David R. Henderson  

How Food Stamps Increase Measured Poverty

Today on the Eurozone Crisis... Technology and Higher Educatio...

The leading story in the Monterey County Herald this morning is titled, "Nearly half in U.S. near poverty or low-income." It is Hope Yen's Associated Press item, and the leading two paragraphs are:

Squeezed by rising living costs, a record number of Americans -- nearly 1 in 2 -- have fallen into poverty or are scraping by on earnings that classify them as low income.
The latest census data depict a middle class that's shrinking as unemployment stays high and the government's safety net frays. The new numbers follow years of stagnating wages for the middle class that have hurt millions of workers and families.

That is bad. But then she quotes University of Michigan professor Sheldon Danziger "who specializes in poverty." She's right; he does. So you would expect a specialist in poverty to know how the U.S. Census Bureau, the data source for this story, measures poverty, right?

Before quoting him, let me quote from the Census Bureau's study on poverty in 2010, "Income, Poverty, and Health Insurance in the United States, 2010," September 2011:

[M]oney income does not reflect the fact that some families receive noncash benefits, such as food stamps, health benefits, subsidized housing, and goods produced and consumed on the farm.

So food stamps, Medicaid, Medicare, and subsidized housing, a substantial part of the welfare state, don't count in measuring people's income. And those items, especially food stamps, are a particularly large part of the poor people's income.

That makes Danziger's quote particularly striking. [Of course, I'm assuming that she quoted him correctly. If she didn't, then my apologies to Professor Danziger.] Here are the next two paragraphs in Ms. Yen's piece:

"Safety net programs such as food stamps and tax credits kept poverty from rising even higher in 2010, but for many low-income families with work-related and medical expenses, they are considered too 'rich' to qualify," said Sheldon Danziger, a University of Michigan public policy professor who specializes in poverty.
"The reality is that prospects for the poor and the near poor are dismal," he said. "If Congress and the states make further cuts, we can expect the number of poor and low-income families to rise for the next several years."

Contrary to Professor Danziger, food stamp programs did not keep the U.S. Census Bureau's measure of poverty from rising even higher. Indeed, even if the effect of food stamp programs on the willingness to earn income is small, any incentive effect at all means that food stamp programs made measured poverty higher. And if the feared cuts that Professor Danziger is referring to are cuts in food stamps [I don't know if that's what he had in mind], those cuts will not cause the number of poor and low-income families to rise.

Comments and Sharing

COMMENTS (16 to date)
Lord writes:

You are assuming the author of the article discussing census measures and Danziger discussing poverty are talking about the same thing. The latter may be referring to more general measures such as the proposed revised poverty measures that incorporate the value of those benefits and show even higher poverty levels.

Tim Worstall writes:

And the EITC! America's largest anti-poverty program is not included in the measurement of who is in poverty.

David R. Henderson writes:

You’re right: I am assuming that.
@Tim Worstall,
Wow! Are you sure?

Thomas Sewell writes:

"An American having the average income of the bottom US decile is better-off than 2/3 of world population." (Milanovic 2002, p. 50)

We're talking here about the world's upper class, right? Comparing them to the world's super-rich?

Danziger sounds like a specialist in getting tax money for pet projects...

I would REALLY like to see a measurement of poverty in the U.S. using absolute wealth levels (not income, which doesn't account for the retired) necessary to live. Those results wouldn't fit the narrative, though...

Vipul Naik writes:

May be Danziger thinks that food stamps increase people's earning capacity by preventing them from starving and improving their health? A capability effect rather than an incentive effect.

Chris T writes:

The United States median income for a family of four is about 49,000 in 2011 and according to the author, 45,000 is low income. If income follows anything like a bell curve, then nearly 1 in 2 should fall below 45,000.

The Federal poverty level is set at 22,350 for a family of four for 2011:

Lauren writes:

For those interested in the way poverty has been measured, there is a recent podcast on EconTalk on the subject:
Bruce Meyer on the Middle Class, Poverty, and Inequality.

Patrick R. Sullivan writes:

I think Chris T is correct. If you read the AP article, it's hardly rigorous. It comes very close to saying the obvious; half the people are below the MEDIAN, and we're calling them poor or low income.

steve writes:

What is missing here is how much food stamps are worth. For a family of three making about $10,000/year, it would add about $1800/year.


Thomas Sewell writes:


I was curious about your numbers, because they seemed low to me. So I went and filled out my states benefits estimator with my zip code.

This isn't a "cadillac" state, nor a high cost of living state, and it calculated for an income of $1000/year and a family of 4, snap benefits would be $952/month, or $11,424/year. That's a bit different than your $1800/year. Sure you don't have something wrong in your calculations?

Thomas Sewell writes:

Correction, meant to post an income of $10,000 was used in the calculations.

Zach Pruckowski writes:

Poverty metrics were overhauled in November. Your linked document and the EconTalk podcast are from September and October, before the new metric was announced. My brief research with The Googles mentioned that the overhauled number accounts for payroll taxes and food stamps. I assume that metric is what Professor Danziger is referring to, since the new metric is the one used by the AP article his quote appears in.

Patrick writes:

Regardless of the cutoff used, how easy do folks in here think it would be to support a family of 4 on $49K per year?

HOI CHI IAN writes:

No matter will the food stamp increase the measured poverty, I think it is important to keep it. As the economy is not good today, many people lose their job and unemployment rate is still high, people may depend on the food stamp to maintain their lives. It is a kind of subsidies to help people to overcome this hard time. Even though the States is having national debt and trade deficit, keeping citizens' living standard is also crucial. Besides, instead of cutting social benefits, government can try reduce other costs. For example, lower the budget of buying weapons and foreign military spending. With healthier bodies and government support, we will have better incentive to work harder.

MG writes:

Poverty thresholds should be based on nominal (not relative) levels of consumption (adjusted for inflation and real hedonic effects). Moreover consumption should try to impute a real ability to consume in an attempt to account for unused, non-cash entitlements such as when people voluntarily choose not to sign up for medicaid, welfare, etc. It should include all consumption, including that of private goods (medical service, education, etc.) where a significant amount of means-tested subsidizing exists; and all consumption of those "public goods" whose provision is specifically aimed at providing access to goods and services that others pay for. Then, lets talk.

Sendvich writes:

Food stamps is a good idea. It will help them, but cannot solve all this problem.

Comments for this entry have been closed
Return to top