Innovation in an Evolutionary Framework
Based on Schumpeter’s invention-innovation-diffusion trilogy, evolutionary economics tackles three basic questions. (1) How is technological variety generated? (2) How, at which rate and in which direction do innovations spread through the system? (3) How do innovation activities and diffusion processes shape the evolution of firms, of industries and of the economy as a whole?
The formal application of evolutionary thinking to economics by Nelson and Winter (1982) is inspired by Schumpeter’s theory and is concerned with the mechanisms involved in the process of technical change.
The approach relies on explicit micro-foundations with heterogeneous, bounded-rational agents who do not engage in optimizing behaviour but follow routines. Further, the economy is not in equilibrium but constitutes the environment in which evolutionary processes of stochastic variety generation and competitive selection interact. Based on this, Nelson and Winter investigated the processes of searching for new production methods and imitation, the dynamics of selection and the consequences for aggregate growth and industry concentration.According to this “bottom-up” approach, technical change at the aggregate level is the outcome of ongoing “search and selection” of firms and can be measured by what is called “evolutionary accounting”: “The fundamental evolutionary idea is that distributions... change as a result of (1) learning by incumbent entities; (2) differential growth (that is, a form of selection) of incumbent entities themselves; (3) death (indeed, a different and more radical form of selection); and (4) entry of new entities” (Dosi and Grazzi 2006: 194-5). See Metcalfe (1997) for the connection between innovation, diffusion and the aggregate measure of total factor productivity. Silverberg and Lehnert (1993) adopted the Nelson-Winter approach and explain the empirical evidence of long waves generated by the introduction and gradual diffusion of new energy technologies.
For an extensive discussion of evolutionary models of innovation-driven growth, see Dawid (2006).In evolutionary selection models diffusion of new techniques is effectuated via differential firm growth: cost and profit rate differentials translate into growth differentials. Similar to the vintage approach, investment is necessary in order to exploit the economic potential of innovations. The formalization of selection dynamics adopts the replicator equation and Fisher’s principle, taken from population theory, and provides insights into the determinants of the direction and velocity of technical change as well as into the impact of diffusion processes on aggregate dynamics (see Metcalfe 1998).
Evolutionary diffusion models provide an explanation of the stylized facts of the diffusion process, namely, that it is a time-consuming process and typically follows an S-shaped curve. Early empirical diffusion studies dating back to the 1940s and 1950s led to the development of a variety of technology diffusion models which are not based on the mechanism of differential growth but on imitation and adoption decisions, including epistemic and probit models (see Stoneman 2002). The evolutionary framework of Nelson and Winter (1982) laid the basis for the development of various further concepts concerned with Schumpeter’s trilogy seen as a non-linear process.