Holdup, Royalty Stacking, and the Presumption of Injunctive Relief for Patent Infringement: A Reply to Lemley and Shapiro

J. Gregory Sidak


“Patent holdup” is described by its critics as occurring when a patent-holder uses a court’s issuance of an injunction (or merely the threat of an injunction) to block an infringer’s use of the patented invention unless the infringer, who has made sunk investments in expectation of using the patented invention, pays a royalty that is, from the infringer’s perspective, excessively high. “Royalty stacking” is described by its critics as occurring when a product sold to end users incorporates many separate patented inputs, and the holder of the patent to one such input – an input lacking immediate substitutes – demands a high royalty from the manufacturer of the end product without regard to the effect that this royalty will have on the total amount of royalties that the manufacturer must pay to all holders of patented inputs and, consequently, the price that the manufacturer must charge end users. Professors Mark Lemley and Carl Shapiro argue that patent holdup and royalty stacking are serious problems, and that legislators or courts (if not both) should limit the circumstances in which a patent-holder may avail himself of the existing statutory right to enjoin the infringer’s use of the patent – essentially only if the patent protects an input that represents a “significant” portion of the final value of the product.

I critique the Lemley-Shapiro model of patent law. I dispute its main finding that the threat of an injunction inflates royalty payments in many cases relative to a hypothetical benchmark royalty rate. The Lemley-Shapiro framework does not properly account for the relevant error costs associated with weakening the presumption of injunctive relief. In particular, Lemley and Shapiro fail to consider how removing the presumption of injunctive relief could decrease dynamic efficiency. Furthermore, even if their framework were correct, Lemley and Shapiro rely on biased parameters that preordain their result. This outcome follows for two reasons. First, because Lemley and Shapiro fail to account for the real option conferred on potential users of the patent when a patent-holder makes sunk investments in new technologies or products, their hypothetically reasonable royalty rate is biased downwards. Second, the Lemley-Shapiro model reaches its result not by deriving a general bargaining model, but by assigning all the bargaining power to the patent-holder and claiming a general result. Both factors bias Lemley’s and Shapiro’s results in favor of the infringing party.

Download as PDF