Defining Poverty: An Exchange

b) Counting regional variations in the cost of living is a bit fishy, no? If I make enough money to live semi-comfortably in Tennessee, but choose to live uncomfortably in New York City, should I really be counted as part of America's failure to eradicate poverty? Keohane cheers Bloomberg's measure for apparently carrying this to ridiculous extremes by adjusting for varying costs of living "even within the city." It's one thing to suggest that New Yorkers shouldn't be expected to seek cheap rents in Tennessee. It's another to say people in Manhattan can't be expected to move to Brooklyn. And there's an obvious pecuniary incentive for a New York pol like Bloomberg to take into account geographic variations in cost of living—it makes New Yorkers look needier and helps him beg for more federal assistance.  

c) The poverty line is just a line—a necessarily arbitrary line. It's mainly useful to show trends—i.e., is there more "poverty" or less? If the line is reformulated so more people fall below it, they are no better or worse off than before. But moving the line serves an obvious propaganda point—if "advocates" can say 23% of the population, not 18%, is officially "poor." Why not avoid the "propaganda" charge by doing what Gary Burtless of the Brookings Institution once suggested to me: refine how we measure income, but then set the poverty line so that, for the first year, there are exactly the same number of poor people under both new and old measures. That would make it harder for those on the left to use the new formula as part of a rhetorical scare campaign. Why do I have a feeling that would also reduce much of its appeal to Keohane?

Georgia Levenson Keohane responds:

Mickey Kaus' critique of my article on measuring poverty—which examined the anachronistic federal poverty metric and New York City's alternative approach—lets us probe this complex issue further: 

• To quibble once more over Patrick Moynihan's place in the history of social policy: the senator's work on poverty is selectively embraced and reviled by the right and the left, depending on the day, the issue, and the circumstance. Ironically, in a November AEI discussion of Nick Eberstadt's book The Poverty of "The Poverty Rate," Doug Besharov's beef with Moynihan ("he wanted a new poverty measure, not because he thought it didn't count right, it didn't do enough for New York") rings similar to the "pecuniary incentives" Kaus ascribes to Bloomberg. At present, the New York formula is not being used as "propaganda" to "beg for more federal assistance" but rather as an analytical tool to parse local demographics and need. This observation is not meant to eulogize the CEO's new measure—which is indeed imperfect—but rather to point out that:

• The poverty line is more than "just a line." Although Kaus may believe that "it's mainly useful to show trends," the reality is that it's mainly useful to avoid hunger. The federal poverty threshold is used to qualify poor people for food stamps and free or reduced-price lunches at school—as well as for Medicaid and SCHIP, supplemental Social Security, Head Start, and other programs that meet real needs. How the line is devised, therefore, matters very concretely for millions of Americans. Why wouldn't we want to reform a 40-year-old, inaccurate measure to assess more precisely who is in need of these services? Furthermore, as the New York experiment shows, improved and more refined data allow us to ask not just whether there is more poverty or less but whether our efforts to combat poverty—through tax credits or child care subsidies, for example—are having any effect.