Early voting, absentee requests & statistics

[Adam Griffin]

At first I thought I understood, but then I realized I didn't. Not sure if the charts are formatted properly for GA...can anyone explain to me what all of this "w/2010 vote", "w/o 2010 vote" and other categories even mean? I can't make sense of all of these numbers for my state...

[Adam Griffin]

At first I thought I understood, but then I realized I didn't. Not sure if the charts are formatted properly for GA with all of the racial breakdowns included...can anyone explain to me what all of this "w/2010 vote", "w/o 2010 vote" and other categories even mean? I can't make sense of all of these numbers for my state...

[Adam Griffin]

Some updated Georgia numbers. I've downloaded the huge absentee list to assess the count of all ballots cast and requested up through Monday (10/20), and also sorted it to show how many of those ballots have been cast and returned. In addition to this, I'm using the public voter file I have access to in order to break down early voters (through Sunday; doesn't include today's voters) by "party" (which is assessed from party primary voting records).

As of 10/20:

Ballot Status
All Ballots Cast/Requested: 241,920
Ballots Cast/Returned: 185,387

As of 10/19 (First Week + Sunday)

Historic Public Voter File Data (Based on Primary Voting History)
All Ballots Cast/Returned: 156,106

Republican 43.1%
Democrat 39.5%
Independent 6.1%
No Data 12.3%

(It should be noted that public voter file data/primary voting history skews Republican these days, due to a near-supermajority of Georgians living in Republican-dominant counties, so the numbers are quite possibly even better than they'd appear.

[Adam Griffin]

Are those numbers good or bad for Nunn?

Hard to say considering that I've never done this sort of comparison before. I did just do some sketchy math that attempts to take into account what I know about Georgia's primary dynamics (which is practically everywhere, although to a greater degree of skewing in some areas).

I started out by assessing areas where one party is the dominant force/primary is equivalent to election, which gives you 63% R / 37% D in terms of percentage of population that lives in an area where that party's primary is more influential. I then took what I know about the open primary voting habits of people from the opposite party in such circumstances and tried to reverse engineer numbers to put people where they actually belong. After all of that, I got this:

Democratic: 41.1%
Republican: 37.5%
Independent: 9.1%
No Data: 11.3%

[Adam Griffin]

Georgia:

Early voting figures by race look good - but more so bad.

Through Tuesday, 2014:

Total Votes Cast: 239,749

Whites: 66.7%
Blacks: 28.3%
Other: 4.0%
Latino: 0.5%
Asian: 0.5%

And to put it into perspective, here are the figures for the entire period of early voting in 2010:

You must be logged in to read this quote.

Get it together, minorities: nobody's going to take you seriously if you don't vote.

[Adam Griffin]

As of 10/20:

Ballot Status
All Ballots Cast/Requested: 241,920
Ballots Cast/Returned: 185,387

As of 10/19 (First Week + Sunday)

Historic Public Voter File Data (Based on Primary Voting History)
All Ballots Cast/Returned: 156,106

Republican 43.1%
Democrat 39.5%
Independent 6.1%
No Data 11.3%

(It should be noted that public voter file data/primary voting history skews Republican these days, due to a near-supermajority of Georgians living in Republican-dominant counties, so the numbers are quite possibly even better than they'd appear.

Numbers appear to be moving in Dems' direction, relatively speaking.

Through 10/22:

Ballot Status
All Ballots Cast/Requested: 348,768*
All Ballots Cast/Returned: 281,185

Republican: 42.1% (-0.4)
Democrat: 39.1% (-1.0)
Independent: 5.7% (-0.4)
No Data: 13.1% (+1.

* Through 10/23


Racial breakdowns are also moving favorably:

Through 10/22:

Total Votes Cast: 281,185

Whites: 66.3% (down from 66.7% on Tuesday)
Blacks: 28.6% (up from 28.3% on Tuesday)
Other: 4.2% (up from 4.0% on Tuesday)
Latino: 0.5% (same as Tuesday)
Asian: 0.5% (same as Tuesday)

And to put it into perspective, here are the figures for early voting for 2010:

Total Votes Cast: 678,939

Whites: 66.5%
Blacks: 29.0%
Other: 3.6%
Asian: 0.5%
Latino: 0.4%

What's your source for those numbers?

Michael McDonald has them (as of today) at 62.9% white, 27.9% black and 8% new registration/unknown race, which would be pretty decent numbers for the Dems. Yours look much worse - as if unknown race EVs are distributed heavily toward whites and 'other'.

It's coming from a public voter file system that I have access to, which uses SoS information. I need to track down SoS' exact file so see why there's such a big discrepancy. Perhaps he's counting ABMs that haven't been returned (which my race breakdowns don't), but I wouldn't think there's that big of a racial difference between those that have been returned and those that haven't to cause such an effect.

[Adam Griffin]

Georgia, ballots cast:

(Figures with "(^)" next to them indicate group as a % of early voters is consistently increasing)
(All 2014 early voting totals are through Sunday, 10/26; 2010 totals are for entire early vote period)

Georgia is looking quite good for Democrats so far. There has been an seven-point swing thus far in early voting turnout by likely party when compared to 2010 (!!), with it being likely that the number will continue to improve for Democrats over the remainder of this week, if historical trends are any indicator. The number of unknown affiliated being slightly higher is potentially an indicator that there are more first-time voters voting early.

By race, blacks are two percentage points more of the electorate than they were in 2010, and that number will also continue to increase. In 2010, blacks were 29% of EVs and the final number was 28%, suggesting that the total black share of the electorate in 2014 could very well be 30-31%.

