Patent Thickets: Strategic Patenting of Complex Technologies
by James Bessen (Research on Innovation and Boston University School of Law)
FULL
TEXT
Varieties of Thickets
One of the main arguments for patents is that exclusive ownership may provide the strongest incentive for innovation. When a prospective innovator owns exclusive rights to a product or process, he or she can earn the largest possible profit and therefore has the strongest incentive to invest in making the innovation. But what happens with complex technologies, when that product or process involves hundreds of innovations and their associated patents? In such cases it is not clear that a patent does in fact grant exclusive ownership of all the relevant assets.
Complex technologies give rise to a variety of strategic patenting behavior, and to a variety of problems collectively called “patent thickets.” Some researchers have identified problems that might arise between upstream and downstream producers and users of a technology. Consider the example of biomedical research tools. A researcher may need to negotiate many licenses for all of the research tools necessary in a project. The resulting transaction costs may contribute to an “anti-commons” problem (Heller and Eisenberg, see also Ziedonis, this issue), making innovation prohibitively expensive. In this setting, patent holders will also tend to charge too much to license their technologies, especially when they can “hold-up” a downstream user who has already made large investments in a project and will agree to pay inflated license fees (Shapiro). But some researchers have suggested that institutions like patent pools and cross-licensing may alleviate many of these problems (Shapiro, Merges).
A Model of Strategic Patenting Between Competing Firms
This paper addresses a different sort of patent thicket, namely one that arises when competing firms may mutually infringe each other’s patents. This situation concerns interaction between competing firms, as opposed to interaction between suppliers and downstream users, and gives rise to problems even when there are no transaction costs or hold-ups. It is more typical of electronics and computer-related industries. Such firms have long been known to patent intensively despite reports from R&D managers in these industries that they actually rely on lead time and related advantages in order to profit from their innovations rather than patents (Cohen et al.).
Why do these firms patent so heavily? The paper answers this question using a model of the strategic interaction between two firms. Firms improve their prospects of winning a legal battle when they have a larger portfolio of patents and are more likely to hold a valid patent that the other firm infringes. The greater its likelihood of winning at litigation, the more a firm can demand in cross-licensing negotiations. Consequently, firms have an incentive to accumulate large portfolios of patents in order to obtain more favorable cross-licenses.
But patents are costly to obtain, so there is a tradeoff. Firms can obtain more patents by doing more R&D, but they can also direct more resources to patenting alone. As the empirical literature on patent propensity shows, firms display great variation in the degree to which they obtain patents for a given investment in R&D.
The cost of patenting, in turn, is affected by patentability standards—nonobviousness or minimum inventive step, among others. Low standards may allow patents on trivial improvements that are easy to conceive and describe, while high standards require patents to be substantial improvements on existing knowledge that require a much greater effort to create.
It turns out that patent standards are key in determining firm behavior. When standards are high, firms obtain a small number of important patents and tend not to assert them against competitors who have their own innovations and patents. This sort of “mutual forbearance” describes the early semiconductor and computer industries. When standards are low, however, firms acquire large portfolios of “cheap” patents and assert them aggressively to obtain cross-licenses. This behavior is more typical of the computer and electronics industries today.
So What?
Is this sort of aggressive strategic patenting socially harmful? One might view it as relatively benign, since cross-licensing serves to overcome the problems of transaction costs and hold-up, as noted above.
But there are two potential problems with such optimism. First, strategic cross-licensing favors incumbents over industry newcomers. Incumbent firms, especially in mature industries, acquire larger portfolios and extract a larger share of profits from industry entrants. To the extent that entrants increase the pool of new ideas and are thus important to innovation, strategic patenting subjects these firms to an undesirable “entry tax.”
Second, and more important, strategic cross-licensing can reduce the incentive to innovate, especially in industries where firms have traditionally used lead time to profit from innovation. Cross-licensing agreements effectively force some sharing of profits made thanks to a lead time advantage. Lower profits mean less incentive to invest in innovation. In a way, strategic patenting turns the logic of patents on its head: instead of providing the strongest incentives for innovation by granting exclusive ownership, strategic patenting reduces these incentives through a form of ownership that is effectively collective, with greater shares of profits going to incumbent firms with large portfolios.
Rather than helping to resolve problems with the patent system, aggressive cross-licensing may be a symptom of deeper problems and an instrument of detrimental policy. This analysis suggests one should look critically at changes in patent law over the last decade or so that have reduced standards of non-obviousness, utility, and enablement in many industries and eliminated subject matter restrictions on software and business methods.
REFERENCES
Cohen, W.M., R.R. Nelson, and J.P. Walsh (2000). “Protecting Their Intellectual Assets: Appropriability Conditions and Why U.S. Manufacturing Firms Patent (or Not),” NBER Working Paper no. 7552.
Heller, M.A. and R.S. Eisenberg (1998). “Can Patents Deter Innovation? The Anticommons in Biomedical Research,” Science, 280: 698-701.
Merges, Robert P. 1999. “Institutions for Intellectual Property Transactions: The Case of Patent Pools,” Boalt Hall School of Law Working Paper, University of California, Berkeley.
Shapiro, C. (2001). “Navigating the Patent Thickets: Cross-Licenses, Patent Pools, and Standard-Setting,” in A.Jaffe, J. Lerner and S. Stern, eds., Innovation Policy and the Economy, v1 (Cambridge MA: NBER)
