Designing Optimal Software Patents

Abstract:     
Patents constitute our foremost policy tool for encouraging innovation. However, because each new technology provides an important input to subsequent innovation, the exclusive rights conferred by a patent may also impose significant costs upon follow-on innovators. Optimal patent policy should seek to maximize the patent incentive effect, while minimizing burdens placed on future innovation by tailoring the scope of the patent to the characteristics of each technological sector affected.

In the case of software, recent scholarship has illuminated the innovation profile of the current industry. Software is characterized by incremental innovation, relatively low development costs, and short, volatile product life cycles. Interoperability and compatibility between complementary products is a major concern, making technical transparency or reverse engineering critical to product development. This suggests a need for relatively narrow patents that are relatively easy to obtain, and subject to the exceptions necessary to ensure interoperation and follow-on development.

However, current software patent doctrine bears little relationship to this industrial profile. The United States Court of Appeals for the Federal Circuit has set an extremely lax standard of disclosure software patents, resulting in patents scope unconstrained by doctrines of enablement and written description. Recent changes that make patent law amenable to software have produced a flood of new applications, allowing firms to adopt a patent thicket strategy for licensing leverage. At the same time, Federal Circuit case law suggests that a stringent standard for patent non-obviousness will be applied to such patents, resulting in relatively few valid software patents. Optimal software patent doctrine would constrain scope to deal with patent thicket while lowering the non-obviousness standard to validate more issued software patents.

Keywords: software, innovation policy, patents

JEL Classifications: O31, O34, O32, L86, L52, K20, D23

Accepted Paper Series