Have we been measuring poverty wrong this whole time?

Nov 21, 2012

The U.S. poverty rate has hardly budged in half a century. The Census says the same share of our country is living in poverty right now as in the 1960’s.

So there’s lots of traction for accusations that programs like food stamps and Medicaid cost too much and don’t work. That criticism is not new, President Regan famously said in his 1988 State of the Union address, “My friends, some years ago the federal government declared war on poverty, and poverty won.”

Why are all the programs not making a dent? Ask James Sullivan,  and he'll say they actually are.

Sullivan and his co-author Bruce Meyer say since we started measuring poverty we’ve been doing it wrong. Sullivan is an economist at the University of Notre Dame. He thinks a simple change in how poverty is measured would have huge implications. Sullivan thinks people should just be asked how they spent their money instead of how much money they earned. It's called a consumption measure. As Sullivan explains,

"The official poverty between 1970 and today has risen by two and a half percentage points. But if you look at consumption based poverty over those same four decades you see that poverty has fallen by 12 percentage points, which is a very different story.”

A 12 percentage point drop in the poverty rate would be a big deal. So if Sullivan’s measure is right, and that is kind of a big if, how could our official measure be so far off?

The official measure is based on people's pre-tax, cash income. To get that number, the Census has to just ask people how much money they made last year. They can’t get it from tax returns, they don’t have access to government program records, so they just have to ask.

Lots of economists say that’s the problem. People just don’t accurately remember how much they made. Many people living in poverty often scrape together hourly work, part time jobs, or work from odd jobs.

There’s evidence those people do a better job remembering how much they spent, and that’s why a consumption measure might improve accuracy. Sullivan thinks his consumption measure would also give better information about government programs because it measures them too.

As Sullivan says, "the official poverty measure it misses a lot of features of our anti-poverty programs. It doesn’t capture food stamps for example which is not cash income, it doesn’t include housing subsidies, it doesn’t include the Earned Income Tax Credit."

Those programs can move somebody out of poverty. Amber Pederson is a high school Spanish teacher and the single mother of two daughters living in Monroe. Until a few years ago she was living below the poverty line of $19,090 for her family of three. Government programs helped her pay for medical care, housing, and food for herself and her daughters. She could then spend some of her income paying her student loans, and getting a used car so she could get outside her rural area to a good job.

Now, she’s a success story. She still has to stretch a dollar, but she lives above the official poverty line.

Even the Census thinks their official measure downplays the importance of government programs. It has a new way to measures poverty that is supposed to take these programs into account called the supplemental poverty measure. But it still uses the official number to measure the change in poverty over time.

So in the upcoming fight over the fiscal cliff we’re sure to hear debate about the value of anti-poverty programs. We’ll probably hear little debate about how we’re measuring these programs, and if we’re doing it wrong.

*This story was informed by the Public Insight Network. Share your story here.