Patent citations as a measure of knowledge flows

Salvatore Torrisi (University of Bologna), Marco Corsini & Myriam Mariani, “Patent citations as a measure of knowledge flows: a replication exercise with extensions”

This paper compares direct information about knowledge interactions to patent citations to assess the “validity” of patent citations as a measure of knowledge flows. For a sample of about 10,000 inventions patented at the European Patent Office, we collect: (1) data about backward patents citations; (2) primary survey data about the use of knowledge flows as reported by the inventors who took part in the inventive process. By comparing the two datasets, we substantiate possible measurement errors associated with the use of patents citations. We replicate and extend a study of Roach and Cohen (2013) based on US patents. Replication is a research endeavor that has proven to be crucial for the advancement of research in ‘hard sciences’. Only recently, social scientists and management scholars have highlighted the role of replication as a key mechanism to the accumulation of scientific knowledge.