Wednesday, October 22, 2008

Swing Voters in North Carolina

We're now down to 4% of voters in North Carolina undecided and 7% of those who do have a preference open to changing their mind. That means the campaigns are spending a whole heck of a lot of money to try to win over 10-11% of the electorate. Who are these folks that still don't have their mind completely made up?

Solid Support

Both John McCain and Barack Obama have extremely solid support, although there are a few more Obama supporters open to changing their mind. 96% of McCain's supporters say they are definitely going to vote for him, while 93% of Obama's do.

Looking at folks leaning toward Obama who might flip, they're folks voting on the economy who probably supported Bush in 2004 and aren't 100% comfortable casting a ballot for a candidate who might be more liberal than they are. 38% are small town voters, compared to just 27% of the population overall. John McCain is going to need to either convince those folks that he will be as good on the economy as Obama, or that they should decide who to vote for based on other criteria.

Although Bob Barr has dropped to the 1-2% range in our polls that remaining support is still pretty weak. Just 37% of respondents expressing a preference for him said that they would definitely vote for him. I think those folks are undecideds 'parking' with Barr, and that his actual final share of the vote will be less than 1%.

Who's persuadable?

-The voters in the middle come heavily from two types of places that supported George W. Bush in 2004 but have been moving Democratic this year: suburban and small town voters. 70% of not completely decided voters are coming from those places even though they only account for 58% of voters overall. What you're seeing there is folks who are unhappy with the Republican Party but aren't totally convinced the Democrats would do a better job, so they're not sure which way to go.

-It's actually a pretty young segment of the electorate. 29% are under 30 and 62% are under 45, compared to just 16% and 44% for those figures in the population as a whole. While the intensity of Barack Obama's support has certainly been much higher among young voters, it appears some still have their doubts about both candidates.

-They are heavily concentrated in eastern North Carolina- more than a third of them are there even though only about a quarter of total voters are. I think this is definitely your registered Democrats who usually vote Republican in the Presidential election trying to decide if the economy's bad enough that they can vote for a candidate they disagree with on a whole bevy of social issues.

Throw out your voters who aren't completely committed and Obama leads McCain 47-42 here with Bob Barr picking up 1%. So Obama only needs to win about a third of the rest of the voters to take the state. He should be able to pull that off if there's not a big shift in the trajectory of the campaign over the next couple weeks.

2 comments:

Tyler said...

CNN/Time has a poll with Obama up 3 in NC.

This thing is done. Even if the final polls show the state a tie, Obama's ground game and huge turnout among AA will turn that into a 2 point advantage.

Rasmus said...

"

Looking at folks leaning toward Obama who might flip, they're folks voting on the economy who probably supported Bush in 2004 and aren't 100% comfortable casting a ballot for a candidate who might be more liberal than they are. 38% are small town voters, compared to just 27% of the population overall. John McCain is going to need to either convince those folks that he will be as good on the economy as Obama, or that they should decide who to vote for based on other criteria."

You aren´t really trying to analyze a subset of 7% of Obama voters, are you? What was the sampling size? If it´s a typical PPP poll, it´s probably around 1200. Obama supporters are maybe 600, then the "might change" Obama supporters are 7% of 600, or 42 people. That´s a margin of error of +-15%!.

If they´re more suburban as the total sample is by chance really irrelevant.

 
Web Statistics