Economic Theories About the Costs and Benefits of Patents
Richard D. Nelson and Roberto Mazzoleni, Columbia University
This paper provides a broad overview of theories about the principal costs and benefits of patents and discusses assumptions about the contexts in which inventions are made or developed. Consideration of different contexts suggests that patents play different roles in different technologies and sectors. In some contexts, several patent theories have a degree of plausibility; in others, none of them is very plausible. The question is under which conditions the theories make sense and under which they do not.
In 1958, Fritz Machlup reviewed how economists view the patent system. He reported that economists tended to be negative about the value of the patent system to society, reflecting their concern that patents generate monopolies and that, in many cases, patents are not even necessary to encourage invention. His own position, however, was that there were no good models to replace the patent system and that it serves some useful purposes.
In this paper we identify four broad theories about the principal purposes of patents:
Invention-inducement Theory: The anticipation of receiving patents provides motivation for useful invention.
Disclosure Theory: Patents facilitate wide knowledge about and use of inventions by inducing inventors to disclose their inventions when otherwise they would rely on secrecy.
Development and Commercialization Theory: Patents induce the investment needed to develop and commercialize inventions.
Prospect Development Theory: Patents enable the orderly exploration of broad prospects for derivative inventions.
The four theories are not necessarily mutually exclusive. The anticipation of a patent might stimulate an invention, and the holding of a patent might stimulate its subsequent development. But some versions of the theories are at odds. For example, one version of the disclosure theory assumes that inventions will occur without patents and that the existence of patents widens their use. That is quite the opposite of the most familiar version of the invention-inducement theory, which assumes that invention is motivated by the anticipation of patents.
In general, however, the theories differ in the assumptions that they make about the conditions under which inventions are made, developed, or commercialized. The assumptions are made about the following conditions:
The nature and effectiveness of means other than patents to induce invention and related activities.
The likelihood of a group of potential inventors to work on diverse and noncompeting ideas or to be focused on a single alternative or a set of closely connected ones.
The transaction costs of licensing an invention with and without patents.
Whether the multiple steps in the invention, development, and commercialization of a new technology tend to proceed within a single organization or several organizations tend to be involved at different stages of the process.
The topography of technological advance—how inventions are linked to each other both temporally and as systems.
We will examine those assumptions to show the strengths and weaknesses of all the theories.
The theory that patents motivate useful invention is the most familiar theory of the benefits of patenting. Indeed, much discussion about the benefits of patents proceeds as though motivating useful invention were the only social purpose served by patents and patents always serve this purpose. In fact, as explained later, the situation can be much more complicated in many cases.
All versions of the invention-inducement theory presume either that if there is no patent protection there will be no invention or, more generally, that without a patent system incentives for invention will be too weak to reflect the public interest. In particular, they assume that stronger patent protection will increase the amount of invention.
Under what might be called the canonical versions of the invention-induce-
ment theory—versions associated with the models of economists Arrow (1962), Nordhaus (1969), and Scherer (1972)—inventors, as a group, are implicitly assumed to be diverse, working on different and generally noncompeting things. Thus in the absence of redundant efforts that might occur if many groups worked on competing things, stronger patent protection results in a greater number of useful inventions.
Most articulations of the invention-inducement theory presume that an invention is used or sold by the firm that made the invention. However, Arrow (1962) and more recently Merges (1995) and Arora and others (1994) address the problem that an inventor has in selling an invention to someone else in the absence of legal property rights to it and have taken the position that strong property rights to an invention reduce the transaction costs of licensing it. Strong patents would then also serve the purpose of providing incentives to invent for parties who are limited in the extent to which they can use the invention themselves, by facilitating the sale of rights to an invention.
