Counting The Poor So It Counts: How To Measure Poverty In The Information Age

by David L. Englin and Traci L. Parmenter
Moving Ideas Network, June 6, 2002

As part of the 2002 welfare reauthorization bill, House Democrats tried to add "reduce the extent and severity of poverty and promote self-sufficiency among families with children" as the stated purpose of welfare reform -- House Republicans rebuffed them. Welfare by its proper name is "Temporary Assistance to Needy Families," or TANF, and it is difficult to understand how one could reject reducing poverty as a major purpose of the system. Maybe it is because, four decades after the federal government began trying to define and measure poverty, it still does so in a way that bears little resemblance to the lives of real people. With Congress about to spend billions of dollars on welfare, without really knowing who does and does not need temporary support, it's time to cut through the politics and reclaim the integrity of our primary system of public assistance. Using the technology on our desktops, we can define and measure poverty in an honest way that will actually make a difference to people struggling for a piece of the American dream.

We know from high-profile efforts like the 1995 National Academy of Sciences study of poverty measures that the current poverty formula is deeply flawed, but all we do is continue to study the problem. The federal government itself has invested significant time and energy in considering a more responsive system -- so why has nothing changed? It is almost self-evident that inaction on this issue is all about politics; all about ignoring a population that lacks influence, and all about fear that changing the system will make poverty appear better or worse or will make spending higher or lower. Statistical exercises like the Census Bureau's study of experimental poverty measures might be interesting academically, and they might hold the keys to our long-term understanding of a variety of socioeconomic issues. But they have so far done little to change our understanding about who is poor.

For nearly four decades, the federal government has derived the so-called "poverty line" (actually issued annually by the Department of Health and Human Services as 'poverty guidelines') by taking a family's most basic grocery budget, multiplying it by three and adjusting for inflation. This formula was established in 1955, when a survey was conducted that found food to be one-third of a family's total expenses. Today, food is less than one-sixth of a family's budget, but nobody has bothered to change the formula. Even if we pretend that three times a family's food budget will cover food, housing and health care, not to mention other minimal necessities, this formula ignores cost of living differences, so that a family of four in New York City is expected to survive on the same budget as a family of four living in rural Kansas. (The exceptions to this are separate poverty thresholds for Alaska and Hawaii.) This formula is so inadequate that TANF, as well as many other major federal assistance programs, ignores federal poverty guidelines entirely.

So let's imagine what a better system, a system that reflects reality, might look like.

First, we should know what we mean by 'poverty.' Adam Smith, the father of capitalism, wrote that "poverty is a lack of those necessities that the custom of the country renders it indecent for creditable people, even of the lowest order, to be without." The National Academy of Sciences report suggests that food, clothing, and shelter top that list. To that we would add transportation, child care, and health care, since it would be would be 'indecent,' not to mention impractical, to expect people to support themselves with no way to get to work, no safe place to leave their children, and nobody to get them back to work quickly when they are sick. Beyond that, we may want to include something for basic utilities (including local telephone service, for the sake of public safety) and modest household expenses. Remember, these are not services we are paying for, but things we would not expect people to live without.

Once we agree on a list of necessities (which, admittedly, is probably easier said than done), we would need to know what they cost in different parts of the country to account for regional cost of living differences. We could perform some kind of statistical adjustment, as others have suggested. Or we could calculate poverty measures for each state based on actual costs in that state. That improvement would complement welfare reform, which gives states block grants to provide welfare services. However, even within state borders, different areas can have different costs of living. Therefore, we ought to get as local as we can, perhaps measuring costs and calculating poverty lines for different zip codes or municipalities or-- better still for democracy -- for different legislative districts.

Most people understand that some places are more expensive than others; but cost of living can also change over time. The Census Bureau's experimental poverty measures and the National Academy of Sciences recommendations each focus on annual figures, but many Americans learned the hard way after Sept. 11 that cost and income figures reported annually might not reflect people's real lives, which can change suddenly. The most accurate system would update poverty lines on a regular basis, perhaps even weekly, to reflect the experience of families trying to get by between checks, whether those checks come from employers, charities, or the government.

We also should consider non-income sources of support. Prior studies have found that this might actually decrease poverty rates among people presently on welfare and increase poverty rates among full-time workers. For example, a family with employer-based health coverage in a low-cost area could have much less income and still not be poor, while a family in a high-cost area who pays all expenses out of pocket, or whose main breadwinner is self-employed, would need a much higher income to avoid poverty.

Finally, as in the current system, we must make adjustments for family size. It goes without saying that a family of one can pay for necessities with a lower income than a family of five.

A different poverty line for every zip code? New numbers each week? Obviously, this new and improved system would be more than a table of figures. It would be a database, into which we would feed cost and income numbers from a variety of public and private sources, and which we could configure to produce targeted and up-to-date poverty reports on demand. More than an advanced statistical model, this system would use the multitude of data recorded each week, and sometimes daily, by government and commercial enterprises to measure costs directly. And as a boon to research already under way, this database could be online and accessible to academics, government leaders, journalists, and voters.

Much, if not all, of the information required to create this system already exists. Federal and state governments collect and maintain data on food, housing, healthcare, childcare, transportation and income. In the private sector, sales and performance research has become a high art. National and regional realtors, supermarkets, department stores, HMOs, telecom companies and utility providers collect and track such data all the time. In some cases, only monthly, or quarterly, or state or regional numbers might be available. But in every case, public and private leaders could work together to give us a system that is dynamic and geographically specific.

When the Johnson Administration adopted the current formula in 1965, in a world without globally connected, high-speed computers on local, state, and federal desktops, it might have been the best we could do. But today we are a nation of technology leading the world into the Information Age, and today we can do better. And we should do better, not because of a narrow political interest in making poverty appear better or worse than it is, not because we want to increase or decrease a particular expenditure, but because integrity and morality demand that we do the best job we can to see the real situation as it relates to the lives of real people.