The Mirage of Food Deserts: Disparities between Stated and Revealed Results

Increasing evidence indicates that the neighborhood food environment has become an important determinant of dietary behaviors in recent years (Giskes et al., 2010; Zenk et al., 2009). For example, a lower intake of fresh foods and nutrients (e.g., fruits and vegetables) may cause unhealthy meal patterns (Branca et al., 2007) and lead to adverse health outcomes (Lebel et al., 2009). A review of U.S. studies by Larson et al. (2009) and a case study of Detroit by Zenk et al. (2009) also show that residents who live near supermarkets are more likely to eat more fruits and vegetables, and have healthier diets compared with residents who live further from supermarkets. Identification of “food deserts”, which responds to the neighborhood food environment, has thus been broadly discussed in the existing literature in terms of access to fresh food (Alviola et al., 2013; Bonanno and Li, 2014; Wang et al., 2014). Studies have conceptualized a variety of methods to capture food desert research and fresh food access measures, and two broad approaches have been proposed. One approach is based on a geographic perspective that uses spatial analysis (e.g., calculate the minimum distance from a neighborhood to the nearest supermarket) and analyzes GIS (Geographic Information Systems) data on neighborhood characteristics (e.g., population density, median income). Associations between the neighborhood fresh food environment and socio-economic characteristics have been revealed (Wang et al., 2014). The second approach relies on individual or household data obtained from surveys (LeDoux and Vojnovic, 2014) and interviews (Freedman, 2009). Questions regarding shopping behaviors and constraints of food purchase or fresh food access have been stated through this method. However, the majority of current studies that use the GIS approach to identify food deserts comprise “aggregate” assessments and reveal information about study areas at the neighborhood level. While most research that base on either qualitative or quantitative surveys focus on “individual” level and state respondent’s perspective of the food environment. To our knowledge, no previous study has investigated residents’ perspective about food deserts, or specifically, whether they consider their neighborhoods to be food deserts, together with food deserts identified by the GIS approach. The objectives of this study are two-fold: Compare respondents’ perspective of the food desert issue relative to the GIS analysis that has been widely adopted; Explore the potential reasons for the disparity between stated (survey) and revealed (GIS analysis) results regarding food deserts.