Washington Post columnist Shankar Vedantam starts his column today with:

About 35 million Americans regularly go hungry each year, according to federal statistics.

But it is simply not true that 35 million are “regularly” going hungry, as I’ve previously noted.

The U.S. Department of Agriculture is the official source for such statistics. Here is what the agency says:

USDA does not have a measure of hunger or the number of hungry people. Prior to 2006, USDA described households with very low food security as ‘food insecure with hunger,’ and characterized them as households in which one or more people were hungry at times during the year because they could not afford enough food … In 2006, USDA introduced the new description ‘very low food security’ to replace ‘food insecurity with hunger.’

O.K., how big is the new group called “very low food security?” If you look at the chart here, you see that it is about 3% of the population (about 9 million people). These are the people with an episode of “very low food security” during even a short period of the year. For those with very low food security in the prior 30 days, the rate falls to about 2%, or about 6 million people.

Shankar Vedantam throws the phrase “each year” into his sentence, which blurs his statistic a bit. But if people go hungry only once a year, how could that be “regularly”? In any event, he has made the same error as many politicians and advocacy groups who use the grossly inflated 35 million figure.

I enjoy Vedantam’s column, which often highlights interesting data and research. But in this case, he should have recognized that more than 10% of Americans “regularly” going hungry didn’t pass the smell test.