Simulation models of technological innovation: A Review

The use of simulation modelling techniques in studies of technological innovation dates back to Nelson and Winter''s 1982 book "An Evolutionary Theory of Economic Change" and is an area which has been steadily expanding ever since. Four main issues are identified in reviewing the key contributions that have been made to this burgeoning literature. Firstly, a key driver in the construction of computer simulations has been the desire to develop more complicated theoretical models capable of dealing with the complex phenomena characteristic of technological innovation. Secondly, no single model captures all of the dimensions and stylised facts of innovative learning. Indeed this paper argues that one can usefully distinguish between the various contributions according to the particular dimensions of the learning process which they explore. To this end the paper develops a taxonomy which usefully distinguishes between these dimensions and also clarifies the quite different perspectives underpinning the contributions made by mainstream economists and non-mainstream, neo-Schumpeterian economists. This brings us to a third point highlighted in the paper. The character of simulation models which are developed are heavily influenced by the generic research questions of these different schools of thought. Finally, attention is drawn to an important distinction between the process of learning and adaptation within a static environment, and dynamic environments in which the introduction of new artefacts and patterns of behaviour change the selective pressure faced by agents. We show that modellers choosing to explore one or other of these settings reveal their quite different conceptual understandings of "technological innovation".