Essays on innovation and finance

This dissertation contains three essays on the economics of innovation, learning and finance. The first chapter studies the effect of having multiple shareholders who share control of a firm in a sample of small and medium-sized companies. Using survey data for Italian manufacturing firms between 1992 and 2003, I focus on the decision to innovate and I show that firms with shared control are less likely to launch new products into the market. In addition, I find that the costs of shared control are higher when shareholders are more involved in running the firm (i.e., family management) and in industries in which the default is to not innovate (since overcoming the status quo becomes more difficult). I also exploit industry shocks and changes in exchange rate volatility over the sample period and find that agreeing to innovate is more difficult when uncertainty is high. I present a model that rationalizes these findings. I show that if shareholders distrust others' ability to interpret market signals (or overestimate their own), they are less willing to compromise. Hence, even if they all have the same objectives for the firm, the result will be slow decision making and inertia. These findings provide an additional rationale for the reluctance of firms' owners to give up control of their firms and raise a trade-off. While concentrated ownership increases innovation for a given firm size, the founders' willingness to maintain control of the firm may hinder innovation by keeping firms smaller than they would otherwise be. The second chapter examines the role of financial institutions in fostering innovation, distinguishing between the effect of bank-based and market-based institutions. Using a differences-in-differences approach, I find that industries that are more dependent on external finance innovate relatively more in countries with higher financial development. I decompose the effect of financial institutions between banks and stock markets and I find that both significantly increase patenting activity. However, I show that only stock markets are associated with higher quality patents, whereas bank credit has no effect. In addition, I study an alternative mechanism by which financial institutions may influence growth and innovation: providing risk sharing opportunities. I present results that suggest that more volatile industries are relatively more innovative in countries with more developed stock markets, consistently with financial markets also playing an important role in facilitating risk taking. In the last chapter, written jointly with Frank Schneider, we empirically analyze the causal nature of the volume-outcome relationship to study the effect of consolidation on survival in the market of matched unrelated donor bone marrow transplants for leukemia treatment. Controlling for average hospital quality during the study period, we find that one additional transplant during the six months preceding a transplant is associated with a 1.5% increase in the patient's survival probability. We address the potential endogeneity due to time variant hospital quality with instrumental variables, using the number of competitors and their distance to a hospital as exogenous predictors of volume. Predicted volume is then used to establish a causal relationship between volume and outcome. We find evidence for a strong and causal relationship between volume and survival. One additional transplant raises survival by 3%. We also find evidence for perfect forgetting, indicating that high average volume is not sufficient but that regular performance of the procedure is also essential. This evidence suggests that some consolidation would be beneficial for patient survival, even at the cost of lower competition.