here's a nice little piece about the distribution of wealth in the states--the data is a bit old (through 2000 it seems, based on other analyses published through 2005 from a quick look) but it's interesting:
Who Rules America: Wealth, Income, and Power
figure 4 is illuminating.
fact is that the statistical indices the united states generates for itself are not necessarily geared around providing a coherent basis for policy formulation, but seem to just as much be an extension of political ideology. while i happen to remember a significant number of fantasy-reality adjustments made by the reagan administration (my favorite is still inflation: want to control it? stop counting stuff that causes it...) but they're not unique to the reagan administration by any means (look into how the us does not count structural unemployment sometime)...
there is census information about poverty rates:
Poverty - Main
but there are alot of problems with it...the most recent controversies have been about the systematic undercounting of poor folk. that's one way to deal with the question: not count it.
it's not that the stats are useless, but rather that they're problematic both in their organization and content. so it's a bunch of work to figure out how to interpret them.
but it's pretty clear that there's not a whole lot of interest in generating accurate images of what class stratification in the united states--or there hasn't been during the neo-liberal period. i would hope this changes under obama and afterward, but it's not obvious that it will.
the reality of class warfare--you know, the everyday routinized violence produced by radically uneven distribution of even the most basic economic. social and cultural resources---is quite different from the language of class warfare.
both are interesting, both important--but they're not the same.