Venture Capital and Innovation
Faculty Advisor: Morten Sorensen
This dissertation delves into the relation between venture capital and innovation. The existing literature usually addresses this question by using industry-level data. In contrast, the analysis here relies on data at the company level on patents invented in venture-backed companies. The dissertation has four parts. The first part, a paper coauthored with my advisors Bruce Kogut and Morten Sorensen, examines the relation between venture capital and the rate and quality of companies' innovative activity. We compare the number of patent filings, and the quality of innovations, before and after companies are first financed by a venture capital investor. As an attempt to control for the endogeneity of venture capital investments we exploit amendment by the Texas Legislature that freed public state pension funds in Texas to invest in venture capital. Our results suggest that venture funding increases the rate of companies' innovative activity. Interestingly, we also find that venture capital is associated with a decrease in the quality of companies' research output. The second part estimates the effect of venture capital on the diffusion of knowledge. I compare citations to patents invented in venture-backed companies to those of comparable patents invented elsewhere. To isolate the causal effect, I exploit time variation in the assets of state pension funds that allocate capital to venture capital. This variation provides a valid instrument if the effect of changes in innovation opportunities within a state is uniform across local patents in the same technology-class and vintage-year. I find that after venture funding annual citations to a given patent increase 19% relative to the citations of comparable patents. Additional results are consistent with two mechanisms: venture capital investors certify the value of patents to the general public and facilitate communication among companies in their portfolios. The third part of this dissertation explores whether the strategic interaction of companies in the same venture capital network affects the direction of companies' innovative activity. Theoretically, this effect is not clear. Whereas the presence of common investors can stir companies' research in the same direction by facilitating knowledge spillovers, competition for the same financial resources may undermine the incentives of companies in the same venture capital network to collaborate, or even work in similar areas. To test this question empirically I use the propensity of patent citations among pairs of companies as a measure of the similarity in companies' research. To reduce concerns of strategic investment by venture capital investors, I control in the estimation for the technological similarity and geographical co-location of companies. Consistent with venture capitalists facilitating the diffusion of knowledge across the companies they finance I find that companies in the same venture capital network produce similar innovations. Interestingly, I also find that this convergence in innovation is only true for companies that are not competing for the same financial resources, specifically, those pairs of companies that are geographically distant or work in different technological areas and industries. Results suggest that the optimal strategy for companies that are competing for the same financial resources is to differentiate and pursue different lines of research. Finally, the Appendix describes in detail the construction of the dataset.