Pennsylvania represents an interesting case, with two polls this weekend showing a tight race. A Muhlenberg College poll conducted for the Morning Call of Allentown shows Obama ahead by 3 percentage points, 49 percent to 46 percent. Meanwhile, Obama and Romney are tied in a new poll conducted by the Republican pollster Susquehanna Research and Polling for the Pittsburgh Tribune Review, each with 47 percent. But that actually represents an improvement for Obama: The previous Susquehanna poll, conducted in mid-October for the Pennsylvania GOP state committee, showed Romney leading by 4 points; Susquehanna is the only pollster to show Romney ahead in the Keystone State in the past few months.
There are some battleground states in which no new reliable, independent, live-caller polls were released this weekend. Polls over the last week have showed the race is neck-and-neck in Colorado, North Carolina and Virginia (an NBC News/Wall Street Journal/Marist poll in that state scheduled for release early on Monday will show a dead heat in Virginia), while Obama is leading in Michigan, Minnesota, Nevada and Wisconsin.
The electoral math still favors Obama. If Michigan, Minnesota, Nevada and Wisconsin are added to the list of states safely in his camp, he is at 233 electoral votes. Add Iowa, Ohio and Pennsylvania to that list, and he could afford to lose Colorado, Florida, New Hampshire, North Carolina and Virginia, where the polls are roughly tied. In that scenario, Obama would still win with 277 electoral votes.
Republicans insist that the public polls are wrong, suggesting that pollsters are making assumptions about the electorate that more-or-less replicate the voters who cast ballots in 2008 and claiming that too many Democrats are being interviewed. But, in most cases, public pollsters make no assumptions. They interview a randomly selected group of voters -- either by randomly dialing phone numbers or selecting random names off voter lists. They make sure they start with a demographically representative sample, but few pollsters make adjustments to that sample to match historical trends with respect to party identification.
It is certainly possible that the polls are less accurate than they have been in the past. Achieving a true random sample of Americans by telephone is harder -- and more expensive -- than ever. In order for Romney to emerge as a winner in the Electoral College, it appears that these polls will have to be proven not just inaccurate but biased toward the incumbent.
Get the latest news and analysis delivered to your inbox. Sign up for National Journal's morning alert, Wake-Up Call, and afternoon newsletter, The Edge. Subscribe here.
Leave A Comment