As goes Pennsylvania, so goes the nation... Not exactly. When Buzz and I started OneVoteCounts, we had planned on doing a comprehensive, in depth, no stone unturned analysis of the Presidential race in Pennsylvania. We figured that Pennsylvania would be close (one or two percentage points,) and we also figured the electoral vote totals would be close, and thus Pennsylvania was a must win for Obama if he hoped to reach 270. As we approach single digits in the election day countdown, it appears that even if McCain wins Pennsylvania, he may still lose the election.
First, here's a little Pennsylvania Presidential election history. Since 1948, Pennsylvania has voted for the Democratic candidate 8 times and for the Republican 7 times. 11 out of 15 times, PA has gone with the winner. PA went with the losing candidate four times. They were Dewey in 48, Humphrey in 72, Gore in 00, and Kerry in 04. George H. W. Bush was the last Republican to win the state in 88.
Today's realclearpolitics (RCP) average in PA is 11.0 in Obama's favor (51.5 to 40.5). This number has been in the double digits for the last few weeks, despite McCain pouring millions of dollars into Pennsylvania. It should be noted that Obama is out spending McCain in PA, as he is everywhere else. What is troubling for those who believe McCain has a chance in PA is the fact that McCain hasn't lead in a statewide poll since April. Only two of the 12 polls this month showed Obama's lead under 10 points, and those leads were 7 and 8 points. There also is no indication of any significant narrowing of the gap by McCain. We do note the number has dropped from a high of 13 a week ago, but if that is a trend, McCain would have to continue that for another seven weeks and he only has 10 days to the election.
The problem McCain faces by drawing a line in the sand with Pennsylvania is that Pennsylvania may not matter to Obama. Obama has solid leads in all the states John Kerry carried in 2004. This puts Obama at 252 (including PA). Starting with the 252 number, Obama has an RCP average lead of 8.4 in New Mexico (5EV), 7.0 in Virginia (7EV), 6.6 in Ohio (20EV), and 5.4 in Colorado (9EV). This would put Obama at 293 (270 is the magic number). If McCain wins PA and all the other toss up states not listed above, Obama would still have 272 electoral votes.
Here's some more bad news for McCain. Five other states Florida (27EV), North Carolina (15EV), Nevada (5EV), Missouri (11EV), and Indiana (11EV) show Obama leading in the RCP average. That's another 69 electoral votes.
Polls are not 100% accurate, but even if we throw out half of them and just give those states to McCain, he still loses.
The Pennsylvania or die strategy is difficult for McCain. The more money and time he spends in Pennsylvania, the less time and money he has to spend elsewhere. Even if by some peculiar set of circumstances McCain does win PA, it just may not matter. Pennsylvania is the only Kerry 04 state (maybe New Hampshire (4EV)) where McCain is waging a battle. All the rest of the fights are on Bush 04 territory. Obama is bringing the battle to McCain home turf, and there are just too many states for McCain to defend with his limited budget. Obama can, has, and will continue to outspend McCain.
The bottom line is this. Obama wins PA. The revised OVC prediction has just been handed to me by Buzz. (I say revised, because Buzz, an ardent Bob Barr supporter, had fudged the original numbers to show Barr winning PA, with Obama and McCain coming 3rd and 4th behind Ralph Nader.) Gore won PA by 4.5% in 00, Kerry won by 2.4% in 04. The OVC revised prediction is 53.5 % for Obama, 45.0 % for McCain and 1.5% for the two other candidates.