TCS Daily economic analyst Arnold Kling proposes a fair use doctrine for patents, somewhat analogous to the fair use doctrine in copyright.

Much as I appreciate Arnold's analyses (usually), I think he is askew on this one -- not in his assessment of the problem, but in his solution. The fundamental problem in the examples he uses is that the inventions should not have received patents in the first place, under an appropriate application of the non-obviousness requirement, as argued in the certiorari petitions (including the one by PFF) in the KSR case..

In short, we agree with Arnold's view that there is a problem, but think that the answer is in a better application of the patentability tests, a view reinforced by the fact that the fair use doctrine in the copyright context is a thorough mess that should not be imported into patent law.