http://www.dailykos.com/story/2006/9/22/ 133046/377

I think its a good diary I pointed out showing that:

Gov. Rendell has a huge 20-25pt lead, and his machine WILL turn out the Philadelphia area (and suburbs) in huge numbers. This race is great all around because the candidate, Swann, is seen as a joke to many conservatives and moderate Republicans, they are not going to be driven to the polls to vote for this guy. However, Rendell will guarantee that the Solid base Democrats will get out and vote.

Dem Senate Nominee Bob Casey has a GOLDEN family name, and has a 10-15pt lead over incumbent Rick Santorum. Normally, Santorum's conservative base would come out in droves for him, however a significant chunk of that base are older voters who very fondly remember Bob Casey Sr. with his staunch pro life views that his son carries very deeply. Casey will win and with him he'll draw out many "Reagan Democrats" and indepedents that Rendell wouldn't drive to the polls on his own. That's two big races drawing out the entire spectrum of the Democratic Party.

Santorum ended this race as soon as he tried to use casey's father against him on Mtp.

As a former Green in PA, I remember well that Nader was consistently at 5% in the 2000 polls in PA and 7% before being kicked off the ballot in 2004. (See http://www.fandm.edu/x2336.xml for 2000 and http://www.swingstateproject.com/2004/03 /nader_hurting_k.php for 2004.)

Romanelli will get 1% or less when people actually go to the booths. The other 7% represented in this poll will go to Casey.

This particular poll did oversample Republicans and also the "T" and the NW, but the Keystone Poll always puts all its numbers out there from each successive poll in their marginal frequency report.
(If you've never read this part of the poll, it's great stuff.) Anyway, you can judge each sample against every other one. Also, the Keystone Poll does weight its result for demographic balance.

The interesting thing is, though, that when the sample was more Democratic, Casey got nearly the same numbers. It tells me there must be a lot of Republicans in Pennsy who are dissatified with
Santorum one way or the other.

I agree that if Romanelli remains on the ballot - and the judge is dragging the case out long enough that absentee ballots will probably go out with his name on them - he won't draw 5%, maybe 2-3% tops and it will come from BOTH candidates. I say this because of the very strong anti-incumbent sentiment in the SW and the south central parts of the state. These voters, who would normally be probable Santorum votes, are likely to vote for Romanelli purely because he's an outsider.