Thesis abstract:

«p»Simulations, also known as agent-based models (ABMs), are used to show how complex systems behave. As a result of its successful application to biology and physics, this modelling method is now increasingly relied upon to understand complex socio-economic behaviour and to give related policy advice. Consequently, ABMs of economic systems are based on analogies to other systems, such as epidemics or particles. While there have been isolated attempts to highlight issues with applying these analogies to specific social phenomena, there has been no effort to present the components of complex economic systems, or the mechanisms that operate between such components, in a way that facilitates the development of more appropriate ABMs. This thesis argues that, for ABMs to effectively represent complex economic systems, it is essential that modelers develop analogies based on a system where the unique internal environments of individuals, the collective institutional environment, and the highly adaptable physical environment all co-evolve. A framework for developing and implementing such ‘system-based’ analogies is presented.«/p»