September 21, 2010
Stanford Social Innovation Review, September 17, 2010
In their SSIR article on improving poverty measurement, Rourke O’Brien and David Pedulla convincingly argue that both the current official poverty measure and a new “Supplemental Poverty Measure” (SPM) proposed by the Obama Administration are flawed because they “reduce the complexity of poverty to a single figure, a line.” This point echoes Amartya Sen’s critique of measuring poverty solely in terms of having income above or below a line. As Sen writes in his most recent book, The Idea of Justice, “the relationship between [income] and poverty is both variable and deeply contingent on the characteristics of the respective people and the environment in which they live—both natural and social.”
Research on the differences in the relationship between income and material hardship for people with disabilities provide some of the strongest support for this idea. Most notably, recently published studies by Susan Parish and Peiyun She of Mathematica Policy Research and Susan Parish of Brandeis University show that families with disabled members need higher levels of income than families without disabled members to avoid food insecurity and various other material hardships. A closely related problem is that the current poverty line—and likely the SPM—is now set so low that excludes the majority of low-income Americans who experience the concrete forms of economic hardship that most Americans associate with the word “poverty.” Consider food insecurity and hunger. One might assume that most people experiencing food insecurity, and especially the most severe forms of it, live below the poverty line. But, in fact, the majority of food-insecure households, roughly two-thirds of them, live above the federal poverty line, with the bulk of that two-thirds falling between 100 and 200 percent of poverty.
Thus, O’Brien and Pedulla are right to conclude that “if our goal is to achieve a better measure of well-being to diagnose human needs and design effective solution, no line—no matter how thoughtful or sound—will do.” But I’d also argue that their proposed solution falls short of this goal because it don’t actually “move beyond the line.” O’Brien and Pedulla propose modifying the SPM to take account of regional variations in the costs of a range of goods and services, not just housing costs as proposed by the Administration. While useful, this still leaves us with an approach that reduces poverty to having income above or below a income line, with the only difference being that the line varies by area. It doesn’t really address the many limitations detailed by O’Brien and Pedulla with using an income line as the sole measure of poverty, including the particularly important fact that it provides incomplete information about specific forms of material deprivation, such as food insecurity, and housing and health related hardships. Moreover, because the SPM will likely have an income threshold only modestly higher than the current poverty threshold, it would still exclude the substantial share of households with incomes between 100 and 200 percent of the poverty threshold that experience various severe forms of material deprivation.
So what to do? O’Brien and Pedulla hint at a better solution in a short sidebar discussion of multi-dimensional indices of well-being. Such indices deserve a much more central place in current U.S. conservations above reforming the poverty measure. In particular, a new child poverty measure adopted by the United Kingdom earlier this year provides the best model for the United States. The new UK measure is actually three measures bundled in one package. The first two “tiers” of the measure are relatively conventional income-based poverty measures, although both improve on the official U.S. measure (and the SPM) because they are set at 60 percent of median disposable income (net of taxes). The real innovation, however, is with the third tier, which combines income, using an income standard that is roughly equal to 200 percent of the current poverty line, and 21 indicators of material deprivation. The indicators of material deprivation are grounded in part on public consensus about the “necessities” of modern living—a majority of the UK public agrees that each of them are necessary—but aren’t limited to bare subsistence indicators. So, for example, one of the indicators measures whether a family is able to regularly save £10 a month or more for rainy days or retirement; another assesses whether parents are able to afford celebrations on special occasions for their children. This income-plus-deprivation measure of poverty shows that just over 17 percent of UK children experienced poverty in 2008. For more on the UK measure, and also on a similar measure that has been used in Ireland for more than decade, see my report: A Modern Framework for Measuring Poverty and Economic Security.
Also worth considering is a measure modeled on the just released Multidimensional Poverty Indicator (MPI) developed by researchers at Oxford University. The MPI uses 10 indicators to measure three critical dimensions of poverty at the household level: education, health and living standards. Of particular note, the MPI goes a step beyond the UK measure to also reflect the intensity of poverty—the index number reflects both the number of people experiencing poverty and the number of specific deprivations they experience. Although the MPI was developed for 106 developing countries, the methodology could easily be adoptedfor use in the United States. For more on this measure, see the website of the Oxford Poverty and Human Development Initiative.
What would it take to implement these kinds of “next-generation” poverty measures in the United States? The key step would be to incorporate a set of deprivation indicators, similar to those used in the United Kingdom, into an annual survey that is large enough to produce reliable state and sub-state estimates. While the Survey of Income and Program Participation (SIPP) includes all of the indicators that would be needed, it isn’t an annual survey and is currently too small to produce state- and local- estimate. The best option would be to incorporate the SIPP hardship indicators into the Current Population Survey and American Community Survey. Foundations could play an catalytic role here by funding U.S.-specific research on multi-dimensional poverty measures as well as public opinion research on the what goods, services, and activities most Americans view as necessities.
Shawn Fremstad is director of the Inclusive and Sustainable Economy Initiative at the Center for Economic and Policy Research in Washington, DC. CEPR conducts research and public education to promote democratic debate on the most important economic and social issues that affect people’s lives.>