Defining Poverty: An Exchange

• Kaus is right to raise the issue of regional variation. It is indeed complex and often normative. For many of us, the decision to live in Tennessee vs. New York (or Manhattan vs. Brooklyn, East L.A. vs. Venice Beach) represents a set of personal and elective trade-offs. Kaus or I mostly live and work where we'd like and allocate our remaining resources accordingly. The presumption, however, that the same set of choices exists for everyone is either cynical or out of touch. Poor people often face barriers to mobility—lack of available jobs, subsidized housing, child care, transportation, the cost of moving itself—which can complicate relocation to cheaper climes.

In fact, a number of federal agencies, including the Census Bureau and the Office of Management and Budget, already employ differentiators (e.g., metropolitan and micropolitan statistical areas) to measure income, earnings, and poverty levels within and across regions. Adjusting the federal poverty measure to account for geography would hardly be groundbreaking. 

The real question is not whether regional variance ought to be reflected in poverty measures; it is where to draw the line. It's clear that an accurate calculation should account for the cost-of-living differential between—as Kaus suggests—Tennessee and New York. It is a harder call to determine whether the poverty metric should incorporate differences between, say, Rochester and New York City, or Manhattan (the Upper East Side? Washington Heights?) and Brooklyn (Park Slope? Brownsville?). Ultimately, this is an empirical question: It depends on where the barriers to mobility really are. The Bloomberg administration may not have struck the perfect balance, but it is studying the issue seriously.