Georgia's Very Own Megathread! (v2)

[Adam Griffin]

RIP Georgia's Very Own Megathread (v1; 2012-2018)

OBLIGATORY WATCHIN'

Griff's 2002-2018 Animated Precinct Map

Know Your History (Election Results):
1988: President
1990: Senate | Governor
1992: President | Senate (General) | Senate (Runoff)
1994: Governor
1996: President | Senate
1998: Senate | Governor
2000: President | Senate
2002: Senate | Governor
2004: President | Senate
2006: Governor
2008: President | Senate (General) | Senate (Runoff)  
2010: Senate | Governor
2012: President
2014: Senate | Governor
2016: President | Senate

Know Your History (Runoffs):

Every candidate in Georgia elections (save for Presidential candidates) must receive 50% + 1 vote to win office. In the event no candidate - whether it be in a municipal, countywide, state legislative or statewide election - receives that amount, a runoff shall be held between the top two vote-getters 4 weeks (in the case of non-federal offices) or 8 weeks (in the case of federal offices) after the general election.

Fun Fact: Runoff requirements have been changed multiple times throughout Georgia history. Until 1993, the threshold for avoiding a runoff was 50% +1 vote. After Wyche Fowler lost his Senate seat to Paul Coverdell in a Senate runoff, the Democratic legislature changed the provision to 45% +1 vote, where it would remain until the GOP took the majority in the General Assembly. In 2005, the GOP majority reverted the runoff threshold to 50% +1 vote, where it has remained ever since.

Another Fun Fact: From 1936 to 1964, Georgia had laws on the books to ensure presidential runoffs would occur if no candidate received 50% +1 vote. In the event of a runoff, the Georgia General Assembly would decide the winner. This provision was never utilized, due to no candidate winning GA with a plurality during this time. This provision was scrapped in 1968 in support of George Wallace, who ended up winning the state with a plurality for the first time in history.

[Adam Griffin]

Something I just noticed - while not particularly meaningful - is that in the PSC race, Democrats outvoted Republicans in the primary:

PSC District 3:
Democrats: 477,571
Republicans: 471,261

[Adam Griffin]

This kind of stuff is nothing new. Notice how it said 10 other counties (most of which are heavily-black) have done the same in effect? The only sigh of relief I can muster is that Randolph County only gave Democrats a net gain of 117 votes in 2014 (1110-993). I don't know if the Chairman is a Republican or just a Dixiecrat but if it's the latter then it wouldn't surprise me one bit that they're trying to gut black representation in favor of white Democrats (see below).

Also, let's not forget that Nathan Deal & Brian Kemp also brought the full force of their offices and the Georgia State Patrol to forcibly remove three black school board members in Brooks County after they organized a campaign in 2010 to encourage vote by mail - which led to a black majority on the school board for the first time, and the ouster of the white Democratic majority who initiated said witchhunt - and went around intimidating every black voter in the county literally on their doorsteps. They were all acquitted four years later.

Oh, and that ignores the intimidation of organizations like New Georgia Project & AALAC by threatening legal action against them for purportedly registering voters incorrectly. Neither investigation found any (legal) wrongdoing.

[Adam Griffin]

Abrams drew huge crowds down in Ben Hill, Echols, and Lowndes counties today. When I hear her speak she is really hitting her stride her focus is squarely on that Governor’s mansion.

Yeah, looks like the VSU event drew 250-300 people easily. The indoor venue was for 160 and there were 100+ outside who couldn't get inside.

With the campaign consistently exceeding capacity at stops in less-than-friendly places all around the state, I wonder if they're ever going to start booking bigger venues?

[Adam Griffin]

I'm going to fill out the OP in this thread with some vital stats and info; any recommendations would be appreciated. Here's what I briefly already added:

RIP Georgia's Very Own Megathread (v1; 2012-2018)

Know Your History (Election Results):
1988: President
1990: Senate | Governor
1992: President | Senate (General) | Senate (Runoff)
1994: Governor
1996: President | Senate
1998: Senate | Governor
2000: President | Senate
2002: Senate | Governor
2004: President | Senate
2006: Governor
2008: President | Senate (General) | Senate (Runoff)  
2010: Senate | Governor
2012: President
2014: Senate | Governor
2016: President | Senate

Know Your History (Runoffs):

Every candidate in Georgia elections (save for Presidential candidates) must receive 50% + 1 vote to win office. In the event no candidate - whether it be in a municipal, countywide, state legislative or statewide election - receives that amount, a runoff shall be held between the top two vote-getters 4 weeks (in the case of non-federal offices) or 8 weeks (in the case of federal offices) after the general election.

