUAL PROPERTY LAW AND SOME LEGACIES OF HISTORY
After a review like the one just completed, there is a temptation to issue an overall evaluation. In regard to intellectual property protections, however, a great array of benefits and costs cannot be quantified readily. Thus, it is difficult to do more than confirm the observation of Fritz Machlup (1958:80) in regard to the patent system:
If one does not know whether a system "as a whole" (in contrast to certain features of it) is good or bad, the safest "policy conclusion" is to "muddle through"— either with it, if one has long lived with it, or without it, if one has lived without it. If we did not have a patent system, it would be irresponsible, on the basis of our present knowledge of its economic consequences, to recommend instituting one. But since we have had a patent system for a long time, it would be irresponsible, on the basis of our present knowledge, to recommend abolishing it.
Modern economic analysis offers little to refute the conclusion that we who would use U.S. institutions of intellectual property protection to accomplish the purposes of a modern industrial society must remain "prisoners" of their particular history. It is all too easy to miss this central message in the intricate economic analysis characteristic of the literature just reviewed. By focusing selectively on specific features of the complex structure of intellectual property protections and pointing to their putatively favorable consequences for social efficiency in resource allocation, one can convey the misleading impression that the law in this area is susceptible to easy and rapid reshaping to enhance economic welfare.
The evolution of the law in Western societies for protecting intellectual property does attest to a great adaptive capacity. Nevertheless, whereas economic efficiency would seem to call for great subtlety and differentiation in the nature and degree of intellectual property protection provided, based on differences among industries in technological and market circumstances, that evidently is not a direction in which adaptation of the law has proceeded very far. Even more evident is the fact that, today, much uncertainty and controversy surround the persistent difficulties of adapting intellectual property law to new technologies. A major source of these difficulties are the problems of achieving the semblance of consistency in the application of legal principles, of preserving the force of precedent, and thereby circumscribing the remaining areas of ambiguity and uncertainty as to the ultimate enforceability of legal claims, and the likely costs of the entailed litigation. Unfortunately, the economist's conventional approach of evaluating specific institutional arrangements and policies in isolation does not naturally accommodate consideration of these sorts of systemic concerns, with which traditional legal scholars so often are occupied. As a
consequence, the literature on the economics of patents, copyrights, and trade secrets rarely takes note of the problems that arise at the interfaces between those regions of the law. Neither has it paid much attention to the interrelationships and connections between intellectual property law and the larger body of property, tort, and contract law (however, see Cheung, 1982, as an exception).
Nevertheless, it is to a large extent for reasons of the sort that conventional economic analysis has tended to overlook the fact that legal institutions evolve incrementally. Legal institutions preserve many aspects of outward continuity even when it has become apparent that the circumstances of many of the economic actors affected by the institution have changed and that a radical transformation has occurred in the inner rationale and motivation for its maintenance. Thus, although the history of intellectual property rights in the West is replete with instances of redefinition and reinterpretation in response to pressures to accommodate or advance the economic interests of those most affected by the laws, many of the structure's gross features continue to reflect the remote historical circumstances in which they originated. These legacies from the past should not be ignored, nor should their problematic aspects in contemporary contexts be minimized. In persisting, they impinge on the search for new technologies and the organization of economic activities based on the exploitation of the resulting additions to the stock of knowledge.
Patents began as instruments used by noble or republican governments in later medieval and early Renaissance Europe primarily to induce the transfer and disclosure of foreign technologies. This bit of history calls into question one causal supposition that the basic economic analysis of the patent system has fostered, that is, that the protection of intellectual property has been instituted where governments recognized there was more to be gained by stimulating indigenous inventive activity than by applying knowledge of techniques and products that could be "borrowed" freely from the rest of the world.
Patent, the English adjective, means open, and the noun form comes from the term letters patent (a literal translation of the Latin litterae patentes), which means simply open letters. These were the official documents by which certain privileges, rights, ranks, or titles were conferred and publicly announced. Hence, they carried the seal of the sovereign grantor on the inside, rather than being closed by a seal on the outside (see Hill, 1924:406). The "openness" involved, thus, had nothing to do with disclosure of an invention—despite misapprehensions on this point that persist today (see, for example, Bugos and Kevles, 1991). Only much later did the granting of
letters patent evolve into social contrivances for stimulating original invention.
Encouraging Technology Transfer
In the fourteenth century, patents were employed to encourage the introduction of foreign technologies through the immigration of skilled artisans from abroad. Letters patent were given, for example, to the Flemish weaver John Kempe by Edward II in 1331, to two Brabant weavers to settle at York in 1336, and to three clockmakers from Delft in 1368 (see Federico, 1929a:293-295). England at this time was technologically laggard in comparison with many regions on the continent of Europe and, understandably, was endeavoring to "borrow" the more advanced industrial practices. It was hoped that the foreign master craftsmen would introduce English apprentices to the "mysterie" of their respective arts. However, because the master was not likely to remain in control of the newly skilled workers once they acquired journeyman's status, he obviously wished to be protected against the cohort of potential domestic competitors he would create.
Many of the basic features of the patent are better suited to its initial purposes and historical contexts than to the subsequent use to which patents have been put. The disclosure provisions of modern patent systems, for example, were an essential and natural aspect of the effort to induce foreign artisans to reveal a "mysterie" and train domestic craftsmen in its pursuit. Making the conduct of the trade or craft—and the consequent training of apprentices and journeymen—a condition for the privilege conveyed by the patent was quite straightforward since that was the object of the patent. Protecting instructors from the competition of their students, by giving them a monopoly of the trade, directly addressed the spillover problem because there was no way those they trained were likely to benefit except by setting themselves up in competition as soon as they learned the "mysterie." Even the duration of early English patents—14 years, with 7-year extensions possible—was not fixed arbitrarily. Seven years was the term of service of an apprentice, so the protection afforded was to last at least for two generations of trainees. Inasmuch as 7 years was the conventional term of apprenticeship irrespective of the trade or craft, there was considerable logic to making the term of the patent award uniform across all branches of industry. (As has been pointed out, however, modern economic analysis finds this aspect of the contemporary patent system difficult to rationalize.)
