Georgia's Very Own Megathread! (v2)

[Adam Griffin]

One of my classmates from college shared her wait in line in Henry County this morning.

*snip*

This seems to be happening all over the state to some degree. After Abrams' Dalton rally on Friday, I spoke with voters who waited an hour to an hour and a half in line. Honestly, it's another worrisome element for me, thinking that even in all these heavily-GOP counties, there are wait times to vote early.

[Adam Griffin]

Does anybody know which counties had early voting today? The state only mandates the second Saturday (next Saturday), so I'm assuming pretty much all of the votes that came in today came from heavily-Democratic turf - or were mail ballots.

[Adam Griffin]

According to Georgia Votes, an additional 32,899 valid votes were cast yesterday (Saturday); today's accepted ballot total stands at 515,399 votes: 20.21% of total 2014 turnout.

Early vote totals are 213% higher than at this point in 2014 (up from 197% yesterday).

The white share of the electorate decreased by 1 point since yesterday. The electorate that voted today was 46% white, 35% black and 19% other (64/28/8 yesterday): a drastic difference; basically enough to reverse yesterday's losses for Democrats.

Code:

White	307619	59.7% (-0.9)
Black	152911	29.7% (+0.4)
Latino	6321         1.2% (+0.1)
Asian	6573	         1.3%  (+0.2)
Other	41975        8.1%  (+0.3)

Female	274536	53.3% (0.0)
Male		230992 	44.8% (-0.1)
Unknown	9871         1.9%   (+0.1)

18-29	 29507	5.7%   (+0.3)
30-39	 36196	7.0%   (+0.3)
40-49	 59665	11.6%   (+0.5)
50-64       162053	31.4%  (+0.5)
65+         218459      42.1%  (-1.7)
Unknown	  9519       1.8%   (+0.1)

[Adam Griffin]

Something I found interesting from local reporting on the Abrams visit: at the same time that Abrams is traversing North Georgia, Kemp is launching his tour of Southeast Georgia. I initially thought it was weird that Tifton wasn't on Kemp's stop, especially since it's on the path of his no-name town tour.

But a closer look at the exact stops makes me wonder...

[Rookie Yinzer]

Does anybody know which counties had early voting today? The state only mandates the second Saturday (next Saturday), so I'm assuming pretty much all of the votes that came in today came from heavily-Democratic turf - or were mail ballots.

DeKalb, Cobb, Gwinnett, Henry, Fulton, Clayton, Paulding, and Muscogee are all I could confirm. None of the other "big" counties had early voting and I know none of the other Metro counties had it either so that should be it.

[GeorgiaModerate]

Griff, is it possible to break down the EV share by both race and age?  Well, I'm sure it's possible, just a question of whether you or anyone wants to do the work.  For example:

White 65+
White 50-64
...
White 18-29
Black 65+
Black 50-64
...

It seems possible that the relative increase of whites in the last few days isn't necessary favorable for Republicans, depending on its age distribution.

[Bidenworth2020]

Something I found interesting from local reporting on the Abrams visit: at the same time that Abrams is traversing North Georgia, Kemp is launching his tour of Southeast Georgia. I initially thought it was weird that Tifton wasn't on Kemp's stop, especially since it's on the path of his no-name town tour.

But a closer look at the exact stops makes me wonder...

not to be naive, but what are you referring to jere?

[wbrocks67]

According to Georgia Votes, an additional 32,899 valid votes were cast yesterday (Saturday); today's accepted ballot total stands at 515,399 votes: 20.21% of total 2014 turnout.

Early vote totals are 213% higher than at this point in 2014 (up from 197% yesterday).

The white share of the electorate decreased by 1 point since yesterday. The electorate that voted today was 46% white, 35% black and 19% other (64/28/8 yesterday): a drastic difference; basically enough to reverse yesterday's losses for Democrats.

Code:

White	307619	59.7% (-0.9)
Black	152911	29.7% (+0.4)
Latino	6321         1.2% (+0.1)
Asian	6573	         1.3%  (+0.2)
Other	41975        8.1%  (+0.3)

Female	274536	53.3% (0.0)
Male		230992 	44.8% (-0.1)
Unknown	9871         1.9%   (+0.1)

18-29	 29507	5.7%   (+0.3)
30-39	 36196	7.0%   (+0.3)
40-49	 59665	11.6%   (+0.5)
50-64       162053	31.4%  (+0.5)
65+         218459      42.1%  (-1.7)
Unknown	  9519       1.8%   (+0.1)

Does GA to Souls to the Polls on Sundays?

[NewYorkExpress]

https://www.cnn.com/2018/10/20/politics/gwinnett-county-absentee-ballots/index.html

You must be logged in to read this quote.

More voter suppression on orders from Kemp (or Deal) or a county going rouge?

