Why do patent applicants pay higher fees when they succeed than when they fail? In a terrific new draft posted last week, "Loser Pays" in Patent Examination, Neel Sukhatme (Georgetown Law) argues that "such pricing is precisely backwards, penalizing good patent applications instead of bad ones." Instead of this "winner pays" system, he argues that the PTO should discourage weak applications by forcing unsuccessful applicants to pay more.
I think most patent scholars would agree that the PTO issues many patents that do not meet the legal standards for validity; see, e.g., work by Michael Frakes and Melissa Wasserman showing that time-crunched examiners have higher grant rates, or work by John Allison, Mark Lemley, and David Schwartz on the large number of patents invalidated during litigation. To address this problem, it may be worth increasing the resources devoted to examination—e.g., examiners could be given more time, and I have argued for some form of peer review. But I think most patent scholars would also agree with Mark Lemley that at some point the costs of increased scrutiny would not outweigh the actual social costs of the invalid patents that slip through (most of which are never asserted or litigated).
Sukhatme's paper builds on a growing body of scholarship that tackles the problem of invalid patents not through increasing the resources spent on examination, but rather through price structures designed to disincentivize socially costly behavior by patent applicants. For example, Jonathan Masur has argued that the high costs of obtaining a patent disproportionately select against socially harmful patents, and Stephen Yelderman suggests a number of more fine-grained way to rationalize application fees. Economists Bernard Caillaud and Anne Duchêne have developed some of these ideas in a formal model, including the possibility of "a penalty for rejected applications" which "would unambiguously encourage R&D" because it would have "no impact on the submission strategy for non-obvious projects." And it seems clear from work by scholars such as Gaétan de Rassenfosse and Adam Jaffe that changes in fees do affect applicant behavior in practice.
Sukhatme does a wonderful job expanding on these ideas to explore how loser-pays rules might be adapted to patent examination. For example, he suggests that applicants could be required to post a bond at the outset of examination, some of which would be returned if the applicant is successful. Continuations could be discouraged by reducing the recoverable bond amount as prosecution proceeds. Additional revenues could be returned to successful applicants, providing stronger incentives for filing valid patents. To reduce unfairness to individual inventors, the PTO could continue the current practice of offering discounts for small and micro entities.
The PTO could implement this system using its fee-setting authority under current law. Indeed, I think Sukhatme could go further and argue that the PTO has not only authority but also some obligation to do so. I've previously blogged about Jonathan Masur's work on cost-benefit analysis (CBA) at the PTO, in which he explains why the PTO's attempt to use CBA for its recent fee-setting regulations was a misapplication of basic principles of patent economics. If the PTO's current fee structure were compared with Sukhatme's under a correct application of CBA, I think Sukhatme's would win.