Competition and innovation in automobile markets

Abstract

Innovation is viewed as critical to fostering the growth of markets, generating efficiencies, and improving welfare. In this study we examine the determinants and the shifting dynamics of innovation in the U.S. automobile market. We use firm-level time-series data over a long horizon (1969-2012) for nine well established firms selling in the U.S. market (GM, Ford, Chrysler, Toyota, Honda, Nissan, Volkswagen, BMW, and Daimler).We examined three aspects related to market competition, innovation, and innovation rivalry in the U.S. automobile industry. First, we examine the relationship between competition and innovation. We use patent counts as a measure of innovation, and use market shares and market concentration as measures of market competition. Some of our key findings are: (1) increase in firms’ market shares result in higher patenting, and the relationship is reasonably non-linear; (2) higher market-wide competition results in an increase in patenting, and the relationship is weakly non-linear; (3) there is relatively strong path-dependence in firms’ patenting behavior. Second, we examine the relationship between knowledge gap and patent rivalry. In particular, we examine how a firm’s current patenting behavior is influenced by the knowledge gap between the leader and the firm, firms’ market shares, and the indirect effects of knowledge gap and market share. Our key findings are: (1) the relationship between patenting and knowledge gap is non-linear, and is U shaped; (2) an increase in market share results in higher current patenting; (3) the interaction between firms’ market share and technology gap does not have a statistically significant effect on their current patenting. Third, we examine the changing composition of patents. Over time, the dynamics in a market can change substantially due to competition in the product market, shifting positions of firms from an innovation standpoint, fundamental shifts in technology, changes in regulatory constraints, among other factors. We study the intertemporal shifts in the composition of innovation, and in particular, we examine how a firm’s current patenting behavior is influenced by the own patent concentration and rivals patent concentration. Our key findings are: (1) the relationship between knowledge stock and current innovation is complex, and depends on technologies categories; (2) an increase in own patent diversification results in higher current patenting in electrical & electronic and mechanical technologies; (3) an increase in rivals’ patent diversification results in higher current patenting in electrical & electronic on mechanical technologies; (4) rivals’ knowledge diversification has statistically higher effects on current patenting than own knowledge diversification, and is the main driver of changing compositions in patenting.