“Yet Republican internal polls are showing a similar slight advantage for Obama from 2 to 4 percent”
That’s strange too, for their internals to show a similar result to this poll.
This poll has a D/R/I sample of 50/37/11 for a D+13
Actual exit polling in 2008 and 2010 for Pennsylvania:
2008 44/37/18 D+7
2010 40/37/23 D+3
2008 was a D blowout
2010 was a R blowout
Splitting the difference, wouldn’t a sensible model this year be D+5?
Yes, it would. All I can tell you is I have two good friends in Republican politics here in PA and they're both telling me that Obama will probably win Pennsylvania. One of them tends to be a pessimist, but the other is an optimist if anything and they're both saying the same thing re: internal polling.