Identifying Food Deserts in Brandon, Manitoba Using GIS
Peter Brandt1, Dylan McNaughton1, Dr. Dion J. Wiseman1,
Dr. Derrek Eberts1, Nancy McPherson2, Erin May3
1Brandon
University, Department of Geography; 2Brandon University, Department of Health Studies;
3Brandon
Neighbourhood Renewal Corporation
Introduction
Food deserts are a significant problem in many urban areas worldwide, and are an issue
for city and food officials. The main purpose of this project was to determine if there are
food deserts in Brandon, Manitoba, Canada. According to Smoyer-Tomic, et al. (2006),
food deserts are "Neighbourhoods where residents have little or no access to stores or
restaurants that provide healthy and affordable foods”. The basic idea of food deserts are
that people may have limited access to healthy food due to socioeconomic constraints or
health issues that limit their mobility. As a result, major supermarkets located far from
where they live may be inaccessible.
This project identifies food deserts using the following criteria: walking or transit distance
to grocery stores, proportion of residents who walk or take the bus, and income level as
an indicator of mobility.
Background
Food deserts are a significant problem in many urban areas worldwide, and are an issue
for city and food officials. The main purpose of this project was to determine if there are
food deserts in Brandon, Manitoba, Canada. According to Smoyer-Tomic, et al. (2006),
food deserts are "Neighbourhoods where residents have little or no access to stores or
restaurants that provide healthy and affordable foods”. The basic idea of food deserts are
that people may have limited access to healthy food due to socioeconomic constraints or
health issues that limit their mobility. As a result, major supermarkets located far from
where they live may be inaccessible.
This project identifies food deserts using the following criteria: walking or transit distance
to grocery stores, proportion of residents who walk or take the bus, and income level as
an indicator of mobility.
Grocery Stores
In this study, the first variable that was considered was which food vendors should be
included in this analysis. The type and location of food vendors in Brandon was provided
by the Brandon Neighbourhood Renewal Corporation. These data were imported as
points to ArcGIS and were categorized into grocers, specialty stores, bakeries/baked
goods, meat markets, health food stores, and convenience stores. For the purposes of this
project we used only grocers and did not consider the other subcategories of food
retailers. A large grocery store was defined as a store greater than 30,000 square feet, with
a produce, deli, dairy, and frozen aisle section (e.g. big box store). An intermediate
grocery store was defined as less than 30,000 square feet, with more than two cash
registers, and two or more aisles found in large grocery stores. A small grocery store was
defined as one with two or less cash registers, and one or more aisles found in large
grocery stores.
This project used Statistics Canada census data from 2006 at the dissemination area level
to determine income level areas in Brandon, Manitoba.
For this analysis, risk (or level of accessibility) was evaluated based on income level and
proportion of residents who walk or take the bus. Risk associated with income level was
evaluated as the percentage of households with an income of less than $15,000. However,
risk based on income is not the only to be considered when identifying a food desert. This
analysis also considered how many people walk or use the bus. This was determined by
identifying the employed labour force 15 years and over who walk or take public transit
and as a percentage of the total employed labour force 15 years and over.
These two metrics were then added to determine a combined food desert risk index. This
produces an index between zero and two, where the higher the index value the higher
risk it has of being a food desert.
For this project, both the area within walking distance and the area within bus distance of
grocery stores were determined, using the ArcGIS Network Analyst toolset.
The walking service area was determined by combining the service area of the small,
intermediate, and large grocery stores using the Brandon road network. The service area
of large grocery stores was established as a road, rather than straight line, distance of 1000
metres, 500 metres for intermediate grocery stores, and 390 metres for small. These
distances were based on the London study by Larsen (2008), which considered
supermarkets of comparable size to our large grocery stores. The service area of
intermediate grocery stores was established as a distance of 100 metres further than one
would be expected to walk to a bus stop, and for small grocers, somewhat less than this
distance.
