Hierarchical organization of cities and nations

TL;DR

This paper explores the scale-invariant distributions of city and nation populations and areas, revealing different power-law behaviors and proposing growth models based on geometric partitioning to explain these patterns.

Contribution

It introduces simple growth models based on plane partitioning to explain the power-law distributions of cities and nations without complex socio-economic assumptions.

Findings

Population and area distributions follow inverse power-laws.

Exponent differs significantly between cities and nations.

Models successfully reproduce empirical distributions.

Abstract

Universality in the behavior of complex systems often reveals itself in the form of scale-invariant distributions that are essentially independent of the details of the microscopic dynamics. A representative paradigm of complex behavior in nature is cooperative evolution. The interaction of individuals gives rise to a wide variety of collective phenomena that strongly differ from individual dynamics---such as demographic evolution, cultural and technological development, and economic activity. A striking example of such cooperative phenomena is the formation of urban aggregates which, in turn, can be considered to cooperate in giving rise to nations. We find that population and area distributions of nations follow an inverse power-law behavior, as is known for cities. The exponents, however, are radically different in the two cases ($\mu \approx 1$ for nations, $\mu \approx 2$ for cities). We interpret these findings by developing growth models for cities and for nations related to basic properties of partition of the plane. These models allow one to understand the empirical findings without resort to the introduction of complex socio-economic factors.