Emergent cities

This model constitutes the first behaviorally-plausible, microscopic explanation of the city size distribution, on the one hand, and the mutual existence of Zipfian firm and city sizes on the other. [...] The number of people is fixed in the model, but the number and sizes of firms and cities is endogenous. [...] This model is a quantitative explanation of several empirical features of U. S. firms, including: (1) the extreme right skewness of the firm size distribution, (2) the non-normal distribution of firm growth rates, (3) the decrease in growth rate variance with firm size, (4) the wage premium associated with working in larger firms, and (5) approximately constant returns to scale at the aggregate le [...] Varying the parameters of the model changes only the quantitative aspects of the model. [...] But where the probabilities in Simon's model governed the disposition of 'lumps' of individuals, here it is a parameter in agent decision-making, mediated through the institution of the firm.