Disclosure as a Strategy in the Patent Race

Scott Baker -- University of North Carolina, Department of Economics, sbaker@email.unc.edu

Claudio Mezzetti -- University of North Carolina, Department of Economics, mezzetti@email.unc.edu

abstract

Research firms disclose a surprisingly large amount of information to the patent office through “targeted” disclosures; that is, disclosures intended to make the patent office aware that potentially patentable information is already in the public domain. Conventional wisdom holds that these disclosures are made for defensive purposes; the disclosing firm does not itself plan to pursue patents related to the disclosed information, so the firm discloses as a way of creating prior art that might stop rivals from patenting. But firms have an incentive to disclose even if they themselves intend to pursue patent protection. The reason is that, by making it more difficult to patent, disclosure in essence extends the patent race. If an invention of a certain quality would have been sufficient to qualify for patent protection before the disclosure, after the disclosure any invention must be that much better before it will represent a sufficient advance over the now-expanded prior art. Extending the patent race can be an attractive strategy for a firm trailing in a given race since a longer race might offer that firm a better opportunity to catch up. Extending the race can similarly be attractive to a leading firm, since making the race longer raises the costs of racing, a strategy that will in certain instances discourage trailing firms from pursuing aggressively. This paper models disclosure strategies and explains how research firms can use disclosure to gain a competitive edge. It also presents empirical evidence that IBM, at least, engages in disclosures consistent with the model developed here.