GA: InsiderAdvantage/Fox5: Perdue +10

[Adam Griffin]

Perdue 50
Nunn 40

MoE +/- 2.9%


Like I said for the Gubernatorial one, these crosstabs look funny. They actually had to take them down earlier today after publishing them, then threw up an altered version a couple of hours later.

[Adam Griffin]

Insider Advantage is garbage, throw in trash.

I've always maintained this (even if I may take their D-leaning polls a bit more seriously, given how they are usually ridiculously R-leaning). For the snafus in this particular poll, here's the jist from PP and IA:

You must be logged in to read this quote.

[Adam Griffin]

You must be logged in to read this quote.

So a InsiderAdvantage poll has a 33% black sample with Perdue winning by 10, while the AJC has a 24% black sample that shows Perdue ahead by 4?

[Adam Griffin]

You know, for a state that gets polled constantly, you'd think we'd have a clearer idea of what's happening here. At least with Kentucky we know that McConnell is leading by exactly 4%.

If you throw out the garbage polls produced by InsiderAdvantage & YouGov, the polling isn't that wacky, but certainly not as consistent as KY's. A fairly wide fluctuation, but most polls are within MoE to suggest that it's statistically tied.

I honestly don't think this poll should even be entered into the database, considering all the hoopla and mistakes IA made in releasing it (you can also inspect the comments on PP).

EDIT: Maybe I forgot about how many polls are truly all over the place in this race. It is pretty bad.

[Adam Griffin]

I don't believe there will be such a huge drop-off this time, if only because we're not dealing with a landmark election year with the highest black turnout in history and historically unstable electoral performance. The people who vote in mid-terms are much more likely to come back out to vote in a run-off when compared to a presidential electorate. I'd expect a 49-46 Nunn lead in November to become a 53-47 Perdue victory in January (although the fact that there'll be two run-offs - one for Gov and one for Sen - could throw a wrench into that).

The national party also won't completely forego Georgia in a 2014 runoff like they did in 2008 (which in hindsight, turned out to be a really stupid decision, no?).

[Adam Griffin]

I don't believe there will be such a huge drop-off this time, if only because we're not dealing with a landmark election year with the highest black turnout in history and historically unstable electoral performance. The people who vote in mid-terms are much more likely to come back out to vote in a run-off when compared to a presidential electorate. I'd expect a 49-46 Nunn lead in November to become a 53-47 Perdue victory in January (although the fact that there'll be two run-offs - one for Gov and one for Sen - could throw a wrench into that).

The national party also won't completely forego Georgia in a 2014 runoff like they did in 2008 (which in hindsight, turned out to be a really stupid decision, no?).

No. Bill Clinton and Al Gore did campaign for Jim Martin, and didn't do any good. Obama would have wasted political capital by investing in a losing Senate race 3 weeks after becoming President-elect.

It wouldn't have cost him any political capital - it was Georgia, after all. The DSCC did spend - I suppose I shouldn't have been so hyperbolic - but they spent as much in the final week of October on Georgia as it did during the entire run-off. The President that put Jim Martin so close to winning in the first place was who was needed to turn those voters back out; Bill Clinton and Al Gore didn't get Jim Martin to within three points of Chambliss.

[Adam Griffin]

Someone please add to database (even though this one is garbage).