Subcenter formation in a neg model of a city system

The development of the monopolistic competition model by Dixit and Stiglitz (1977) started as a key innovation in industrial organization and has influenced urban economics. The extension of the Dixit-Stiglitz model into the new economic geography took off with Krugman (1991). We apply the Dixit-Stiglitz model in a new way that differs from Krugman's by proposing three general equilibrium models: a monocentric equilibrium model of a single city economy, a tricentric equilibrium model of a single city economy, and a model of an economy of a system of cities, where the force of urban agglomeration within a city is caused by the taste for product variety, and the force of agglomeration within a city diseconomies is caused by the demand for land. In the model, commuting cost is ignored but inter-city and intra-city product transportation costs are considered. Lot size is variable and spatial equilibrium is determined by the trade-off between accessibility to product and the demand for land.; This dissertation explores some fundamental questions. Why do manufacturing firms concentrate in a few centers rather than disperse? Why do subcenters emerge suddenly at a critical time, and then grow gradually? Do the CBD and the subcenters grow in parallel or does the CBD shrink? What is the optimal city size to maximize workers' welfare? What are the factors influencing the stability of the monocentric equilibrium? The models we propose are different from the existing models. I focus on the formation of the growth of edge cities/subcenters and the factors influencing the monocentric equilibrium. The general equilibrium model of a system of cities may be further extended to study of the economy of hierarchical cities or diversified and specialized cities; to study the formation and growth of central cities; and to test whether it is efficient for city sizes to decrease or increase as total population grows.