In most versions of the invention-inducement theory it is assumed, generally implicitly, that the social benefit of a particular invention is strictly its final use value; the social benefit of patent protection stems, therefore, from the additional invention induced by the prospect of a patent. And the social cost of a patent is the restriction on the use associated with the monopoly power lent by a patent. That formulation of the invention-inducement theory leads naturally to analysis of optimal patent ''strength,'' defined as duration (Nordhaus 1969; Scherer 1972), or breadth (Klemperer 1990), and the tradeoff between the amount of increased invention induced by greater patent strength and the increased costs to society associated with the stronger monopoly position of the patent holder (see also Gilbert and Shapiro 1990).
The issues of the consequences of greater patent duration or scope are more complicated if an invention is not only useful as is, but also provides the basis for second-generation inventions. Arrow especially called attention to the possibility that the principal use of some inventions is as input for further inventions (such as PCR, see Chapter 5). Van Dijk (1994) considers what he calls the height of a patent, by which he means the extent to which the patent controls later improvements and variations in the initial invention. Those considerations lead us to the development and commercialization theory, and prospect-development theory, which are discussed later.
We stress that the version of the invention-inducement theory that we have been considering up to now presumes that more inventive effort and more inventors mean more useful inventing. The theory takes on a different look if, instead, all inventors are assumed to be focused on the same set of paths to invention. This assumption gives rise to the "patent race models" of Loury (1979) and Dasgupta and Stiglitz (1980a) and, if the assumed common focus is on a broader but still limited "pool" of invention prospects, to the "overfishing" models of Barzel (1968) and Dasgupta and Stiglitz (1980b). Under either model, patents no
longer provide an unambiguous benefit when there are increases in the total inventive effort exerted at any one time or in the number of persons engaged in inventive activity. If inventors perceive that other inventors are in the game, they will see that their returns depend not simply on whether they achieve an invention, but on whether they achieve it first. That might induce them to invest their resources faster or more widely than would be appropriate if the objective were defined simply in terms of achieving a particular invention most efficiently. If inventors are unaware of or ignore the presence of other inventors in the game and calculate their expected benefits as though they had no competition, then from the vantage point of standard welfare economics there will be "too many" inventors playing the game. Under either model, patents will constitute a winner-take-all system. The lure of a patent, therefore, induces inefficient inventive effort in a competitive context. Another possibility suggested by Dasgupta and Stiglitz (1980a) is that the recognition that others are likely to be running toward a particular objective will deter parties from engaging in inventive work in a given field.
It seems that a consequence of that kind of invention inefficiency induced by strong patents would be to shift the tradeoff between the benefits and costs of stronger patents so as to increase the costs. Thus, society ought to opt for stronger patents in fields in which stronger intellectual property protection yields a larger flow of valuable inventions, rather than in fields in which stronger patents lead largely to more hounds barking up the same tree. And that is the case in the model of optimal patent duration developed by McFetridge and Rafiquzzaman (1986).
In any case, under the invention-inducement theory, the basic presumption is that if the award of a patent is not necessary to induce an invention, then it is not in the social interest to offer or grant a patent. That raises the question, "How important is the anticipation of a patent in the inducement of inventing?"
The best empirical evidence suggests that among firms engaged in R&D, patents are an important part of the inducement for invention in only a small number of industries (for example, see Levin and others 1987). Mansfield (1986) asked chief R&D executives of 1,000 US manufacturing firms to identify the fraction of inventions developed by their firms between 1981 and 1983 that they would not have chosen to develop if they had been unable to obtain patent protection. For electrical equipment, primary metals, instruments, office equipment, motor vehicles, and several other industries, the fraction would have been less than 10%. Executives in those companies rated what they could gain by patents as much less important than the advantages that go with a head start in reaping returns from their inventions. Pharmaceuticals and fine chemicals were important exceptions: executives in the pharmaceutical industry reported that without patent protection 60% of their new pharmaceuticals would not have been developed, and the reduction in "other chemicals" would have been about 40%.
