When it comes to polling for Democrats, Mark Mellman is at the top of his field. He worked for Sen. John Kerry during his 2004 presidential campaign, polls for the majority leaders in the House and the Senate, writes regularly for the Hill and is generally just plain respected for his work. So the gauntlet he threw down in the New York Times on Thursday, which came in the form of an Op-Ed that suggests fears about Barack Obama's troubles with white working-class voters are overblown, is bound to get some notice.
"Mr. Obama certainly seems to have trouble with these crucial swing voters. In almost every Democratic primary, he has lost white voters who didn’t graduate from college to Hillary Clinton," Mellman admits. "But," he says, "that doesn’t mean those voters will snub him in the fall.
"First, there is no relationship between how candidates perform among any particular group of voters in primaries and how they do with that segment in the general election. In 1992, Bill Clinton lost college-educated voters to Paul Tsongas in the early competitive primaries, but he went on to win that group in November by the largest margin any Democrat ever had."
Mellman argues that whether Democrats lose white working-class voters may not ultimately be as important as the magnitude of the loss, and notes the wide margins by which both Kerry and former Vice President Al Gore lost the white working-class vote in their respective runs. "Mr. Gore lost them by 17 percentage points while winning the national popular vote. Mr. Kerry lost them by 23 points and the country by fewer than two and a half points," Mellman says, going on to observe that, in the latest New York Times/CBS News poll, in the disputed demographic Obama trails presumptive Republican nominee John McCain by just 2 points. (Another poll, Mellman says, has him down 7, but even that means he's running well ahead of Gore and Kerry.)
Ultimately, Mellman says, Obama "cannot afford to take any constituency for granted. But he plainly has a path to victory. And the white working class does not seem poised to block his way."
Mellman didn't include a disclaimer about the relative value, or really the lack thereof, of polls conducted this early, but I will. In just one example I happened to see earlier today, Ben Smith posted an Electoral College map that was produced based on state-level polling of the 2004 election done around this time; the map shows Kerry winning in a landslide. State-level polling is considered less reliable in this case than national polling, but early polling generally shouldn't be considered gospel.