A Random Walk Approach to Quantifying Racial Segregation and Food Access in Atlanta, Georgia

Abstract

Racial segregation continues to persist in urban areas across the United States, contributing to racial inequality across domains such as education, housing, and the focus of this thesis: food access. Existing food justice research has focused on examining the relationship between race and food access, with conflicting results. Exploration of the influence of segregation itself on urban food landscapes is much more rare, and only measures segregation through metrics that are dependent on the scale of the regions of interest, preventing accurate comparison between multiple geographies. In search of a multiscalar method for quantifying segregation, we find an existing random-walkbased approach and apply it to the Atlanta, Georgia metropolitan area. We relate our segregation results to the Atlanta food landscape by both exploring traditional food access measures and introducing a new metric related to spatial clustering of food store locations. We find that in Atlanta, segregated, predominantly Black regions experience the worst food store access. We also discover that at the census-tractlevel, higher segregation is correlated with worse proximity to food stores. Finally, we find that more segregated regions have better access to convenience stores than to grocery stores. By treating segregation itself as a significant factor in food access disparities, we bring nuance to current understandings of racially segregated urban food landscapes.