However, the studies mentioned above focused on large and medium firms
with R&D laboratories, which are for the most part in a position to exploit their inventions by using them themselves. They do not depend on sale or license of their inventions for returns. For inventors who must sell or license to reap returns, patents might be far more important. It appears that in some sectors patents are an essential part of the inducement for inventing.
There is a proliferation of empirical work on what patents are about, who uses them, and how important they are. This work is motivated entirely by the invention-inducement theory; that is why a lot of what was learned in the work might not be relevant to research tools.
The case can certainly be made that for many university inventions that were funded with public money, the policy implication of the invention-inducement theory is that a patent should not be granted if it is not necessary to grant a patent to get an invention. The results of research would be published in any case. In many instances, firms will have ample incentive to work with and develop what comes out of university research. They can usually patent their developments, gain the advantage of a head start in the market, or both. No grant of an extensive license based on anticipated economic changes is needed to motivate this work. Moreover, the presence of a patented invention, with a requirement for would-be developers to get a license to do further work on the original idea, restricts the number of parties that will engage in that work. The latter argument against patenting seems particularly strong if potential developers are diverse in the directions that they might follow and if the licensing arrangement of preliminary ideas, whose ultimate commercial value is unclear, are not easy to work out.
The primary issue raised by the disclosure theory is not so much whether strong patents encourage more inventing, but rather how inventors reap the returns from their inventions. It presumes that secrecy is possible and sufficient to induce invention but that society is better off granting intellectual property rights and getting disclosure in exchange. A patented invention would thus be available for uses that the inventor did not know about or was not in a position to implement. Under this theory, a patent both advertises the presence of an invention and facilitates licensing. That argument, in effect, turns the invention-inducement theory on its head: patents are not necessary to induce invention, but rather what patents do is encourage disclosure and, given some assumptions about the transaction costs of licensing the invention, enable it to be used more widely than it would be without a patent.
Figuring out what makes a newly-marketed product work is substantially easier than the initial invention and development of that product. Thus, as suggested by Levin and others (1987), secrecy would seem more effective for process than for product inventions. From that perspective, the most relevant domain of the disclosure theory might be process inventions. Various studies have
shown that in some industries, firms customarily engage in general cross licensing of their process technologies, a sharing of technology that might not occur if patents on processes were not available.
However, a broader view of the disclosure theory opens the door to a wider possible range of applicability. The possession of a patent, rather than simple reliance on nonpatent measures to reap returns, might make a firm more willing to advertise its inventions and to contract to give technological information and assistance to a noncompeting firm to help it to adapt the invention to its own uses. This variant of the theory is a close kin to the version of the invention-inducement theory in which inventors have limited ability to exploit an invention themselves. Under this variant, the possession of a patent might make the firm more receptive to proposals by firms in other lines of business to develop the invention for different uses. Such possibilities lead to the development and commercialization theory and the prospect-development theory.
DEVELOPMENT AND COMMERCIALIZATION THEORY
In its simplest version, the theory that patents induce the development and commercialization of inventions seems to be a variant of the invention-inducement theory, but with patenting occurring early in the process of inventing and with much additional work needed before the crude "invention" is ready for actual use. A patent at an early stage is seen as providing the assurance that if the development is technologically successful, its economic rewards will be capturable, thus inducing a decision to develop it.
Rebecca Eisenberg has called our attention to a supplement to this argument: that the possession of a patent enables the patent holder to go to capital markets for development financing. That capability might be important for a small firm faced with large development costs before it can get its invention to market.
The development and commercialization theory is different from the invention-inducement theory for circumstances in which one organization does the early inventing work but is not in a position to do the development work. The original inventor's possession of a patent then facilitates handing off the task to an organization better situated for development and commercialization. Years ago, Willard Mueller (1962) pointed out that many of DuPont's product innovations were based on inventions bought from smaller firms. Similarly, in the 1920s, General Electric bought and developed many inventions made by private inventors or small firms (see Reich 1985).
