Everyone is entitled to his own opinion, but not to his own facts.
--Daniel Patrick Moynihan

November 8, 2011

The Blurry Poverty Line

By David K. Shipler

Now that the Census Bureau has offered a more realistic way of calculating poverty, the outdated method should be discarded instead of retained as the “Official Poverty Measure” used to determine eligibility for government benefits. It was designed in the 1960s, when the average family spent about one-third of its budget on food, a proportion that has fallen to one-seventh as housing and other costs have soared. So it makes no sense to take the price tag of a minimum food basket and multiply it by three. But that will continue to happen unless Congress and the administration act—a hard act to perform with federal and state governments in fiscal straits.

The revised method uses food, shelter, utilities, and clothing costs as the basis for determining the poverty line. But what is most useful about the approach is not its tally of the poor—49 million (16 percent) under the new calculation compared with 46.6. million (15.2 percent) under the old. It is, instead, the ability to drill down into the complex of government programs to assess how much they help, and into families’ everyday expenses to see how much they hurt. This is accomplished by examining both sides of the household ledger: income and spending.

While the “official” method treats the country as if it were homogeneous, the new calculation—called the Supplemental Poverty Measure—includes regional adjustments to such expenses as housing, whose costs vary markedly from place to place. In-kind government benefits (food stamps, etc.) that are not considered in the official calculation of household income are part of the “supplemental” method. In this way, the Census Bureau has built on two decades of study by experts to develop a sophisticated tool for policymakers, if they will use it.

We learn, for example, that without the Earned Income Tax Credit, the overall poverty rate would jump to 18 from 16 percent, and for children to 22.4 from 18.2 percent. Without housing subsidies, 0.9 percent more of the general population, and 1.3 percent more children, would reside below the poverty line. Eliminating food stamps would raise the overall poverty rate to 17.7 percent. But a much smaller difference—just 0.1 percent—is made by WIC, the nutrition program for women, infants, and children.

Expenditures such as transportation and child care for working parents, and out-of-pocket medical payments, are omitted from the official measure but included in the supplemental, with illuminating results. It turns out that the costs associated with having a job raise the poverty rate by 1.5 percent, and medical expenses raise it by 3.3 percent.

These and other findings point to the leverage over poverty that could be exercised by changes to benefits and expenses. The number of poor would be reduced significantly by expanding government medical insurance, such as Medicaid and the State Child Health Insurance Program. Poverty would be eased by broader access to subsidized child care and by sufficient mass transit to reduce the necessity of owning a car to get to work. More generous housing subsidies would have a major impact, especially since housing is typically the largest item in the family budget, soaking up cash that could otherwise be spent on food. Child malnutrition is higher among low-income families who do not have housing subsidies.

The new data also confirm what specialists have long believed: that the Earned Income Tax Credit has a direct, efficient effect on poverty, since it puts cash straight into the hands of low-wage workers. It also has the rarest of attributes: bipartisan appeal in Washington.

Only someone who has a job and files a tax return is eligible, which is why Republicans have traditionally supported it. The payment from the IRS is more than just a tax refund. It can exceed taxes withheld from a paycheck, and can add up to several thousand dollars if a person has a low enough income and at least two children. “When I get my taxes” has become a familiar line in poor neighborhoods. Obama’s stimulus package raised the payments, but only temporarily, leaving millions likely to fall again below the poverty line as the funding wanes.

The “poverty line” is artificial, whether it is the “official” $22,113 in annual income for a family of four or the new $24,343 calculated by the “supplemental” method. The line is imaginary. It defines no true boundary or threshold. It tells something about the material conditions of American families, but it does not mark a dramatic difference between misery and well-being.

Impoverished families struggling to move up and over the poverty line find that it’s more complicated than showing a passport and crossing a frontier. It’s more like walking through a minefield, and even those who get to the other side discover how little their living standards have changed.

Fulltime work is still hard to come by, wages are still inadequate, housing is still expensive, and the unquantifiable ingredients of powerlessness and stress and uncertainty remain. In territory that stretches some distance above the poverty line, there is still not enough to cushion a family from a reversal or a mistake. There is little to stop the chain reaction of problems that can cascade from a minor car breakdown to a day missed from work, to a lost job, to rent unpaid, to eviction, and on and on.

These are not terrible fantasies. They are real experiences. It will take compassionate perception for both voters and officials to read the human stories behind the new numbers and act accordingly.


  1. It makes a difference when someone with David Shipler’s public recognition for caring about and understanding poverty shows us the importance of the new Supplemental Poverty Measure (SPM). Shipler, more adroitly and vividly than the SPM report, demonstrates why we need to know that in-kind benefits such as Food Stamps or the Supplemental Nutrition Assistance Program (SNAP) and refundable tax credits such as the Earned Income Tax Credit have a profound effect on the lives of low-income families. He explains how the wildly varied costs of work expenses and out-of-pocket medical expenses affect the resources families have at their disposal for providing the steadier and universally required costs of food, clothing, shelter, and utilities. Unless the citizenry understands the realities reflected in these supplemental statistics, we cannot fathom what policies have succeed in reducing poverty. These “subtleties” are lost in our official poverty statistics that show no progress in reducing poverty overall from 1973, when the official poverty rate was 11.1 percent, until 2010, when it was 15.1 percent.

    The SPM, as Shipler shows us, reveals that SNAP and EITC reduce income poverty, especially for households with children, substantially, even as other variables have prevented us from reducing the overall poverty rate. We can also see clearly how support for childcare and especially the Affordable Care Act of 2010 can reduce costs that the working poor or near poor do not control. These data reflecting the reality of individuals seeking to participate at modest levels in society are all the more important in light of recent criticism of progressive tax policy and current levels of social spending. Legislators and other politicians demand that low-income households pay income taxes rather than receive tax credits and that we reduce social spending and repeal the healthcare law that offers hope for reducing the out-of-pocket medical expenses of these households. If the citizenry understands the reality disclosed by the SPM data, it will not so quickly consent to claims that we have no choice but to reduce inefficient social spending that has done nothing to diminish poverty.

