Barack Obama has risen to his highest-ever level in both our electoral college and popular vote projections, principally on the strength of his commanding lead in the national tracking polls. Gallup, Rasmussen and Hotline each have Obama ahead by 7 points, and Research 2000 has him up by 12 (Battleground, which has generally had the most conservative numbers for Obama, does not publish on the weekend). Whether or not the McCain campaign's new round of attacks will have a significant impact on Obama's numbers we shall see, but they're going to have to knock him off a fairly high pedestal.
There is state polling out today in Minnesota, Colorado and Ohio. In Minnesota, the Star Tribune has Obama ahead by 18, quite a contrast from SurveyUSA's contemporaneous poll which had McCain up by 1. Yesterday, I discussed the disparities between these two polls on the senate side, and it is not surprising that the presidential numbers have followed suit. Our model projects Obama to win Minnesota by 8-9 points, roughly in between the SurveyUSA and Star Tribune estimates.
In Colorado, Mason-Dixon -- polling for the Denver Post -- has the race tied at 44-44. Mason-Dixon's polls have had a statistically significantly Republican lean thus far this cycle, and so it's not terribly surprising to see their numbers a couple of points to the McCain side of other recent polling of the state. Nevertheless, there have now been a couple of different polls coming out in Colorado -- ARG, Ciruli, and last Monday's Rasmussen number -- suggesting that the race there may have tightened a bit.
Lastly, in Ohio, the Columbus Dispatch has Barack Obama ahead by 7. It's a good number for Obama, but not one that should be taken very seriously, as the Dispatch poll is conducted by mail and has not been very reliable in the past. Still, the notion that Ohio was somehow immune from Obama's recent bounce is rapidly losing credibility.