Cliff May reports in The Corner:

"We simply do not defeat an incumbent president in wartime."

The quote above is from Mark Mellman. Mark is a Democratic pollster. And he’s very partisan. And he’s been working for Kerry this year.

But Mark is no fool and he has a reputation to maintain.

He explains his reasoning at some length in a column in The Hill this morning. He predicts that Bush will get 51.6% of the two-party vote.

By contrast, Jay Cost is predicting Bush will get less than that: 48.63% to Kerry’s 46.24% [Ed: see update below -- these number are wrong].

The Corner also includes this report by Larry Kudlow:

That’s the call made by Ed Goeas, co-author of the blue-chip, bipartisan Battleground Poll. I spoke to Ed this morning, at 7 a.m. He is using an ultra-conservative, worst-case, 2000-like, partisan sample of 44.3% Democrats and 42.3% Republicans. He tells me Bush has bounced up in the past 48 hours on leadership and Iraq. Also, on the economy -- on keeping America prosperous, Bush has moved up from a 45% to 47% deficit to a 48% to 46% lead...Ed also believes that white, conservative, Christian Evangelicals could carry the day for Bush, where 83% look to be voting, giving a 69 percentage-point margin to Bush, and seven percentage points higher intensity support for Bush than labor unions have for Kerry. Goeas also weights Hispanic support for Bush at a conservative 35%, even though his “unaided” tally actually produces a 43% Hispanic vote for Bush. Goeas also believes he is underweighting black support for Bush and male support.

Developing...

UPDATE: Here's Mellman's column: "The uphill fight." Several readers have written to point out that Cost's final prediction is Bush 51 to Kerry 47.3. Guys, I was just quoting...