WASHINGTON – The actual number of Marylanders who received food stamps in 2001 was almost twice what the Census Bureau reported, an undercount that could have far-reaching implications for the poor, the nonprofits who defend them and the state that helps support them.
A report co-authored by analysts from the U.S. Department of Agriculture, Maryland Department of Human Resources and even the Census Bureau itself, found that 157,857 recipients got food stamps in 2001 — almost double the 87,429 figure given by the federal government.
“That’s not wrong by a little, that’s wrong by a lot,” said Richard Larson, study co-author and Maryland DHR policy and research director. “If it’s like this for food stamps, what’s it like for any piece of Census data you read?”
Larson stressed the report was not created to criticize the bureau or get more funding, but merely detailed what the agency acknowledges in its own surveys — room for error.
“Surveys are based on people’s responses to questions, not necessarily the truth,” said Charles Nelson, an assistant division chief at the Census Bureau and not a report author. “There’s nothing you can do if somebody doesn’t want to tell you something.”
The questions’ wording also could have skewed the data, Nelson said. The traditional “Did you get food stamps?” may no longer work because most people now use food cards, not stamps.
The Census Bureau plans to test a modernized question early next year.
“It’s always your aim to get as close as possible to matching the real world,” Nelson said. “In this case (the analysts) gave us some hints as to why the numbers didn’t match. It’s valuable research — the kind of thing we should do more often.”
The report’s six authors — two of whom still work for the Census Bureau — compared the national survey with Maryland records and found that federal underreporting contributed most to the sharp contrast in numbers.
About 68 percent of the time, the federal agency misreported that no one in the survey household received food stamps when, in fact, Maryland issued benefits to someone in that household.
Such underreporting occurred most often for recipients who had acquired the stamps in only one month during the survey period.
Other factors were discrepancies about what actually constitutes a household, an inability to account for households that moved in and out of Maryland and differences between the total numbers covered by the national and state records.
“This has enormous implications for how we talk about data,” Larson said. “It should cause (people) to pause and at least ask the questions, ‘How valid is this number?’ ‘How was it derived?’ . . . When you talk to the Census or read the fine print, they’ll tell you it’s always an estimate, but all of that gets lost in a haze of academic and bureaucratic gobbledygook.”
Such estimates could have consequences.
The U.S. Agriculture Department provides bonuses to states that demonstrate the best or most improved access to their food stamp programs, as measured by the number of people receiving food stamps and the amount in poverty.
The department awarded $18 million on Sept. 29 to 12 states, including Virginia, West Virginia and the District.
Maryland has never received a bonus, and the report’s authors said data discrepancies could lead to an uneven playing field among states when competing for awards.
“It’s important to have our policies grounded in fact,” Larson said. “Census data is used for everything — sometimes to distribute money. It may be a warped ruler people are using.”
Others were a little more blunt.
“We know there are these poor people, and I think this is another example of how the Census tends to undercount them,” said Valerie Coffin with Maryland ACORN, a community organization of low-and moderate-income families. “Therefore it’s difficult to prove the need (for more funding).”