Abstract:

Basic science ultimately improves living standards through the innovation process. Technological know-how acquired through research experience is embedded in the scientist's human capital. Sometimes this knowledge may be available to an external organization only through hiring or collaborating with the scientist. While social scientists agree on the importance of the basic science and the diffusion of technology, little is known about the transmittal mechanism.

This dissertation project investigates the productivity of scientists over their lifecycle, the impact of academic science on industry innovation, and factors that influence an innovating firm's interaction with basic science. We then apply a new empirical approach to one of the central questions in Industrial Organization: the returns to scale to firm R&D using an alternative measure of R&D rather than the traditional R&D per dollar measure.

Patent data has been vastly used in economic research but inventor information contained in the patent data has not been utilized. Patent inventor data has great potential in the micro-level studies on many issues, including our dissertation project. Using systemic methods to identify inventors in the U.S. patent data and to combine inventor education and patenting firm information from five other data sources, we have created a unique longitudinal patent-inventor-firm-matched dataset. The dataset contains patent productivity of inventors in the pharmaceutical and semiconductor industries most prolific in generating innovations, and thus is able to look into those issues.

We find that employment changes for inventors have a positive effect on their own productivity, i.e., the inverted U-shaped productivity profile of an inventor shifts upward after moving. We find that the research personnel played a pathway for the diffusion of scientific knowledge across institutions, specifically, from university to industry. We find that industries have greatly used the technological know-how of inventors with university research experience over the decade of the 1990s. Using an alternative measure of R&D activity, we find no relationship between patenting productivity and firm size in the pharmaceutical industry. In the semiconductor industry, we find that patenting productivity increases with firm size, consistent with Schumpeter's hypothesis; this is the opposite of the results obtained using R&D expenditures.