One of the major stories in the end-game of the 2012 presidential elections has been the late decision by the Romney campaign to “play” (via both paid advertising and significant candidate resources) in Pennsylvania, a state that always seems to tempt Republicans in presidential elections while going routinely Democratic (at least since 1988).
Some over-enthusiastic (or spin-mad) Republicans initially interpreted the move into Pennsylvania as portending The Romney Landslide: Mitt was so confident of victory in what we all assumed to be the “battleground states” that he was now able to “expand the map” and threaten “enemy turf,” much as Obama did in 2008 when he did some last-minute campaigning in Indiana.
A more common interpretation was that Team Mitt was performing a “feint:” with money to burn, why not put Pennsylvania into play in hopes of diverting Democratic resources into a state outside the true “battlegrounds,” improving GOP odds in those? This is reportedly what the McCain campaign was trying to do in 2008 with a similar maneuver.
But I think the most reasonable explanation is that Romney is serious about trying to win Pennsylvania, but not because he’s so confident of victory; au contraire, he’s seeking an unorthodox and very narrow path to 270 electoral votes, and the Keystone State makes sense in that context.
Nate Silver lays out a scenario where Romney gets 273 electoral votes despite losing Ohio, Iowa, Nevada and New Hampshire (and also Wisconsin, the other state where the Romney campaign is making a late bid):
Mr. Romney could not afford to lose Virginia, where he is narrowly behind in the polling average, or Florida, where he is narrowly ahead. He could also not afford to lose Colorado, unless he won New Hampshire.
But Florida, Virginia and Colorado are traditionally slightly Republican-leaning. If Mr. Romney overperforms his state polls across the board — something he will need to do anyway to win the election — they may come along for the ride.
Add in PA to the “red states” plus FL, VA and CO, and you get 273 electoral votes. And then there’s the turnout wild card: Pennsylvania has limited early voting, and it’s reasonable to think Sandy could have a greater effect on GOTV efforts in eastern than in Western Pennsylvania. So with a higher percentage of the electorate voting on November 5 than in OH, IA, FL, NV and CO, and with southwest Pennsylvania trending red in recent cycles, there might be a way to eke out an upset–again, assuming the national dynamics don’t doom any road to 270 for Romney.
Calling the Pennsylvania gambit rational doesn’t mean it’s plausible, much less likely to succeed. The RCP polling average for PA gives Obama a four-point lead in the state, and that’s with three of the seven relevant polls being from firms with a significant history of a pro-Republican “house effect” (Susquehanna, Gravis and Rasmussen). The Romney invasion of Pennsylvania could well turn out to be like Lee’s in 1863: a desperate gamble by an army on the brink of defeat. We should know relatively early tomorrow night if it has any chance of working.