Hi All,

A few observations on Amalie's questions and some thoughts to add to Jill's response (pasted below).

As food desert definitions were a topic of conversation at our last sub-committee meeting- here is the definition that was advanced in the 2008 Farm Bill:

…an area "with limited access to affordable and nutritious food, particularly such an area composed of predominantly lower-income neighborhoods and communities."

Are these food deserts defined differently in Urban vs. Rural areas?  Jill – did you have a specific definition of Rural Food Deserts that you are using you could share? How would assessment of such areas differ? At the August 10^th meeting it was decided to try and identify a model that would be advanced for urban food desert assessment for local food advocates around the state.  Please share your thoughts on these questions and any models you feel are exemplary (those that use readily accessible data).

All of the key words that take a lot more work to define, and which we spent a lot of time thrashing around, are present:  access, affordable, and nutritious.   For access, as Jill notes there is distance and availability of various public and private transportation methods (walk, bike, bus, car, etc. which require network analysis and sometimes beyond road networks) and cultural preference that can be important (i.e. distance to halal or kosher meat not just to meat).  Affordability becomes an issue of both household income, cost of living and food price.  Nutritious brings up nutritional quality, safety, quantity, price, balance, cultural preference, and probably more.  

Jill provided some excellent references.  I've attached another by Samina Raja at University of Buffalo, who comes at it from an urban planner's perspective.  She uses pretty readily available data, addresses a lot of the access and affordabilty issues, and suggests a means of incorporating nutriion, and proposes an index that could be used for comparative purposes within and between cities. 

We have a new GIS application developer on board with AMP who would be able to put mapping applications into the hands of people in Ohio communities, including an opportunity for them to add data and groundtruthing.  That may be a way to get citizen involvement in collecting more of the primary data in urban areas, perhaps coordinated by local councils.

What are your suggestions on the four target commodities to use in a regional impact analysis for local food systems?

 I think the grains and oilseed crops are at 100% because we grow more than enough corn and soybeans to satisfy our thirst for corn and soybean oil.  I'd want to focus on a staple grain (wheat or spelt, for flour) rather than oilseeds.

The others are good choices, although vegetables will need to be narrowed down and fish (aquaculture I assume) may need to be as well.  As part of the NEO inventory we are starting to work with production capacity for specific vegetable crops starting with lettuce, cole crops, potatoes and asparagus (no real method to this selection of initial foci other than that we know all of them are currently produced and we have specs for soil conditions etc.). 

Jill Clark's response from email:

Here are some initial thoughts….

I say that two of the target commodities should be ones that we already grow in Ohio and could potentially meet demand if we sold in Ohio too.  I got the following from this policy brief:  http://cffpi.osu.edu/docs/estimate_capacity.pdf

--Grains, oilseeds, dry beans, and dry peas – We could meet 100% of Ohio demand by current production.

--Poultry and eggs – We could meet 72% of Ohio demand by current production.

Then, pick two industries that have the potential to grow.

--Fish – could be doing a lot more on the fish front – I know the director is keen on this.

--Vegetables – could be growing a lot more veggies

The standard definition used by the farm bill is our general starting point.  The differences between urban and rural food deserts, in my mind, are born out when operationalizing what is meant by geographic (versus economic or informational) access (for example, distance and transportation type). Further, if you are attempting to conduct a macro analysis (all rural areas of Ohio or all urban areas of Ohio) versus a micro-analysis (one or a couple case studies) that will dramatically change what kind of analysis you would be reasonably capable of doing.  I made some notes below regarding our methodology and I would be happy to discuss at our next meeting.

We are examining the following four aspects of access:

o   Ability to get to the food, geographic access (travel to the store, access to buses, and car ownership).

o   Ability to afford food, economic access (median income of household).

o   Ability to choose among various stores, economic access (competition)

o   Ability to choose among fresh food varieties, healthy access (size of food outlet)

Methodology:

If you haven’t seen the following resource, it is very useful.  This appendix “Measures of Access Used in Food Desert and Related Studies” reviews access methodology: http://www.ers.usda.gov/Publications/AP/AP036/AP036k.pdf

1.  Most of the rural food desert work I have read is merely just developing buffers (as the crow flies) around the center of the zip code where the store is located.

--We have the advantage of having the actual location.  Which provides a more accurate assessment. 

--With the location, we can do an analysis of competition, which generally results in greater affordability.

--Because of our ability to do network analysis, we are able to do driving time and not just a distance buffer around the store.  We can also do walking distance.

2.  Some interesting urban food desert methodology:

Gallagher (2006) estimated a “Food Balance Score” which is done by dividing distance of every city block to the nearest fresh food outlet over distance to the nearest fast food restaurant. Interesting concept because analyzing the distance to the closest store relative to the closest fast food restaurant would give us an idea of relative food accessibility rather than measuring absolute food accessibility.

-- Gallagher, M. (2006) Examining the Impact of Food Deserts on Public Health in Chicago. Chicago, Mari Gallagher Research & Consulting Group

Rose et al (2009) used a 1 km radius to estimate food desert in New Orleans. They identified the number of people under the poverty line who live in and out of the supermarket service areas, and identified whether they have a car or not. Since their study was looking at a small geographical area, they also included shelve space from in-store surveys.

-- Rose, J., Bodor, N., Swalm, C., Rice, J., Farley, T., Hutchinson, P. (2009). Deserts in New Orleans? Illustrations of Urban Food Access and Implication for Policy. Manuscript prepared for University of Michigan National Poverty Center/USDA Economic Research Service Research "Understanding the Economic Concepts and Characteristics of Food Access"

Apparicio et al (2007) used three measures of accessibility to food outlets to identify food desert in Montreal, Canada. They used proximity (distance to the nearest supermarket), diversity (number of supermarkets within a distance of less than 1000 meters) and variety in terms of food and prices (average distance to the three closest different chain-name supermarkets). They later concluded that Food desert is not an issue in Montreal.

--Apparicio, P., Cloutier, M., & Shearmur, R. (2007). The case of Montreal’s missing food deserts: evaluation of accessibility to food supermarkets. International Journal of Health Geographics, 6(1), 4.