Defining ‘metropolitan’ poverty: Isolation gradients in major US urban areas

Abstract

As suburban poverty has become more prevalent in recent years, low-income neighbourhoods have emerged outside of many central cities. In many regions, however, there remains a stark divide between the core city and outlying areas. Measuring poverty concentrations, therefore, depends on the choice of geographic scale. Widespread urban poverty results in low measures of poverty concentration inside core cities, and whether this increases or decreases once suburbs are considered depends on specific regional characteristics. This study calculates ‘isolation gradients’ for the population in poverty in the 50 largest US Metropolitan areas (MSAs), extending from the core city to beyond the MSA level. A variety of patterns emerge, including sharp increases in isolation scores outside the city (as in Detroit) as well as decreases (for Atlanta). An econometric analysis of the ratio of MSA-level to core-city isolation scores finds that, controlling for land area and core-city poverty rates, much of the variance is explained by per-capita income and differences in racial segregation. A comparison of gradients in 2000 and 2015–2019 for a subset of cities shows additional patterns of suburban and urban convergence and divergence.