Wednesday, August 6, 2008

Crossover Voters

One thing that could be a problem for both Barack Obama and Pat McCrory as they try to succeed in North Carolina this year is folks in their parties crossing party lines to vote for their opponents.

Typically Republican candidates get a higher percentage of the vote from folks who identify with their party than Democratic candidates do- and that's a virtual necessity in a state where Democrats have a significant registration advantage. But our most recent poll found McCrory leading just 72-12 with Republicans while Perdue leads 76-12 with Democrats. It's not a big difference but still one that makes it close to impossible for McCrory to succeed unless he changes it.

So who are these Republicans for Perdue? They're disproportionately middle aged, urban women. In addition to supporting Perdue they're also going for Beth Wood, June Atkinson, and Elaine Marshall. At the same time they support Elizabeth Dole over Kay Hagan and Cherie Berry over Mary Fant Donnan overwhelmingly. So it appears this crop of Republican voters is choosing gender first, then using party as the tie breaker. It certainly helps to explain the success of female candidates in the state.

How about the Democrats for McCain? They don't fit the profile of former Hillary supporters, the way crossover voters have in other states like Florida and Michigan. They are white men, disproportionately senior citizens. Obama's race is probably not the issue for them because while they identify as Democrats they are voting Republican in almost every race with the exception of their support for Walter Dalton, Roy Cooper, and Elaine Marshall. These are the folks who are registered 'D' either because they have been since the 1960s and 70s, or because their parents were before them but don't tend to actually vote Democratic.

Folks often wonder why a state with a 12 point Democratic registration advantage isn't more supportive of Democratic candidates, particularly at the federal level. Those folks are the reason why.

No comments:

 
Web Statistics