Georgia's Very Own Megathread! (v2)

[Adam Griffin]

I don't think yard signs do a whole lot (or we'd have Congressman Ossoff), but I agree that they may at least get the candidate's name out there.  Something I've complained about recently is that Abrams seemed to have gone quiet -- no TV or radio ads, signs, or other outreach that was visible to me.   Kemp has been mostly unopposed with TV ads during the times/stations I watch, although last night I did finally see an Abrams ad on channel 11.

Pretty much from the beginning, Abrams herself has said that this campaign wouldn't be one won or ran based on TV ads, radio spots and so forth. The reason so many areas are getting field organizers and energy pumped into them is because that money isn't being pumped into expensive ATL media market ad buys. There will come a day where both can happen, but a win statewide will be needed before that happens. Much like with the sign stats I posted above, TV and radio spots are essentially a never-ending black hole of resource entrapment for campaigns; as long as you can keep the disadvantage to 2:1 or less, there's arguably no negative impact (or positive impact if you're on the winning side of those odds). Even if you're outside those odds, effective field and abundant enthusiasm can easily overcome deficits created by it.

Furthermore, it seems pretty obvious that the campaign is also pursuing (or at least benefiting from) a stealth approach alongside this; they don't want the extent of their investment to be immediately known or visible to every GOP operative and voter. Based on the universe of voters being targeted, it's not difficult for that to be the case. While I still disagree with the strategy of effectively abandoning persuasion efforts, they are reaching out to large numbers of registered voters.

[Adam Griffin]

I don't think yard signs do a whole lot (or we'd have Congressman Ossoff), but I agree that they may at least get the candidate's name out there.  Something I've complained about recently is that Abrams seemed to have gone quiet -- no TV or radio ads, signs, or other outreach that was visible to me.   Kemp has been mostly unopposed with TV ads during the times/stations I watch, although last night I did finally see an Abrams ad on channel 11.

Pretty much from the beginning, Abrams herself has said that this campaign wouldn't be one won or ran based on TV ads, radio spots and so forth. The reason so many areas are getting field organizers and energy pumped into them is because that money isn't being pumped into expensive ATL media market ad buys. There will come a day where both can happen, but a win statewide will be needed before that happens. Much like with the sign stats I posted above, TV and radio spots are essentially a never-ending black hole of resource entrapment for campaigns; as long as you can keep the disadvantage to 2:1 or less, there's arguably no negative impact (or positive impact if you're on the winning side of those odds). Even if you're outside those odds, effective field and abundant enthusiasm can easily overcome deficits created by it.

Furthermore, it seems pretty obvious that the campaign is also pursuing (or at least benefiting from) a stealth approach alongside this; they don't want the extent of their investment to be immediately known or visible to every GOP operative and voter. Based on the universe of voters being targeted, it's not difficult for that to be the case. While I still disagree with the strategy of effectively abandoning persuasion efforts, they are reaching out to large numbers of registered voters.

Abrams deciding on no persuasion gives me serious pause

How many persuadable voters are there in GA though? It's a state where the overwhelming majority of voters are either evangelical whites or black voters.

I mean, the argument can be made in almost every state: persuadable voters are a tiny minority of voters overall. Certainly they're a smaller share in GA than in most other states, but, to abandon persuasion in GA in 2018 essentially means that:

1) You're fine with the fact that Clinton did substantially worse than both Carter and Obama in a large portion of the state, with said under-performance yet to be baked into a midterm

2) You're taking at face value that slippery, treacherous GA suburbrons who have been strong GOP supporters consistently until Clinton aren't going to stab you in the back after a one-time defection; that they're more valuable and reliable than various rubes who have generally voted D until 2016

3) You're betting on low-propensity voters (who are much more the definition of "low-propensity voter" in GA than in many other states) to make or break your campaign

4) You're going to need twice the number of irregular and first-time voters to cancel out the effects of those lost via persuasion, since "+1 D is only half of "+1 R, -1 D"; only after negating all of that can the reduction of the 250k vote deficit truly begin

Perhaps the data suggests that a huge gender bias is going to raze Democratic support in some of these places and it's impossible to overcome via persuasion: it wouldn't be the craziest thing I've heard (especially after comparing Nunn and Carter support in some areas), but I don't think racial bias alone would merit abandoning persuasion (again, both Obama and Carter carried tens of thousands of - if not more than 100k - Georgia Democratic voters who Clinton lost).

