LA Times/Bloomberg has Obama +12. Adding Nader and Barr pushes it to +15. Obama runs even with McCain among whites.
FiveThirtyEight.com has details. A commenter there suggests that the daily tracking polls from Rasmussen (and maybe Gallup as well?) assume fixed partisan proportions within the electorate, and reweight the results of their daily polls to make the sample match the assumed “true” proportions of Democratic and Republican identifiers. Neither LAT/Bloomberg nor Newsweek does that, and both show voters tilting strongly Democratic. That may explain the difference between Obama +6 (Rasmussen), Obama +3 (Gallup) and these much bigger margins.
Either approach is legiimate. But there’s lots of external evidence that the Republican brand this year has about the same cachet as the Edsel had; if Rasmussen is really sticking to a 10-point identificaiton margin for the Democrats, the Rasmussen results will tend to understate the true gap among candidates.
Footnote More than four out of five Obama voters, about 40% of the entire electorate, say they’re “enthusiastic” about voting for him; less than half of McCain voters say the same about their candidate. That’s a pretty big cult. Maybe we’re not crazy after all?
Second footnote Republican incumbent Senator Gordon Smith of Oregon has a TV spot up bragging about his work with Barack Obama, complete with two screengrabs from the Obama website showing the logo.
Update Andy Sabl points out that in a crisis year party ID, usually rather stable, may shift faster than pollsters or political scientists are wont to believe. And “Prup” points out that this year, unlike most years, the population that goes to the polls may be even more Democratic than the registered voter population now being sampled. Some Democratic-leaning demographics may turn out to vote in droves, and Democrats (regardless of demographics) are more energized than Republicans.