Problems with the Definition of Food Deserts

by Brandon McFadden

Food deserts are often used to define areas that have low access to food.  In fact, many people are now referring to food deserts as low access, low income areas.  A Food Access Research Atlas is a map that shows tracts that are defined as food deserts throughout the U.S.  The Atlas is made available by the USDA and can be accessed here.  According to the USDA, “The Food Access Research Atlas maps census tracts that are both low income (li) and low access (la), as measured by the different distance demarcations. This tool provides researchers and other users multiple ways to understand the characteristics that can contribute to food deserts, including income level, distance to supermarkets, and vehicle access.” 

However, the current definition of low access may be too general.  A tract is defined as low income if: 1) The tract’s poverty rate is 20% or greater; or 2) The tract’s median family income is less than or equal to 80% of the State-wide median family income; or 3) The tract is in a metropolitan area and has a median family income less than or equal to 80% of the metropolitan area's median family income.  The original food desert measure defines low access as living one mile from a supermarket in urban areas and 10 miles in rural areas.  For more information about how the USDA defines food deserts read this. 

To illustrate that the current definition of low access may be too general, allow me to use Gainesville, FL as an example.  Below are two maps of Gainesville.  The map on the left is from USDA and the map on the right is a map from a Google search (the scaling for the two maps is not exact).  The green tracts in the USDA map represent the original food desert measure and the brown tracts represent a more stringent measure of access—0.5 miles from a supermarket in an urban area. 

From the Google map you can see that there are many Publix grocery stores in or near these green and brown tracts.  Moreover, there are many other supermarkets in the map area that are not shown.  Also in this map area are 3 Winn-Dixie grocery stores, 3 Wal-Marts, Target, Lucky’s Market, Earth-Fare, Trader Joes, Fresh Market, Ward’s Supermarket, Earth Origins, several ethnic specialty stores, and a weekly farmer’s market.  Something not captured by the Atlas, is the availability of public transportation.  For example, there is a bus system in Gainesville that increases the access to supermarkets.

The high number of supermarkets in this map area makes me wonder how access could be reasonably increased in Gainesville.  Consumers obviously need supermarkets, but consumers also need housing, green spaces, medical services, shops, etc.  The point of this is not to trivialize the effects of access to food.  Rather, the point is that the current measure of food deserts appear to be too liberal.  If we are interested in the effects of access and income on diets, we need more realistic measures of low access and income.