This isn't to try and discount the other polling, but just to give folks some polling that they can feel a little better about this evening. Rasmussen's daily tracking polls for OH, FL, MI, and PA. Data from September 2-8 (yesterday was Sep 1-7)
OHIO:
Kerry 49 (48)
Bush 46 (46)
Bush Favorables:
49 Favorable, 50 Unfavorable (49, 49)
Kerry Favorables:
54 Favorable, 44 Unfavorable (53, 45)
Right Track/Wrong Track: 41 right, 52 wrong (42, 52)
MICHIGAN:
Kerry 48 (46)
Bush 42 (47)
Bush Favorables:
47 Favorable, 52 Unfavorable (49, 51)
Kerry Favorables:
54 favorable, 45 unfavorable (52, 45)
Right Track/Wrong Track: 39 right, 56 wrong (38, 56)
PENNSYLVANIA:
Kerry 48 (48)
Bush 46 (46)
Bush Favorables:
51 Favorable, 48 Unfavorable (51, 48)
Kerry Favorables:
51 Favorable, 48 Unfavorable (51, 48)
Right Track/Wrong Track: 39 right, 55 wtrong (39, 55)
FLORIDA
Kerry 47 (47)
Bush 47 (47)
Bush Favorables:
49 Favorable, 50 Unfavorable (50, 50)
Kerry Favorables:
52 favorable, 47 Unfavorable (53, 46)
Right Track/Wrong Track: 43 right, 53 wrong (43, 53)
*Rasmussen's 3 Day Rolling Average for 16 Battlegrounds:
Kerry 51.3, Bush 45.8
Bush Battleground Favorables:
48.2 Favorable, 50.7 Favorable
Kerry Battleground Favorables:
53.3 favorable, 44.8 unfavorable
*Fox News Battlegrounds: Kerry 48, Bush 43
*Gallup Battlegrounds: Kerry 50, Bush 45
I think there's little doubt that Kerry is behind nationally, and I'm concerned about the internals, particularly Kerry's favorables. Juat wanted folks to realize that not everyone has Bush surging; note that Ras does weigh for party turnout, whereas most of these other polls have not. I'm not making claims about which method is better, but many of the recent polls do appear to have oversampled Repubs. That's not entirely their fault; there are, probably, more folks identifying as Repub right after the convention. But there's little to no evidence that there's suddenly been some permanent shift in Party ID.
My basic feeling on the state of the race: I think that neither candidate has very good internals. The highest job approval that Bush seems able to muster is 52, and that's lousy for a bubble period. It leads me to believe that Bush's equilibrium approval is probably under 50. I'd also expect that Bush's equilibrium re-elect number is under 50. That leads me to my belief that the majority of folks who vote on Nov 2 will NOT APPROVE of Bush and WILL NOT BELIEVE HE DESERVES RE-ELECTION. That, of course, is not the same as predicting a Kerry victory.
Kerry needs to bring his favorables back up. And he needs to make dents on the C-i-C questions. I don't think he needs to lead on those questions, but he has to get back into the ballpark. That will mean not only continuing a more direct attack on Bush's Iraq policy (and he needs to continue this, as well as continuing the linkage of Iraq to the domestic issues, because that reframes the issue in a way that's favorable to Kerry) but a much more aggresive and concise explanation as to why Bush's policies have left us less safe. That may now include a direct attack on Bush's handling of 9/11, both pre-and-post. It's risky, but it's time to take the risks.
Basically, I don't think that Kerry is the favorite anymore, but I'm not quite ready to declare Bush the likely winner. His fundamentals are lousy, and that'll be more evident once the bubble period ends. He's also got some unfavorable stories in the news cycle. But Kerry has much work to do.