    Rebecca Blank, currently Deputy Secretary of Commerce and earlier Undersecretary of Commerce in charge of census data has played an instrumental role in developing the SPM. She has been writing on poverty measurements and laboring for these revisions for years. See her article, “How to Improve Poverty Measurement in the United States.” We now need informed-journalists like David Shipler to show us their significance for the lives of millions of our fellow citizens.

    Please see additonal installment

  2. There are two areas in which we should push the SPM and David Shipler to go further.

    First, the SPM, while it recalculates the threshold or income poverty line for determining which households are in poverty, does not account for how changes in the standard of living can marginalize low-income households. Even if these households have minimally adequate income for food, clothing, shelter, and utilities plus 20% more, they may be impoverished based on an income threshold relative to the standard of living. As median income levels differ over time or from society to society, access to goods and services required for even minimal participation may be inadequate independently of provision for necessities. The cost of minimal communication and information and of minimal transportation have, for example, changed dramatically since the mid-sixties in the U.S. when phones, computers, televisions served very different purposes and residential location was typically closer to the workplace and grocery stores.

    For these reasons, students of poverty data like Amartya Sen contend that poverty thresholds should be set relative to standards of living rather than calculated on the basis of necessities. The SPM stands between the relative poverty measurements of the OECD and the Eurozone and the absolute poverty line used for the official poverty measurement in the U.S. The SPM adjusts for the changing cost of the market basket of necessities and for the cost of living across regions, but, as far as I can tell, it makes no adjustment for the changing patterns of consumption required for marginal participation in society. While using 50 or 60 percent of median income, as the OECD and Eurozone do, may not be the best formula, Sen is correct to notice that persons with adequate food and shelter may not have the capability to participate minimally in a society where most persons rely on costly means to interact with their fellow citizens. Persons with the similar resources may have been well integrated into society in 1960s and yet currently blocked from the same level of integrated functioning with their fellow citizens. It is important to note that the official poverty line has decreased from nearly 50% of median income at the beginning of the sixties to approximately 30% of median income today. The official poverty threshold does not provide a contemporary family the same capability for functioning as it did for a family in the 1960s, and the SPM, while adjusted somewhat, does not take into account the concern of Sen and others that changing patterns of consumption affect the capability to function in society.

    Please see separate installment.

  3. Second, the SPM considers only income. It does not consider the capability to function mentally, socially, or physically. Conservatives would replace the income measurements of the official poverty rate and the SPM with consumption levels or possession of commodities. Some liberals prefer measuring hardships such as going hungry or having the heat turned off. These other measurements have value, but none of them measures the freedom persons have to function minimally in society. Amartya Sen introduce the concept of capability and has offered some appropriate measurements of capability such as functional literacy, infant mortality, a reasonably long life, or, negatively, levels of long-term unemployment. Measurements of the capability or effective freedom to function in society could enrich our knowledge of whether persons and households are severely deprived independently of their levels of income, consumption, or hardships. These various measurements reinforce each other, but they can move in different directions or in the same direction at a different rate.

    The implication for effective policy is enormous. For example, David Shipler notes, based on the SPM, that eliminating the Women Infant and Children (WIC) program would have increased the income poverty rate by .01 percent in 2010. However, that datum does not tell us the affect WIC has on the health of women and children and the cognitive achievement of children in early education programs. We do have data that demonstrate that WIC improves the health status and cognitive functioning of its recipients. We need more authoritative data that reveal whether investment in this program produces these benefits systematically.

    The benefit of these measurements is not limited to WIC. While the SPM shows us that public healthcare insurance could and likely does have a huge influence on out-of-pocket medical expenses and thus on income poverty, it does not measure the effects SCHIP, Medical, or (prospectively) the Affordable Care Act have or will have on the health status of recipients. Various studies, including a recent randomized study of Medicaid recipients in Oregon, show health benefits from public insurance.
    It is not enough to know the income benefits for low-income families in order to assess the value of investments in assistance programs like WIC and a variety of public insurance. We need to know if these programs increase the capability of low-income households for functioning apart from their income effects.

    I am confident David Shipler agrees that such capability measurements have merit. We also know from the SPM report that funds are lacking for additional measurements. Both conservatives and liberals contemplating the merits of investing in social programs should want to know if WIC enhances the educational functioning of children and if public insurance reduces morbidity as well as saves families expense. Public investment in these measurements would go a step further than the SPM in informing us what works to reduce poverty and what does not.

    Harlan Beckley
    Director, Tom and Nancy Shepherd Program on Poverty and Human Capability
    Washington and Lee University
    Lexington, Virginia

    Program Director
    Shepherd Higher Education Poverty Consortium

  4. Rebecca Blank's article, to which Harlan Beckley refers in his first comment, can be found at http://www.npc.umich.edu/news/events/pdf/Blank-PovertyMeasure%20(2).pdf.

    His point about WIC's effectiveness in enhancing cognitive abilities is a good illustration of how much bang you can get for the buck in such a program. In sheer money terms, it's not very expensive and therefore doesn't have much impact on the poverty rate, but its longterm impact on children's nutrition, and therefore their brain development--and consequently their school performance and ability to grow intellectually--is significant. Indeed, many elements of poverty are hard to quantified and are not taken into account by income/expense measurements. I plan to come back to this issue in a later post. Thanks to Harlan for his very useful comments.