The previous chapter presented the basic endogenous technological change models based on expanding input or product varieties.
The advantage of these models is their relative tractability. While the expansion of the variety of machines used in production captures certain aspects of the economics of innovation, most process innovations in practice either increase the quality of an existing product or reduce the costs of production.
Therefore, typical process innovations have a number of distinct features compared to the “horizontal innovations” of the previous chapter. For example, in the expanding machine variety model a newly-invented computer is used alongside all previous vintages of computers, though, in reality, a newly-invented computer often replaces existing vintages. Thus in some fundamental sense, models of expanding machine variety do not provide a good description of innovation dynamics in practice because they do not capture the competitive aspect of innovations. These competitive aspects bring us to the realm of Schumpeterian creative destruction in which economic growth is driven, at least in part, by new firms replacing incumbents, and new machines and products replacing old ones. For this reason, the models discussed in this chapter are often referred to as Schumpeterian growth models. My purpose in this chapter is to develop tractable models of Schumpeterian growth.As Chapter 12 discussed, innovations that involve quality improvements or cost reductions will feature the replacement effect, which implies that entrants should be more active in the research process than incumbents. Schumpeterian growth therefore raises a number of novel and important issues. First, in contrast to the models of expanding varieties, there may be direct price competition between different producers with different vintages of quality or different costs of producing the same product. This will affect both the description of the growth process and a number of its central implications.
For example, market structure and anti-trust policy can play potentially richer roles in models exhibiting this type of price competition. Second, competition between incumbents and entrants brings the business stealing effect discussed in Chapter 12 to the fore and raises the possibility of excessive innovations.This description suggests that a number of new and perhaps richer issues arise when in the context of Schumpeterian growth models. One may then expect models of Schumpeterian models to be significantly more complicated than expanding varieties models. This is not necessarily the case, however. This chapter presents the basic models of competitive innovations, first proposed by Aghion and Howitt (1992) and then further developed by Grossman
and Helpman (1991a,b) and Aghion and Howitt (1998). The literature on models of Schumpeterian economic growth is now large and an excellent survey is presented in Aghion and Howitt (1998). My purpose here is not to provide a detailed survey, but to emphasize the most important implications of these models. I will also present these models in a way that parallels the mathematical structure of the expanding varieties models, both to emphasize the similarities and to clarify the differences. A number of distinct applications of these models are also discussed later in the chapter and in the exercises.
14.1.