The recent Fortune interview with Brad Smith and Horacio Gutierrez from Microsoft left me thinking about patent licensing markets, and whether patent agreements between Microsoft and other firms benefit the technology industries. The significance of patent licensing markets is that they affect the diffusion of innovation, interaction among small/ large firms and commercialization of inventions. More importantly, patent licensing markets may also affect industrial structure, and consequently, patent policy should be informed by potential impact on the shape of the industry.

An important article on patent licensing markets comes from industrial organization literature. Years ago, Professors Deepak Somaya and David Teece addressed the issue of how the US patent system, that was developed in view that inventions and patents would generally correlate 1-1, could provide innovation incentives, and ameliorate transaction costs, when modern inventions often contain numerous patent claims. Combining Inventions in Multi-invention Products: Organizational Choices, Patents, and Public Policy (2000). Haas School of Business CCC Working Paper No. 99-4. Their conclusions contradict some seemingly intuitive views on patent reform.

In response to transactional problems in multi-invention contexts, the instinctive remedy sought is often a weakening of patent rights. Although this solution does address the transaction cost problems, it also means a weakening of incentives for innovation and the reinforcement of integrated modes of production... it may not be the optimal response in many cases... stronger patents can make integrated modes less effective, component and licensing modes benefit from the resulting incentives, and in some cases, even from lower transaction costs.

Somaya/Teece use a theoretical framework that particularly suits analysis of patents in the technology industries. They develop a transaction cost model where commercialization occurs by: 1) combining patent claims via licensing, 2) sales in components, and 3) integrated development (in-house development). Somaya/Teece argue that their model leverages two policy insights: 1) that patent policy should be informed by how it affects organizational choice for innovators, 2) that the impact of patent policy on organizational transaction costs should be considered alongside innovation incentives in crafting patent policy. Patent policy affects transaction costs and innovation incentives in the licensing, component and integrated models of organization, thus influencing the choice of organization among innovators.

From their framework, Somaya/Teece find that a healthy patent licensing market benefits small and specialized firms, whereas a system of weak patents favors large vertically integrated firms- the former situation of relative advantage for small firms is what academics and policy makers see as the proper functioning of the patent system.

…maintaining the strength of patent rights is worthwhile for at least three reasons. First, stronger patents may induce a shift in organization mode, away from integration. The resulting licensing or component modes may provide overall organizational advantages, despite higher transaction costs. Second, even with an integrated mode, the innovation engendered by stronger patents may make higher organizational costs worthwhile. [Third], stronger patents may actually reduce some forms of transaction costs in component and licensing markets – by safeguarding transactions from leakage of tacit know-how – which may compensate for the other transaction costs created by stronger patents in some cases.

For several reasons, the optimal organizational model for the technology industries is the Somaya/Teece conception of licensing. In licensing market dominant landscapes, firms can specialize in technological spaces and leverage comparative advantage, rather than pour resources into complementary technologies. The result is the prominence of small firms.

Industries where licensing is dominant rely on strong patent rights. Current innovation in the technology industries has resulted, in part, from the decentralized and non-concentrated structure of the sector, enabled by the exchange and licensing of patents. Stronger patent rights benefit small firms by giving them tools with which to bargain with larger firms. Weaker patents favor large firms that have more resources and other means of appropriation, over small entities. Without patents, or with weak patents, large firms are the primary beneficiaries of public policy- a fate for the technology industries few would support.

Without patents... market-based advantages of licensing and component modes for innovation are severely compromised due to the weak bargaining position of inventors... with stronger patents, these “incentive” advantages grow more quickly for licensing and component modes than integrated modes, reflecting their market based advantages. Without patents, we observe a net advantage for integrated modes, but as patent strength increases, licensing eventually becomes the superior mode.

Somaya/Teece argue that in a well-functioning licensing market, patents enable modularity in the innovation process, allowing entities to appropriate returns from incremental inventions in the innovation chain. The result is an innovation landscape that provides for the viability of small firms to enter, compete and remain viable in new markets by solving complex technical problems.

First, specialization creates targeted incentives for making and commercializing inventions in an autonomous fashion. Second, decentralized innovation results in greater experimentation and trial-and-error learning, increasing the chances of discovering different market needs and finding superior solutions. [Third], licensing and component trade can save on costs of excessive reinvention, and make best-in-class inventions available to the entire industry.

Licensing modes facilitate the independent commercialization of inventions… However, these benefits can only be reaped by licensing modes if inventions, once made, can be protected from imitation... Absent such appropriability guarantees, licensing modes are unable to access their potential market-driven advantages, while at the same time being burdened with transaction costs.

Somaya/Teece provide persuasive arguments for the importance of patent licensing markets in sectors such as the technology industries. If maintaining conditions suitable for the competitiveness of small firm furthers innovation, then strong patent policies and the licensing of patents should be encouraged. Although the work by Somaya/ Teece dates back to th early era of software patents, their policy recommendations generally align with more previous patent policy work by Professor Ronald Mann and PFF's J.V.Delong, as well as interpretations of the Patent Clause by Professor Lawrence Solum. The work of Somaya/Teece should be taken seriously when evaluating Microsoft Corporation's current patent licensing efforts.