Firm Creation and Local Growth

Abstract

Firm creation is central to many theories of economic growth. I show using U.S. Census microdata that new firms play a dominant role in the growth of local areas, such as cities and counties. Entry is very persistent at the local level, and variation in this extensive margin accounts for most of long-term employment growth. In contrast, the firm lifecycle is invariant across space. I rationalize these findings in a simple theory of variety-led growth, in which firm creation amplifies and propagates local shocks. New firms raise wages and attract new workers, whose spending spurs the creation of more new firms. I estimate the model exploiting changes in local demand driven by aggregate structural change, and find that this feedback is quantitatively powerful. Using the model, I show that the employment effects of the decline of manufacturing in the U.S. Rustbelt have been greatly amplified by a substantial startup deficit.