Just 2.2% of U.S. households both lack a car and live more than a mile from a supermarket, USDA estimated in a much anticipated report to Congress released today.
An estimated 5.4% of households lack a car and live more than half a mile from a supermarket.
Household surveys give similar estimates. About 5.7% of respondent households in the nationally representative Current Population Survey reported not always having enough food or the right kind of food, in connection with lack of food retail access.
Low-income people are more likely than higher-income people to have close walking distance access to a supermarket. According to the Census Bureau data used by Michele Ver Ploeg and her colleagues at USDA's Economic Research Service, 29% of low-income people live within half a mile from a supermarket, while 22% of higher-income people live this close to a supermarket.
The debate over supermarket deserts has been much on my mind in recent months (see also many posts under the food retail tag). The USDA results will be seen as out of step with the advocacy literature on supermarket deserts. I hope advocates for low-income people read the report with an open mind. The ERS report readily acknowledges the clear fact that, in particular locations, such as some devastated urban cores and some poor rural areas, lack of retail access is significant. Yet, for a variety of reasons, I have been worrying that the way this issue had been framed as a national concern has been muddied by incorrect assumptions and generalizations. The data-heavy USDA report offers an important opportunity for clarifying the food retail agenda.
To understand how the estimates in the USDA report could differ from one's prior expectations, it helps to look closely at the maps. Take East Saint Louis (below) or Washington (see report), for example. One can find very poor neighborhoods without supermarket access, but they are not where most people -- even most poor people -- live. Most densely populated neighborhoods, including densely populated poor neighborhoods, can boast a supermarket.
The important policy issue at stake here is a choice over the proper remedy for supermarket deserts. I think that government incentives to attract supermarkets may be good policy in rare and limited retail situations. Efforts to improve fresh fruit and vegetable marketing make sense in a wider variety of low-income neighborhoods. In many cases, the best policy is to think of food retail improvement as one component of a longer-term neighborhood anti-poverty agenda, rather than something that can be typically achieved with tax breaks to supermarkets.
I imagine this report may generate some discussion, so I'll save more thoughts for the comments.
Update: Ezra Klein discusses this report.