Three days does not an election make, but the pattern emerging in the Georgia Senate runoff is looking more and more ominous for Jim Martin and better and better for Saxby Chambliss. Why? Well, as the data below clearly suggests, women and African Americans -- two groups that typically favor Democratic candidates -- are just not turning out at levels that will reverse the preference ordering from the first round on November 4. If anything, the fact that women and blacks are turning out at lower levels suggests that Chambliss will do even better on December 2 than he did earlier.
But let's parse this out a bit. The one thing that cannot be stressed enough is that early voting trends don't necessarily portend how the ultimate election outcome will look. Back before the election, I made the case that the 2008 presidential election could play out in a way similar to the 2004 election, just over an extended period of time. If you'll recall, early exit polling on election day in 2004 showed that John Kerry was ahead and likely to be the next president. But it didn't turn out that way. As the actual returns came in, the outcome was the opposite of what the early exit polls had suggested. Earlier, I argued that early voting could cause a similar effect to play out, but instead of across just election day, across the last month or so of the campaign. In other words, the early voters could be Democratic (which they were throughout much of the nation) while election day voters were more Republican. Again, it didn't turn out that way...exactly.
The argument has been made that the reports on early voting and the direction is appeared to be heading may have pushed election day turnout down. The perceived direction of the count, then, caused voters to question whether their vote was necessary in achieving their desired outcome. In the presidential election, this was more likely to affect Democrats. The most extreme example of this phenomenon was what happened in Alaska on November 4. Early voting that favored Obama and other down-ballot Democrats in the state when coupled with the 9:30pm (EST) reality in the presidential race, made it much more difficult for some Alaskans (who were getting off work around that time -- four hour time difference) to justify turning out to vote. This is part of the reason it looked -- at least on election night -- as if Ted Stevens had been reelected to his Senate seat. Those election day voters were more Republican. It wasn't until all the early votes and provisional ballots were counted that that changed.
So why am I taking this discussion to the Last Frontier and to the national level? Again, early voting may not necessarily help us to determine the ultimate outcome of the race. But here is why early voting is likely to be a fairly strong cue as to what will happen on December 2: enthusiasm. The enthusiasm that caused so many to turn out to vote early for the November 4 election doesn't seem to be there in this runoff. Now, we would expect turnout to be lower in the runoff, but it isn't proportionally lower across the various demographic groups listed in the data from the Karen Handel's office here in Georgia. It is lower among the groups that would be expected to help Jim Martin.
But couldn't we see an opposite effect from what we saw in Alaska? In other words, Republicans see that the early voting is going well for Chambliss and don't come to the polls on election day. Meanwhile Democrats, knowing they are behind, are motivated to come out to their polling places on December 2 to vote for Jim Martin.
That is possible, but it is not as likely as it would be if Georgia were not as conservatively tilted as it. If the Peach state was more competitive between the parties, then I'd be more inclined to listen to that argument. [Yes, Alaska is just as conservative as Georgia, if not more so, but it is a different kind of conservative, shaped by a completely different set of circumstances.] And this certainly works with the enthusiasm angle posited above. Georgia Republicans are on the defensive while Democrats in the state already have something they wanted out of this election: an Obama victory. Sure the talk after November 4 was that Republicans would be depressed because of what had happened cumulatively on election day, but it may be that Democrats are too elated to care instead.
An Obama appearance is still the wildcard here. But what would that signal? That he cares about the 60 seats potential in the Senate (Why, when he's going to be working across party lines anyway?) or that Martin is in danger of falling short in his effort to unseat Chambliss (Is that a good way to expend this political capital everyone is talking about?). If Obama came to Georgia to stump for Martin and Martin still lost, it likely wouldn't look good for the president-elect. And you don't want to look bad before you are even president.
As promised, below is a map that shows the different dates on which the counties throughout Georgia started (or will start) the early voting process. Today another five counties kick off early voting with five more to follow tomorrow. The counties in light purple will have advance voting starting next Monday and those in white have yet to inform the Secretary of State's office of their plans (Of course, it is also possible that the counties have shared that information with SOS but SOS hasn't updated their online information yet.)
That sounds like a hit on the good folks at the Secretary of State's office. It isn't one. The folks in Atlanta have been very helpful to me in putting the data I've been using the last few days together. And they have also been good about updating the early voting totals. So a heart-felt thank you is extended to them.
Georgia Senate Runoff: Early Voting (Day 2)
Georgia Senate Runoff: Early Voting (Day 1)
Hillary Clinton vs. John McCain