John McCain 50
Barack Obama 48
Bob Barr 2
Barack Obama still has a decent chance of pulling off the upset in Georgia. He's leading 52-47 among voters who have already cast their ballots, which accounts for 57% of those surveyed. John McCain is up 54-43 with those planning to vote tomorrow.
The demographics of early voters and of those who are election day voters are quite different. 35% of early voters were black, 56% were women, and when it comes to party identification 46% describe themselves as Democrats while 40% are Republicans.
For likely voters who have not yet gone to the polls just 24% are black, 42% are Republicans compared to 38% who are Democrats, and a slight majority are men.
Obama probably needs to crack 30% of the white vote to take Georgia and right now he is falling just short, trailing McCain 70-28 within that demographic. Obama's solid lead with young voters is being offset by McCain's overwhelming lead with voters over 65.
In the state's US Senate race a runoff still seems like the most likely outcome with Saxby Chambliss receiving 48% to Jim Martin's 46%. Libertarian Allen Buckley is the guy who could cause the campaign to continue another month, receiving 4% of the vote. Chambliss would have to win over the late deciders overwhelmingly to get a majority tomorrow.
Full results here.
Subscribe to:
Post Comments (Atom)
18 comments:
Holy Cow!! Turn Georgia Blue!
Wow. That's damned close. Been waiting for Georgia to turn towards Obama all year, and it finally seems to have happened starting two weeks ago. No idea if he'll pull it off, but this is amazing.
If Obama wins 28% of the white vote, then I think he wins. There will be more AA voters in this election than in 2004, hopefully over 30%. And this number bodes well for Florida, as well. Would love to see Obama win the entire eastern seaboard, save SC.
Obama is likely to fall just short here. I wonder what would have happened if Obama put in ad money (I know he put in significant organizing nmoney) in GA?
Blue Georgia in 2012.
Hey ant,
Late last week on a conference call, David Plouffe said they were going to buy ads in GA. So, my guess is they may have started running ads in GA this past weekend.
How about Jim Martin? Does he have a decent shot of kicking Saxby Chambliss out of the Senate?
Any findings on whether either campaign has been successful on getting "unlikely" voters to the polls in early voting? or where they all people who would have likely shown up on election day?
I'm Atlanta and I saw a couple Obama ads this weekend. It would be nice to see him pull it out here.
Does this mean it's almost certain that blacks will not come close to 30 percent of the voters on election day?
The Gender cross tab is peculiar. Both M and W go 50-48 McC-O. Other available cross tabs in this and other southern states show a distinct gender bias. Gender could be equalized if higher than expected AA Men vote.
My back calculation shows the early vote for Obama (fractions of total cast)==
.317 Black
.190 White
.012 Other
This means Obama has been polling 31.5% of Early White votes. As early voting includes liberal whites this is possible.
If the overall vote holds, his Election Day white share must drop to 0.233 in order to "solve" the overall 28% prediction found in the poll sample.
Obama only gets 91% of the black vote? I've always thought it would close to 95%.
antmatic Obama was up on the airwaves in GA for the entire summer into September. He went off the air for a couple of weeks in late Sept to mid October and he is now back on the air there.
antmatic Obama was up on the airwaves in GA for the entire summer into September. He went off the air for a couple of weeks in late Sept to mid October and he is now back on the air there.
Are there more states coming?
"Obama only gets 91% of the black vote? I've always thought it would close to 95%."
This is where I think the reverse Bradley Effect might come in.
SUSA has a new Georgia poll. It has very similar results for Early voting.
SUSA predicts an overall 26% Black turnout. However, early voting is about 35% black, so a residual 44% election day share of the total electorate would need to drop to 14.5% Black to erode the overall AA turnout.
SUSA shows a 57 to 39 McCain edge in the "not-yet-voted" cohort. Getting to 39 with only a 14% black share would require higher than sampled white Obama vote.
Several alternative interpreations of the inconsistency in the SUSA cross tabs can be presented. Over-reporting by respondents of white early would account for some of the descrepancy.
Is there some data to support the supposition that voters over-report their participation and this process happens preferrentially to some classes of candidates.
The GA Sec of State website is reporting over 2 million have voted early, but only 57 of the 1,253 likely voters surveyed reported voting early.
How can that be? Shouldn't the number of people reporting they already voted to you be much higher?
Nevermind - I see it's 57%
My bad (My eyes are blurry from reading all these poll numbers)
Post a Comment