Granting monopolies also made sense fiscally for sovereigns whose powers of taxation and borrowing were very circumscribed. It shifted the market risks to the foreign artisan and transferred to him also the bother of collecting the excise tax in the form of the markup over his production costs. Finally, there was no need to ascertain that the grantee had origi-
nated anything, only that at the time of the grant the practice was not being carried on, and hence could be presumed to be unknown, within the sovereign's domains. The criteria of originality, novelty, and nonobviousness that have emerged as definitions of what qualifies as an invention at the U.S. Patent Office, and elsewhere, might well be seen as the makeshift results of a 200-year struggle to use the granting of patent privileges to accomplish a purpose for which it was not originally designed (see Lubar, 1990).
Most historical accounts place the origins of systematic state protection of intellectual property firmly in Renaissance Italy, from where it spread first on the continent of Europe and eventually to England. In the fourteenth and early fifteenth centuries, however, the property rights in question typically took the form of grants for the exclusive exploitation of locally unfamiliar processes or devices that had been originated elsewhere, and more likely than not by individuals other than the one seeking the privilege. Venice took the lead in these developments. As early as 1332 the Venetian Grand Council established a privilege fund for providing loans and other rewards for a foreign constructor of windmills who offered to bring knowledge of this art to the city (see Prager, 1944:713). In 1416 the council awarded Franciscus Petri, from the island of Rhodes, a patent for a superior device for the fulling (shrinking and thickening) of fabrics, which gave Petri and his heirs exclusive rights for 50 years to build, alter, and reconstruct the apparatus he would erect for that purpose (see Mandich, 1958:115116, 149-150; Prager, 1960:379; Long, 1991:877).
In this era the practice of granting privilegi, which was hardly confined to Venice, sought the revelation and application of "secrets"—whether of foreign provenance or native genius. When, in 1421, the Florentine commune awarded a patent to Brunelleschi for a new design of ship he claimed could haul loads more cheaply on the Arno River (to the benefit of merchants and others), the nature of the bargain for disclosure was spelled out candidly in Brunelleschi's petition (Prager, 1946:109-110):
He refuses to make such machine available to the public in order that the fruit of his genius and skill may not be reaped by another without his will and consent, and that, if he enjoyed some prerogative concerning this, he would open up what he is hiding and would disclose it to all.
From about this time forward, the issue of patent privileges for various devices became increasingly frequent, and by 1460, the Venetian Senate in its administrative practice was differentiating between grants of exclusive monopoly to sell products incorporating an "invention," and awards that forbade use of the device without permission while obligating the holder to grant licenses to others when "reasonable royalties" were offered (see Kaufer, 1989:4). Technology importation continued to figure as a primary objective: in 1469 a German, Johann von Speyer, received an exclusive mo-
nopoly of the trade of printing in the Venetian domain in exchange for introducing the craft.
Protecting Intellectual Property
Most modern historical accounts of the development of intellectual property protection in the West assign great significance to the Venetian Senate's passage, on March 19, 1474, of the first general patent law. This is quite understandable given the correspondence between contemporary preoccupations with stimulating invention and innovation and the language of the famous preamble (translated by Gilfillan, 1964:11):
We have among us men of great genius, apt to invent and discover ingenious devices .... Now, if provisions were made for the works and devices discovered by such persons, so that others who may see them could not build them and take the inventor's honor [sic] away, more men would then apply their genius, would discover, and would build devices of great utility to our commonwealth.
Yet, most authorities view this statute as having codified prior practice rather than enunciating any novel principle (see Frumkin, 1945; Prager, 1948; Phillips, 1982; Long, 1991). The law required the registration of any "new and ingenious" device not previously made within the Venetian domain, and it prohibited all private parties except the inventor from making it for 10 years, on pain of penalties for violation of the code. Further, it appears that between 1474 and 1490, very few patents actually were issued under the Venetian code, despite the fact that right through to the middle of the sixteenth century many patent privilegi continued to be granted, conferring exclusive production rights for terms varying between 5 and 80 years, as well as monopolistic trade privileges (see Kaufer, 1989:6).
Despite the rising interest in invention and the spread on the continent of Europe of the use of patent grants to encourage the development of new industrial practices as an instrument of mercantilist policy in France during the mid-sixteenth century, in England the first clear provision for "patents of invention"—as distinct from technology transfer franchises sometimes referred to as "import patents"—did not emerge until the seventeenth century—and it did so then rather as an afterthought, in the course of a movement to free the economy and polity from the abuses of royal grants of monopoly privileges.
