SurveyUSA (9/28-29, likely voters, 9/14-16):
Jim Martin (D): 44 (36)
Saxby Chambliss (R-inc): 46 (53)
Allen Buckley (L): 5 (8)
(MoE: ±3.8%)
When we saw this internal poll for the DSCC yesterday that showed Chambliss leading by only three, many of us were rightly skeptical. Now SurveyUSA is out with a poll confirming a tight race.
I took a quick glance at the crosstabs and couldn’t find any significant demographic differences between this poll and SUSA’s last one (in terms of the sample’s composition), but it’s worth noting that the same poll finds McCain ahead of Obama by 52-44 — a much more respectable margin than the 57-41 blowout of two weeks ago.
SSP currently rates this race as Likely Republican, but it definitely is worth watching.
Update: Here’s one tidbit that I missed — of those who have already voted (9% of likely voters), Martin leads Chambliss by 61-36 an Obama leads McCain by 64-35.
Oh DSCC please put some money into this race. Imagine winning a Senate seat in GEORGIA!
I can’t think of any other reason for why all of these southern Senate races are swinging our way so much.
The Libertarian is polling at 5%, and he won’t get anywhere near that much on election day; it’ll be more like 1%, and most of the rest of them will probably end up voting for Chambliss.
Still, I really like this trend.
Both Obama and Martin are doing better with men than they are with women; that’s the reverse of how it usually goes for Democrats.
Can Martin get himself about 10% more of the white vote? If so, he can win this.
I would not like to be in charge of spending at the NRCC now.
– African-Americans are probably undersampled – 26% instead of the actual number 29-30%; 26% may be reasonable for previous elections, but I’m guessing it will be 29-30% now.
– As has been mentioned, male/female crosstabs are weird. Females are also slightly undersampled (in 2004, was a 44/56 split)
– One thing that I’ve noticed is that all SUSA polls seem to be undersampling low-income voters; this is probably due to their likely voters screen, but I haven’t seen a single SUSA poll that had similar or higher low-income turnout than in 2004, which seems a bit suspicious considering the massive GOTV drives. Georgia’s median income is $43k, and in 2004, 42% of voters made less than $50,000/yr. SUSA has the number at 32%, which is a small factor as low-income voters tend to vote Democratic. If we adjust it to the 2004 numbers, this is a 1-point race (Chambliss-46, Martin-45.) Not a very large difference, but significant in a close race.
Was Chambliss not exactly loved in GA to begin with?
I suspect that the gas shortages in Atlanta are hurting Chambliss’s poll numbers.
As I said yesterday. You don’t get these kind of chances often, in the deep South.
When are we going to see new numbers for MS and ME??
End of Quarter Jim Martin Money Bomb
Can we do $500 tonight?
I understand why everyone has written off the Cornyn-Noriega race, but it stands to reason that if Chambliss is in trouble then so might Cornyn.
The early voting news is incredibly good (and surprising!), particularly if Obama’s lead widens such that overall turnout is depressed. I think this news is incontrovertible evidence that Obama’s ground game is working, and it’s even working in states where he hasn’t gone all in (like Georgia).