I want to return to the new Washington Post/Kaiser/Harvard poll discussed below for a moment to talk about how well Obama needs to perform among white voters this fall in order to win -- or, I should say, to win a majority of the national popular vote.
By my back-of-the-napkin calculus, Obama's national target among white voters is around 40 percent. Here's how I derived that: First, assume -- and I think this is a safe assumption -- that this will be the first election in American history in which nonwhite voters account for one-fourth (or more?) of all votes cast. Given recent Democratic performance among key nonwhite voting blocks, adjusted slightly for an Obama candidacy, that means the Illinois senator is likely to get about 95 percent of African-American votes, as well as solid but not quite as strong levels of support from Native Americans (say 80-85 percent), Hispanics (65-75 percent) and Asian-Americans (55-60 percent). Weighting for group size -- African-Americans and Hispanics account for most of the nonwhite vote -- figure conservatively that Obama gets about 80 percent of this quarter of the electorate, or 20 percent net toward a majority threshold of 50 percent nationally. He thus needs to glean 30 more percentage points from the remaining 75 percent of white America. And 30 divided by 75 yields 40 percent.
Of course, that is the mark to forge a minimum winning coalition. But if Obama, back from his European trip, can successfully steer the conversation to the domestic-economic situation and tackle head-on the underlying pessimism/realism voters feel about the futility of that situation, hitting that 40 percent mark will not be a problem. In 2004 John Kerry, a far less compelling candidate who got a late start and offered a weak economic message against an incumbent president whom he let beat the tar out of him for several weeks in August four years ago, still managed to get 41 percent of the white vote.