At the same time Congress allowed a recession-era boost of $5 billion to the federal food stamps program to expire in early November, the U.S. Census Bureau released new data that suggests more than 5 million people stayed out of poverty in 2012 because of the program. But that's not stopping Congress from considering deeper cuts to food stamps in the next farm bill, ranging from $4 billion in a Senate bill to $40 billion in a House version.

That's wrongheaded, says Timothy Smeeding, director of the Institute for Research on Poverty at the University of Wisconsin-Madison. "No other program for the nonelderly does such a great job preventing poverty or alleviating poverty's weight on those who remain poor," he says. Smeeding and many other antipoverty researchers decry the recent interest by Congress in cutting the food stamps program -- officially known as the Supplemental Nutrition Assistance Program (SNAP). They say the cuts would hurt the poor while doing little to address concerns about the federal debt.

But conservatives argue that food stamps warrant a major overhaul, in part because they may create a dependency on government programs and reduce people's incentive to work. "If anything, the compromise SNAP reductions are far too modest," wrote Michael Tanner, a senior fellow at the Cato Institute, a libertarian think tank, in a blog post earlier this month. Tanner has called for a variety of reforms to SNAP, such as strengthening work requirements and eliminating broad-based categorical eligibility -- where people qualify for food stamps because they are eligible for another federal public assistance program.

This appears in our free Human Services e-newsletter. Not already a subscriber?

"This is a really good program and it's getting hammered for all the wrong reasons," Smeeding says. Food stamps reduced poverty for 8 million people in 2011, according to a report Smeeding co-authored on food stamp benefits in September. His estimate is higher than the 2012 Census figures, largely because it corrects for under-reporting in Census surveys both in terms of the number of people who self-identify as being on food stamps and the monetary value of the benefits they receive.

Some 46.8 million people were in poverty, on average, between 2010 and 2012, according to official estimates by the Census. That's about 15.1 percent of all Americans. The official poverty definition varies by household size and the age of individuals, but for a single person under 65 the poverty threshold last year was $11,945 in annual pre-tax income. The historical method of measuring poverty in the United States, however, fails to capture the true number of people who can't afford a basic standard of living as defined by the federal government. Social scientists have long criticized the official poverty measure for failing to incorporate the full breadth of non-cash benefits (such as food stamps and refundable tax credits) and costs (such as payroll taxes and out-of-pocket medical expenses) that impact Americans' personal wealth.

The November Census report corrects for that by examining what the poverty rate would look like if the official measure was more comprehensive. Under the improved measure, the Census found that 16 percent of Americans, meaning an additional 2.6 million people, would be counted as poor. If people did not have the value of food stamp benefits, however, another 5.4 million would have been counted as poor, ratcheting up the national poverty rate to 17.6 percent, the report said.

But the Census estimate for calculating food stamps' impact on poverty is a blunt instrument in some ways, says Sarah Bohn, a California researcher who this year calculated an alternative measure of poverty for the state. She says a more sophisticated study would need to model a hypothetical world in which food stamps no longer existed. Would the absence of those benefits compel poor people to find jobs they weren't willing to take before? Would food stamp recipients shift to other public assistance programs? Under both scenarios, it's conceivable that the poverty rate wouldn't be as affected as the Census data seem to indicate. "It is a simple calculation," concedes Kathleen Short, an economist with the Census who wrote the most recent poverty report. If food stamps no longer existed, she says, "people would have behaved differently. We're not capturing that."

In the absence of food stamps, however, even a more sophisticated model could show poverty getting worse. "What would people do if they didn't have SNAP? They would probably allocate other money to food," Smeeding says. "But in general they would end up with less, and they would have to spend other money on food, so something else would lose." But Dustin Cable, a poverty researcher at the Weldon Cooper Center for Public Service at the University of Virginia, points out that another possible scenario is that the "the lowering of taxes and economic growth that would result could outweigh the deleterious effects of losing the food subsidy."

The advent of a revised poverty measure and the information it provides about the relationship between public assistance programs and poverty is relatively new. This is the third year for which the Census has published estimates using the supplemental poverty measure and only the second time it has published state-by-state figures. Under the revised measure, some states saw their poverty rates increase. For instance, California had the highest poverty rate at 23.8 percent, even though its official rate was 16.5 percent in 2012. Several of the nation's poorest states actually look better under the alternative measure, such as Mississippi, which had an official rate of 20.7 percent, but a supplemental rate of 16.1 percent. The Census does not publish state-level estimates on what impact SNAP and other public programs have on state poverty rates because the sample sizes would be too small for statistical validity, says Short.

The following table shows differences in state poverty rates measured using official counts and the supplemental poverty measure (SPM). Figures shown represent average annual totals for 2010-2012, shown in thousands.

Source: U.S. Census Bureau, Current Population Survey, 2011-13 Annual Social and Economic Supplements.