With the advent of the Tudor dynasty (1485), the use of open letters as a means of encouraging national industry gave way to the negotiation by the Crown of secret agreements designed to attract skilled foreign artisans into its service. For example, German armorers, Italian shipwrights and glassmakers, and French ironworkers were enticed to cross the English Channel in this
fashion. With Elizabeth I's accession to the throne (1558), however, the previous policy of general encouragement of technology transfers was reinstituted. Between 1561 and 1571, many patents were issued by the Crown under this policy, starting with a grant to two foreigners to introduce the manufacture of hard white Spanish soap and one for the manufacture of saltpeter, an item previously imported from Antwerp (see Federico, 1929a:293297). The royal prerogative of awarding monopolies of all sorts was exercised so extensively on behalf of Court favorites and the Crown's fiscal needs, however, that by 1601 Elizabeth was compelled to promise reforms in order to deflect a parliamentary challenge to her authority in this regard. Nevertheless, this only deferred the conflict. The abuses and retaliatory efforts to curtail the royal prerogative increased under James I, until in 1623 Parliament passed the Statute of Monopolies, which declared all Crown monopolies, charters, and patents thereafter contrary to law. An exception was allowed, however, for royal patents conferring a monopoly for 14 years or less "to the first and true inventor" of a new manufacture (see Federico, 1929a:299).
It is on the above exemption that the British patent system and its derivatives elsewhere have been erected. Even so, the modern reading of the Statute of Monopolies as "the Magna Carta of the rights of inventors" (Machlup, 1958:2-3) is somewhat anachronistic. The verb to invent carried far more extensive connotations at that time than it does today. For example, in a famous patent for a pump, granted by James I to Robert Crumpe in 1618, the sense of invent included "bringing into use, find, establish or institute manufacture" (Hill, 1924:416). In short, originality of use in England alone might be a sufficient basis, since technology transfer, commercialization, and industrial development were also seen as worthy public purposes that could be served through the award of patent monopolies.
Creating a U.S. Patent System
Patent institutions in the United States were derived from those of Britain's North American colonies, dating back to early seventeenth century grants of an ad hoc nature that resembled import franchise contracts. The first such grant, awarded in 1620 by a general court of the Virginia Company's stockholders sitting in England, went to a Mr. Somerscalls for a tobacco-curing process that was not clearly an original invention (see Bugbee, 1967:58). In 1641 the General Court of Massachusetts Bay adopted a number of provisions, including one patterned on the Statute of Monopolies and its exemption, that created a statutory basis for granting future patents individually for "such new inventions that are profitable for the Country" (Bugbee, 1967:61). Importation of inventions from the Old World was a natural enough proposition for New World settlers. Thus, while British courts
during the eighteenth century increasingly construed the purpose of patents to be the encouragement of indigenous invention, American courts continued to consider the potential utility of providing incentives for technology transfers. Moreover, even at a later stage, in respect to the conditions of economic ''openness" and competition for mobile resources, the situation of the American colonies and their successor states under the Articles of Confederation resembled those of the city-states and principalities of Renaissance and early modern Europe. In an address to a joint meeting of Congress on January 8, 1790, President George Washington, who on previous occasions had concerned himself with the subject of intellectual property, called attention to various matters requiring legislative attention, such as "the advancement of agriculture, commerce, and manufactures, by all proper means," including "giving effectual encouragement, as well to the introduction of new and useful inventions from abroad, as to the exertions of skill and genius in producing them at home."
The constitutional era ushered in a decisive shift toward preoccupation with protecting national inventive and literary activities, swept away the disparities of treatment that had arisen among the former colonies, and cemented into the structure of federal law the distinctions between patent and copyright protection that today are taken to be fundamental. Despite the considerable attention to patent-related policy issues in the American colonies during the latter seventeenth and the eighteenth centuries up to the hiatus during the Revolution itself, the first systematized patent provision in America emerged only in 1784, and then as a footnote to the copyright provisions in South Carolina's Act for the Encouragement of Arts and Sciences. This statute's purpose was to establish literary property protection for a renewable 14-year term, but it included the following interesting rider (Bugbee, 1967:93):
The Inventors of useful machines shall have a like exclusive privilege of making or vending their machines for the like term of 14 years, under the same privileges and restrictions hereby granted to, and imposed on, the authors of books.
What makes this provision rather intriguing today is that it so closely coupled patent protection with copyright protection, assigning the former as most appropriate to "machines" and the latter to "books," but otherwise barely distinguishing the treatment of the one from the other. The language adopted by the Constitutional Convention in 1787 was influenced strongly by previous state laws and so spoke also of securing exclusive rights for "Authors and Inventors" to "Promote the Progress of Science and Useful Arts." Copyrights and patents for invention were not mentioned explicitly, nor were import franchises explicitly rejected, as the means for accomplishing this purpose.
The formal creation of a patent system during the early days of America's nationhood was thus shaped most strongly by the experience of the former British colonies and was little influenced (except in rhetoric) by the actions of their revolutionary French contemporaries. The law of 1791 that formally established a patent system in France continued the practices of l'ancien regime, under which inventors received royal privileges freeing them to exploit their inventions outside the confines of existing guild controls. What the French law rejected was the legal justification of the practice based on the assertion of royal prerogative. The new, revolutionary dispensation provided, instead, for the issuance of brevets d'inventions (commission, or equivalently, patents of invention) grounded in the "natural rights" of citizens to the fruits of their creative genius (see, for example, Hilaire-Perez, 1991; MacLeod, 1991:889-891). Americans were not so ready to accept this rationale in place of the rather different English legal theory with which they had grown up, however sympathetic in other respects they might be to the French children of the Enlightenment. Recall how disparaging even Thomas Jefferson was of the argument for natural rights in intellectual property on behalf of authors and inventors.
