Saxby Chambliss Holding, But in Trouble

We have gotten three new polls of the Georgia Senate race between incumbent Republican Saxby Chambliss and Democratic challenger Jim Martin which basically put the race exactly where it has been for the last month: Chambliss is clinging to a two-point lead, but both men are mired in the mid-40s with just 11 days until November 4. Let's quickly go through all three.

Poll #1: Rasmussen (conducted 10/22)

Saxby Chambliss (R) 47 (50)
Jim Marin (D) 45 (44)
Allen Buckley (I) 1 (?)
Undecided 7

The numbers in parenthesis represent Rasmussen's findings from its last poll from October 7. Rasmussen further found that what was a 15-point Chambliss lead with men is down to six points.

Perhaps more importantly, Chambliss' favorable/unfavorable split currently sits at 51/43 (+6), down from 57/37 (+20) just over two weeks, a drop of an aggregate 14 points. Martin's split is now 48/43 (+5), and was 45/38 (+7) in the last poll. Therefore, the challenger' personal numbers are about the same.

Poll #2: Strategic Vision (10/20-22)

Saxby Chambliss (R) 46 (47)
Jim Marin (D) 44 (44)
Allen Buckley (I) 5 (4)
Undecided 5 (5)

The numbers in parenthesis represent Strategic Vision's findings in their last poll which was taken October 5-7. Remember, this is a Republican firm.

Another two-point lead for Chambliss. Here, however, the Libertarian Party candidate gets five percent, and I think this is right, or at least closer to what Buckley will get than the one percent Rasmussen finds for him. The one percent found in the Rasmussen poll seems a bit low. Strategic Vision found Chambliss' approval split at 48/40 (+8), basically the same as the 49/41 (+8) number found in the last poll.

This poll seems suggest an equally tight race, but it finds much less movement from its last poll a few weeks ago. I think the reason for this is that Rasmussen last poll, which put Chambliss up six, was kind of an outlier, and this month's version has the race where it should be.

Poll #3: Insider Advantage (10/23)

Saxby Chambliss 44 (45)
Jim Martin 42 (45)
Allen Buckley 2 (2)
Undecided 12 (8)

The numbers in parenthesis are from IA's 10/9 poll of this race. It is important to note that Insider Advantage is based in Georgia, so the firm may have much better experience in polling the state than other reputable national outfits.

Maybe Buckley really is around 1-2 points in the polls. Nevertheless, interestingly, Insider Advantage, while also finding a two-point contest, finds a lot more undecided voters than any of the other recent polls. They provide no cross-tabs, but Matt Towery a GOP columnist, former aide to Newt Gingrich and Georgia State Represenative, and current political analyst provides some Georgia-centric thoughts:

“The senate race appears in my judgment either headed for a runoff or a Martin win, and here’s why I say that: over 15 percent of blacks say they still are undecided. The largest undecided segment among party affiliations are Democrats, who mirror that black percentage – 14.5 percent – and independents at 18 percent, who are leaning toward Martin by a margin of 45-33 percent. Generally speaking, at this point in a race, unless something were to turn it around, we would treat the undecideds two ways: we would either lop it off and redistribute it, or we would assign it based on the relative positions of the candidates as they stand today. Under either of those scenarios, Chambliss and Martin would be in a runoff.”

Conclusion

Certainly no Democrat, Towney's analysis appears largely correct to me. For the last three weeks or so, this race has been incredibly stable, both in its head-to-head numbers, and the lack of real movement. Check out the last ten polls of this contest over this period:

46-44 Chambliss (SUSA, 9/28-9/29)
45-44 Chambliss (R2K, 9/29-10/01)
47-44 Chambliss (Strategic Vision, 10/5-10/7)
50-44 Chambliss (Rasmussen, 10/7)
45-45 Tie (Insider Advantage, 10/9)
46-43 Chambliss (SUSA, 10/11-10/12)
47-45 Chambliss (R2K, 10/14-10/15)
46-44 Chambliss (Strategic Vision, 10/20-10/22)
47-45 Chambliss (Rasmussen, 10/22)
44-42 Chambliss (Insider Advantage, 10/23)

