Redefining the food desert: combining GIS with direct observation to measure food access

As public and private resources are increasingly being directed towards the elimination of food deserts in urban areas, proper measurement of food access is essential. Amelioration has been approached through the use of farmers markets, virtual grocery stores, and corner store programs, but properly situating these assets in neighborhoods in need requires localized data on both the location and content of food outlets and the populations served. This paper examines the reliability of current techniques for identifying food deserts, and identifies some of the flaws in those approaches. Information derived from geographic information systems (GIS) mapping is the predominant means of determining food availability. In this study, food access in Bridgeport, CT, is examined utilizing both computer-based GIS mapping and on-the-ground observations. While the GIS output indicates generalized food accessibility issues, supplementation by survey data reduces the geographic extent of the food desert problem. Still, nearly 60,000 people (40 % of the population) reside in neighborhoods served only by small retailers who provide few healthy food options, and those at inflated prices. The high opportunity cost of travelling by bus to a major grocery store may outweigh the direct cost savings, and residents choose to consume locally available but unhealthy foods.