The U.S. Senate complied with Washington's recommendation in his address of January 8, 1790, by appointing a committee charged with considering provisions for the granting of technology importation franchises, patents for invention, and copyright protection, all within a single act. Only the latter two provisions, however, emerged from the congressional deliberations of 1790-1791 (see Bugbee, 1967:125-148). Indeed, the response of the legislators to mounting pressure for grants of copyright led to the rapid passage of the Copyright Act (1790) first, which then made it necessary to pass a separate Patent Act in the following year, thereby creating two distinct statutory bases for intellectual property protection in federal law. It is the perpetuation of this legal separation—one body of law having developed to protect inventors of "machines" and the other to protect the authors of "texts"—that causes contemporary difficulties when new technologies are found not to fit neatly into either mold. Computer software, for instance, has posed awkward problems inasmuch as this class of technology is well described as "machines which are implemented in the form of text" (for further discussion, see Samuelson, 1984, and Chapter 12 in this volume). This is not to suggest that the separation between patent rights and copyrights that developed in U.S. law sprang simply from the accidents of the legislative history of the first Federalist administration. Quite the contrary. The readiness of members of Congress to deal separately with petitions for grants of copyright, as state legislatures before them had done, reflected the long antecedent evolution of the law of copyright.
In its late medieval origins, as noted, the copyright privilege had nothing to do with the encouragement of intellectual creativity or originality of expression. Indeed, the very notion of claiming originality of authorship was a Renaissance departure from the scholastic tradition of seeking to cloak one's own ideas with the authority of Aristotle and the other ancients (see, for example, Long, 1991). Rights of literary property involving published works remained legally unprotected in Europe until the fifteenth century, when the introduction of the printing press made the rewards of publishing, or plagiarism, far greater than ever before. The new technology of printing also transformed the economics of the copying business by substantially increasing the disparity between the cost of the first (printed) copy and the unit cost of subsequent copies. Copyright law, from the beginning, has been shaped more by the economics of publication than by the economics of authorship (see Patterson, 1968; Plant, 1974:Ch.4).
Like the earliest patents of invention, the first known copyrights appeared in Renaissance Italy. By the end of the 1460s the craft of printing had been introduced in Rome and Venice, and, with the issuance during 1469-1517 of a series of privileges relating to books and printing by the Venetian Cabinet, Senate, and other governmental bodies, Venice quickly assumed the lead in Italian printing. These privilegi included importation franchises, the first of which (1469), as noted, awarded the German printer Johann von Speyer the exclusive privilege of conducting all printing in the city for five years in return for establishing the craft (see Prager, 1944:715). There soon followed monopolies in the form of exclusive licenses to print or sell an entire class of books for a stipulated term, prohibitions of the importation of books printed abroad, and patents for the improvement of printing and typography (see Bugbee, 1967:43-44). The question of rights of authorship was largely disregarded because much of the demand was for extant works (such as the Bible) that were in the public domain and whose authors, even when identified, were long since dead.
Toward the end of the century, however, some privileges were awarded for the protection of authors that did have the character of modern copyrights: In 1486 the historiographer of the Republic was granted exclusive control over the publication of his work. In 1493 the Venetian Cabinet gave Daniele Barbaro an exclusive 10-year grant of proprietary rights to the publication of a book authored by his deceased brother (Bugbee, 1967:45). More typical were the copyrights issued to editors and publishers for individual works written by others; these were petty monopolies prohibiting publication of the work without permission of the grantee. Publishers were soon flocking to the government to reserve well-known titles for themselves, in the hope of either publishing themselves or selling the right later
to another printer. By 1517 the resulting shortage of available titles caused the Senate to restrict all such copyright privilegi henceforth to "new and previously unprinted works."
What was probably the first general copyright law in the world came in the form of a decree issued by the Council of Ten in Venice (1544-1545) that prohibited the printing of any work unless written permission from the author or his immediate heirs had been submitted to the Commissioners of the University of Padua. No provision was made, however, for maintaining a register of protected works (Bugbee, 1967:46). This decree was prompted by the continued unauthorized printing of works for which copyrights had been granted. A further measure directed toward more complete regulation of the printing business came in 1548-1549 with a Council decree establishing a guild into which all Venice's printers and booksellers were to be organized. An added motivation was to assist the Church in suppressing heretical literature. The same concern with censorship of a potentially dangerous new medium of communication, rather than securing the rights of authorship, prompted the royal officials of sixteenth century France to issue licenses, or privileges, for the publication of acceptable books. The French Crown, however, proved better able than the Italian city-states to resist the Church's efforts to share control of the printing business.
In the Netherlands, privileges resembling those of Venice, but without censorship provisions, were issued to publishers by state and central governments, but the primary means of regulating destructive competition involving the pirating of texts was a system of informal noninterference agreements among Dutch printers. Similar arrangements had developed among leading German publishers and were exercised through a guild and the book fairs of Frankfurt and Leipzig (see Bugbee, 1967:48). When the German book trade was interrupted during the Thirty Years' War, the Dutch quickly assumed leadership of the publishing industry in Europe. Although the flourishing printing business of the Netherlands benefited from the attraction of scholars to the comparatively free intellectual atmosphere of the Dutch towns in this era, protection of local authors' rights was not a concern, nor were the rights of foreign authors and publishers. At this time, throughout Europe, imported books, pamphlets, and pictorial material were subject to reprinting and sale without compensation for their originators. The highly successful Elzevir family of Leyden and Amsterdam was especially notorious in this regard. According to Henry Haven Putnam (quoted by Bugbee, 1967:178, n.150):
As far as the foreign authors were concerned, the Elzevirs appear to have followed simply the dictates of their own convenience and advantage. They took what material they thought they could use, without troubling themselves to make either requests or acknowledgements. They were, in fact, the most extensive piratical publishers that the world had as yet seen, and may be said to have reduced piracy to a business system.
