Identity politics run amok

According to John Derbyshire, various bloggers have engaged the question of why Barack Obama is winning in states with large or small African-American populations, while Hillary Clinton does best in states where that population is neither large nor small. It’s no mystery, of course, why Obama does well where there are lots of African-Americans. With the possible exception of college professors at elite schools, there’s no more monolithic voting group in America than blacks.
As to why Obama does well in places like Iowa and North Dakota where very few blacks reside, but not in places with a middling blacks presence, the working hypothesis seems to be that “where the proportion of black citizens is big enough to cause tension, but not big enough to swing an election, the white majority will not support Obama, and he will do badly.” However, it seems to me that the explanation does not lie in racial tension per se. Clinton has strong support among Hispanics, Jews, and white urban working class voters. These groups are underrepesented in Iowa, North Dakota, etc. Thus, Clinton struggles.
It’s possible, of course, that Clinton’s support from these constituencies has something to do with racial tension — e.g., between Hispanics and/or Jews on the one hand and African-Americans on the other. But I see no reason to assume that this is the case. The Democrats sometimes divide along class, racial, and ethnic lines even when no African-American is in the top-tier — e.g. Mondale vs. Hart. The big difference this time is that the non-establishment guy who appeals to upscale voters and kids happens to be black. Thus, African-Americans combine with the upper income and younger voters, and their candidate wins both in states where many blacks live and in states where other traditional Democratic constituencies are not well represented.