| In this work, I argue that the large majority of the models of economic geography possess a very consistent regional and urban theory. However, they have shown a limited capacity to incorporate economic diversity. In Chapter II, I evaluate some recent models of regional economics and conclude that most of the theories restrict the introduction of micro asymmetries, which might be the base for asymmetric regional economies. In Chapter III, I present an agent-based model in which cities and regions are spatial agglomerations of firms and consumers with asymmetric characteristics. I depart from the mainstream models based on representative agents proposing a self-organizing system populated by heterogeneous agents. To simulate the self-organizing economy introduced in Chapter III, I employ cellular automata in Chapters IV and V. Several simulations are presented to illustrate the ability of the model to mimic some stylized facts of industrial organization and regional economics. Two basic models are tested: a closed economy with one industry and a closed economy with four industries. The presence of several industries allows the study of industrial agglomeration patterns of and urban structures. Moreover, the model can generate the empirical regularity called the rank size rule, and in some cases the Zipf's Law, which has been seen as a minimum criterion of admissibility for any model in urban economics. |