The Venetian printers' guild was the model for England's Stationers' Company, which was chartered by Mary Tudor in 1557. The object was to provide the Catholic sovereign with the instrumentality to control what could be printed for widespread circulation. Masters of the Company were empowered to search the premises of any printer or bookseller for works not printed in accordance with the licensing laws, and whether censorship was obnoxious or desirable in their opinion, they had a strong economic motive to enforce their monopoly by suppressing publications not licensed by the Crown. Indeed, it has been suggested that censorship in England, particularly in the mid-seventeenth century, was more a product than a cause of the Stationers' monopoly (see Patterson, 1968:101; Plant, 1974, on the Stationers' petition of 1643).
Thus, in England, copyrights began with a monopoly franchise granted for the purpose of regulating the business of printing and publishing. They had nothing to do with the encouragement of "freedom of expression," nor were they intended to promote authorship per se. Nevertheless, authors in England had personal property rights in their unpublished manuscripts, as well as contractual protections under the common law. These protections extended to a recognized interest in the integrity of the form and content of the work for which publication permission had been given, which restrained printers from making arbitrary alterations in texts once they were published and from dispensing with the need to recompense the author. In short, under these arrangements a stationer (i.e., a printer-copyist) had to obtain the author's permission to publish his manuscript even though the author did not hold the copyright (see Patterson, 1968:65-69).
The modern statutory protection of authors' copyrights in the United States and Britain arose in the early eighteenth century, almost as an accident. In England during the closing decades of the seventeenth century, the passing of the era of political and religious censorship made it increasingly difficult for the Stationers' Company to interest the government in the control of the new printing presses that were springing up throughout the country. When the Licensing Act that had given teeth to the Stationers' monopoly was allowed to lapse in 1694, the competition intensified as country booksellers openly flouted the doctrine of perpetual copyright that the Stationers' Company had sought to establish on the evidence of assignments registered in its record books. After 15 years of increasingly chaotic conditions of unregulated competition, the London printer-booksellers at last managed to secure new legislation, in the form of the 1709-1710 act of Queen Anne. This, the first copyright statute, did not give the publishers the perpetual rights they had sought; instead, it limited the exclusive right to printing new books registered with the Stationers' Company to a term of 14 years (following the precedent established in the case of patents under the Statute of Monopolies of 1623); and it gave the holders of copyrights on existing books the sole right to print for 21 years. Moreover, to open up the trade,
the Act of Anne eliminated the guild monopoly on the holding of copyrights. Anyone could now hold the copyright for a new work—printers, bakers, cobblers, and even authors.
From the foregoing brief account of the origins of copyrights, it is evident that the signal distinction between the protection of ideas under patent law and the protection of expression under the law of copyright owes a great deal to the fact that copyrights arose in response to internal and external interests in regulating the nature of competition in the printing and publication business, an industry in which, at an early date, decreasing costs were thought to be a source of instability. Copyrights, therefore, were inherently concerned with the security of property rights in the expression of ideas—whether old ideas or new ones. Only much later did they come to be enlisted in the cause of stimulating the production of new knowledge. Is it so surprising, then, that in this new role they sometimes are found to perform rather awkwardly? Consider, as a simple case in point, the recent assignment of copyright law to the task of protecting intellectual property rights in computer software. Observers have noted that the protection afforded to original expression in copyright law offers no security for originators of novel algorithms and concepts for applications programs (such as spreadsheets and relational data bases). Yet at the same time, the opportunities that the law creates to protect original expression have had the effect of encouraging an excessive degree of variety in the "look-and-feel" of software, whereas some greater degree of standardization of the machine-user interface is widely thought to be desirable from the standpoint of economic efficiency (see, for example, Farrell, 1989; David and Greenstein, 1990).
Historical studies reveal that although patents, copyrights, and legal protection of trade secrets have been recognizable institutions in Western societies for centuries, policies bearing on the protection accorded to intellectual property, and the juridical-institutional arrangements used to implement them, have been a mutable thing, adapted over time and across societies to the perceived needs and advantages of interested parties. The adaptations in each form of protection, moreover, have occurred within the historical context of other, related institutional arrangements affecting the costs and benefits of maintaining specific intellectual property rights. Thus, the effort to institute a uniform international regime for the protection of intellectual property rights is almost certain to cause conflict and controversy. Even though a new intellectual property regime could be Pareto improving in some situations, the need to align domestic and international laws adds further constraints that tend to render such solutions impractical. As a
result, discussions of the "correct" international system for protecting intellectual property are more likely than not to degenerate into rhetorical efforts to impose institutional arrangements that may be well adapted to the national purposes and legal contexts of one country (or several similar countries) on societies that are quite different in those respects.
The supposed trade-off between promoting technological progress and technology diffusion has led to the view that strong protection of intellectual property rights must serve the former goal, at the expense of the latter. This has been a rationale for the conflicts between the technologically advanced and the developing nations over intellectual property issues: An interest in weak or minimal protection of intellectual property is imputed to the developing countries' limited capacity to innovate technologically and their comparative advantage in imitating the products and processes originated elsewhere. Yet that is not necessarily the case. Indeed, just the opposite point may be made on modern theoretical grounds and by reference to historical experience: legal protection of intellectual property rights in the form of state-sanctioned monopoly franchises can have seriously detrimental consequences for the processes of discovery and invention, whereas it may be instrumental in bringing about the successful transfer and commercial application of new scientific and technological knowledge. The arguments supporting this unorthodox contention are summarized below.
First, because invention is often a cumulative process, as scientific inquiry more generally is recognized to be, the enforcement of patent rights can interfere with further discovery. It deflects resources into "racing" for the priority prize and into inventing around the basic patent. It discourages complementary inventions, because the returns may be extracted by the patentee whose work has been built upon. Note the distinction made here between inhibiting progress rather than discouraging investment in R&D.