The development and commercialization theory was widely cited in the discussions that led to the Bayh-Dole act, which gave universities the patent rights on inventions that emanated from their government-funded research projects. The proposition was that although the inventions had been achieved with public funding, they would serve no economic purpose until they were developed to a point where they were commercial, and only companies were able to undertake
such development. That constituted a separation of the site of invention from the site of development. Under the version of the development and commercialization theory most clearly articulated in the discussion that led to the Bayh-Dole act, a company would be unlikely to engage in development of a university invention unless it held proprietary rights. If universities held strong patent rights, they would be in a position to sell exclusive licenses. In contrast, if there were no patents, or if the government held them with a commitment to nonexclusive licensing, companies would be unlikely to invest in the necessary development work.
Another interpretation of the development and commercialization theory is that the possession of a patent gives the original patent holder—a university or small firm—an incentive to push its inventions out to firms that can develop and commercialize them. That is basically an extension of the version of the disclosure theory discussed above. It is a different view of the development and commercialization theory from the one that implicitly assumes that without a strong initial patent a firm will not undertake the development work necessary to lead to a profitable product or process innovation.
PROSPECT DEVELOPMENT THEORY
A number of years ago, Edmund Kitch proposed a prospect-development theory of the societal benefits of patents. Like the development and commercialization theory, it proposes that the utility of a patent comes after an initial invention is made.
Kitch's theory was that having a broad patent on an initial invention enabled the patent holder to orchestrate development of a technological prospect in various dimensions, whereas development of an initial invention that was freely available to all would be chaotic, duplicative, and wasteful.
The theory that patents enable orderly development of broad technological prospects differs from the development and commercialization theory in suggesting that a wide range of developments or inventions might become possible if the initial invention is available as an input—through either development or modification in different directions. Many university inventions, particularly research tools, are of this sort.
An implicit feature of various versions of the development and commercialization theory is that although important resources and risk-taking might be needed to develop an invention, there is essentially one product at the end of the commercial rainbow. In its most common formulation, the development and commercialization theory makes the assumption that the work based on an initial invention is not patentable or otherwise appropriable. In the prospect-development theory, Kitch (1977) assumed that there is an abundance of appropriable inventions to be made by using the initial invention as input but suggested that it is problematic. That is, many inventors share knowledge and see the same
potential inventions, and they know that their competitors also see them, so there is a lot of racing for specific targets of opportunity and general overfishing in the prospect pond. Thus, a broad patent on the initial invention is necessary if the "mining of the prospect" or the "fishing of the pool" is to proceed in a less-wasteful, less-duplicative fashion.
If you reflect on the prospect-development theory, you will immediately recognize that it depends on a view that is almost antithetical to the notion about what makes progress in science—it depends on the view that it is good to have many people doing different types of things because different ones will see different things and different ones will be more skilled than others at doing different types of things. Trying to make orderly or rationed access to innovations is likely to be socially very costly.
Indeed, under this version of the prospect-development theory, there might be very high social costs to granting a broad initial patent that gives monopoly rights to exploration of the prospect. It would reduce the number of diverse inventors who would be induced to work on the prospect by the lure of a patent down the road, inasmuch as their ability to work on that patent would be constrained by their ability to negotiate a license with the holder of the original prospect-defining patent.
This theory suggests that an important issue defining the benefits and costs of granting patents on broad prospects is what is assumed about the market for patent licenses. If one assumes that, in general, the transaction costs of patent licensing are small, then one may take a relatively relaxed view of the costs of granting a large prospect-controlling patent, even when one believes that potential explorers of the prospect have diverse ideas of what they would do. But if one assumes that transaction costs are high, one is less sanguine about this outcome.