[Adam Griffin]

Why Abrams is making a beeline for GOP territory in Georgia

You must be logged in to read this quote.

[Adam Griffin]

I actually started a project a week or two ago, trying to map out what Abrams would need in terms of benchmarks in each county for a majority (of course, one has to make some baseline assumptions in regard to turnout and persuasion). Maybe I'll get around to finishing it soon, but for:

Fulton: 69.5-28.6
Dekalb: 81.0-17.3

[Adam Griffin]

So in regards to my "50% +1" model: here's the problem I'm running into (and one I've expressed on here before).

Starting with 2014 as the baseline (in terms of margins & total votes; however, this is based on margins predominantly, meaning the raw vote isn't relevant except in making my initial spreadsheet formula calcs easier), Abrams would be down 262k votes (or 10 percentage points; in terms of getting to 50%). She must overcome not only the 202k vote deficit that Carter had, but an additional 60k in third party votes.

I can see easily how she can net 200k votes and actually pull off a plurality. After stretching the 30 counties in green (i.e. most-populated and/or regional metros; 52-44 Clinton) to what I believe are the realistic maximums margin-wise, she nets 203k over Carter - which is coincidentally almost the exact amount by which Deal beat Carter. But she still needs another 59k net votes to win a majority...out of the areas that remain (counties in red & blue; 69-29 Trump).

In fact, the areas in green are ~75% of the state's population and Abrams has pulled 75% of her needed majority margin from those counties, meaning...she'd need swings in that heavily-GOP remainder of the state to be just as big in her favor as in the already-calculated area (a huge segment of which is Metro ATL & urban GA). This is where the math gets difficult - unless you start assuming that she can pull 75% in Fulton, 85% in Dekalb, 60% in Chatham, etc.



Again, the raw vote totals here aren't exactly important - the margins are what matter. In the above raw totals, I simply assumed that for every vote Abrams gained through turnout, Kemp lost 1 via depressed GOP turnout (I think this is an acceptable method, given that Abrams is focused on turnout rather than persuasion). Nevertheless, you could model this with identical GOP 2014 totals paired against an all-turnout Democratic effort, a modest Democratic increase in turnout and substantial metro R-to-D defections, or depressed GOP turnout coupled with Democratic enthusiasm being through the roof (or any combination). However, my formula's primary focus is built on margins, not raw vote; there's multiple ways to express the latter, but only one way to express the former.

Nevertheless, the important thing to remember is that every single one of these green counties was projected to be better - substantially in most cases - for Abrams than for Carter, and that still only gets her 75% of the way to a majority in a segment of the state that is 75% of its population, and is much more friendly to her than the remainder.

[Adam Griffin]

^^^ This, I believe, is why Abrams herself is - as the AJC put it - "making a beeline for GOP territory" (including some very hostile turf). Let the field organizers and volunteers hammer away at the urban clusters - it's needed - but without serious gains elsewhere, it's a plurality instead of a majority. It's also probably why places like Dalton and Rome are suddenly getting paid staff in them.

[Adam Griffin]

I love how bronz is implying there are whites in Atlanta. Or blacks in Georgia outside Atlanta for that matter.

[Adam Griffin]

I don't get the Yates thing. Literally the only reason anybody knows who she is is because Trump fired her. I'm all for political outsiders if I think they can win, but I see nothing about her that is uniquely inspiring or noteworthy that couldn't be found in a variety of fundamentally better candidates. She's basically Stacey Evans if you took away her ability to relate to rural voters (and probably would end up being the female version of Jim "I'm Not a Politician: I Wear a Hat!" Barksdale).

[Adam Griffin]

Alright, so I've finished my first draft of Georgia 2018 Abrams win benchmarks. I've decided for the time being to settle on a nominal win rather than a majority win, simply because I'm not sure where else it's feasible to stretch the numbers to get to 50%. Nevertheless, I did manage to get Abrams to 49.3% when all was said and done; in the end, this is equivalent to closing 223k of the 262k deficit Carter had in 2014. Obviously the easiest way for Abrams to get a majority with these figures as a baseline is if Kemp's raw turnout is lower than Deal's was in 2014.

I'm open to suggestions or recommendations in some parts of the state. Below is a map and the list with each county's benchmarks. I'll probably turn this into an interactive map in the next day or two.

County-by-County Benchmarks (2014 and 2018 Results + Swings)

[Adam Griffin]

Did somebody really merge both of bronz's GARBAGE THREADS into this sacred ground?

