Georgia's Very Own Megathread!

[Adam Griffin]

A few years back, I made this similar graph which at the time had projections for 2016-2024. Those projections were quite optimistic (in large part because there was heavy weight on the Obama years in terms of demographic shifts, and also because black voters specifically were better categorized as black by SoS prior to 2015 or so).

Nevertheless, I have updated and revised, adding the most recent two elections and including the 2017 1-Year ACS & 2019 Census as well.

For all previous elections/years, Census data and race breakdowns via SoS are used. Much like the SoS pivoted from an all-encompassing "Other" category in 2004 that made tracking specific non-white, non-black groups difficult until then, new registrants (i.e. disproportionately non-white registrants) for the past 5 years or so have been disproportionately categorized as "other" or "unknown". This plays a major role in why the black share of the electorate is smaller in 2016 & 2018 than in 2012 & 2014 (though not 100% of the cause) and why "other" jumps over the same period. This is also why the 2020-24 projections may seem "gentler" than some might expect.

Full-sized image

What would 2020 look like with these breakdown differentials vis-à-vis 2018 GA Gov?

Assuming a replication of Abrams' margins (particularly among black voters), Democrats would receive 49.8% of the vote (±0.2 points). The variance is due to the potential difference in "other" composition between the two elections and/or how exactly you calculate Abrams support among that group.

[Adam Griffin]

Didn't see a post about this, but it's (tentatively) a big deal:

Judge rules Georgia ballots mailed by Election Day must be counted

A federal judge on Monday extended the deadline for absentee ballots to be returned in Georgia, ruling that they must be counted if postmarked by Election Day and delivered up to three days afterward.

The decision will likely result in tens of thousands of ballots being counted after Nov. 3 that would have otherwise been rejected, enough to swing close elections. The ruling invalidates Georgia’s requirement that ballots had to be received at county election offices by 7 p.m. on Election Day.

U.S. District Judge Eleanor Ross wrote that voters must be protected during the coronavirus pandemic, when record numbers of Georgians are expected to cast absentee ballots.

“Extending the deadline would ensure that voters who receive their ballots shortly before Election Day are able to mail their ballots without fear that their vote will not count,” Ross wrote in her 70-page order.

The secretary of state’s office plans to immediately appeal the ruling.

[Adam Griffin]

Assuming a replication of Abrams' margins (particularly among black voters), Democrats would receive 49.8% of the vote (±0.2 points). The variance is due to the potential difference in "other" composition between the two elections and/or how exactly you calculate Abrams support among that group.

GA for Democrats currently can really be summarized like this:

I should also point out that the numbers above are not a two-way model, which means that (at least for President), it'd be a win for Democrats of 0.4-1.8 points. Obviously for any other statewide races, it would very likely signify a runoff...

[Adam Griffin]

A few years back, I made this similar graph which at the time had projections for 2016-2024. Those projections were quite optimistic (in large part because there was heavy weight on the Obama years in terms of demographic shifts, and also because black voters specifically were better categorized as black by SoS prior to 2015 or so).

Nevertheless, I have updated and revised, adding the most recent two elections and including the 2017 1-Year ACS & 2019 Census as well.

For all previous elections/years, Census data and race breakdowns via SoS are used. Much like the SoS pivoted from an all-encompassing "Other" category in 2004 that made tracking specific non-white, non-black groups difficult until then, new registrants (i.e. disproportionately non-white registrants) for the past 5 years or so have been disproportionately categorized as "other" or "unknown". This plays a major role in why the black share of the electorate is smaller in 2016 & 2018 than in 2012 & 2014 (though not 100% of the cause) and why "other" jumps over the same period. This is also why the 2020-24 projections may seem "gentler" than some might expect.

Full-sized image

Does this mean that the "Other" category includes a sizable share of newly-registered Black voters?

Yes (especially post-2015). While I haven't had unabashed statewide voter file access in some time, based on past access and relative black flat-lining in population percentage growth since then, I feel fairly confident in simplistically approximating the "other" category as displayed here as 30% Latino, 25% black, 25% white and 20% Asian (not including other racial rounding errors). Also, it has seemed that over time, at least some of these improper individual racial classifications get corrected via other state data-sets updating their info, though I am not 100% sure about that phenomenon. However, I'm not really looking to take an otherwise series of "official" data-sets and inject my own approximations into the mix.