Second, weak and narrow patents, as in the modern Japanese system, encourage firms to cross-license and thus disseminate findings rapidly. They encourage the collective invention process—in the direction of elaboration and adaptation to particular markets, although they may discourage efforts to achieve radical, fundamental inventions (see Ordover, 1991, and references therein). This is consistent with one aspect of Kitch's (1980) "prospect" argument that broad, strong patents encourage fundamental innovations and their orderly development, but it contests the premise in the latter that a monopolist can identify and efficiently contract for the performance of cumulative, elaborative research.
Third, although it is arguable that weak patent protection regimes encourage exchanges of patent licenses among firms that are symmetrical in their technological capabilities, the opposite is more likely to be the case in regard to transfers of technical know-how from more to less capable organizations. Much of a firm's capability for absorbing and implementing pat-
ented innovations depends on its access to tacit knowledge that is complementary to the patent. If it does not possess such knowledge already because it has no experience base in the area and cannot readily hire skilled personnel from firms that do, it must contract for the required information. However, tacit information is extremely difficult to contract for, due to the problems of informational asymmetry and monitoring costs. This problem bedevils North-South technology transfers that do not involve the mediation of multinational organizations. Nevertheless, a regime of strong intellectual property protection of codified knowledge in the receiving country provides a basis—as Arora (1991) has recently demonstrated—for structuring contracts that would accomplish the transfer of uncodified and tacit knowledge that is necessary for the profitable operation of industrial processes yet remains undisclosed to the public by the patent or copyright licenser. Further, contractual arrangements to transfer tacit knowledge as part of the terms for the licensing of the use of codified and published information (e.g., in a patent) generally will require enforcement of legal protection for trade secrets, in the interests of both the licenser and the licensee.
Fourth, intellectual property rights in the form of exclusive franchise guarantees can overcome failure to exploit a patent through modifications to local market conditions due to the problem that learning of this kind will not be appropriable and, hence, there is less interest in generating learning-by-doing gains, as David and Olsen (1991) have shown. History reinforces the implications of this line of theoretical analysis by revealing extensive early use of patent privileges to encourage technology importation, both in medieval and early Renaissance Europe and in late nineteenth century Latin American countries.
A reading of the historical chronicle of the evolution of intellectual property institutions underscores several further propositions. First, the protections accorded intellectual property by nation-states have not manifested any great consistency in adhering to pure principle. Rather, they have been pragmatically altered over time in response to changing perceptions of the way the creation and dissemination of information and information products affect "national interests." They also have been tinkered with periodically to remedy unanticipated problems in the workings of institutional arrangements due to changes in the technologies employed to produce and distribute information products. Much of the late nineteenth century "reform" in national and international copyright law, for example, was provoked by developments in the technology of printing that underlay the cutthroat competition for mass markets (by the standards of the day) in cheap editions of popular novels.
Second, to be effective, statutory protections and judicial interpretations of laws defining intellectual property rights must fit within—and be compatible with—the principles of the larger framework of a society's legal
institutions. Formal legal precedents and informal conventions impose historical constraints on a country's ability to fine-tune its intellectual property institutions to suit currently perceived needs. These constraints would remain even if there were widespread agreement as to the needs of the moment.
The two foregoing observations imply, in my view, a third conclusion: Proposals now being advanced to establish a uniform international regime of intellectual property protection are not practical, even though careful economic analysis would indicate that there may be considerably more points of agreement between the interests of the technologically advanced and the economically developing countries than often has been supposed.
Finally, U.S. assertion of the justice of striving to protect the "natural" ownership rights of creators of intellectual property, and its unwillingness to grant other nations any quid pro quo for accepting a uniformly strong international regime for protecting international property production, reflects confusions of French and British legal doctrines that are part of the American heritage concerning the subject of intellectual property. It is also quite inconsistent with some aspects of the past conduct of the United States and that of other economically advanced countries—in the enforcement of intellectual property claims in the international arena.
In preparing this paper I have benefitted from the able research assistance of Weston Headley and Phillip Lim. Editorial suggestions by the National Research Council staff contributed significantly to improving the exposition. I can claim full rights only to any errors and omissions that remain.
Arora, Ashish. 1991. The transfer of technological know-how to developing countries: Technology licensing, tacit knowledge, and the acquisition of technological capability. Unpublished Ph.D. dissertation, Stanford University.
Arrow, Kenneth J. 1962. Economic welfare and the allocation of resources for invention. Pp. 609-625 in The Rate and Direction of Inventive Activity: Economic and Social Factors. Princeton, N.J.: National Bureau of Economic Research.
Beck, Roger L. 1981. Competition for patent monopolies. Research in Law and Economics 3:91-110.
Beck, Roger L. 1983. The prospect theory of the patent system and unproductive competition. Research in Law and Economics 5:193-209.
Bender, David. 1986. Protection of computer programs: The copyright/trade secret interface. University of Pittsburgh Law Review 47:906.
Benko, Robert P. 1987. Protecting Intellectual Property Rights: Issues and Controversies. Washington, D.C.: American Enterprise Institute for Public Policy Research.
Berg, Jeff. 1991. Moral rights: A legal, historical and anthropological perspective. Intellectual Property Journal 6:341-376.
Besen, Stanley M., and Leo J. Raskind. 1991. An introduction to the law and economics of intellectual property. Journal of Economic Perspectives 5(1):3-27.
Bugbee, Bruce. 1967. Genesis of American Patent and Copyright Law. Washington, D.C.: Public Affairs Press.
