The power law within a metropolitan area

While theoretical reasoning emphasizes the complexity in metropolitan spatial structures, empirical efforts still focus on measuring spatial structures using employment centers. This paper measures a metropolitan spatial structure as a system of employment clusters for 361 metro areas in the United States. We apply a log-log model to assess the relationship between a cluster's size and its rank in a metro area. We found that among the largest 50 metro areas: (1) cluster sizes in a metro area follow a power law distribution and (2) larger metro areas tend to have evener spatial structures. The results suggest that policymakers can better predict urban growth locations and sizes; may invest in the largest clusters for the biggest economic payoffs; and should consider holistically all elements' (e.g., clusters, infrastructure, socioeconomic interactions) respective scaling laws in the city for its healthy urban growth.