Association for Public Policy Analysis & Management

Annual U.S. federal investment in research and development (R&D) has surpassed $130 billion for the past decade, 40% of which has been allocated to R&D centers funded exclusively by the federal government. For over thirty years national policies have required commercially relevant inventions discovered in federally funded R&D centers to be transferred to the private sector to diffuse knowledge and to promote private sector follow-on innovation. However, there is limited empirical evidence demonstrating the effectiveness of these policies. I quantify the effect of technology transfer on innovation spillovers in the context of patent licensing at the U.S. National Laboratories using data from over 800 licensed patents since 2000. I demonstrate that licensing increases the annual citation rate to a National Lab patent by 31 – 48%, signifying that licensing is an effective mechanism for inducing innovation. Further, I find that over 75% of follow-on innovation after a patent is licensed occurs outside of the licensing firm, indicating that knowledge spillovers induced by licensing diffuse broadly. These estimates rely on a novel matching algorithm based on statistical classification of the text of patent abstracts. I explore possible mechanisms for the effect of licensing on knowledge diffusion by examining the quality of the patents that cite a licensed patent and rule out the possibility of a strong strategic patenting effect. These results demonstrate that transactions over formal intellectual property enhance the benefits of publicly funded R&D in the “market for ideas.”