Monday, July 18, 2011

Diagnosing supermarket deserts

USDA's new Food Desert Locator offers a lot to think about.  At least in U.S. cities, I think the reaction of many viewers will be surprise that food deserts appear so few and far between.  Most poor neighborhoods in most cities do not appear to be food deserts.

For example, here is my home town of Washington, DC.  Even Ward 8, across the Anacostia River in Southeast Washington (in the bottom middle of the map, near the southern point of the DC diamond, north of the Maryland border), has only one fairly small census tract colored in pink.  Ward 8 was highlighted as a problem area in a report from D.C. Hunger Solutions on the "Grocery Gap."


Faced with surprising data, two good responses are: (a) to read the data definitions carefully, and (b) to see for yourself.

(a) The data definition for a "food desert" in USDA's mapping utility is "a low-income census tract where a substantial number or share of residents has low access to a supermarket or large grocery store," meaning more than a mile to such a store.

(b) The part about "seeing for yourself" is more fun.  On a business trip to USDA this week (disclosure: some of my funding for research on food economics comes from USDA), I checked out a bike from the super-terrific new Capital Bikeshare program in Washington, and took it on a tour of food retail store fronts.  I was impressed that the Capital Bikeshare kiosks are located all over town, including low-income neighborhoods as well as tourist destinations.

This corner store does not count as a "supermarket or large grocery store" under the USDA definition, and yet I always give these non-chain retailers some thought when I do this type of food retail tour.  I would have mixed feelings if new supermarkets, supported by tax incentives, put out of business these retailers that stuck with a low-income neighborhood even in the toughest years.  Coincidentally, when I passed by on my shiny red rental bike, this corner store had a lovely bright poster advertisement for ... Capital Bikeshare.


View Larger Map


Although I did not visit them on this trip, here are the three larger food retailers in this part of Washington, and the reason why most of Ward 8 did not show up as a food desert on the USDA mapping utility.


View Larger Map


View Larger Map


View Larger Map


There is no conclusion to this post.  Just material for longer contemplation, while following the current policy debate about food deserts.

6 comments:

Stephanie said...

I'm sure you saw this, but just in case: http://www.latimes.com/health/la-he-food-deserts-20110712,0,537936.story

Stephanie said...

You make a really good point about the smaller retailers. Did you go into that smaller store to see what they had?

I think these kinds of monitoring systems are potentially prone to the problem you mention...they can obscure the important details, the things that really give the character to a place. In this case, the system can obscure those little Mom and Pop businesses that are so important to making places/specific neighborhoods attractive and interesting (if I'm understanding that the little shops didn't show up on the USDA Food Desert thing)

Anonymous said...

I'm not convinced the smaller retailers are actually that great for the neighborhood (in terms of availability of what I'd call quality food); I'm only familiar with one corner store in DC that offers fresh produce for instance, and that one's within a mile of Eastern Market. That said, my sample is totally nonrandom and non-scientific. But I'm also not entirely convinced that a chain supermarket would displace the corner stores - they might serve different niches. There certainly seems to be growing evidence RE: the benefits of the big supermarkets in low income neighborhoods vis-a-vis economic development, in any case.

I'd also probably nitpick the definitions they're using; census tracts are likely the wrong unit of analysis, as I imagine its hard to make an apples to apples comparison between a huge swathe of southwestern Idaho and the Oxon Hill neighborhood in DC, for example.

usfoodpolicy said...

I didn't go into this store, because of being on a rental bike, but I have in the past visited the Anacostia Warehouse Supermarket (also pictured), which is the opposite of fancy, but which is a real and legitimate source of food for an urban neighborhood.

Stephanie said...

I wonder that sometimes decision-makers think that just because it doesn't look like they think it should look, they think it is less valid or less efficient or less good. Thanks for the post.

BTW, my friend is the guy in charge of developing the DC Bikeshare program. I passed along your blog to him today.

StephanieB said...

I agree that the unit of analysis is possibly not the best option, but, honestly, unless you must consider the transportation system to get a valid picture. Unfortunately, I can't imagine a simple way of doing that and since I've never seen it done, there may not be one!
How people reach stores, and how long it takes may matters more than a "by the crow" measurement.