[Adam Griffin]

I do disagree that she is abandoning persuasion efforts. She is banking on issues like Medicaid Expansion, public education, and jobs being things that people of different ideologies can line up behind while also not abandoning progressive views on guns, women's right to choose, and immigration.

I'm just going off of what I've seen and been told. Abrams herself told me this back in July or August of last year at an intimate gathering in Dalton. I was basically asking her about our area specifically ("this area's Democratic base doesn't look like the rest of the state: it's pretty much all blue-collar whites and low-propensity Latinos, and needs tending from cycle to cycle to maintain it, and so any potential message focused at galvanizing 'Georgia's Democratic base' might not work as well here; we used persuasion to great effect in 2014; what is your plan for this part of the state campaign-wise?") and she essentially said that it wasn't her plan to persuade voters to vote for her but that it was her job to turn out those who already support us. The broader message from here ever since has generally matched that.

Along with that, her campaign manager (who was nice enough to make a contribution to our county party - glory!) has said in one of those interviews for the campaign that turnout in South Georgia and Metro ATL is the focus. That of course has been a repeated point from the campaign many times. Individual conversations with various staff have all been focused on turnout, turnout, turnout. Additionally, the field work is entirely - at least right now - based on turnout.

I'd point out that Q3 in the current script (what I believe you're referring to about the issues) isn't true persuasion in that it's not designed to persuade voters to vote differently, but is information being gathered to regurgitate to those voters at a later date to convince him to turn out.

[Adam Griffin]

But alas, even though I've always been worried about the effects that such a campaign focus/method/style means for my own area, where the lack of extensive persuasion could really end up hurting us is in the suburbs. I don't trust those fickle suburbanites to stick with us after just one election, and they have been the bane of Democrats' existence since dirt was formed. I wager they're just looking for a reason to go back to the GOP.

[Adam Griffin]

and I'm in one of the more friendly and civilized parts of the state.

Haha, I meant in this part of the state; certainly not one of the most friendly or civilized statewide!

[Adam Griffin]

[snip]

Just woke up so I'm not ready to analyze, but I went back and look at some of the comparison maps I made after the 2014 election. Looks like the Nunn underperformance was largely concentrated to CDs 8 & 14 (Perdue probably had some home-turf advantage in 8, and Carter had some ancestral home turf in 14, for what it's worth). At the county level, generally, there wasn't anywhere where Nunn did more than 2 points better than Carter.





Some other maps I thought might be relevant to the discussion from there:

(White Vote Performance; Who Did Better)

[Adam Griffin]

1) You're fine with the fact that Clinton did substantially worse than both Carter and Obama in a large portion of the state, with said under-performance yet to be baked into a midterm

One more map that might illustrate this point is a simple comparison of vote share between Carter and Clinton. This is among all voters (unlike the Carter-Obama comparison) and compares the percentage of the vote each candidate received.

In other words, this isn't a swing map of margins; if Clinton received 24% and Carter received 31%, then that'd be a 7 percentage-point advantage to Carter on the map. More or less, you can double the figures shown in the map to get an approximate swing in margin - though given bigger third-party support in 2016, it's not an exact situation.

[Adam Griffin]

I was walking through the living room last night and heard some kind of campaign commercial. Practically everything we hear here is TN Senate and Governor campaigns, so I wasn't paying much attention...but something caught my ear. Unfortunately, people were yakking in there, but I gathered that it was an attack ad against Brian Kemp over massage parlors and it sounded really nasty (great job!).

So apparently there are two of them? I can't find the first one (which from the looks of it was pretty cheesy), but here's the other one:

https://www.youtube.com/watch?v=5BXpgFxVsyU
https://politics.myajc.com/blog/politics/kemp-demands-stations-pull-attack-ads/aOOMsG7Y0cmcGOOnMHvUwN/

You must be logged in to read this quote.