The female percentage of the vote is still a bit below 2010 numbers but steadily increasing, and is on track to meet or surpass 2010 numbers. The current breakdowns by age when compared to 2010 are the least optimistic, but thankfully in Georgia (at least in this case), voting preference by age is relatively uniform across the board.

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

[Adam Griffin]

Georgia, ballots cast:

(Figures with "(^)" next to them indicate group as a % of early voters is consistently increasing)
(All 2014 early voting totals are through Sunday, 10/26; 2010 totals are for entire early vote period)

Georgia is looking quite good for Democrats so far. There has been an seven-point swing thus far in early voting turnout by likely party when compared to 2010 (!!), with it being likely that the number will continue to improve for Democrats over the remainder of this week, if historical trends are any indicator. The number of unknown affiliated being slightly higher is potentially an indicator that there are more first-time voters voting early.

By race, blacks are two percentage points more of the electorate than they were in 2010, and that number will also continue to increase. In 2010, blacks were 29% of EVs and the final number was 28%, suggesting that the total black share of the electorate in 2014 could very well be 30-31%.

The female percentage of the vote is still a bit below 2010 numbers but steadily increasing, and is on track to meet or surpass 2010 numbers. The current breakdowns by age when compared to 2010 are the least optimistic, but thankfully in Georgia (at least in this case), voting preference by age is relatively uniform across the board.

...

I just realized with the most recent update that I had some suppression features enabled in the voter file system that was excluding roughly 40,000 people from the counts (LOL, who am I: the Secretary of State?). This explains the 100,000-vote jump between yesterday's update (Monday) and today's (Tuesday). These numbers are insane compared to four years ago, and look much better than even yesterday (in part because the suppressions are now gone).

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

[Adam Griffin]

Louisiana: With early voting finished, 236K votes were cast. Thats double the voters that were cast early in 2010, 80% 2008 and about 2/3 of 2012.

The overall race total was 64.6% white 32.8% black. The partisan split was 52.5 D 34 R 13.5 I.

How do the stats by race compare to 4 years ago in early voting?

[Adam Griffin]

I just posted an update on GA, but the voter file just updated on my end with Tuesday's totals, so here's another summarized version:

607,569 voters
Black share steady (32.0%)
White share steady (62.3%)

Democrats slightly down (40.0%)
Republicans steady (39.2%)
Unknowns rising strongly (14.5%)

Females up (54.9%)

Large increase of 31-50 year-olds (18.8%)

[Adam Griffin]

Georgia, through Wednesday:

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

[Adam Griffin]

Georgia, through Thursday. We have broken the 2010 early voting record by more than 115,000 votes, and my original projections of 750,000 - 800,000 seem to have been too conservative. It's likely that we finish this up closer to 900,000 early votes. The black share of the vote continues to increase and is on track to be 3 to 3.5 points higher than 2010, but we're probably near or at the peak of it in terms of percentage of the electorate. Females are also a slightly larger share of the early electorate than in 2010. Both likely Dems and Reps continue to decrease as a % of the identifiable electorate, but Dems maintain a overall lead in this category by close to 1 point. This early vote electorate as a whole seems to skew a bit older than 2010's, but in Georgia, ideology and voting preference is fairly uniform across age groups.

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

[Adam Griffin]

Interactive Georgia Early Vote Map: BK and I worked on collecting 2010 data by county tonight in order to bring you this - a map that compares 2010 early voting totals to the current 2014 early voting totals for each Georgia county. We still have yesterday (Friday's) totals to update in the VAN (hopefully today), but by and large, the picture is painted.

To make everything relative until it is updated with the final data, counties with 2014 vote totals of 106% or greater are out-performing the state as a whole in terms of the increase in early voters between 2010-2014; those less than that are obviously under-performing.

Counties shaded red are <95% of their 2010 totals; white is 95-105% of 2010 totals; green is >105%. There's a legend on the actual map below (for example: "0.98" equals 98%). You can click on each county for more details/percentages.

2014 Early Vote Turnout as a Percentage of 2010 Early Vote Turnout

[Adam Griffin]

Georgia, Final Early Vote Totals and Analysis:

Wow. I'm not sure what it looks like in every other state (I'll compare later), but the total share of early voters in Georgia jumped from 678,000 in 2010 to 930,000 in 2014.

The black share of the electorate jumped by close to four percentage points (from 29.0% in 2010 to 32.7% in 2014). The white share of the electorate shrank by more than five percentage points (from 66.5% in 2010 to 61.1% in 2014).

When we go by party affiliation based on primary voting record (on a statewide level, this is more accurate than it'd appear and I've explained in detail before why this is), the early vote electorate in Georgia effectively swung from R+6 in 2010 to D+1 in 2014.

By age and by gender, we saw a fairly steady hold between 2010 and 2014. Females were 55% in 2010; closer to 56% in 2014. The youngest and oldest brackets grew and shrank very modestly, respectively, while the 51-64 age group saw the largest growth between 2010 and 2014.

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

You must be logged in to read this quote.

Here are two maps that show 2014 early vote turnout compared to 2010 early turnout. The first one shows it in sheer numbers; the second one shows it in a "trend-like" format (counties that saw EV percentage growth greater than the state as a whole are green; those that saw less or shrank are in red).

2014 Early Vote Turnout as a Percentage of 2010 Early Vote Turnout



TREND - 2014 Early Vote Turnout as a Percentage of 2010 Early Vote Turnout

[Adam Griffin]

Griff,

With numbers like this, do you think Nunn can win on election night? Most polls have been showing a slight Perdue lead with a likely runoff, but these are very impressive figures for the D's at this stage.

If she (or Carter) can hit 29-30% with whites and blacks hold at 90% D, then they get a majority when combined with this math; perhaps with even a bit less of the white vote, as I haven't done calculations since blacks were at 31% of the electorate.