The question of transaction costs is particularly important when technological advances within a prospect are strongly connected. Advances in technology can be connected to each other in two ways. First, technological advance can be cumulative, in that today's advance lays the basis for tomorrow's. Although the simplest notion of a prospect is that of an initial node with a large number of potential add-on branches, in fact any of the branches can take the form of a long chain. In such a long chain, ability to operate the most advanced version will require the ability to do things that were the subject of earlier inventions. Second, the operative products or processes can form a system, in that they incorporate a number of components. Ability to use the most advanced system might require access to a collection of components. Some of the most important technologies have both attributes. For example, aircraft and computers may be called cumulative system technologies. Merges and Nelson (1990) propose that the historical record show that granting broad patents in cumulative-system technologies is often counterproductive (for instance, research tools). Unless licensed easily and widely, the presence of such patents tends to limit the range of potential users
who have access to all components of the technology. In a number of instances, the consequence was to make technological advance difficult and costly.
We suggest that in most fields of technology, when one or a very small number of parties essentially control the inventive effort, not much happens—with the possible exception of Bell Laboratories and AT&T in bygone days.
If you look at the history of aircraft or radio in the United States, strong broad patents that were selectively enforced by the patent holder interfered with the development of the technology. The situation was resolved only when a system for relatively low-cost licensing emerged in the industry or the technology in question, as it did after a while with radio, aircraft, semiconductors, and computers.
In any case, particularly in the prospect-development theory but more generally whenever an invention is viewed as contributing to further invention potential, as well as creating a new or improved product or process of immediately final use, it can be asked whether strong patents enhance or hinder technical advance over the long run. The question of how strong a patent should be, or whether a patent should be granted at all, no longer turns on the analysis of a tradeoff between the positive effects of stronger patents on inventing and the restrictions in use of the technology associated with a regime of strong patents, as in the invention-inducement theory. Rather, a good part of the argument is about whether the long-term effect of strong patents is to encourage or discourage innovation.
ISSUES IN PATENT REFORM
There is a good deal of argument about what theory, or what version of a theory, is appropriate to which context, but this question often is glossed over. Almost all empirical work on the role of patents has been oriented by invention-inducement theory, and almost all patent policy issues are argued out on the same terms. Many of today's most important patent policy issues seem not to have been adequately viewed through this theoretical lens.
Consider the debate in the late 1970s that led to the Bayh-Dole act. As we noted, the argument that carried the day was that patents were required if inventions, already achieved under federal funding, were to be developed and commercialized. That is a development and commercialization theory argument with a touch of disclosure theory. We have argued that the particular version the development and commercialization theory put forth most vigorously in these debates—that companies would not develop an invention unless they had a patent on it—probably was not widely relevant. But Bayh-Dole undoubtedly has led to a substantial increase in university entrepreneurial activity. Whether that is good or bad is a complicated question. However, the evaluation of Bayh-Dole, like the arguments that lay behind its genesis, must be cast in terms of all four theories.
The same is true of the issue of patenting the codes on gene fragments
identified in the human genome project. Here, too, at least in the early days of the discussion, the issue was not that the prospect of a patent was needed to get the research work done; the research work was being funded by government. Rather, it was felt that patents on the coded gene fragments were needed if companies were to be induced to take that information and use it to achieve commercial products. That is, the arguments were those of the development and commercialization theory and perhaps the prospect-development theory.
As matters have turned out, the belief that codes for gene fragments will be patentable has led to the birth of private for-profit firms, whose business is to discover the codes, in anticipation of profiting from licensing to larger companies that would take on the development work. Interestingly, several large pharmaceutical companies have argued that gene fragments should not be patented, but rather that their identified codings should be in the public domain. Their case is that progress from coded gene fragments to useful final products will cost more if gene-fragment codes are patented than if they are in the public domain. They are essentially arguing that the standard version of the development and commercialization theory has it backwards. And several of these companies are supporting research to identify gene codes, on the condition that the information be put in the public domain. The public-policy issues here are very complex.