Bugos, Glenn E., and Daniel Kevles. 1991. Plants as intellectual, property: American practice, law, and policy in world context. Mimeo (forthcoming in Osiris, 1992).
Cheung, Steven N.S. 1982. Property rights in trade secrets. Economic Inquiry 20(1).
Chisum, D. 1989. Patents. New York: Matthew Bender.
Cooter, Robert, and Lewis Komhauser. 1980. Can litigation improve the law without the help of judges? Journal of Legal Studies 9.
Cooter, Robert, and Thomas Ulen. 1988. Law and Economics. Glenview, Ill.: Scott Foresman and Company.
Dasgupta, Partha, and Paul A. David. 1987. Information disclosure and the economics of science and technology. G. Feiwel, ed., Kenneth Arrow and the Ascent of Economic Theory, New York: MacMillan.
Dasgupta, Partha, and Paul A. David. 1988. Priority, Secrecy, Patents, and the Economic Organization of Science and Technology. Center for Economic Policy Research, Publication no. 127, Stanford University, forthcoming in Science in Context.
Dasgupta, Partha, and Paul A. David. 1990. The New Economics of Science. High Technology Impact Program Working Paper, Center for Economic Policy Research, Stanford University.
Dasgupta, Partha, and Joseph E. Stiglitz. 1980. Industrial structure and the nature of innovative activity. Economic Journal 90:266-293.
David, Paul A. 1991. Reputation and Agency in the Historical Emergence of the Institutions of "Open Science." Center for Economic Policy Research, Publication no. 261, Stanford University.
David, Paul A., and Shane Greenstein. 1990. The economics of compatibility standards: An introduction to recent research. Economics of Innovation and New Technology 1(1-2):343.
David, Paul A., and Trond E. Olson. 1989. Innovation Diffusion, Learning-by-Doing, and the Optimum Patent Life. Center for Economic Policy Research, Technical Paper no. 170, Stanford University.
David, Paul A., and Trond E. Olson. 1991. Technology Adoption, Teaming Spillovers, and the Optimal Duration of Patent-Based Monopolies. Working paper, Stanford University, forthcoming in International Journal of Industrial Organization.
Davis, Michael. 1989. Patents, natural rights and natural property. In Owning Scientific and Technical Information, Vivian Weil and John Snapper, eds. London: Rutgers University Press.
Dworkin, Gerald. 1981. The moral right and English copyright law. International Review of Industrial Property and Copyright Law 12(4).
Farrell, Joseph. 1989. Standardization and intellectual property. Jurimetrics Journal 30(Fall):3550.
Federico, P.J. 1929a. Origin and early history of patents. Journal of the Patent Office Society 11:292-305.
Federico, P.J. 1929b. Colonial monopolies and patents. Journal of the Patent Office Society 11:358-365.
Folster, S. 1985. Government coordination of competitive R&D projects. D. Phil. thesis, University of Oxford.
Friedman, David D., William M. Landes, and Richard A. Posner. 1991. Some economics of trade secret law. Journal of Economic Perspectives 5(l):61-72.
Frumkin, Maximilian. 1945. The origin of patents. Journal of the Patent Office Society 27(March): 143-149.
Gilbert, Richard, and Carl Shapiro. 1990. Optimal patent length and breadth. RAND Journal of Economics 21(1):106-112.
Gilfillan, S.C. 1964. Invention and the Patent System, Materials Relating to Continuing Studies of Technology, Economic Growth, and the Variability of Private Investment. Joint Economic Committee, Congress of the United States. Washington, D.C.: U.S. Government Printing Office.
Goodman, John. 1978. An economic theory of the evolution of the common law. Journal of Legal Studies 7.
Gould, Stephen Jay. 1980. The Panda's Thumb. New York: W.W. Norton.
Hilaire-Perez, Liliane. 1991. Invention and the state in 18th century France. Technology and Culture 32(4):911-931.
Hill, Thomas A. 1924. Origin and development of letters patent for invention. Journal of the Patent Office Society 6:405-422.
Hirschliefer, J., and J.G. Riley. 1979. The analytics of uncertainty and information: An expository survey. Journal of Economic Literature 17:1375-1421.
Jager, Melvin F. 1991. Trade Secrets Law, vol. 1. New York: Clark Boardman Company.
Jefferson, Thomas. 1972. The Life and Selected Writings of Thomas Jefferson, Adrienne Koch and William Peden, eds. New York: The Modem Library.
Johnson, W.R. 1985. The economics of copying. Journal of Political Economy 93:158-174.
Kaufer, Erich. 1989. The Economics of the Patent System. Chur, Switzerland: Harwood Academic Publishers, 1989.
Kitch, Edmund W. 1977. The nature and function of the patent system. Journal of Law and Economics 20:265-290.
Kitch, Edmund W. 1980. The law and economics of rights in valuable information. Journal of Legal Studies 9:683-724.
Klemperer, Paul. 1990. How broad should the scope of patent protection be? RAND Journal of Economics 21(1): 113-130.
Landes, William M., and Richard A. Posner. 1989. An economic analysis of copyright law. Journal of Legal Studies 18:325-366.
Liebowitz, S.J. 1985. Copying and indirect appropriability: Photocopying of journals. Journal of Political Economy 93:945-957.
Long, Pamela. 1991. Invention, authorship, ''intellectual property," and the origins of patents: Notes toward a conceptual history. Technology and Culture 32n .4:pp.846-884.
Lubar, Steven. 1990. New, useful, and nonobvious. American Heritage of Invention and Technology 6(1):8-16.
Machlup, Fritz, and Edith Penrose. 1950. The patent controversy in the nineteenth century. Journal of Economic History 10(1):1-29.