[Adam Griffin]

A post on the 2020 board reminded me of the straw poll we conducted at this year's Whitfield Kennedy-Carter Dinner. Approximately 150 people voted in the straw poll; here are the results:

You must be logged in to read this quote.

[Adam Griffin]

Alright, so I've finished my first draft of Georgia 2018 Abrams win benchmarks. I've decided for the time being to settle on a nominal win rather than a majority win, simply because I'm not sure where else it's feasible to stretch the numbers to get to 50%. Nevertheless, I did manage to get Abrams to 49.3% when all was said and done; in the end, this is equivalent to closing 223k of the 262k deficit Carter had in 2014. Obviously the easiest way for Abrams to get a majority with these figures as a baseline is if Kemp's raw turnout is lower than Deal's was in 2014.

I'm open to suggestions or recommendations in some parts of the state. Below is a map and the list with each county's benchmarks. I'll probably turn this into an interactive map in the next day or two.

County-by-County Benchmarks (2014 and 2018 Results + Swings)

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


Abrams - County Plurality Benchmarks (Margin)

Abrams - County Plurality Benchmarks (Swing)

[Adam Griffin]

Why does Brooks County flip but not Lowndes?

Blacker by current registration figures, remains Democratic at the local level (unlike Lowndes), is more elastic (voted for Barnes in 2010), had half the margin of loss of Lowndes in 2014 and I believe Abrams is exactly the right person to motivate voters here, given what happened in 2010 (both good and bad).

[Adam Griffin]

Is it possible Lowndes could flip, given there is an urban area (Valdosta) in the county?  (Granted, Valdosta isn't some mega city or anything, but enough voters could maybe change things?)

I mean, it's possible, but as you already pointed out, it's not really urban in a true sense; Lowndes is basically Whitfield's southern equivalent/cousin in a lot of ways. Coupled with the fact that both whites and blacks tend to be essentially 90% R/D, making up the difference in a county with a base electorate of 65k RV @ 60% white/35% black is a tall order.

Historically, Lowndes tracked well with the state; its margin tended to be 1-5 points more GOP than the state as whole barring a couple of exceptions, but beginning in 2014 (and definitely in 2016) it's been drifting to the right. In most recent past elections, I would consider it a bellwether-like county in that you could look at its results and get an idea of what would be the state's margin, but not so much in the past few years. Whether that continues to hold or not, I don't know...but I imagine Abrams would need to win by at least 5 before Lowndes is seriously in contention.

[Adam Griffin]

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


Abrams - County Plurality Benchmarks (Margin)

Abrams - County Plurality Benchmarks (Swing)

Insufficient. Your figures don't get Abrams to 50% plus one, so the race goes to a runoff which she almost inevitably loses.

The problem is that I didn't want to stretch counties beyond what I could see as being realistic, which...shows what I believe to be the underlying problem of this race. Nevertheless, one can look at these figures and - with the statewide total being within less than a percent of a majority - compare on Election Night; if she's hitting or exceeding these in most areas, then she has a chance. If not, then no.

The path of least resistance (at least in terms of simplicity and compactness) for a majority win would involve her getting 2 percentage points more of the vote in Fulton, Dekalb, Gwinnett & Cobb (71%/83%/55%/52%) and getting 1 percentage point more in Henry, Cherokee, Forsyth, Douglas, Rockdale & Newton (53%/27%/28%/56%/65%/53%).

Still, this is built on a 2014 baseline in terms of vote distribution, meaning that each county is the same share of the statewide electorate as it was in 2014. Nevertheless and even though that is likely not to be the case, the shift in margins in these counties isn't built on persuasion - meaning that the swings in favor of Abrams in the benchmarks essentially reflect higher Democratic turnout from unlikely & first-time voters indirectly, but distribute it more evenly than it may end up being. Still, I imagine the movements in terms of counties' shares of the statewide electorate will be small in relative terms; Fulton isn't going from 10% of the electorate to 20%, Cherokee's not going from 3% to 2%, etc.

Perhaps the biggest point to take from all of this is that Abrams is going to need to improve by more than half as much over Carter outside of broader Metro ATL (8 points) as she needs to improve within it (14 points).

[Adam Griffin]

Alright, so I went ahead (even though I think it's a stretch) and added the enhanced metro figures I mentioned above to get to 50%. There - happy?

[Adam Griffin]

When Abrams loses the majority by 1 vote after DTC votes for Kemp

[Adam Griffin]

Lol y’all, go to briankemp.com

Lolololol

[Adam Griffin]

Remember that Channel 2 and AJC have unfortunately moved to using university polling (UGA specifically) and that their poll of the Democratic primary in late April was something like 32-15 Abrams. Probably still better than that garbage InsiderAdvantage polling, though.