[Rookie Yinzer]

Georgia Votes now has a feature that breaks down the vote by Congressional District as well as State House and Senate Districts.

Unsurprisingly, GA-6 and GA-7 have the largest increase in votes vs. 2014 as well as the largest percentage of non-2014 voters.

[Adam Griffin]

not to be naive, but what are you referring to jere?

The expansive area where Kemp is having his rallies has fewer than 500,000 people and was 3:1 Trump. I actually need to expand the above map because Abrams has/had been on the road for several days (what was shown above was just 1-2 of those days), but in the non-Democratic areas, her 1 day of visits was also in an area that has 500,000 people and was 3:1 Trump.

The fact that Kemp is spending time rallying his base in very sparsely populated areas at the same time that Abrams is targeting similarly Republican but higher density areas could mean something. It's not inherently an indicator that Kemp is struggling (after all, maybe he has his North GA base all sewn up and is focusing elsewhere), but one would expect that if Abrams is targeting heavily-GOP areas surrounding the metro at the same time that Kemp is trying to rally his base in sparsely-populated GA during early voting, there might be a reason for why that is...after all, a standard GOTV tour between the two parties would see Kemp barnstorming the counties Abrams was in North GA, while Abrams barnstormed ATL & all the other urban clusters.

[Adam Griffin]

Griff, is it possible to break down the EV share by both race and age?  Well, I'm sure it's possible, just a question of whether you or anyone wants to do the work.  For example:

White 65+
White 50-64
...
White 18-29
Black 65+
Black 50-64
...

It seems possible that the relative increase of whites in the last few days isn't necessary favorable for Republicans, depending on its age distribution.

Without a comprehensive statewide voter file to use like I had in 2014 (i.e. NGPVAN with statewide access) or paid access via SoS files, I'm not sure it's possible. The absentee spreadsheets freely available for download don't provide race or age information with each entry, so GeorgiaVotes must be using that and/or a more comprehensive voter file solution to aggregate the information.

We could infer some trends based on how age and race groups have been shifting from day-to-day, but that is of course nowhere near exact (or even guaranteed to be right).

[Rookie Yinzer]

https://twitter.com/gtryan/status/1054202123771101184

[Adam Griffin]

I've went ahead and updated the GIF with Saturday/Sunday shifts in ballots (didn't really want Monday's trends to be tainted by any weekend #s): no absolute flips, but a couple of gradient shifts.

[Adam Griffin]

And in relation to what I was talking about earlier, here's a completed map showing all the visited/planned stops over the past few days. The only overlapping area was/is Kingsland.

[ON Progressive]

JMC has Abrams up 47-46 in GA-07: https://www.scribd.com/document/391316947/GA-07-Polling-Memo-Rob-Woodall-vs-Carolyn-Bourdeaux-October-2018-Bold-Blue-Campaigns-with-JMC-Analytics-and-Polling

[Adam Griffin]

JMC has Abrams up 47-46 in GA-07: https://www.scribd.com/document/391316947/GA-07-Polling-Memo-Rob-Woodall-vs-Carolyn-Bourdeaux-October-2018-Bold-Blue-Campaigns-with-JMC-Analytics-and-Polling

At first I thought this said GA-06 (and I thought "that's good...but not as good as it needs to be"), but wow. If true, that's pretty great.

I know GA-07 is a weird place where on paper, the district should be more Dem than GA-06 - yet a closer look at the fundamentals suggest Democrats would do more poorly there. If this is a situation where the latter scenario is the more accurate one, then this suggests Abrams is clearing 50 in GA-06. I think GA-06 is going to be basically a bellwether (or at the least the closest thing we have to one in terms of CDs).

I would love to know what these figures look like in GA-1 & GA-12; both districts are more GOP than the state as a whole (yet Barrow held on in GA-12 until 2014) but were "historically" the 2 most competitive CDs in the state prior to this suburban surge for Democrats. I still think that if Abrams is going to pull this off - based on the fundamentals - much of the territory overlapping with CDs 10 and 12 are going to be her most-improved areas outside of the metro (as shown in my benchmarks). Those counties are not heavily-populated areas, but given the threshold, it'd be nice to know if she's able to improve by a large amount in areas like that.

So here are the interactive maps in case anybody's interested:


Abrams - County Majority Benchmarks (Margin)

Abrams - County Majority Benchmarks (Swing)

[Adam Griffin]

So, relating to what I mentioned yesterday:

On the Georgia trail: Kemp doubles down on rural strategy

You must be logged in to read this quote.

[Rookie Yinzer]

Didn't Hillary lose GA-7 by 7 or 8 points?

I also just read Stacey will be in Thomaston, Columbus, Americus, Albany, Thomasville, Valdosta, Carrollton, Newnan, and Fayetteville on Wednesday and Thursday.

