Whew. Another semester is just about over (only 20 more papers to read!). But I am back!
The Economic Research Service plans to release a report (later today, I think) on how to measure prices of healthy food, and I will have more on that in the coming days. But coincidentally, the April issue of Agricultural and Resource Economics Review arrived in my mailbox yesterday, containing a paper on a related topic: “Does Healthy Food Cost More in Poor Neighborhoods? The authors (Hatzenbuehler, Gillespie, and O’Neil) analyze the cost of two baskets of food: one based on the Dietary Guidelines and the other on USDA’s Thrifty Food Plan, which forms the basis of the food stamp program. The study area is Baton Rouge, LA.
This work contains several novel elements. One, the market basket based on the dietary guidelines incorporates foods that are culturally appropriate for the region. Two, spatial competition among food stores is explicitly considered. Lastly, different store formats are included in the study (chain supermarket, supercenter, and independent supermarket).
The authors pose two related research questions: (1) do low income households face high prices because low-cost food stores do not invest in their neighborhoods and (2) is there less retail competition in low income neighborhoods. Food cost and store location data were collected from 60 stores over three weeks, in January 2009.
The average market price for the thrifty food plan basket was $259.61 (standard deviation was not reported), while the average price for the dietary guideline basket was $270.11.
A two stage econometric model was estimated, where in the first stage, spatial competition is considered to be function of demand factors ( population density, income, household size, race). Spatial competition is approximated by a gravity index that measures how many competitors each store faces, and the proximity of the competition; thus, a higher competition index results when stores are close, and when there are more competitors. In the second stage, the prices consumers face depend on results from the first stage re: competition, the store type (chain, independent, supercenter), store size and the services each store offers.
The model, as estimated, suggests that supermarkets located in low income areas do not charge more for food. However, the main factor determining food costs was whether the store was a supercenter or a chain (others have found this result as well; see Leibtag, 2005).
Perhaps the most interesting finding is that competition among food stores is related to neighborhood income: the higher the income in a census tract, the more competition a store faces (consequently the lower the price to consumers). This finding brings us full circle to the food desert literature, which finds that low income households (when compared to higher income consumers) tend to have low access to supermarkets.
The research is another piece of evidence that supports policies that increase the likelihood of a store locating in a low-income neighborhood. Of course, store location is based on a profit maximization decision, so the policy needs to tip the scale of profits in favor of the low-income neighborhood (topic for another day!) Policymakers are thinking about this, and the Healthy Food Financing Initiative is a policy that is designed to encourage stores to locate in underserved communities.