Machlup, Fritz. 1958. An Economic Review of the Patent System. Study no. 15 of the Subcommittee on Patents, Trademarks, and Copyrights of the Committee on the Judiciary, United States Senate, 85th Congress, Second Session. Washington, D.C.: U.S. Government Printing Office.
MacLeod, Christine. 1991. The paradoxes of patenting: Invention and its diffusion in 18th- and 19th-century Britain, France, and North America. Technology and Culture 32(4):885-911.
Mandich, Giulio. 1958. Principi riconoscimenti veneziani di un diritto di privative agli inventori. Rivista di diritto industriale 7:101-155.
Marx, Karl. 1970. Capital (translated into English by Moore and Aveling). London: Lawrence and Wishart.
Merges, Robert P., and Richard R. Nelson. 1990. Market Structure and Technical Advance: The Role of Patent Scope Decisions. CCC Working Paper no. 90-100, University of California at Berkeley.
Mody, Ashoka. 1990. The new international environment for intellectual property rights. Pp. 203-239 in Intellectual property rights in science, technology, and economic performance: International Comparisons, Francis W. Rushing and Carole Ganz Brown, eds. Boulder, Colo.: Westview Press.
Nelson, Richard R. 1990. What Is Public and What Is Private About Technology? CCC Working Paper no. 90-9, University of California Center for Research in Management.
Nordhaus, William D. 1969. Invention, Growth, and Welfare: A Theoretical Treatment of Technological Change. Cambridge, Mass.: MIT Press.
North, Douglass C. 1991. Institutions. Journal of Economic Perspectives 5(1):97-112.
Novos, I.E., and M. Waldman. 1984. The effects of increased copyright protection: An analytical approach. Journal of Political Economy 92:236-246.
Office of Technology Assessment. 1986. Intellectual Property Rights in an Age of Electronics and Information. Washington D.C.: U.S. Congress.
Ordover, J.A. 1991. A patent system for both diffusion and exclusion. Journal of Economic Perspectives 5(1):43-60.
Patterson, Lyman. 1968. Copyright in Historical Perspective. Nashville: Vanderbilt University Press.
Pavitt, Keith. 1987. The objectives of technology policy. Science and Public Policy 14:182-188.
Phillips, Jeremy. 1982. The English patent as a reward for invention: The importation of an idea. Journal of Legal History (May):75-76.
Plant, Arnold. 1934. The economic theory concerning patents for inventions. Economics l(February): 167-195.
Plant, Sir Arnold. 1974. Selected Economic Essays and Addresses. London: Routledge & Kegan Paul. (Chapter 4 reprints the author's essay "The Economic Aspects of Copyrights in Books," 1934.)
Polanyi, Michael. 1944. Patent reform. The Review of Economic Studies 1(1):61-76.
Prager, Frank. 1944. A history of intellectual property from 1545 to 1787. Journal of the Patent Office Society 26(November):711-760.
Prager, Frank. 1946. Brunelleschi's patent. Journal of the Patent Office Society 28(February):109-135.
Prager, Frank. 1948. Translation of Giulio Mandich, "Venetian Patents (1450-1550)." Journal of the Patent Office Society 30(March):166-224.
Prager, Frank. 1960. Translation of Giulio Mandich, "Venetian Origins of Inventors' Rights." Journal of the Patent Office Society 42(June):378-382.
Priest, George L. 1977. The common law process and the selection of efficient rules. Journal of Legal Studies 7.
Priest, George L. 1986. What economists can tell lawyers about intellectual property: Comment on Cheung. Research in Law and Economics 8:19-24.
Rosenberg, Nathan. 1990. Why do companies do basic research with their own money? Research Policy 19:165-174.
Rubin, Paul. 1977. Why is the common law efficient? Journal of Legal Studies 6.
Rushing, Francis W., and Carole Ganz Brown. 1990. Intellectual Property Rights in Science, Technology, and Economic Performance: International Comparisons. Boulder, Colo.: Westview Press.
Samuelson, Pamela. 1984. CONTU Revisited: The case against copyright protection for computer programs in machine-readable form. Duke University Law Journal.
Scherer, F.M. 1972, 1984. Nordhaus' theory of optimal patent life: A geometric reinterpretation. American Economic Review 62:422-427. Pp. 130-141 in Innovation and Growth: Schumpeterian Perspectives. Cambridge, Mass.: MIT Press.
Schmitz, James A. 1989. On the breadth of patent protection. Mimeo, SUNY-Stony Brook.
Scotchmer, Suzanne. 1991. Standing on the shoulders of giants: Cumulative research and the patent law. Journal of Economic Perspectives 5(1):29-41.
Scotchmer, Suzanne, and Jerry Green. 1990. Novelty and Disclosure in Patent Law. RAND Journal of Economics 21(1):131-146.
Siebeck, Wolfgang E. 1990. Strengthening Protection of Intellectual Property in Developing Countries: A Survey of the Literature, Robert E. Evenson, William Lesser, and Carlos A. Primo Braga, eds. Washington, D.C.: World Bank.
Smith, Adam. 1776, 1976. The Wealth of Nations, R.H. Campbell, A.S. Skinner, and W.B. Todd, eds. Oxford: Clarendon Press.
World Intellectual Property Organization. 1988. Existence, Scope and Form of Generally Internationally Accepted and Applied Standards/Norms for the Protection of Intellectual Property. WO/IN-F 129. Geneva: WIPO.
World Intellectual Property Organization. 1989. General Information. Geneva: WIPO.
Wright, Brian D. 1983. The economics of invention incentives: Patents, prizes, and research contracts. American Economic Review 73(4):707.