Today, the Washington Post's Jon Cohen offered a criminally misleading analysis of exit polls, imagining a racial conflict where non exists.
"Education has been a key divider among white voters in a contest marked by an evident racial divide," Cohen writes. His analysis presents numbers showing that Clinton does better with white non-college graduates than Obama and that Obama does better with white college graduates than Clinton.
Cohen concludes that racism may explain this gap. "Just a third of whites without college degrees were that sure the country is prepared for a black president," he says.
The fact is that Hillary Clinton's race and Barack Obama's race have little to do with these numbers.
To understand that, you have to look at some recent history: the 2000 primary in New Hampshire between Al Gore and Bill Bradley.
In that contest, Gore was seen as Bill Clinton's heir apparent -- much like Hillary Clinton is today. Bill Bradley was the outsider, non-establishment candidate, the brilliant Senator -- much like Barack Obama is today. Heck, they both even like basketball, though Bradley (the white guy) was evidently a better player.
In that primary, Bradley did 13 points worse than Gore among non-college grads, who made up 45% of the electorate. (Although this isn't broken out by race, New Hampshire's Democratic primary voters are 95%+ white, so there isn't much need to do so.)
Meanwhile, among college graduates, Bradley did 9 points better than Gore.
In 2008, in a 3-way race, Obama did 9 points worse than Clinton among non-college grads, who made up 46% of the electorate.
Meanwhile, Obama did 5 points better than Clinton among college graduates, who made up 54% of the electorate.
2000 and 2008 aren't entirely comparable; Clinton has an appeal to women that Gore did not, and Obama has an appeal to blacks that Bradley did not. A three-way race (in NH) is different than a two-way race.
Still, on balance, the remarkable similarity between the 2000 education gap and the 2008 education gap seriously undermines the claims advanced by the Post that today's numbers are indicative of racism or racial divison.
They are not. They indicate a preference gap, seemingly tied to education. And in elections, that happens. Most of the time.
© Jed Lewison