Patentable Subject Matter and Institutional Choice

Abstract

Given the existence of fine-grained requirements for patentability such as nonobviousness, the utility of a separate requirement of patentable subject matter has sometimes been questioned. The courts‘ fumbling efforts to regulate patentable subject matter have helped stain the enterprise with suspicion and even disrepute. This Article first defends limitations on subject-matter eligibility by showing that they provide a categorical filter that can improve patent-system performance. The Article then argues that the enterprise of regulating patentable subject matter should be primarily entrusted to the USPTO, rather than, as it is now, to the courts. Two mathematical models illustrate (1) how more individualized tests for patentability can fail to ensure that patents improve social welfare and (2) how a particular form of subject matter - fundamental principles having a very high number of potential uses - can generate particularly high social costs and thus qualify as a form of subject matter that the patent system would best fil-ter out. With respect to the proper locus for rulemaking authority, the USPTO‘s capacity and incentive to respond promptly and meaningfully to questions of subject-matter eligibility make it the best candidate. Moreover, giving the USPTO rulemaking authority with respect to subject-matter eligibility does not require giving it rulemaking authority on all matters of patent-law substance. Just as other regimes of U.S. law have divided tasks of adjudication and enforcement between different institutions, the patent system can divide areas of primary interpretive authority between the USPTO and Article III courts. Such an institutional innovation appears the best way to leverage the relative institutional competences of the USPTO, the courts, and Congress.