If you want to do such in determining true black vote share from 2012 onward, however, then:

2012: 32.2%
2014: 30.7%
2016: 30.6%
2018: 31.9%

So based on these corrections, real black voter share in 2012 is still the strongest black showing in GA's history, and 2014's was identical to 2016's. This means that even prior to 2016, the "other" category still included a meaningful amount of both black and white voters, but circa 2015 is when SoS race classifications went off the deep end and large segments of all newly-registered voters began being classified as "unknown/other" by SoS.

Of course and if you want to be reflective of both sides of the political coin, you also need to correct for real white vote share as well:

2012: 63.6%
2014: 65.4%
2016: 63.6%
2018: 62.0%

Not as steep of a decline as it might appear on paper...basically, Latino and Asian voters were responsible for all of the non-white demographic shift between the 2012 and 2016 elections, and were responsible for half of the non-white shift between 2014 and 2018.

[Adam Griffin]

I understand. Indeed it felt a bit strange to me that the official Black share of the vote never crossed 30% in those years. Thank you!

About the bolded part: any hope this gets solved in the near future?

I updated my original post with some more information. I can't be for sure, but it does seem that in between the heat of each election cycle, the percentage of other/unknown voters drops and the percentage of black voters increases, which makes me think that other data-sets via the state get incorporated and update this information. Especially now that GA has opt-out AVR via the Department of Motor Vehicles (since December 2016), every time somebody gets or renews a license, there is an explicit update to their voter registration.

(Also keep in mind that in comparison to my graph, SoS has had categories for Asian, Latino, Native-American, etc since 2004, so the explicit SoS percentages of "other/unknown" are lower than displayed above).

[Adam Griffin]

If it is successful long-term, we just might get a Democratic Wilkinson County back!

19 families buy nearly 97 acres of land in Georgia to create a city safe for Black people

(CNN)"Welcome to Freedom!" exclaims real estate agent Ashley Scott as she surveys the nearly 97 acres of land that she and a group of 19 Black families purchased in August.

"I'm hoping that it will be a thriving safe haven for people of color, for Black families in particular," Scott says.

The land sits just East of Macon in rural Wilkinson County, Georgia. Scott and her friend, investor and entrepreneur Renee Walters, didn't initially plan on buying a large plot of land, but they had a vision that was clear -- to create a safe space for their Black families.

...

Scott and Walters plan to develop the land in phases. The first phase involves clearing the land, farming, and creating a man-made lake for sustainable fishing.

At a ribbon-cutting ceremony, Scott said while they didn't know the exact history of their plot of land, the symbolism of reclaiming this land gives them an opportunity to write their own story.

They hope to grow within a few years. By the end of their development plan, they hope to have a fully operational, self-sufficient city -- putting Freedom, Georgia, on the map.

Location of Toomsboro, GA

Wilkinson County:
2008: McCain +1.1
2010: Barnes +2.6
2012: Romney +1.5
2014: Carter +0.2
2016: Trump +10.3
2018: Kemp +11.6

[Adam Griffin]

I'm hoping either tonight or tomorrow night, the absentee voter file will be available via SoS:

https://elections.sos.ga.gov/Elections/voterabsenteefile.do

I'm thinking maybe since the state is processing & mailing all online mail ballots requested by September 15 (and mailing them out on 9/18), that is why it has taken so long for this data to appear (usually it shows up in early September). After today, the counties will begin processing & mailing all mail ballot requests themselves, so perhaps we'll see the absentee file finally appear sometime this week at minimum.

[Adam Griffin]

I'm hoping either tonight or tomorrow night, the absentee voter file will be available via SoS:

https://elections.sos.ga.gov/Elections/voterabsenteefile.do

I'm thinking maybe since the state is processing & mailing all online mail ballots requested by September 15 (and mailing them out on 9/18), that is why it has taken so long for this data to appear (usually it shows up in early September). After today, the counties will begin processing & mailing all mail ballot requests themselves, so perhaps we'll see the absentee file finally appear sometime this week at minimum.

It's finally up! Unfortunately, there's no statewide file - just 159 individual county files.

[Adam Griffin]

I'm obviously not digging through the individual files and I've never had much luck at merging them in the past, but based on packed/unpacked file sizes and the number of voters in the largest counties, it appears we're somewhere in the vicinity of 875,000 - 950,000 requested mail ballots as of Tuesday (this includes people who voted in the primary who were 65+, who automatically receive ballots for all elections within a cycle).

