Essays on Innovation, Firm Dynamics, and Productivity Growth

Abstract

The United States has been experiencing a secular decline in the pace of business formation and young firm activity shares in recent decades. U.S. productivity growth also slowed down during the same period. This thesis studies two questions. First, what are the driving forces and long-term growth implications of the observed trends? Second, how is the creative destruction process translated into the measured cost of living?Using a longitudinal worker-firm matched dataset from the U.S. Census Bureau, I document that in the innovation intensive high-tech sector, the decline in young firm activity shares is accompanied by: 1) a decline in the growth rate of the demand for skills, and 2) a flattening of the life cycle of skilled labor accumulation of high-tech firms. By developing an innovation-based firm dynamics model that is consistent with the micro-level skilled labor accumulation over the firm life cycle, I show that rising frictions in skilled labor adjustment can explain the joint evolution of young firm employment shares and demand for skills. These frictions influence productivity growth through affecting the stock of human capital firms possess. A calibrated model shows that a rise in skilled labor adjustment costs lowers productivity growth by 75 basis points in the high-tech sector. A rise in entry costs, on the other hand, is not likely the main driver for declining young firm activities, as it implies an increase in demand for skills. Finally, productivity gain (loss) from reallocation can be offset by the general equilibrium effects of reallocation on aggregate demand for skills. The impact of innovation on welfare depends critically on taking into account the impact of innovation on the cost of living. Building upon the framework of Redding and Weinstein (2020), I estimate the exact cost of living in the U.S. consumer goods sector using the Nielsen Retail Scanner data over the period of 2006 to 2015. The estimated inflation rate considering product turnover and taste shocks is one percent lower than the tradition CPI measure. Furthermore, I show that the direction of the bias in traditional price indices is determined by the correlation between the initial period market share of products and relative taste shocks. Finally, the exact price index based on a nested CES demand structure can be used to study the contribution to the cost of living by firms of difference sizes.