Title: Competition, innovation, and the number of firms
Abstract: Theories of endogenous markups, firm entry, and innovation consistently predict a positive relationship between barriers to competition and firm-level innovation, in stark contrast to a large empirical literature that documents a negative relationship. I argue that this puzzle can be resolved by considering two distinct types of barriers – barriers to starting a firm, and barriers to competing in a product market. Barriers to starting firms discourage startups, resulting in higher markups, higher market shares, and more innovation. Barriers to product-market entry also increase market shares within each market, and thereby encourage innovation. But firms also enter fewer markets, which discourages innovation. On net, market-entry barriers decrease firm-level innovation and the number of competitors per market, while increasing the number of firms per industry. To test the implications of the theory, I consider an episode of product market deregulation (the Single Market Programme in the E.U.), and find that lower market-entry barriers indeed resulted in fewer firms. Interpreted through the lens of the model, I estimate that productivity increased by 8.4% in deregulated industries, and show that this estimate would be 30% lower if deregulation was mistakenly modeled as lowering firm startup costs. Across countries, I provide evidence that market-entry barriers are quantitatively much more important than startup barriers. As evidence for the central mechanism in the model I show that across countries, higher market-entry barriers are associated with fewer products per firm.