[Adam Griffin]

My estimate was very approximate and based on file sizes for each county, but Ryan Anderson (the guy who runs georgiavote.com) says:

https://twitter.com/gtryan/status/1306209379956535297

Hopefully his site will be functioning in the next couple of days (though he usually only did returned votes + the sheer number of mail ballot requests will make this a nightmare for him).

[Adam Griffin]

Does anyone know how long it takes to receive your absentee ballot? I requested mine the day I made the below post.

Very refreshing to see Georgia in the news for a positive (& I don't mean a positive COVID test).

We also got our new absentee ballot request site on line on Friday.  Very quick and easy to use!

Thanks for the link!

I'm not sure if it applies to applications submitted by mail and/or directly to the county election boards, but everybody who used the state's website and applied by Tuesday will have their ballots sent out by the SoS starting today (Friday, 9/18). Not sure if they have the bandwidth to send out all requested ballots on the same day (especially if they're handling all pre-9/16 requests regardless of method), but it's definitely possible you'll have it by early next week.

[Adam Griffin]

For those who are curious:

Similar to Crumpets' map for NC, except this one is for GA (two color schemes):

VBM Requests as of 9/19, % of Total 2018 Vote:
Abrams Counties: 33.68% (726462/2156818)
Georgia, Overall: 29.89%
Kemp Counties: 25.32% (453819/1792128)

 

38 votes cast thus far in Georgia.

White 21
Black 13
Other 3

Male 21
Female 14

Voted in 2016: 21
Didn't Vote: 16


For ballot requests, here are the altered racial figures (that attempt to account for & redistribute the large "other" contingency) for those who voted in '16 and those who did not.

Race	2016 Voters	Non-16 Voters	Total VBM
White	62%	51%	59%
Black	32%	37%	33%
Lat/Asn	6%	12%	8%

[Adam Griffin]

I could speak a lot on what happened in GA-14 since I'm the Democratic Chair of the second-largest county in the district (doesn't really matter now: Kevin remains on the ballot, but needless to say that he wasn't "pushed out" or bullied into his decision), but instead of being a big ol' leaker, I am going to speak about what I've seen in my mailbox in the past 3 days!

Can't blame the Georgia Republican Party for trying (actually this is the second mailer I've received from them; threw the first away). The Democratic Party of Georgia, however, is just trying to raise unreasonable sums of money from me per usual. I'm obviously not a normal voter, but I wonder if this is what/all that people in non-urban environs in GA are seeing?

(I should also point out that I've been getting at least 4 of these boilerplate DPG donation pitches by mail each year for the past 3-4 years asking me for $375-750, so I'm not even sure this can be counted as Election Year mail)

[Adam Griffin]

Sending off tomorrow (once I fill the rest of the races in with "Democrat" write-in votes):

[Adam Griffin]

VBM Requests as of 9/26, % of Total 2018 Vote:
Abrams Counties: 37.90% (817453/2156818)
Georgia, Overall: 33.28%
Kemp Counties: 27.72% (496804/1792128)

[Adam Griffin]

I thought this might be one illustrative indicator beyond simple VBM requests, since the youngest cohorts are the least likely to vote by mail in most circumstances. It's also probably indirectly somewhat of an indicator (statewide) where young people are either fleeing in droves (i.e. parts of South GA) or being swamped by out-of-state retirees (NE GA mountains).

[Adam Griffin]

I wanted to analyze preliminary VBM data specifically for GA Latinos after my earlier map today, given my county (Whitfield) is the most Latino in the state & my county party has been aggressively pushing early & VBM campaigns for Latino voters in our area (our digital operations regularly focus on a radius within 20 linear miles of Dalton - especially for Latinos, especially given the cluster of Latinos on our border in Gordon and Murray specifically). This is because if left to their own devices, many Latinos (along with Asians and young voters at large) tend to wait until Election Day to vote, and often get told they are not registered, don't have proper ID on them or skip most of the down-ballot - which lead to large drop-off and turn-away rates at the polls.

Of the 10 most-Latino counties in the state, I still was surprised to see 3 of the top 4 in terms of equity were in our "turf" (Chattahoochee, #2, is a military base predominantly, so lots of military ballots/VBM isn't unusual).