Isn't this remarkable? Averaged out we get

Chambliss 46.3
Martin 44

If subtract the results of the 10/7 Rasmussen poll, which I firmly believe is an outlier, we get these averages

Chambliss 45.9
Martin 44

Another razor-close finding. Martin has basically been stuck at his 44 or so percent for the last month, which is both good and bad. It is bad because it might indicate that he may be at or near his ceiling in such a red state. It is good because a lot of undecided voters just can't go firmly into the incumbent's camp, a bad sign for a sitting Senator.

As for me, I think these results are much more good for Martin than for Chambliss, even though the incumbent is ahead in nine of these polls, and tied in the last one. I have several reasons for my viewpoint.

First, the initial fear for Democrats was that as the headlines of the economic crisis and the bailout receded, Chambliss would regain his footing, and in turn his electoral strength. Clearly, that has not happened at all. While Martin has not been able to really break the 44 percent plateau, Chambliss too is stuck in the mud and vitally, several points below the magic 50 percent mark. This race is tight to stay.

Second, Chambliss might actually be losing his footing. While Rasmussen's latest poll has the race at 47-45 in sort of correction of its outlier poll, it finds that Chambliss' personal split is a mediocre 48/43. Strategic Vision reveals a 48/40 split. Chambliss is not gaining back favorability numbers, and that's bad for an incumbent.

Third, Obama is actually getting stronger in Georgia. Let's look at the presidential contest in Georgia in those same polls we summarized above

52-44 McCain (SUSA, 9/28-9/29)
51-43 McCain (SUSA, 10/11-10/12)
From +8 to +8

54-45 McCain (Rasmussen, 10/7)
51-46 McCain (Rasmussen, 10/22)
From +9 to +5

50-43 McCain (R2K, 9/29-10/01)
49-43 McCain (R2K, 10/14-10/15)
From +7 to +6

50-43 McCain (Strategic Vision, 10/5-10/7)
51-45 McCain (Strategic Vision, 10/20-10/22)
From +7 to +6

49-46 McCain (Insider Advantage, 10/9)
48-47 Obama (Insider Advantage, 10/23)
From +3 to -1

The movement here is a little less perceptible than in the Senate contest, but we see very slight movement in Barack Obama's direction. Insider Advantage in particular showed Obama now ahead by a single point, though IA has had the best numbers for Obama of the five pollsters.

Regardless of where the presidential race is exactly, if Obama is gaining strength in Georgia -- and all indications are that he is getting strongly nationally -- this can only help Jim Martin.

Fourth, and perhaps most importantly, regards the black vote. In the analysis by Matt Towery of the latest IA poll, he notes that the polling found a large 15 percent or so of black voters are still undecided. This is kind of fascinating, in that it mirrors the trend we have been discussing in Mississippi, where a large number of black voters have remained undecided all cycle in the R2K polls, with 12 percent declaring themselves undecided in both the presidential contest in the state and the Wicker-Musgrove Senate grudge-match in the latest poll from last week.

Because most of the other polls do not provide cross-tabs, we can't check to see if all the other pollsters are finding the same split. The only one we can peruse, R2K, found in its last poll that blacks were six percent undecided in the Senate race, and four percent undecided in the presidential. It's hard to say who's right.

If IA's data is accurate, and there are so many undecided African American voters, Jim Martin is in the driver's seat. These are voters who will break heavily for both Barack Obama, and in turn for him. If this poll is right, Martin is in strong shape, as it would mean that Martin's base is either undecided (or at least not telling the pollsters what it will do, while Chambliss' base might be almost all set). Indeed, as we have noted several times, black turnout in the early voting stages is very high -- today around 36 percent overall, while the state's black population is around 29 percent. This could thus bode very poorly for Senator Chambliss.

A big question will be whether either Chambliss or Martin can clear 50 percent, because if not, there will be run-off in December. The question of who a run-off would favor is a whole other question I will write about this weekend.