Author ORCID Identifier

Matthew Sag 0000-0003-2381-1028

Document Type


Publication Date



Differential impact, Patent reform, Patent litigation, Patent Office, Post-grant review, Licensing fees, Infringement


The structure of the article is as follows. Part I provides an introduction to the problems created by bad patents and introduces the differential impact test for evaluating patent reform proposals.

Part II examines the origin of bad patents and applies two different economic models to explain their persistence. The first model focuses on a potential infringer’s incentives to challenge a bad patent; the second model focuses on a patent holder’s incentive to assert a patent. We explain bad patents as an emergent phenomenon: they are the product of the apparently low quality of patent examination and the complex, uncertain, expensive and time-consuming nature of patent litigation.

Part III then assesses the major patent reform proposals currently under consideration against the test of differential impact in light of the economic models developed in the previous section. From this analysis, we conclude that Congress’ highest priority for patent reform should be the adoption of a system of post-grant review of patent validity. Post-grant review will have a differential impact by significantly lowering the cost of challenging the bad patents. The differential impact test supports the adoption of post-grant review which will provide a low cost method of challenging patents which appear highly likely to be invalid or hyper-asserted. A well designed system of postgrant review will not impose a significant burden on good patent holders because those few good patents that are occasionally subject to review are unlikely to be found invalid. Furthermore, post-grant review will actually benefit good patents by reducing uncertainty and information asymmetries relating to patent quality.

Part IV demonstrates that in addition to supporting the idea of post-grant review in general, the differential impact test may also be applied to build a better model for post-grant review. Looking at post-grant review through the lens of differential impact leads us to propose a system that is very different from those embodied in current legislative and other proposals. In short, we propose: (i) adopting a variable presumption of validity depending on the level of review that a patent has been subject to; (ii) implementing a multiple stage system of post-grant review with two distinct stages in order to balance the goal of greater scrutiny for bad patents with the need to minimize potential harassment of good patents; and (iii) bringing questions of claim construction into post-grant review such that both sides of the bad patent phenomenon can be addressed. These proposals are all designed to ensure that post-grant review provides a low cost way to challenge bad patents, but does not undermine the value of good patents.

First Page


Publication Title

Minnesota Journal of Law, Science & Technology


© 2007 Matthew Sag & Kurt Rohde.