Image below shows the 10 most-Latino counties, along with their Latino population percentages, the percentage of total VBM requests coming explicily from SOS-identified Latinos, and the "equity" (i.e. the percentage of Latino VBM requests relative to population).

[Adam Griffin]

Is there any reason why the votes cast so far are extremely old? Two thirds of the accepted ballots so far are people over the age of 65!

Anybody 65 or over (or anybody who is disabled, a veteran, or living overseas) who requested a mail ballot during the primary (or primary runoff) automatically receives mail ballots for all future elections in said cycle (primary runoff, general election, state general election runoff & federal general election runoff). As such, several hundred thousand mail ballots were baked in for seniors due to COVID leading to higher requests for mail ballots during the primaries (along with the state mailing VBM applications to everybody).

This is why the senior mail vote is so high (it's not abnormal at all - if anything, their % of overall requests is lower than it otherwise would be). If you look at my map above, it's honestly astounding that the statewide 18-29 VBM request share is now above 10%.

Seniors do seem to be returning them faster - as they are only 46% of requests thus far - but these voters are also disproportionately D compared to past cycles (remember that the primary was 55% D in GA, and since the primary is largely responsible for most VBM requests being mailed out to seniors as explained above...).

[Adam Griffin]

Is there any reason why the votes cast so far are extremely old? Two thirds of the accepted ballots so far are people over the age of 65!

Traditionally, wasn’t the black vote on the day of the election?

Black voters have tended to vote early in-person. Election Day vote has generally been several percentage points less black than in-person EV. As far as VBM is concerned, it has generally resembled ED vote more so than EV vote. Prior to Abrams and her pushing of VBM in '18, mail ballots were generally the GOP's best vote type in Georgia ('16 less so, but still not shabby for the GOP).

Just as an example, 2014-GOV:

[Adam Griffin]

^^^ Let's remember a couple of things:

It's not uncommon at all for anywhere from 50-60% of the 18-29 crowd in a given election to be first-time voters or voters who didn't cast ballots in the previous presidential election. Again, remember that this metric is counting everybody who voted in 2018 but not in 2016: given 2018 was a de-facto presidential election turnout-wise in GA, it's very much worth keeping in mind (especially for such a young group).

The reason for this broader dynamic is somewhat simple: relative to other age groups, there tends to be a small uptick in voters between the ages of 18-21, followed by a drop-off among those 22-25, followed by growth once again in the late-20s.

This means that the youngest spike in voters are first-time voters because they literally weren't able to vote 4 years (or in some cases, even 2 years) prior; more people are interested in voting when it's the first election where they can do so. Among the age range of those who possibly could have voted four years prior (22-25), for whatever reason, interest seems to be lost to a certain degree (maybe because they're no longer in college or living at home, which can complicate voting). Among the 26-29 group, it's a combination of first-time voters finally starting to come "online", as well as those who may have voted 8 years prior (but not 4 years prior), as well as those who have voted every time since they were eligible.

At any rate, when you plug these phenomena into the proverbial equation, it leads to a hefty share (often a majority) of 18-29s being first-time voters. I highly doubt we'll see the figure be anywhere close to 70% once everybody has voted - not only because of usual trends, but also because a lot of younger people turned out for both 2016 and 2018.

[Adam Griffin]

It's hard to say what'll happen with Fayette, but in terms of its growth and primary turnout, the odds would suggest less than a 50% chance of flipping when taking a 30,000-foot view:

'20 Primary, % of '16 Pres GE Turnout:
Fayette: 47.95
Georgia: 49.11

'20 Primary, % of '18 Gov GE Turnout:
Fayette: 50.86
Georgia: 51.63

This is despite the fact that Fayette County is growing at a faster clip than the state as a whole and has the third highest MHI of any county in the state; had real turnout been equal in Fayette relative to the state, its share along these lines should have been higher, not lower. This tentatively suggests a turnout differential was in play here that may have benefitted Democratic vote share in the primary.

[Adam Griffin]

The final pre-EV voting figures (total votes cast: 439018).

[Adam Griffin]

Voter suppression is real but I wonder how COVID restrictions play into these long lines as well. It’s an awful situation all around.

