GA-Landmark Communications/Rosetta Stone: Romney+11

[Adam Griffin]

Garbage. Did they mix up our numbers with Alabama's? Also, was this a robo-survey like the one you linked to above? As Tender pointed out, Landmark is absolutely terrible at polling within Georgia. Cope's right: the floor here - barring a total collapse of AA turnout - is 45% now. Also, AA turnout is pretty solid as a % of the population in both presidential and mid-term years (mainly because all statewide elections take place in mid-term years).

Worst-case scenario turnout in 2012:

White (64%) @ 20% for Obama: 12.8%
Black (28%) @ 95% for Obama: 26.6%
Other (8%) @ 60% for Obama: 4.8%

Obama: 44.2%, and that's with white support dropping to unseen levels in GA.

Best-case scenario turnout in 2012:

White (60%) @ 25% for Obama: 15.0%
Black (32%) @ 95% for Obama: 30.4%
Other (8%) @ 60% for Obama: 4.8%

Obama: 50.2%.

I'd say somewhere in the middle - with the white vote being around 62% and the AA vote being around 30% - is the most likely, giving Obama around 47%. The demographic gains in the state will be cancelled out by mild drops in turnout and some exurbanites switching back to R, leaving him where he was in 2008. Come 2016, however, this state will go Democratic (as long as the nominee is white).

[Adam Griffin]

And I also want to predict that Cobb and Gwinnett counties will be unbelievably close this year. Whoever wins will do it by 1-2 points. Conservative suburbanites are fleeing Cobb and Gwinnett and being replaced by minorities, immigrants, and left leaning yuppies (especially in the more urban areas of Cobb).

Cobb is becoming more like Fairfax county, VA and Gwinnett is becoming more like Middlesex County, NJ - in terms of demographics

What about the northern part of Fulton County around Alpharetta?

Demographically, it's similar to a lot of the democratic leaning upscale suburbs up north, but it's still pretty strongly Republican. It has a lot of Jewish people, for instance, but most I've spoken to about politics are also Republicans. There has been some sort of a shift among the white residents up there, so the vote isn't as uniformly GOP as it is elsewhere in the state, but they still skew strongly Republican.

The Asian population is BOOMING up there, so when they become politically active, the area could swing to the Democrats.

North Fulton has a long way to go before it becomes Democratic. While much more diverse than the surrounding exurbs of Cherokee and Forsyth, it's less Democratic than areas of west Gwinnett like Norcross. North Fulton has a promising future with about 35% minority at the moment (12/12/12, AA/L/AS), but there's a decent amount of the Latino community there that is ineligible to vote and highly entrenched WASPs that are not leaving nor are they voting Democratic.

I tend to think that the vast majority of northern expansion for Democrats in the Atlanta area has manifested and is hitting the proverbial brick wall known as Cherokee, Forsyth and Hall. All three of these counties are near 80% Republican and have extremely high voter registration rates. I do think Cobb and Gwinnett will both flip within the next few years, but it will be primarily from added growth and diversification in the southern parts of these counties. Even though growth in Forsyth, Cherokee and Hall over the past 10-20 years has been phenomenal, it's mainly growth among demographics that favor Republicans as they flee Atlanta.

The Atlanta Democratic machine will find expansion to the south, east and west (Douglas, Henry, Spalding, Rockdale, Newton) to be the most advantageous in the coming years.

Do you think the secession plans of the northern part of the county, to recreate Milton County, will be successful?

I hope not. They've been yakking about it for years to no real avail. We already have 159 counties, geez.

[Adam Griffin]

I think you're right about the constituency of North Fulton but BK's also right: they're going to vote their economic interests. Even if Romney and the Republicans are going off the deep end, it's not that much of a worry to many of them, so long as they vote economics. I think because it's part of Fulton, we automatically think it can vote like Fulton. Look at it on a map, however, and it clearly is in the latitude of the Brick Wall. You could split it down 400 and the parts would mesh almost seamlessly into Cherokee's and Gwinnett's demographics.  It probably should be its own county, if for no other reason than different constituencies and it's the largest county.

Also, this occurred to me: were the questions asked in the order stated here? I would assume yes, but if the same-sex marriage question was asked before the presidential preference question, I could see that skewing the result to favor Romney.

 

[Adam Griffin]

No way Obama only gets 40%. 44-45% is absolutely his floor in the state, and I should know, I live here for gods' sake. I think pollsters are underestimating the minority vote in Georgia. The black vote was 30% in 2008, and it only dropped 2 points, to 28%, in the 2010 midterms.

And despite "only" dropping to 28% the Dems could only muster 39% in the Senate race and 43% in the Gov race. 

I see Obama at 43.5% in Georgia with 55.5% for Romney

First of all, race plays a factor here. Black Senate candidates have a DPI at least 5 points lower than white Senate candidates on average.

2002: Cleland (White) - 45.9%
2004: Majette (Black) - 40.0%
2008: Martin (White) - 46.8%
2010: Thurmond (Black) - 39.0%

This can also be seen in how Roy Barnes - in 2010 - received a higher percentage of the white vote than Obama did in 2008. When you consider the fact that the 2010 GA electorate was far more conservative throughout + the national trends, it shows how much race plays a role here. Thurmond received around 19% of the white vote in 2010, which honestly could have been worse; Obama only got 23%. So did Kerry. The black turnout combined with suburban flips and changing demographics cancelled out the race factor for Obama, but there hasn't been a scenario in which high black turnout occurred alongside a Senate race with a black candidate.

Secondly, the drop from 30% to 28% is negligible when compared to many other states with measurable black population. Georgia has a strong black electorate that generally maintains its representation both in mid-terms and presidential elections. However, with blacks voting nearly exclusively Democratic, the 2 point drop from 30% to 28% equates to a 2 point drop in statewide DPI. Two-thirds of Georgia Democrats are black, so a ~7% in black representation at the polls has a pretty drastic effect on DPI.

Thirdly, white liberal youths were not motivated to vote in 2010. We'll see what happens this year, but mid-terms lead to some pretty sharp drop-offs in DPI here. As cope said, 2006 was probably the worst year for Democrats in Georgia, despite being an excellent year nationally. It was also the point at which the demographics and the collapse of Democratic Party in Georgia post-2000 were both at their worse. From there on, the demographics have been growing in our favor and the Democratic Party is (albeit slowly) recovering.