UC Berkeley

One of the main purposes of the patent system is to facilitate the flow of knowledge by ensuring access to information related to inventions. In the United States, the patent applicant has a legal duty to disclose knowledge of relevant prior art in the form of citations, thus placing the innovation in context and enhancing knowledge of this prior art. However, the applicant might miss relevant prior art due to ignorance, or conceal it for strategical reasons, which could reduce the value of public information generated by the applications and also reduce the accuracy of patent grant decisions. Patent examiners can add citations not known to the applicant, because they supposedly have less barrier in accessing knowledge from different regions. The examiners, specialized in a given technical area, might also identify and add prior art that are deliberately omitted by applicants thus tend to render the patent obvious or non-novel. Prior literature generally agrees the examiner citation is more important than applicant citation as a signal of patent value or knowledge flow. Using US patents granted since 2001, I test the hypothesis that examiner citations contain more important information on patent value and knowledge flow than applicant citations. Surprisingly, on a variety of measures, I find evidence to support the contrary. I also show that the evidence supporting the current hypothesis was attained by a model mis-specification. The findings shed light on many innovation studies that rely on patent citations, and also have important policy implications regarding the operation of the patent office and in general the patent examination system.