COGS Honors the Late Arthur Getis at NARSC 2022
In a special session honoring SDSU’s renowned Professor of Geography Art Getis, COGS researchers will present a new paper based on one of Getis’s most famous contributions to the field: the $G_i$ statistic.
In this paper we examine the evolution of urban spatial structure in U.S. metropolitan areas over nearly two decades. Unlike seminal contributions in the field, which focus on the relationship between transport systems and agglomeration economies in shaping the allocation of jobs and industries, the growing body of contemporary research on metropolitan employment centers focuses on the development of novel statistical strategies at the expense of stronger theoretical grounding. Here, we provide a stronger wedding between location theory and spatial analysis techniques. Using annual block-level data from the Longitudinal Employment Household Dynamics database, we introduce a novel technique for identifying regional employment centers that adheres to both urban economic theory and pays homage to classic contributions in local spatial statistics. Our identification process recognizes that theories of agglomeration economies are founded on the basis of accessibility to jobs, not simply their concentration into arbitrary boundaries. Centers are then defined as local spatial statistical outliers on the network-based cumulative accessibility surface. We proceed by identifying the location and employment makeup of centers for each metropolitan region in the USA from 2002 to 2019 and discuss emergent trends across time and space. Critically, we not only explore empirical patterns, but we discuss the relationship between polycentricity, the evolution of urbanization and localization economies, and regional specialization.