Basically most of it. People don't realize how much of an impact COVID is having. Just like in the primary, there are tons of cutesy videos on social media saying "look how these lines wrap around the block!", ignoring that prior to this year, people standing in line were spacing themselves 1-2 feet apart instead of 5-10 feet. It should also be mentioned that there's a subtle phenomenon here just like with car traffic: when people are further spaced apart, as they're served at the front of the line, there is a cascading delay effect the longer the line is (physically) in terms of people moving forward when able. When you have thousands of people in line over the day at a voting site, this actually starts adding up. Lines in my county were 30-40 minutes consistently today, which is relatively unreasonable...until you consider that regular wait times of 60-90 minutes were common throughout most of 2016 EV in Whitfield.

Add to that the fact that the first day of early in-person voting is always a clusterf[inks]k in GA, as tons of people (relatively) show up on the first day. The reality is that these wait times were largely created by high-info Democrats wanting to "send a message" or whatever, when they could have just submitted an application in 90 seconds online and received their ballot (whether they dropped it off in-person or cast it by mail directly would be up to them).

County election boards likely haven't ramped up staff as much as might have otherwise occurred because 35-40% of 2016 voters have already requested a mail ballot, which takes some pressure off in-person voting (both EV & ED, but especially EV given a metric s[inks]t-ton of GOP voters are still going to insist on voting on ED). Additionally, it's not exactly easy to hire lots of temporary staff when you're having to put people at multiple locations throughout a county, which usually means EV sites in metro areas are understaffed (to contrast, at my county's 1 EV site, there were 6 people checking in voters alone because the site is the county elections board and these people are full-time staff). As an example, Metro ATL had more in-person early voting sites than ever before yet record wait times - because they're spreading their resources too thin in the name of "easier access".


126,876 votes were cast on the first day, which...isn't that impressive. Sure, it's the biggest number for any election in terms of the first day of EV, but:

https://twitter.com/gtryan/status/1315801909744238593

https://twitter.com/gtryan/status/1315819283629441024

[Adam Griffin]

After today's update of voters, the electorate got quite a bit blacker (33.6 -> 34.4) and substantially younger (18-29: 5.2 -> 6.1; 65+: 60.0 -> 52.5).

601,247 voters have cast ballots as of Monday.

[Adam Griffin]

Well, everybody (effectively) is registered in GA now: number has to be at least 97% of VEP at this point.

GEORGIA EASY VOTER REGISTRATION OPTIONS BREAK NEW RECORDS

(ATLANTA) – Georgia has hit a new record high registration level. As of October 6, Georgia had 7,587,625 registered voters throughout the state. Over 5 million of those came through automatic registration at the Department of Driver Services, a program of the Office of the Secretary to make registering new voters and updating voter information simple and easy.

“The continued growth of Georgia’s registered voting population is a testament to the simple and easy registration options the Secretary of State’s office provides to Georgia voters, including automated registration through DDS,” said Secretary of State Brad Raffensperger. “As Georgia’s strong economy continues to attract the best and brightest from around the country, even during the pandemic, my office has the infrastructure in place for anyone who wants to register and participate in the democratic process.”

Georgia voters have taken advantage of the numerous convenient and easy ways to register. Of the 7.6 million registered voters according to preliminary data, 5,002,856 of them have registered through automatic registration at the Department of Driver Services. Another 734,000 of those registered voters did so online through the Secretary of State’s website. 1,850,464 registered with paper registration applications.

The final numbers will likely be higher as counties continue to process registration applications, including those submitted online and registrations completed by DDS.

The 7.587 million total registered voters for the November 2020 general elections represents a more than 600,000 registered voter increase since November 2018.

Historically, Georgia has had incredibly high voter registration rates. According to the 2016 U.S. Election Assistance Commission Survey, 95.4% of Georgia’s eligible voting population was registered to vote in 2016. The number of overall registered voters increased from 6.657 million in November 2016 to 6.944 million in November 2018.

Georgia is recognized as a national leader in elections. It was the first state in the country to implement the trifecta of automatic voter registration, at least 16 days of early voting (which has been called the “gold standard”), and no excuse absentee voting. Georgia continues to set records for voter turnout and election participation, seeing the largest increase in average turnout of any other state in the 2018 midterm election and record primary turnout in 2020, with over 1.1 million absentee by mail voters and over 1.2 million in-person voters utilizing Georgia’s new, secure, paper ballot voting system.