Backward Patent Citations and Inventors' Recognition of Differential Influences (Japanese)

While numerous empirical investigations of cumulative innovation have utilized patent citation as quantitative metric, most of the existing studies assume away differences in the relative importance of citing relationships at the micro-analytic level. In other words, no weight has been given to each citer-cited pair with respect to an individual patent. This paper takes advantage of the RIETI Inventor Survey (supplementary survey, conducted in 2008) in order to identify the determinants of relative importance, as recognized by the inventors, between their technological predecessors. The results reveal that the number of forward examiner citations (not inventor citations) which a cited patent generally receives is positively correlated with the importance recognized by its inventor for the specific invention. In addition, only the maximum number of (not the total of) forward examiner citations that cited patents receive have a consistently positive correlation with the recognition of the existence of preceding patents by its inventor. The findings imply that examiner citations are more useful than inventor citations to estimate relative significance between previous patents. Also, a very skewed distribution of influences from various cited patents appears to be the cause of seemingly uneven impacts from within-the-firm and out-of-the-firm technological sources, which was found in the Inventor Survey conducted in 2007.