Tuesday, October 14, 2008

Obama maintains small NC lead

Barack Obama 49
John McCain 46
Bob Barr 1

Barack Obama continues to hold a small lead in North Carolina, although it is down from six points last week.

The McCain campaign has just recently started to really contest the race here, with the candidate making his first appearance in the state in five months yesterday. He has also ratcheted up his advertising, which seems to be having a mixed effect on his poll numbers. Independents, who may be turned off by the negativity of his campaign, have moved more toward Obama in the last week. What was a 46-40 lead with them is now a 52-37 advantage for the Democrat.

They may be helping on some fronts though. There has been six points of movement in McCain's direction among Democratic voters since the last poll, with what was an 82-15 advantage for Obama now down to 79-18. Obama's standing in the state had improved with each passing week as more conservative Democrats who sometimes choose Republicans for Preisdent had moved toward his camp. With McCain's campaign now trying even harder than usual to paint Obama as an extremist that may be having the effect of helping to get more of those folks to cross over to the Republican candidate.

Obama's unusually strong standing here continues to be fueled by voters in suburbia. They voted for George W. Bush in 2004 but Obama leads with them 57-39 this time around. That's the fastest growing group of the electorate in North Carolina, and if they continue to lean Democratic the state is going to be competitive at the national level for cycles to come.

Democrats had hoped that Bob Barr might play the role of spoiler here and take enough support away from John McCain that Obama could win the state with just 48 or 49% of the vote, but the Libertarian candidate has pretty much ceased to be a factor here.

The economy is polling as the top issue for a new record high of North Carolina voters: 65%. Obama leads 59-36 with those voters.

Full results here


Anonymous said...

Tom - North Carolina has early voting. Did you encounter any voters that said they had voted early, and who were they supporting?

p smith said...

Antmatic beat me to it.

SUSA's poll in NC last week indicated that 5% of respondents had already voted and that they were going for Obama 2 to 1.

As we get closer to polling day, the early voting figures will become more and more relevant and I do hope that you have plans to question respondents as to whether they have voted.

Anonymous said...

Something seems strange. If I'm reading PPPs cross tabs right, you have McCain ahead in the research triangle? That just does not sound right.

Tom Jensen said...

Guess you're not reading it right. Obama is up 65-32 in that region this week.

Anonymous said...

Okay I stand corrected.

Anonymous said...

Thanks Tom. When will the Missouri poll be out?

Anonymous said...

In short here are the changes from PPPs last poll on 10/6: Huge swings to McCain in Western NC, Southern NC, and the coast; things pretty stable in Winston and Charlotte, and huge swing to Obama in research triangle area (Raliegh).

Chris said...

re: Early voting

The SUSA poll on that last week was laughable.

5% said they had already voted, but less than 3% of registered voters had requested absentee ballots. So, we know already some people are lying.

Also, something like 20% of those who said they had already voted were "undecided" on the race for Governor.
So, you say you've voted already, but you're not sure which candidate you're going to vote for?
Ok, some more people are lying.

PLUS... the margin of error on a sample size of 5% of a 500 person poll (25 people) is something around +/- 30%. So to say Obama is winning early voters based on a sample of 25 people who we know some are lying is quite a stretch.

Anonymous said...

Chris -

I agree that making projections on such a small portion of a sample (5%) is quite silly.

I just want to point out that there is nothing to suggest that "people are lying". It's entirely possible that SUSA's sample simply included a larger proportion of early voters than in the population at large. This happens all the time, right? The proportion of men, women, AA's etc. all vary somewhat from poll to poll.

I mean, if the census says that a state is 52% women, but our sample has 54% women, would that mean that some people are lying about their sex?!

Web Statistics