Whatever damage the perceived "softness" of the recovery is doing to Obama's numbers nationally over the weekend, it has not extended to Ohio, a state that contrary to Beltway wisdom about the dreary job prospects of White Working Class Voters (tm) in the manufacturing rust-belt, actually has a jobless rate significantly below the national average.
PPP has released a new poll (pdf) that shows Obama up 50-43 over Mitt Romney in the Buckeye State.
Like any poll, this one gets really interesting when you dive into the cross tabs.
This poll has Obama only getting about 75% of the African American vote to Romney's 17% (remainder undecided). It also has the black vote pegged at 11% of the sample (in line with 2008 exit polls). The New York Times' Nate Silver has warned against over-interpreting demographic cross-tabs because of the small sample sizes. I think that's fair, and thus its fair to extrapolate that Obama will probably replicate his 2008 performance amongst Black voters (97%).
If he does so, he would pick up at least another 2 percentage points, making his margin that much stronger.
Put another way, if PPP is accurate, Obama is beating Mitt Romney 47-46 among white voters. if Obama gets 47% of the white vote in Ohio on election day, I'm not sure he can lose.
A lot of that white margin is going to have to come from voters who didn't turn out in 2010. Unmarried women, college students, and voters under the age of 30 remain strong demographics for him. Excitement is obviously down from 2008, so turning these voters out in in 2012 is going to be about the strength of field operations. While some might think that Rasmussen tinkers with his polls to drive Republican media narratives, the biggest difference between him and PPP is that the former users a likely voter model and PPP polls registered voters. Winning or losing this election will come down to ensuring that the electorate looks more like PPP's data than Rasmussen's.