Power-law distributions of corporate innovative output: evidence from US patent data

Abstract

This study aims to examine the existence and the characteristics of power laws in the distribution of corporate innovative output. Using a dataset containing information on 1,102,839 U.S. patent applications by 94,103 U.S. private firms during the period of 1976-2000, we find that corporate innovative output, as measured by either simple or quality-adjusted patent counts, follows power-law distributions that theoretically have infinite variances and, in some cases, infinite means. In addition, we find that corporate innovative output is power-law distributed in all technological fields. We further find that the power-law distribution of corporate innovative output tends to be stronger (i.e., the concentration of corporate innovative output in a few firms is more pronounced) in technological fields with more abundant technological opportunities and for firms with higher technological competence.