As we move deeper into the current recession, the number of poor Americans is rising. Poverty data arrives with a lag; we won’t know the 2008 poverty rate until August, 2009. But poverty and unemployment are closely linked, and unemployment has been rising rapidly. More than 11 million Americans currently report they are unemployed. That means the number of families with seriously reduced income is growing.
Unfortunately, even when the poverty data arrives, we will be missing key information needed to understand who’s poor and who isn’t. And, if the stimulus legislation under consideration in Congress is adopted, it could play a significant role in helping low-income families, but the official measure of poverty won’t do a very good job of reflecting the impacts of these policies.
The current approach to measuring poverty hasn’t been updated since it was established over 40 years ago. No other regularly reported economic statistic has been unchanged for four decades.
The current measure determines whether a family is poor by comparing their pre-tax cash income to a poverty threshold. For example, in 2007, the threshold for a family of four was $21,203.
The thresholds were originally set in the mid-1960s. In 1964, a Social Security Administration economist named Mollie Orshansky looked at data from 1955 — the best available at that time — and found that a family spent, on average, one-third of its income on food. So, the government took the cost of a low-cost food diet, multiplied by three, and the resulting number became the official poverty line. Since then, it has only been adjusted for inflation. But, while food was about one-third of family budgets in the mid-1960s, it’s closer to one-eighth of family budgets today, and the costs families face have changed in many ways over the last 40 years.
Poverty status is calculated by looking at family pre-tax cash income, and comparing family income to the poverty line. Pre-tax cash income may have been a good measure of family resources in the early 1960s, but now it overstates family resources in some ways and understates them in other respects. And, it does a poor job of showing the effects of a host of non-cash policies and programs that are intended to improve well-being. For example, there’s no effect on the official poverty measure if taxes go up or down or if tax credits are expanded. There’s no effect if food stamps or housing subsidies are expanded or cut back. There’s no effect if additional child care subsidies or health care coverage affect a family’s work expenses or health care costs.
To appreciate what the official measure misses, it’s worth considering some of the significant provisions in the recovery package pending in Congress. The pending bill would create a Making Work Pay tax credit worth $500 to most workers, and would expand the federal Earned Income Tax Credit and Child Tax Credit in ways that will help low-income workers and families. The bill would provide a temporary increase in food stamp benefits. It would provide an additional $2 billion in child care subsidies over the next two years. It would expand health care coverage for unemployed workers. All of these will provide real, important benefits to workers and families, but none will affect the official poverty measure.
The nation needs a more accurate poverty measure, with a threshold based on data about current family expenses and a more comprehensive measure of family resources. Doing so would give us a far better picture of who is poor, and of what difference the recovery package and other efforts are making to affect the numbers of poor Americans during and after the recession.
In recent decades, a considerable amount of work has gone into efforts to develop an improved poverty measure. In 1995, a panel of scholars at the National Academy of Sciences (NAS) recommended a set of changes. Under the NAS approach, thresholds would be based on family expenditure data for a bundle of necessities, including food, shelter, clothing, and utilities. Thresholds would vary geographically, reflecting significant cost differences in different parts of the country. Family income would take into account tax liabilities and credits, near-cash benefits like food stamps and housing subsidies, and the fact that money that is spent for work and health care costs isn’t available to meet other basic needs.
Legislation introduced in the last Congress by Rep. Jim McDermott in the House and Sen. Chris Dodd in the Senate would direct the federal government to update its definition of poverty based on the NAS recommendations. The new Administration could decide to make this change even without legislation. There’s a compelling case for doing so now: we should begin a new Administration with a measure that gives a more accurate picture of poverty before and after the recovery package and gives a clearer picture of what happens as unemployment goes up and down.
Political leaders may disagree on the most effective ways to reduce poverty, but they ought to agree on the need for a clear, reliable measure of it. An improved measure won’t, in itself, reduce poverty, but it will give us a far better understanding of the economic conditions of millions of Americans, and of what difference government policies make for large numbers of struggling families.
For a fuller exposition of the themes in this piece, please see the paper entitled “Improving the Measurement of Poverty,” by Rebecca M. Blank and Mark H. Greenberg, published in December 2008 by the Hamilton Project at the Brookings Institution.
Rebecca M. Blank is the Robert S. Kerr Senior Fellow at the Brookings Institution, a member of the Spotlight Advisory Council and served on the Council of Economic Advisers from 1997-1999. She was also a member of the 1995 National Academy of Sciences panel referenced in this piece.
Mark H. Greenberg is Executive Director of the Georgetown Center on Poverty, Inequality, and Public Policy, a member of the Spotlight Advisory Council and a Senior Fellow at the Center for American Progress.