Ling Zhou
As an important policy instrument for guiding innovation, the patent system involves a long-standing tension between creating economic rewards for the original innovators and stifling subsequent R&D activities. The availability of new data allows us to provide unique sights into understanding how effective patents are in providing incentives for innovation and how firms' disclosure of proprietary assets affects the dissemination of knowledge. On the other hand, innovation per se is the fruit of human intellect. The knowledge workforce is, therefore, a crucial component of a country's innovative capacity. The persistent gap in the participation in innovative activities between men and women reflects a misallocation of talented human capital and leaves opportunities for policy implementation. This dissertation ties the questions mentioned above and presents novel evidence on the incentives, diffusion, and disparity in innovative activities. The first essay (chapter 2), in collaboration with Gaétan de Rassenfosse, provides empirical evidence on the real effect of patent expiration on monopoly prices and explores how this effect varies with product market competition. We compile a novel dataset that links patents to consumer products and retail prices for these products from Amazon.com. We find that patent expiration leads to a 7--8 percent drop in product prices. This effect is heterogeneous by patent type and importance and starts one year in advance. We also find a more substantial decline in prices in product markets where competition is intense, consistent with evidence from the pharmaceutical industry. We argue that incumbent innovators lower prices preemptively in the face of generic competition. The second essay (chapter 3), in collaboration with Gaétan de Rassenfosse, studies how firms' disclosure of innovative assets affects the diffusion of inventions. We focus the empirical setting on firms' provision of constructive notice through virtual patent marking (VPM) and exploit patent-level variation in the marking dates. With data collected on marked patents from the VPM web documents of 16 firms, we find an overall effect of VPM is small. However, conditional firms' strong reliance on patents as an appropriability regime, our findings suggest that while VPM induces firms to build subsequent internal inventions on the ones that are commercially valuable, it fends off diffusion to external firms and reduces similarity between inventions. Last but not least, in the third essay (chapter 4), collaborated with Mary Kaltenberg, we explore the role of family policies in closing the innovation gender gap. By matching inventor data to web-scraped inventor age information, we exploit the staggered passage of maternity leave policies across the U.S. and evaluate its impact on the retention and productivity of female inventors. Our findings suggest that maternity leave policy supports women of reproductive ages stay longer in patenting but has little impact on their productivity.
EPFL2021
