11.1% of U.S. households were food insecure in 2007

The U.S. Department of Agriculture this week estimated that 11.1% of U.S. households were food insecure at some point in 2007. In a 30-day period, 6.3% of households were food insecure.

USDA estimated that 4.1% of U.S. households experienced "very low food security" at some point in 2007. USDA used to call such housheolds "food insecure with hunger." A simpler single survey question showed that 3.3% of respondents reported being "hungry" at some point in the previous 12 months, because of not being able to afford food.

The estimates were based on a set of questions about symptoms of food hardship on the Current Population Survey in December, 2007. Anti-hunger groups noted that the new USDA estimates do not capture likely recent increases in poverty and hunger due to the financial crisis.

To put the statistics in context, the United States will fail to achieve national targets for reductions in food insecurity.

Household food insecurity, 1995-2007

Data source: USDA. Graphic: Parke Wilde.

As part of the department's criteria for evaluating food assistance programs, USDA strategic plans set targets for 2005 and 2007 in the rate of "very low food security" among households with incomes below 130% of poverty. In a 2004 paper (.pdf) for the Center for National Statistics at the National Academies, I discussed some ways these evaluation targets might or might not work as intended. After establishing these targets, USDA has not followed up and reported progress toward meeting them, and is considering changes to its evaluation approach in light of the CNSTAT report on food security measurement.

One complication is that the prevalence of "very low food security" is much worse for households that participate in the Food Stamp Program (now called the Supplemental Nutrition Assistance Program or SNAP), than among low-income nonparticipants, presumably because people who face greater hardship are more likely to participate. The detailed tables of the USDA report this week show, furthermore, that the prevalence of "very low food security" among participants actually appears to be increasing over time.

At the very least, this trend makes it difficult to use such data to demonstrate the beneficial effect of SNAP in reducing food insecurity and hunger. I have been using every opportunity to encourage USDA to pilot and evaluate possible program changes, such as twice-monthly benefit delivery, which might increase the effect of SNAP on measured rates of household food insecurity by reducing the episodes of hardship toward the end of the month.

Data source: USDA. Graphic: Parke Wilde.