In the London study, Larsen (2008) reasoned that a ten minute bus ride would be an
appropriate amount of time to travel on a bus, which is about 3 kilometres. In addition, is
was assumed that an individual would walk up to 500 metres to a bus stop. This resulted
in a 3500 metre service area for supermarkets. However, for this project an alternative
method was used to determine the bus service area based on the assumption that a round
trip of 3000 metres would be the maximum distance travelled by bus as measured from
the bus stops. To determine the bus stops that were 1500 metres from the grocery stores,
a network service area was created around the large and intermediate grocery stores using
the bus route network. The bus route network was created using Brandon Transit data
from Google Maps. Small stores were not included in this analysis because it was assumed
that if a person were to ride a bus to get groceries they would travel to a large or
intermediate store. Due to the one way nature of the bus routes, some bus stops were
within a 1500 metre trip to the grocery store, but not a 1500 metre trip back, or vice versa.
To account for this, a “closest facility analysis” was performed on the bus stops within the
bus service area which calculated the distance from each bus stop to the nearest grocery
store. A sort was then performed to determine which bus stops were within 1500 metres
of the stores. The walking service area of these bus stops was then determined using the
road network and a distance of 400 metres rather than 500 metres, which is the
approximate service area for a city bus stop according to Foda (2010).
Results
The food deserts were then determined by displaying the combined walking and bus
service areas with the food desert risk index. The areas that have the highest index values
(red) and are outside the combined walking and bus service areas of grocery stores
(green) are at highest risk of being a food desert. Conversely, areas with the highest index
values (red) but are within the combined service areas (green) are not considered at risk
because of the assumed ease of access.
A Closer Look
The result of displaying the combined walking and bus service areas with the food desert
risk index showed a number of high risk areas which were not covered by the combined
service areas. Figure 10 shows two areas, one on the eastern edge of downtown and the
other in the southern portion of the city, with high risk index values which fall just
outside the combined service area of grocery stores. Those residing in these potential
food deserts may be forced to purchase their groceries at a convenience store nearby or
travel the additional distance to purchase them at a grocery store. Figure 11 shows the
area around Brandon University, which has a moderate risk index and lies just outside the
combined service areas of grocery stores. However, students who live on campus may not
have been included as residents of this area in the census data. This may have skewed the
food desert risk index result, as the income level data would haven been altered by the
inclusion of students living on campus. Consequently, this area may potentially be a food
desert as well and students living in the university residences without vehicle access are
left to rely on the cafeteria for nutritious food.
References
Foda, Mohamed A., and Ahmed O. Osman. "Using GIS for Measuring Transit Stop
Accessibility Considering Actual Pedestrian Road Network." Journal of Public
Transportation 13.4 (2010): 23-40. Web.
Glanz, Karen, James F. Sallis, Brian E. Saelens, and Lawrence D. Frank. "Nutrition
Environment Measures Survey in Stores (NEMS-S)Development and
Evaluation." American Journal of Preventive Medicine 32.4 (2007): 282-89. Print.
Larsen, Kristian, and Jason Gilliland. "Mapping the Evolution of 'food Deserts' in a
Canadian City: Supermarket Accessibility in London, Ontario, 1961–2005." International
Journal of Health Geographics 7.1 (2008): 16. Print.
"Planning / Baltimore Food Policy Initiative / Food Deserts." Planning / Baltimore Food
Policy Initiative / Food Deserts. Johns Hopkins Center for a Livable Future, 2012. Web. 25
Apr. 2014.
<http://www.baltimorecity.gov/Government/AgenciesDepartments/Planning/BaltimoreF
oodPolicyInitiative/FoodDeserts.aspx>.
"Routes & Schedules." Routes & Schedules. Brandon Transit, n.d. Web. 25 Apr. 2014.
<http://brandontransit.ca/route-schedules>.
Smoyer-Tomic, Karen E., John C. Spence, and Carl Amrhein. "Food Deserts in the
Prairies? Supermarket Accessibility and Neighborhood Need in Edmonton, Canada*." The
Professional Geographer 58.3 (2006): 307-26. Print.
Statistics Canada. 2007. Population and Dwelling Counts, for Canada, Provinces and
Territories, 2006 and 2001 Censuses, 100% Data (table). "Population and dwelling count
highlight tables, 2006 Census." "2006 Census: Release topics." Census. Statistics Canada
Catalogue no. 97-550-XWE2006002. Ottawa, Ontario. March 13.
http://www12.statcan.ca/english/census06/data/popdwell/Table.cfm?T=101 (accessed
November 3, 2008.)