Making a Poor Measure Better
Ongoing efforts to improve measures of poverty in the United States must thread the needle between pragmatism and politics.
Academics for decades have been pushing to rewrite the way the federal government measures poverty, a movement repeatedly stymied by both political reality and the sheer difficulty of the task. A new poverty measure, for example, would require not only setting an updated threshold, but reconsidering every way in which a family might fall under it.
Should the measure account for out-of-pocket medical expenses? The difference between families who pay rent, own homes or have no housing expense at all? What about the income derived from food stamps or the money saved by the earned income tax credit? If we incentivize parents to work, shouldn’t we consider the cost of paying day care? And what about the fact that it costs twice as much to live in New York City as it does in Tulsa, Okla.?
The current measure, which has been in existence since the 1960s, takes none of these factors into account.
The Census Bureau, which produces statistics on poverty in America, announced this month that it finally plans to roll out a new measure that will consider such nuances, although the plan comes with one caveat that has disarmed most controversy: The existing poverty measure will remain the official standard, while the new one — “a work in progress” — will be merely “supplemental.”
“A bunch of us have been working on this for a long time, and we’ve decided the perfect should not be the enemy of the good, and we should go ahead,” said Timothy Smeeding, director of the Institute for Research on Poverty at the University of Wisconsin.
The promise to continue studying and refining the new measure, which will first be introduced with the Census Bureau’s 2010 poverty report in the fall of 2011, should please academics. And the promise that nothing will change about the federal distribution of dollars tied to the existing, “official” poverty rate should keep politicians calm, for now.
The biggest benefit of the new measure, then, won’t necessarily be felt directly by low-income families, but by the policymakers who’ve struggled to measure the impact of programs designed to help them. The existing measure, by excluding the impacts of non-cash benefits, by definition makes it impossible to tell if such programs are helping to lift people out of poverty.
“Food stamps and the earned income tax credit are the two biggest anti-poverty programs that we have,” Smeeding said. “We’ll spend $50-$65 billion this year on both of them, and we know nothing about how they’re helping people. Nada, nothing, zip. It’s not there.”
To overcome this, the new measure will offer changes on two fronts: The income a family receives will be recalculated to include non-cash benefits like food stamps, and the annual expenses a family should realistically be expected to pay will be updated to include needs like child care and health costs. The number of elderly in poverty, for example, might be expected to increase due to their disproportionate health care costs. But some of this change will be offset by the fact that many elderly own their own homes outright and have no housing expenses.
The new measure will largely be based on recommendations from a 1995 report of the National Academy of Sciences, which also proposed that an improved measure should take into account the varied cost of living in different parts of the country. (The federal government already does this on a monthly basis in establishing federal per diem rates.)
The Census Bureau’s announcement has drawn little in the way of partisan criticism (somewhat remarkable given Washington’s climate today). Should the new measure paint a picture of considerably larger numbers of Americans in poverty (and in need of social services), some conservatives may object. But the more significant battle will be along geographic lines, precisely because of the new cost-of-living adjustments. The number of people in poverty in Oklahoma will likely go down relative to the figure in New York, and if these statistics were being used officially, that would mean less federal money for Oklahoma.
For this reason, Smeeding is skeptical that the new measure will ever entirely replace the existing one. Statisticians at the Census Bureau and Office of Management and Budget may make great progress creating a nuanced poverty measure that better reflects a family’s ability to pay for the cost of supporting itself. But Congress would decide if the federal government should change the way it allocates money reflecting the new statistics.
“Part of the problem is that we’re talking about money – who gets what,” Smeeding said. “And you don’t want to be campaigning as the senator or representative who got less federal dollars for your state.”
Are you on Facebook? Become our fan.
Follow us on Twitter.