Wealthy hubs and poor chains: Constellations in the U.S. urban migration system

Abstract

Flows of people connect cities into complex systems. Urban systems research focuses primarily on creating economic models that explain movement between cities (whether people, telecommunications, goods or money), and more recently, finding strongly and weakly-connected regions. However, geometrically graphing the dependency between cities within a large network may reveal the roles of small and peripheral city agents in the system to show which cities switch regions from year to year, which medium-sized cities serve as collectors for large cities, and how the network is configured when connected by wealthy or deprived agents. We propose a network configuration method called ‘best friend’ networks, where a node attaches to one preferential node, so that edges = nodes = n. Our case study is 20 years of migrants, sourced from the U.S. Internal Revenue Service, traveling between U.S. cities. In our networks, an edge is created to link a city to its most popular migrant destination city for a given year. The resulting configurations reveal closely connected “constellations” of cities comprised of chains, trees, and hub-spoke structures that show how urban regions are configured. We also show routing behavior within these networks to reveal that high-income migrants tend to flock to hub cities, while low-income migrants form local city chains via nearby movements.