When does income inequality matter? Less often than Paul Krugman thinks, according to Dan Drezner.
While we're on the subject though, in Dan's earlier post on inequality, he notes briefly, in a postscript, a problem with frequent claims about U.S. income mobility you hear in conservative and libertarian circles. The most optimistic numbers always seem to come from a 1992 Treasury Department survey that Glenn Hubbard did—I know Tom Sowell cites it with some frequency, as have Bruce Bartlett and Heritage. Bizarrely, some of the same sources cite an Urban Institute report that explains precisely why the Treasury study is pretty much useless. By restricting its sample to persons who filed tax returns, the Treasury study is, most obviously, skewed in favor of those who did well enough to keep filing returns over that period and, slightly less obviously, filters out those who died over the course of those ten years—disproportionately the elderly and those with less access to quality medical care. This is a pretty glaring methodological problem, and it's bugged me for some time that "our side" nevertheless continues to use those figures, when there are other less spectacular but less flawed data we could reference.