[Adam Griffin]

^^^ Just wanted to point out the genius of it in that most of it is worded just vaguely enough to make you think Kemp was personally the one doing the molestin' (especially near the beginning)

[Adam Griffin]

What makes it so funny is that just recently, Kemp decided to go after Abrams over matters of sexual assault, now suddenly are wilting flowers when the favor is returned:

[Adam Griffin]

An internal for Abrams released these results(Likely Voters):

Abrams(D)- 48%

Kemp(R)- 42%

Metz(L)- 3%

https://secure.staceyabrams.com/page/-/2018_September/Email/Polling%20Memo%20from%20Garin%20Hart%20Yang.pdf

[Adam Griffin]

Just got my absentee ballot in the mail today, and will be sending off in the next day or two:

[Adam Griffin]

https://twitter.com/ElectProject/status/1045495813873573888

Me and my entire family are shifting from voting early in person to voting absentee. Interested in seeing how these numbers hold up through the next few weeks.

Glorious News! And I see he added the comparisons in the tweet below:

At this time in...
2014: 4,163 ballots; 78% white
2018: 6,585 ballots; 48% white

In the spirit of discussions in the early vote thread, here are some up-to-date figures on returned ballots* by county (as of Thursday). Statewide, there have been 6,933 ballots returned as of last night. Obviously, some counties started mailing these things out before others; the figures definitely don't reflect their respective shares of the electorate. I picked out a list of a dozen or so based on their usual size and then scoured through the sheet for some other counties that had the largest number of ballots returned at this point.

*Returned in this case also includes ballots that were found to be ineligible, are being re-issued or that need further clarification before their ballots are formally counted

62% of returned ballots are in the following counties:

Code:

603	DEKALB
507	COBB
488	GWINNETT
434	RICHMOND
278	CHEROKEE
247	HOUSTON
241	FULTON	
216	FORSYTH
179	HENRY
161	PAULDING
141	COWETA
139	LOWNDES
129	ROCKDALE
114	CHATHAM
108	TROUP
99	COLUMBIA
96	FLOYD
48	WHITFIELD
21	CLARKE
14	CLAYTON
13	MUSCOGEE
4	HALL
0	BIBB

[Adam Griffin]

Updating county figures (compared to yesterday).

Excluding the 2 new counties I've added for today, the original counties from yesterday now comprise 63% of returned ballots (up from 62% yesterday); these counties also comprised 63% of Georgia's 2010 population (probably more like 65% as of today) and 72%/75% of Obama/Clinton's statewide vote bloc:

Code:

788 (603)	DEKALB
623 (488)	GWINNETT
526 (434)	RICHMOND
507 (507)	COBB
495 (241)	FULTON	
347 (278)	CHEROKEE
318 (114)	CHATHAM
298 (247)	HOUSTON
285 (216)	FORSYTH
257 (179)	HENRY
218 (99)	COLUMBIA
215 (161)	PAULDING
189 (141)	COWETA
186 (14)	CLAYTON
176 (139)	LOWNDES
154 (129)	ROCKDALE
137 (108)	TROUP
115 (96)	FLOYD
89            DOUGLAS
56 (48)	WHITFIELD
34 (21)	CLARKE
27 (0)	BIBB
13 (13)	MUSCOGEE
4 (4)	        HALL
0              DOUGHERTY

[Adam Griffin]

Looking at the vote preferences of these two segments of counties in 2016, comparing that to 2014 ballots cast at this time and then comparing to 2018 ballots cast, one would expect this electorate - based solely on geography - would be a 54/46 split in favor of the GOP. However, when factoring in the shift along racial lines, you're looking at a 30-35 point shift in favor of Democrats.

If I had to approximate based on the figures available right now, I'd wager the current mail votes banked are at least 62-38. That's on par with what provisional ballots tend to be, which are otherwise the most Democratic segment of votes.

It's worth noting that we've seen these same kinds of trends in recent elections; either at the onset (in 2016) or at the tail-end (in 2014) of early voting, non-white participation suddenly surges and exceeds all previous EV records by a large amount, but it has historically amounted to little given it was merely a cannibalization of ED votes. We'll see if it's different this time. At minimum, this time, it's a huge shift - many times that of what we've seen in the past.

