Market Structure and the Direction of Technological Change∗

We study a model where innovation comes in two varieties: improvements on existing products, and new products that expand the scope of a technology. We make this distinction in order to highlight how market structure can determine not only the quantity of innovation but also its direction. We study two market structures. The first is the canonical one from the endogenous growth literature, where innovations can be developed by anyone, and developers market their own innovations. We then consider a more concentrated industry, where all innovation and pricing for a given technology is monopolized. We study the implications of the different market structures for both types of innovation, focusing on differences they induce in the direction of technological change. We apply our model model to the case of a hardware/software technology and analyze which market structure offers greater profits to a monopolist who can monopolize either hardware or software. We compare social welfare across the market structures, and discuss whether one type of innovation should be subsidized over another. ∗We thank Philippe Aghion, Paul Romer, and participants and the 4th Duke/Texas/Northwestern IO theory conference for suggestions. †University of Iowa. ‡Stanford GSB.