The issues surrounding Bayh-Dole and gene fragments differ in important ways but they also have important common elements and raise common questions. Perhaps the most basic question that they raise is whether the presence or prospect of patents stimulates or interferes with technical advance in a field. Obviously, it does not under the invention-inducement theory. But under a more complex theory, the answer is not always apparent. The appropriate domain of patents is badly in need of open examination today. The argument that strong intellectual property rights in a field can smother technical progress is, of course, connected to assumptions about several of the context conditions discussed earlier. To understand better whether our current patent policies help or hinder the achievement of our societal objectives, we need to examine those assumptions rigorously.
The following three questions are put forward as possible starting points for inquiry: In what fields of technology are technical advances so strongly connected to one another, either temporally or in a system of use, that effective inventing today requires access to prior inventions? What are the fields of inventing in which progress generally requires the effective interaction of a number of different organizations? Do patents in fact contribute to or hinder the access and cooperation needed for technical advance in such contexts? As indicated earlier, little empirical research has been aimed at this cluster of questions. Our lack of knowledge limits our ability to analyze intelligently the current pressing issues of patent reform.
Arora A and Gambardella A. 1994. The changing technology of technological change: general and abstract knowledge and the division of innovative labor. Research Policy Sept. 94: 523–532.
Arrow KJ. 1962. Economic welfare and the allocation of resources for invention. In: Nelson RR, editor. The Rate and Direction of Inventive Activity. New York: Princeton University Press.
Barzel Y. 1968. Optimal timing of innovation. Review of Economics and Statistics 50: 348–355.
Dasgupta P and Stiglitz JE. 1980a. Uncertainty: Industrial structure and the speed of R&D. Bell Journal of Economics 11: 1–28.
Dasgupta P and Stiglitz JE. 1980b. Industrial Structure and the nature of innovative activity. Economic Journal 90: 266–293.
Gilbert R and Shapiro C. 1990. Optimal patent length and breadth. RAND J Econ 21: 106–112.
Kitch EW. 1977. The nature and function of the patent system. J Law Econ 20: 265–290.
Klemperer P. 1990. How broad should the scope of patent protection be? RAND J Econ.: 113–130.
Levin RC, Klevorick AK, Nelson RR, and Winter SG. 1987. Appropriating the returns from industrial research and development. Brookings Papers on Economic Activity 14: 783–820.
Loury GL. 1979. Market structure and innovation. Qtly J Econ. XCIII: 395–410.
Machlup F. 1958. An economic review of the patent system. Washington, U.S. Govt. Printing Office.
Mansfield E. 1986. Patents and innovation. Mgmt Sci 32: 173–181.
McFetridge DG, and Rafiquzzaman M. 1986. The Scope and duration of the patent right and the nature of research rivalry. Res Law Econ 8: 91–120.
Merges R. 1995. Contracting into liability rules: institutions supporting transactions and intellectual property rights, [manuscript], Berkeley: University of California Law School.
Merges R, and Nelson RR. 1990. On the complex economics of patent scope. Columbia Law Review 90(4): 839–916.
Mueller, WF. 1962. The origins of the basic inventions underlying DuPont's major product and process innovations. In: Nelson RR, editor. The rate and direction of inventive activity, NBER, New York: Princeton University Press.
Nordhaus WD. 1969. Invention, growth, and welfare. A theoretical treatment of technological change. Cambridge, MA: MIT Press.
Reich LS. 1985. The making of American industrial research: science and business at GE and Bell, 1876–1926, New York: Cambridge University Press.
Samuelson P, Davis R, Kapor MD, and Reichman JH. 1994. A manifesto concerning the legal protection of computer programs. Columbia Law Rev 94(8): 2308–2431.
Scherer FM. 1972. Nordhaus's theory of optimal patent life: a geometric reinterpretation. Am Econ Rev 62: 422–427.
Van Dijk TWP. 1994. The limits of patent protection. Maastricht, The Netherlands: Universitaire Pers Maastricht.