Unless you're prepared to do some serious regression analysis, it's hard to isolate race as a variable in any construct. For example: a new Allstate/National Journal poll out today finds a stark racial divide on the question of whether President Obama's policies are providing more opportunity "for people like you to get ahead." Only 31% of whites say yes; more than 75% of blacks say yes. 52% of Hispanics say yes; 70% of Democrats say yes; 10 % of Republicans say yes; independents are split down the middle -- a third say the policies won't change the level of opportunity they're afforded, 34% say "yes" and 30% say "no."

Well -- black people tend to be Democratic; a plurality of white people are associated with the Republican Party, so one cannot look at the data and find racial division, per se, without segregating the effects of partisanship, which exerts the strongest pull on our allegiance. Democrats tend to embrace government; Republicans tend to be skeptical of it.  Actually, the attitudes of white people toward President Obama mimic the attitude of independents, with 26% people remaining ambivalent.

Given our history, it'd be naive to assume that attitudes about race are absent when people are asked to consider Obama's policies. I'd bet that certain questions can elicit sharper reflections of the differences. And race -- and culture -- are reasons why a good number of Republicans became Republicans in the first place -- and a paramount reason why black people became reliable Democratic voters. There is evidence in the Allstate data that beyond Obama specific questions, there is some race-based aversion in the most Democratic-leaning group -- white women, as Ron Brownstein points out. White America -- and independent America -- may be moving in a Perot-esque direction that transcends attitudes about Obama. If the economy gets better but the divide remains, there may be more to say.