Food Deserts in the Prairies? Supermarket Accessibility and Neighborhood Need in Edmonton, Canada*

*Research funding was provided by the Social Sciences and Humanities Research Council of Canada (SSHRC) and GEOIDE. The authors thank postdoctoral fellow Dr. Nairne Cameron, and student research assistants Vladimir Yasenovskiy, Julia Healy, Nicoleta Cutumisu, Jared Hewko, Mark Pickersgill, Sherry Diehlman, and Kris Ridell for their assistance in the research. We wish to note that the term “unsupportive local food environments,” which we use in this paper, was kindly suggested by one of the referees. We also thank the five anonymous referees for their insightful comments on this paper.

Abstract

The U.S. and U.K. literatures have discussed “food deserts,” reflecting populated, typically urban, low-income areas with limited access to full-service supermarkets. Less is known about supermarket accessibility within Canadian cities. This article uses the minimum distance and coverage methods to determine supermarket accessibility within the city of Edmonton, Canada, with a focus on high-need and inner-city neighborhoods. The results show that for 1999 both of these areas generally had higher accessibility than the remainder of the city, but six high-need neighborhoods had poor supermarket accessibility. We conclude by examining potential reasons for differences in supermarket accessibility between Canadian, U.S., and U.K. cities.

Key Words:

• accessibility
• Edmonton
• food deserts
• supermarkets

Notes

^a25^th and 75^th percentiles (given in italics) are provided for all residential neighborhoods only.

Notes: Lower values of minimum distance and higher values of number of stores indicate higher access. p-values are given in italics.

^aFor each accessibility indicator, low accessibility neighborhoods are defined as those neighborhoods with accessibility scores in the lowest quartile.

^bFor each accessibility indicator, high accessibility neighborhoods are defined as those neighborhoods with accessibility scores in the highest quartile.

Note: Z and p-values computed using Wilcoxon rank-sum test of means.

Notes: Neighborhoods are identified by number rather than name for anonymity.

^aAll neighborhoods had zero stores within 1 km and fell within the top distance quartile.

^bAbove city-wide median for that variable.

^cWithin the top quartile for that variable.

¹Low-income levels were based on Statistics Canada's (1999) before-tax low-income cut-offs for urban areas with 500,000 people, along with 1999 Edmonton Civic Census cross tabulations of household size by household income.

²We selected the combination of methods used here, and used local knowledge and ground-truthing in uncertain cases to ensure as complete an enumeration as possible of all full-range food stores (i.e., supermarkets) in Edmonton as of 1999. However, we acknowledge that we may have inadvertedly missed some independent supermarkets.

*Research funding was provided by the Social Sciences and Humanities Research Council of Canada (SSHRC) and GEOIDE. The authors thank postdoctoral fellow Dr. Nairne Cameron, and student research assistants Vladimir Yasenovskiy, Julia Healy, Nicoleta Cutumisu, Jared Hewko, Mark Pickersgill, Sherry Diehlman, and Kris Ridell for their assistance in the research. We wish to note that the term “unsupportive local food environments,” which we use in this paper, was kindly suggested by one of the referees. We also thank the five anonymous referees for their insightful comments on this paper.