An Explanatory Model of City-size Distribution: Evidence from Cross-country Data

Abstract

This paper utilises data from files of the World Bank to investigate the relevancy of a wide set of variables in explaining systematic variations of city-size distribution across countries. The Pareto exponent is employed as a measure of population concentration among cities of different sizes. The empirical results allow us to confirm a number of hypotheses. High degrees of economic development as measured by per capita GNP foster urban dispersal. Conversely, high degrees of government involvement in the economy and prevalence of significant scale and agglomeration economies promote urban concentration. Like most other cross-sectional studies, this study does not provide support to a dominating view in the economic development literature that at low levels of development the relation between economic development and population concentration is negative while at high levels the relation reverses and becomes positive.