[Adam Griffin]

Just for reference, here are the total numbers/percentages of people who voted early in each election going back to 2002 (note: 2004 & 2010 are approximations based on references I could find; for 2002, all I could find was that "7% of votes cast were early votes"):

[Adam Griffin]

Another fun interactive map I just made:

Percentage of Presidential Voters to Vote Early (2016, by County)

[Adam Griffin]

If the Abrams campaign doesn't contact me by October 15, I will select Brian Kemp on my absentee ballot

You're a white Republican residing in Kemp's home turf expecting persuasion contact from a turnout-oriented campaign: might as well send the ballot off today!

[Adam Griffin]

And a more relevant ED/EV split interactive map by county for comparison to 2018:

Percentage of Gubernatorial Voters to Vote Early, 2014

(36.8% Statewide EV in 2014)

[Adam Griffin]

I find it interesting that the northeast corner of the state (Towns/Rabun/Habersham/Stephens) are mostly early votes.

Do you have any idea of why that might be?

Wealthy, old, out-of-state conservatives buying up mountain tracts for their tacky log cabins (and to a lesser degree, pretty much every county with large transplant conservative populations have high early voter participation rates). See the previous page for the presidential version; you can see it at work in places like Forsyth as well. Northeast mountainous GA is the most concentrated area in terms of all of these factors converging, however.

Rabun County specifically also consolidated down to 1 precinct, so early voting isn't any harder than voting on Election Day. Probably explains why it had the highest share of ballots cast early both in 2014 & 2016.

[Adam Griffin]

Updating county figures (compared to Friday). According to the SoS, 14,970 ballots have been returned.

Either there's a reporting error or the county boards of election in Macon and Columbus continue to be incompetent. We saw a huge return of ballots over the weekend in Clarke, big surges in Cobb & Dekalb (tripled and doubled the # of returned ballots, respectively) and a few oddities continue to remain visible (such as the disproportionate number of ballots in Troup, and the relatively small #s from Fulton).

These counties now comprise 66% of returned ballots (up from 63% on Friday); these counties comprised 63% of Georgia's 2010 population and 72%/75% of Obama/Clinton's statewide vote bloc:

Code:

1556	(507) 	COBB
1385	(788)	DEKALB
1088	(623)	GWINNETT
658	(526)	RICHMOND
550	(318) 	CHATHAM
511	(495) 	FULTON	
489	(347) 	CHEROKEE
487	(257) 	HENRY
395	(298) 	HOUSTON
387	(285) 	FORSYTH
351	(215) 	PAULDING
311	(189) 	COWETA
271	(218) 	COLUMBIA
213	(34) 		CLARKE
196	(137) 	TROUP
186	(186) 	CLAYTON
179	(176) 	LOWNDES
167	(89)    	DOUGLAS
155	(115)	FLOYD
154	(154) 	ROCKDALE
75	(56) 		WHITFIELD
27	(27) 		BIBB
23	(4)    	HALL
13	(13) 		MUSCOGEE
0	(0)		DOUGHERTY

According to Michael McDonald's stats:

Code:

White	6651	46.2%
Black	6033	41.9%
Latino	312	2.2%
Other	571	4.0%
Unknown	819	5.7%

Female	8353	58.1%
Male		5974	41.5%
Unknown	59	0.4%

18-29	  759	5.3%
30-44	1149	8.0%
45-60	2286	15.9%
60+	      10154	70.6%
Unknown	   38	0.3%

[Adam Griffin]

Mail ballot update (compared to yesterday). According to the SoS, 18,158 ballots have been returned (14,970 were returned as of yesterday; 3,188 returned today).

Macon/Bibb finally started to report, and Clayton had its first huge dump today as well. Fulton finally started to catch up to where it should be, vaulting into third place. Paulding is an interesting one here, given its relatively strong GOP support and relatively small population compared to other metro counties.

