A USDA report this week suggested that U.S. families spent less on food as their housing costs increased from 2000 to 2007, leading to increased risk of food insecurity.
The report by USDA's Economic Research Service said that inflation-adjusted median food spending fell by 6 percent from 2000 to 2007. Meanwhile, ERS said, the national prevalence of very low food security (the condition formerly called "household food insecurity with hunger") increased from 3.1 percent of households in 2000 to 4.1 percent of households in 2007.
In other words, people were spending less in real inflation-adjusted terms, and they were experiencing higher rates of food-related hardship.
When I first read this study, I wondered if ERS was correct to imply a connection between these two trends. The report's author, Mark Nord, strengthens the argument by disaggregating the data according to income strata. It turns out that the second-poorest fifth of households experienced both the greatest fall in food spending and the greatest increase in food-related hardship.
Moreover, the food spending and food insecurity trends were connected in a plausible way to overall spending trends: "The declines in food spending by middle- and low-income households were accompanied by increases in spending for housing and, in the two lowest income quintiles, by declines in income and total spending."
Still, one should resist the temptation to think of food insecurity as primarily a problem of low average food spending. The food security survey that is used for classification asks households about their experience of particular hardship events, such as skipping meals or going a whole day without food, in the preceding 12 months. If a family runs out of food occasionally, it can seem to have adequate average food spending, but still experience food insecurity. Participation in the SNAP (food stamp) program seems to be associated with higher food spending, holding other factors constant, but associated also with higher rates of food insecurity (presumably because people with greater needs for food are more likely to take the trouble to participate). It is not clear whether increased average food spending is itself a cure for food insecurity.