[Adam Griffin]

Didn't Hillary lose GA-7 by 7 or 8 points?

I also just read Stacey will be in Thomaston, Columbus, Americus, Albany, Thomasville, Valdosta, Carrollton, Newnan, and Fayetteville on Wednesday and Thursday.

Yep: 6.4 point loss in GA-07 & 1.5 point loss in GA-06 (and for fun, 15.5 point loss in GA-01 & 16.2 point loss in GA-12).

Still, I could see this election (maybe not the House race, but definitely the Governor's race) being one where GA-07 catches up with - or maybe even surpasses - GA-06 just because of its sheer non-white population...especially after the primary results in Gwinnett.

[Gass3268]

I love John Barrow's ads:

https://twitter.com/AnthonyAdragna/status/1054406227013394434

[Adam Griffin]

I love John Barrow's ads:

https://twitter.com/AnthonyAdragna/status/1054406227013394434

"I'M JAHN BARRA AND LONG BEFORE I WAS BORN, MY GRANDDADDY USED THIS SMITH AND WESSON HERE TO HELP STAHP A LITERACY TEST"

[GeorgiaModerate]

Saw my first Abrams sign in Forsyth County this morning (first one in the county, I've seen others elsewhere.)

[Adam Griffin]

According to Georgia Votes, an additional 116,884 valid votes were cast yesterday (Monday); today's accepted ballot total stands at 632,283 votes: 24.79% of total 2014 turnout.

Early vote totals are 203% higher than at this point in 2014 (down from 213% Saturday).

Despite today being the single biggest vote dump of the cycle, the racial percentages remain virtually unchanged from the weekend. The electorate that voted Monday was 59% white, 29% black and 12% other. These numbers may finally be stabilizing, with a potential non-white surge (like in 2014) just around the corner.

Code:

White	377266	59.7% (0.0)
Black	187277	29.6% (-0.1)
Latino	7716         1.2% (0.0)
Asian	8062	         1.3%  (0.0)
Other	51962        8.2%  (+0.1)

Female	337691	53.4% (+0.1)
Male		282158 	44.6% (-0.2)
Unknown	12434         2.0%   (+0.1)

18-29	 36714	5.8%   (+0.1)
30-39	 45548	7.2%   (+0.2)
40-49	 74852	11.8%   (+0.2)
50-64       200140	31.7%  (+0.3)
65+         263037      41.6%  (-0.5)
Unknown	  11992       1.9%   (+0.1)

[Adam Griffin]

I decided to drill down a bit more into the white vote, stacking its current distribution against the 2014 white Democratic support for Carter in each county. My independent analysis built off of county-by-county models also came to the same conclusion as the exit polls: Carter got 22.9% of the vote in 2014 (among all voters).

Stacking the current proportions of the white vote by county against each county's white Democratic support from 2014 gives us virtually the same number in the early vote right now: 22.7% Abrams. Obviously there are going to be counties where she does better or worse than Carter among whites, but for it to average out almost to the decimal point in comparison to 2014 suggests there is no statistically-significant shift in terms of which types of white communities are voting.

Granted, the early vote is likely more Democratic among whites than it'll be once ED is counted (making this an uneven comparison in some respects), but this is mainly an exercise to see if there are any real proportional shifts from areas with differing levels of white support. It's also a demonstration of what the likely margin would be if the current racial composition ended up being the final totals on Election Day.

Of course, rural whites in one part of the state could be voting at higher numbers, offsetting lower voting rates in another rural white part of the state (and so forth), but I would expect to see some shift of significance if there was something to that. The only way I could see this being reasonably off is if white Democrats in every part of the state are off-setting white Republicans in every part of the state at a 1:1 ratio.

Plugging that 23% into a few different scenarios of support (including 2014 baseline; first scenario) produces the following results (assuming 2% for Metz):

Black %	Latino %	Asian %	Other %	Result
90%	65%	60%	60%	Kemp +4.4
95%	65%	60%	60%	Kemp +1.5
95%	75%	75%	70%	Abrams +0.8

I went back and re-ran these numbers with the updated votes from Saturday, Sunday and Monday.

Not much has changed; Abrams in this baseline has gained another 0.2 points of the white vote based on this calculation (from 22.7% to 22.9%), aligning perfectly with Carter's white vote share in 2014 among all white voters (EV + ED). We'll see if any continued shifts occur, but there has been no statistical shift in neither the projected white preference nor the racial makeup of the electorate overall since Friday.

There are several inconsistencies with the way I'm projecting this (mentioned above) and so this shouldn't be taken as a prediction or projection of the final outcome. Mainly, it's useful in determining whether white voters are fundamentally distributed differently in geographic terms in the 2018 EV when compared to 2014 overall, to the point that it would make a substantial difference in white support on its own (they're not). That of course doesn't automatically mean that Abrams will do better or worse among whites.