(De) Spatializing Food insecurity in Providence, RI
September – December, 2018
As my semester-long project for the course “Introduction to GIS” at Brown University, I interrogated the presumed spatial nature of food insecurity, using the case study of Providence, RI and focusing specifically on the concept of food deserts. Previous coursework and professional experiences have furnished me with an appreciation of the complexity of food insecurity as well as an understanding of the variety of interventions necessary to tackle it (including growing the number of healthy food retail locations, making healthy food more affordable, and expanding nutrition and general food education opportunities). Because Providence is a place with many organizations and government initiatives partnering to address food insecurity, I became interested in analyzing the impact of existing initiatives through GIS mapping to better understand why food insecurity remains a challenge faced by one in eight Rhode Islanders. My research question thus emerged: How can we re-conceptualize the concept of the “food desert” in Providence, RI to more accurately represent the spatial components of food insecurity and how people experience it? Additionally, I sought to identify the different spatial components of food security, as well as to understand whether physical access to food security proxies necessarily produces food security.
I spent much of the semester accumulating mappable data sets of food insecurity proxies in Providence, including SNAP and non-SNAP grocery stores, Farm Fresh RI’s in-school nutrition education programs, food pantries, and community gardens. Through weekly course labs, I experimented with diverse methods for analyzing the spatial relationships of these data. I ultimately discovered two best-fit methods for mapping my data: kernel density analysis and euclidian distance analysis. Both methods helped me to depict the density of food security proxies and the areas of Providence most likely to be able to access them. I ultimately calculated that a strong majority of city residents reside an average of one kilometer or less from a given food security source.
The limited time allowed by a single semester of research combined with the challenge of finding mappable data for my research question left me with many unanswered questions and ideas for future research. Mainly, comparing my results with self-reported levels of household food insecurity would allow me to draw stronger conclusions around the question of whether physical access to food security sources necessarily produces food security. Additionally, writing more nuance into my data (for instance the number of students served by various in-school nutrition education programs and the proportion of fruits and vegetables sold by grocery stores) would also produce more nuanced maps of healthy food access in Providence.