Author(s)
Anne Layne-Farrar
Source
AEI-Brookings Joint Center Working Paper 05-14, August 2005
Summary
This paper looks at four key methods of defining what a software patent is empirically and assesses the pros and cons.
Policy Relevance
Scholars disagree on how to define software patents. Policy makers should beware, as some studies use less reliable methods to assess problems associated with granting these patents.
Main Points
- To define software patents, patent office classifications are hard to rely on:
- The patent office reclassifies patents sometimes, using different systems.
- Applicants avoid certain classifications for strategic reasons.
- Looking at patents individually takes a long time and is not objective.
- Searching patents for keywords includes many computer-related non-software patents.
- Studies of software patent quality, importance, and numbers have different results depending on how software patents are defined.
- Bessen and Hunts’ influential 2004 paper, “An Empirical Look at Software Patents” uses a less reliable method.
