Firm Size, Innovation Dynamics and Growth

This paper investigates the relationship between the size of the \u85rm and the quality of innovations of the \u85rm. Much of the previous literature on innovation focuses on innovation frequency with an economy-wide uniform innovation quality. In contrast to the previous liter-ature, this paper allows \u85rms to choose not only the stochastic innovation frequency but also the innovation quality and focuses on how this heterogeneity in innovation quality is a¤ected by the size of the \u85rm. This paper has three distinct contributions: First, using Compustat \u85rms and their patent applications, I document the following three reduced form facts: i) Firm sales growth (both short-run and long-run) is negatively related to the \u85rm size. ii) Firm R&D intensity, de\u85ned as R&D expenses over sales is negatively related to the \u85rm size. iii) The quality of innovation, proxied by the number of citations that a patent receives, is negatively related to the \u85rm size. Second, I build a tractable general equilibrium growth model that is rich enough to investi-gate these empirical results. I prove the existence of the equilibrium, characterize its properties and show that the predictions of the theoretical model are consistent with the reduced for