These counties comprise 66% of returned ballots (unchanged from yesterday); these counties comprised 63% of Georgia's 2010 population and 72%/75% of Obama/Clinton's statewide vote bloc:

Code:

1618	(1385)	DEKALB
1581	(1556) 	COBB
1235	(1088)	GWINNETT
1119	(511)	FULTON
811	(658)	RICHMOND
607	(550)	CHATHAM	
505	(351)	PAULDING
497	(487)	HENRY
496	(489)	CHEROKEE
482	(387)	FORSYTH
455	(186)	CLAYTON
404	(395)	HOUSTON
311	(311)	COWETA
295	(271)	COLUMBIA
239	(213)	CLARKE
216	(179)	LOWNDES
209	(196)	TROUP
208	(154)	ROCKDALE
194	(167)	DOUGLAS
157	(155)	FLOYD
145	(27)	        BIBB
94	(75)	        WHITFIELD
23	(23) 	        HALL
14	(13)	        MUSCOGEE
0	(0)	        DOUGHERTY

In composition of the electorate, the white share continues to decline, and is potentially on track to drop below 45% by tomorrow.

Code:

White	7943	45.6% (-0.6)
Black	7395	42.4% (+0.5)
Latino	357	2.0% (-0.2)
Other	679	3.9% (-0.1)
Unknown	1062	6.1% (+0.4)

Female	10172 58.3% (+0.2)
Male		7179   41.2% (-0.3)
Unknown	85	  0.4% (+0.1)

18-29	  950	5.4% (+0.1)
30-44	1384 7.9% (-0.1)
45-60	2789 16.0% (+0.1)
60+	      12255	70.3% (-0.3)
Unknown	   58	0.3% (0.0)

[Adam Griffin]

So it's continuing to get less white and more female.

If you had to guess what the margin is among these votes, what do you think it is now?

Maybe 1-2 points different from what I estimated prior to Monday (62/38 in the two way, minimum) - but without comprehensive access to a statewide voter database, it's impossible to know for sure. Just because certain counties are comprising certain percentages doesn't mean that their electorates come close to matching the ballots that have been returned.

It's even difficult to wager what the situation is along racial lines. Even among white mail ballots, I'm guessing the composition is much more Dem-friendly than usual (since Democrats are pushing mail voting so strongly). We could be looking at a 3:1 advantage for Democrats right now for all I know.

[Adam Griffin]

*snip*

Here is something very interesting that I can report at the granular level, however. Looking at my county's voter file, I can see who all has voted (along with a plethora of details about them).

Of the 75 people who had voted by mail in Whitfield County (70-26 Trump) as of yesterday, the most recent scoring models suggest that it's a 53/47 electorate. That's a bit of a simplification (the scoring model rates likely 2018 DNC support in terms of percentage likelihood; technically, 53% of the voters have a greater than 50% chance of supporting Democrats), but still unprecedented.

Furthermore, half of the people to vote by mail did not vote in any of the last 3 primaries. Only 1 in 7 of these voters pulled ballots in all of the most 3 recent primaries. Forty percent of them voted in no more than 1 of the past 3 general elections (2012, 2014, 2016).

And the racial stats (also unprecedented, particularly for Latino midterm participation):

Code:

59% White
18% Latino
15% Black
8% Other

Those stats basically look comparable to our county's Democratic presidential electorate rather than the electorate at-large (a little less Latino and a bit more black than in a presidential general, but still).

It's obviously a small sample size, so salt shakers and all, but still very out of the norm for my county. This pretty much proves that the vote by mail campaign is having an effect, though.

[Adam Griffin]

Intriguing info. It’s still pretty early but I’m loving what I’m seeing so far. 34 more days of hard work!

I'm gonna quit spamming the thread now, but I just ran the same numbers on the requested but unreturned ballots. If we combine the two groups, we have approximately 500 mail voters in Whitfield County who are:

PARTY:
61% Democratic
39% Republican

RACE:
45% White
30% Latino
15% Black
10% Other

PAST VOTING (2012, 2014 & 2016)
* 42% Voted in 0 or 1 General Elections
* 62% Voted in 0 Primaries

Honestly pretty incredible (especially the Latino percentage; hopefully they return their ballots!). But if the combined returned/received mail voting electorate in Whitfield County is 61% Democratic, then we can surely and safely assume that the mail electorate statewide is much more Democratic than my initial minimum estimates.