ABSTRACT
Cities are spatial agglomerations of individual agents acting in particular facilities and buildings. There are at least two camps of such explanations: deterministic and path-dependent. On the one hand, traditional models of urban economics assume homogeneous agents that share some common behavioral assumptions, such as rationality in terms of profit or utility maximization. On the other hand, the market sharing processes of competing technologies are subject to increasing returns resulting in multiple equilibriums. The path-dependence models of describing the market competition of technologies can readily be transposed into models depicting spatial competition of locations. The result derived from Arthur's model simply argues that there is a dominant location depending on the entry order of firms to form the industry, but no claim is made as to what properties the emergent locational patterns have. The reviews current work on the power law distribution of cities and depicts the relationship between increasing returns and spatial evolution.
