美国智库-2024美国专利质量的政策报告

Policy ReportPatent Quality in the United States: Findings and Suggestions for Policymakers#128974S E P T E M B E R 2 0 2 4Ani Harutyunyan, Matthew Chervenak, Mark Schankerman, William Matcham, Nishant Shrestha2www.sunwater.org • 9/2024 • #128974 •Executive SummaryBackgroundDebates surrounding patent quality in the United States have historically focused on the notion that the United States Patent and Trademark Office (USPTO) is letting too many “bad patents” slip through. Yet the rejection of “good patents” is more widespread. The authors hope that by presenting a clearer, more nuanced picture of patent quality in the United States, this report will help facilitate a more balanced and fact-based policy dialogue. MethodologyThe definition of “quality” used in this report aligns with the statutory definition, as adopted by the USPTO. Under this definition, patents must satisfy four requirements codified in Title 35 of the U.S. Code. These are: Section 101: to satisfy the criteria for patent subject matter eligibility Section 102: to be novelSection 103: to be non-obviousSection 112: to be clearly and sufficiently claimed and describedThe authors’ analysis focuses on two types of errors. Type 1 errors occur when a patent application is granted despite including invalid claims. Type 2 errors occur when a patent application containing valid claims is improperly rejected. The authors use three distinct methods to measure patent quality: analysis of the overall pool of patent applications, analysis of a random sample of patent applications, and analysis of patent applications submitted to multiple patent offices.Key Findings• The USPTO’s Type 1 error rate is in the single digits (less than 9%), and low relative to other patent offices around the world.• The USPTO’s Type 2 error rate is significantly higher than its Type 1 error rate. The analysis suggests that the USPTO rejects valid claims more often than it grants invalid claims.• Due to selection bias, patent invalidation rates at the Patent Trial and Appeal Board (PTAB) and at district courts do not accurately reflect patent quality in the United States. 3www.sunwater.org • 9/2024 • #128974 •ConclusionUsing three different methodologies and datasets, this report refutes the claim that there is a crisis of “bad patents” plaguing the U.S. patent system. Type 1 errors are rare, and the USPTO improperly grants patents at a lower rate than its counterparts around the world. Type 2 errors occur at a higher rate than Type 1 errors. Despite this fact, policy dialogue in the United States tends to focus primarily on Type 1 errors, when both kinds of errors are harmful and undermine innovation. Policy debates should account for these realities.Recommendations• Policymakers must recognize that there is a tradeoff between Type 1 and Type 2 errors. Efforts to reduce the prevalence of Type 1 errors would likely result in more Type 2 errors, and vi