[Adam Griffin]

Harry Enten thinks Abrams could win in a run-off. This seems to make sense abstractly (drop-off from a Presidential election to a runoff is not the same as drop-off from a midterm to a runoff). But does it make sense in practice, my dear Gawgans?

https://twitter.com/ForecasterEnten/status/1037832535286456320

You must be logged in to read this quote.

The primary problem with this analysis is that we don't need to look at 1992 or 2008 (or a hypothetical midterm...or Louisiana) for comparable climates - though they are potential data-points in reinforcing a broader trend/likelihood (and ones which I've mentioned in the past).

Democratic turnout in special elections in Georgia in 2017-18 has been lower than in 2016 in real terms (in other words, margins swinging to the GOP). The only other state where this has occurred is CT, which has some pretty unique and obvious trends afoot there causing Democrats to underperform (i.e. voters there hate CT Dems right now). I'd also point out that with the data that was publicly available (which was incomplete), LA looked pretty bad, too - though Enten et al probably have access to more detailed and complete data-sets regarding turnout and presidential performance by legislative district.

Essentially, when broken down by HD/SD, you see a tale of two states in Georgia: that Democrats are actually improving in GA in areas where whites and/or old voters still make up a reasonable share of the electorate, but plummeting in areas where blacks and/or young voters are a huge share.

* Incomplete results; do not have data for Clinton margins in all special elections

Is it possible for Abrams to do better in a runoff than in a general election? Yes, absolutely - it could very well happen, based on her specifically and the power of her campaign. But to assume before anything else that all of the evidence that points to a clear track record and likelihood of underperformance should be dismissed in favor of what we'd like to see happen simply because one candidate is good is reckless IMO.

[Adam Griffin]

Abrams needs to crack 30% with white voters in order to have a shot at winning, which she isn’t doing so far.

No she doesn't.

Except she does... According to this poll, she’s getting 45% of the vote statewide and only 27% of white voters. Yknow how well Michelle Nunn did statewide 4 years ago? 45%. She won 27% of white voters according to exit polls. Those numbers sound familiar?

Abrams needs that white vote total higher if she wants a shot at a majority in November. Even if she does boost black turnout higher, I don’t think she’s got enough support for an outright majority with that 27%.

Technically, both Nunn and Carter got the standard 23% (as did Obama 08 and Kerry) that GA Democrats have tended to hit in recent cycles. To be fair, she could probably win with 27% of whites if the electorate looks more like a presidential one than a midterm one:

Whites: 60% @ 0.27 D = 16.2
Blacks: 31% @ 0.93 D = 28.8
Other: 9% @ 0.60 D = 4.5

Total 100% @ 0.504 D = 50.4%

I recall my projections for potential presidential outcomes in 2016 having Clinton being able to pull off a win with those kinds of white figures; basically, if Clinton had done as well as Obama 08 with the 2016 electorate, she would have won.

[Adam Griffin]

I’ve been on my way to Dahlonega today and have yet to see an Abram’s sign.

You're not going to see many (if any) signs in Deliverance Country because people are literally terrified that their homes, properties and selves will be targeted. When I was putting out signs for Clinton in 2016, I had multiple instances of people yelling "F[INKS] THAT B[INKS]H" out of car windows passing by - and I'm in one of the more friendly and civilized parts of the state. I was pretty much expecting to be hit in the head with a bottle or ran over by some bubba truck driver. I know organizers who insist on carrying guns whenever canvassing or doing visible Democratic work in public in North Georgia. The fact that it's a black woman running will only heighten these fears by the minority and these aggressive behaviors by the majority.

Besides, the Abrams campaign has put the least amount of effort into signs of any campaign - presumably because they understand it's a waste of money. There have been periods where people couldn't even buy their own overpriced signs for $10 through the website. Signs are good for identifying potential hardcore supporters, volunteers and donors (and may be good at giving people reassurance/knowledge of like-minded people at a granular level), but if you're already accomplishing the former via field work, it's pretty unnecessary. I hate the sign game and I hate being nagged about keeping them in supply, but I do admit that for county parties and the like, they provide more benefit than for statewide campaigns with massive resources.

And for those wondering, past precedent/studies suggest that having a 2:1 sign advantage in a particular area can affect the margin by around 1 point in urban areas, 2 points in suburban areas and 3 points in rural areas. It's tangible, but not the most efficient way to achieve those gains.