The evolution of technology and adaptive economic behaviour

This thesis studies the role of learning as a mechanism of economic change. Two areas are considered where this would seem to be important. First, how firms learn about new technology; and secondly, how agents learn to behave in interactive situations. A model of research and development is presented which models the process by which firms solve specific design problems. This may be by individual experimental search or by partial imitation. In the latter case, a close parallel is drawn between biological evolution, based on genetic reproduction, and technological evolution, based on firms blending existing technologies. Some economic implications of these processes are explored, including their application to stochastic learning curves, patent design and the transfer of technology to developing countries. The thesis continues by critically assessing the analogy between biological and cultural evolution often used to model how agents learn to behave in interactive situations. It is argued that the methods used by economists exploiting this analogy are often ill-suited to an economic context. Models are presented which deal with specific issues in the transition from a biological context to an economic context, including models of partnership formation, models of imperfect imitation, and models without payoff-monotonic dynamics. The issue of imperfect imitation is expanded upon in an evolutionary model of the infinitely repeated prisoners' dilemma, where it is shown that the problem of inter-generational copying fidelity may allow one to restrict attention to strategies with a very simple stochastic structure.