I’m no fan of the American Enterprise Institute, the D.C. policy shop that harbors Michael “Every ten years or so, the United States needs to pick up some small crappy little country and throw it against the wall, just to show the world we mean business” Ledeen.  But as far as I know, AEI’s Nicholas Eberstadt is trying to understand reality, and his latest article aims a powerful destructo beam at the way the U.S. has defined poverty since the mid-60s.

The poverty line, which is now built into all kinds of government helping programs, was drawn up by a woman, Mollie Orshansky, who had first-hand experience with hunger. The line is essentially three times a minimum food budget, a level below which “everyday living implied choosing between an adequate diet of the most economical sort and some other necessity.” Though adjusted for inflation, it’s meant to measure absolute poverty (eating vs. not eating), not relative poverty (eating hamburger vs. foie gras). A War on Poverty wouldn’t make much sense otherwise.

The percentage of Americans in poverty, as officially defined, dropped sharply during the 60s, but since 1973 it’s either leveled off or risen a bit. “We have had a generation with basically no progress against poverty,” the University of Michigan’s Sheldon Danziger told the New York Times in 2004. “The economic growth is not trickling down to the poor.”

Well, that’s what the official definition would lead you to believe. About one American in eight (12.7 percent) lived below the official poverty line in 2004, compared to one in nine (11.1 percent) in 1973. So all those people must be just as bad off in real terms as they were three decades before, right?

Well, embarrassingly enough, no. They’re better off. Eberhardt writes:

  • “According to the National Pediatric Surveillance System of the CDC, the percentage of low-income children under five years of age who were categorized as underweight (in terms of BMI for age) dropped from 8 in 1973 to 5 in 2003.”

  • “In 1970, almost 27 percent of poverty-level households were officially considered overcrowded [i.e. more than one person per room]. By 2001, according to the AHS [American Housing Survey], just 6 percent of poor households were ‘overcrowded’–a lower proportion than for nonpoor households as recently as 1970.” Since 1980 the average heated square footage per household member among the poor has risen from 371 to 472.

  • In 1972, about 40 percent of households in the bottom fifth of income had a car; in 2003, more than 60 percent did. “For all forms of motor transport, U.S. poverty households’ ownership levels in 2003 matched overall U.S. families’ auto ownership levels from the early 1960s.”

  • In the early 1970s, 60 percent of poverty-level adults had untreated cavities; by the late 1990s, the rate was 40 percent.


If the poor were really just as poor in 2004 as in 1973, none of this would be happening–and it’s easy to imagine an alternative present in which it hadn’t been. The poverty line isn’t measuring what we think it is.

As Eberstadt firmly states, none of this implies that everything is fine, or that relative deprivation doesn’t matter. It does imply that our yardstick for measuring absolute poverty is badly broken.

He suggests several reasons why it isn’t doing the job Mollie Orshansky had in mind.  The most interesting is that people’s incomes are a lot more variable than they used to be, especially at the low end. A Census Bureau survey found that “at some point during the four years 1996-1999, fully 34 percent of the [non-institutionalized] population spent two months or more below the poverty line. On the other hand, just two percent of the population spent all 48 months of 1996-99 below the poverty line.”  (That rate was over five percent for blacks and Hispanics, something Eberstadt probably wouldn’t mention if he were acting as a shill for conservative policies.) Since almost all the blog posts I can find on this subject are conservatives jumping to conclusions and yukking it up, note that this bit of evidence might well argue against lifetime limits on welfare assistance.

Anybody know of any evidence that the poverty line really is measuring what it’s supposed to be measuring?