And then it just comes down to how well Team Obama has Philadelphia County organized. If he can drive turnout there as he did in 2008, netting 470,000 votes, the state will go his way, albeit narrowly. If his turnout machine is lacking a little bit – due to diminished grassroots enthusiasm, lack of preparation by Team Obama, or just the inevitable drift of marginal Republican voters in the county to the GOP banner – Obama will be in trouble. After all, John Kerry broke all the records in Philadelphia County in 2004 by netting about 410,000 votes, then Obama smashed them to reach 480,000 even as the county’s population was stagnant. If Obama “only” has the county as well organized as Kerry did, then in this scenario Romney would win the Keystone State, albeit it narrowly.
You can tweak various assumptions here in reasonable ways to see the state flip as well. For instance, assume Romney does as well as Toomey did in the rest of the state, where turnout is still up 4 points, then Obama will have to exceed his 2008 haul from Philadelphia. Obviously, tweaks can go in the other direction to yield an Obama hold on the state as well. The point is that Team Romney has a reasonable shot here, one worth an investment, especially given the level of saturation in the other swing states.
Final point: I wonder if Romney has caught Team Obama flat-footed. For months the Obama campaign has bet that the map would basically look like it did in 2008, then after the debates there was a definite shift. Now, the president is left fighting not in the 2008 battlegrounds, but in the 2004 battlegrounds, which included Pennsylvania. I do not think the president’s campaign was fully prepared for this, and I bet he wishes he could have all the money poured into North Carolina back, to redirect it into Pennsylvania.