Combining Inventions in Multi-invention Products:
Patents, Organizational Alternatives, and Public Policy
by Deepak Somaya (Maryland)
FULL TEXT
Summary
Critics of the patent system argue that it is increasingly creating a tragedy of the anticommons—a situation where a resource is owned by too many and hence used by none. Excessive transaction costs for access to patents, they say, interfere with innovation and commercialization in many hi-tech industries. The proliferation of patents in some technologies, they conclude, has helped to create “complex obstacles...that arise when a user needs access to multiple patented inputs to create a single useful product” (Heller and Eisenberg, 1998, p. 699).
This paper deals with the intersection between patents and organization in industries where end-products incorporate many inventions and thus pose anticommons concerns. The anticommons problem arises from a fundamental feature of the patent system. A patentee is legally empowered to exclude others from a patented invention but has no affirmative right to use that invention if, say, the invention also infringes patents owned by others. Multi-invention technologies where very large numbers of patented inventions—tens, hundreds, or even thousands—need to be combined into products challenge this legal logic. The owner of each of these patents has an effective veto on the commercialization of the final product. How can public policy balance the tension between the incentives for innovation inherent in patents and the transactional challenges raised by such anticommons concerns?
This paper presents a theoretical framework to address such questions. This framework rests on two pillars. The first holds that the commercialization of technology in multi-invention contexts entails an implicit choice from among a range of organizational arrangements. For example, a firm with a genomic invention can use it internally to develop end products, or license the technology to other firms who will undertake the complementary innovation required to do the same. Alternatively, the firm might be able to embody the invention into a product sold as a component to the firm developing the end product. We group these alternative organizational arrangements into three modes: an integrated mode, a licensing-market mode, and a component-market mode, respectively.
The second pillar is a detailed description of how patent rights help to determine the relative effectiveness of these organizational modes. In assessing organizational effectiveness we should be concerned with both the organizational costs of combining inventions in each mode and the incentives provided for innovation. While patents increase transaction costs, and thus impair organizational effectiveness, in some contexts, they arguably reduce transaction costs in others. In addition, patents can be valuable for fostering innovation, which must also be weighed against the problem of transaction costs.
The paper’s theoretical framework provides two policy insights. First, while patents may make integrated modes more cumbersome, as critics of the system suggest, they can also make component and licensing modes so attractive as to induce an industry-wide shift to these approaches. Hence, anticommons concerns that seem insurmountable when one considers integrated approaches alone may be mitigated by the real-world variety of non-integrated organizational alternatives. Second, public policy can help alleviate transaction costs in multi-invention contexts by selectively supporting remedial institutional structures or private arrangements, such as patent pools, where multiple patents from different owners are licensed in a package. The framework allows us to assess the pro-competition effects of such public policy solutions.
This paper also helps to explain how organizational workarounds have mitigated anticommons concerns in some multi-invention contexts. In semiconductors, firms have long used cross licensing to solve the problem of transacting for large portfolios of their rivals’ patents (Grindley and Teece, 1997). More recently, a fledgling licensing market for semiconductor design modules has helped to facilitate access to innovations for building increasingly complex and multi-functional integrated circuits (Linden and Somaya, 2003). In biotechnology, a variety of organizational and institutional solutions, including licensing and the market for research tools and reagents, have ensured that the presumed patent anticommons has not significantly impeded innovation (Walsh, Arora, and Cohen, 2003). While the anticommons problem remains an important concern for patent policy, this paper suggests that its effects are more pronounced in specific contexts and can be mitigated in others.
