Census Bureau releases alternative measures of poverty

Washington Post Staff Writer
Tuesday, January 4, 2011; 6:57 PM

The Census Bureau took a baby step toward redefining what is considered poor in America on Tuesday when it released several alternative measurements of poverty, fundamentally revising a one-size-fits-all formula developed in the 1960s by a civil servant.

Under a complex series of eight alternative measurements, the Census Bureau calculated that in 2009, the number of Americans living in poverty could have been as few as 39 million or as many as almost 53 million. Under the official calculation, the census estimated that about 44 million were subsisting on incomes below the poverty line of about $21,750 for a family of four. The alternatives generally set the poverty threshold higher, as much as $29,600 for a couple with two children.

In September, the census estimated the nation's poverty rate in 2009 was 14.3 percent. Under the alternatives, it could have been as low as 12.8 percent or as high as 17.1 percent.

For the time being, the government will continue to use the original poverty definition to determine eligibility for federal programs. The alternatives are experimental and will be revised every year, eventually winnowing them to one.

The bureau's move is expected to reignite a debate over whether to replace the current measurement, as was recommended in 1995 by a blue-ribbon panel from the National Academy of Sciences (NAS). The alternatives are offspring of the NAS report.

"Our interest was in getting a better measurement," said Robert Michael, a University of Chicago dean who chaired the panel. "That's politically difficult to do because of the entrenched benefits to those that are currently getting them. Our interest was in understanding how our nation is doing in terms of serving the need."

The current formula was devised by Mollie Orshansky, a civil servant in the Social Security Administration who took the cost of a "thrifty food basket" for a family of four and multiplied it by three. Her formula has been updated for inflation. It continues to harbor a number of quirks traceable to attitudes of a half-century ago, such as a $1,000 reduction in the poverty line for people older than 65, largely because Orshansky, an economist and statistician, believed older people eat less.

The alternatives reflect a growing consensus among experts in the poverty field that the old formula does not adequately measure poverty in the 21st century.

For example, it does not take into account the impact of governmental anti-poverty programs, such as Medicaid, welfare, food stamps, school lunches, subsidized housing or income tax credits. Nor does it include a host of expenses typical in families with two working adults, such as child care, multiple automobiles and professional wardrobes. It also does not make any consideration for regional differences in the cost of living.

Although the poverty measurement is largely of interest to academics today, it has the potential to alter our perceptions of who is poor, how persistent a problem poverty is and whether policies should be reordered.

The AARP, for one, recently noted that under the alternatives recommended by the NAS panel, the number of elderly people considered poor would rise by more than 3 million, largely because expenses would include large out-of-pocket medical costs more typical of senior citizens.

"We often congratulate ourselves on the big drop in poverty levels that have taken place since the '60s, as we should," said John Rother, vice president of the AARP. "But this shows we haven't made as much progress as we like to think. Particularly because of the increase in health-care costs, we're at risk of losing the progress we've made."

Many experts say they expect that a refined alternative measurement eventually will replace the current measurement.

"If both are being published 10 years from now, I'd hope no one would pay attention to the official measure anymore," said David Betson, a Notre Dame economist who also was on the NAS panel.

"This is not to stomp on the grave of Mollie Orshansky. She was a nice person. We all liked her. Her measurement was probably right back in 1963, but that measurement's not right for today's society. Statistics have to evolve. You have to make them relevant," Betson said.

© 2011 The Washington Post Company