Early & Absentee Voting Megathread - Build the Freiwal

[Adam Griffin]

Here's Griff's patented Election Day Hysteria Generator!

Have fun

You must be logged in to read this quote.

[Adam Griffin]

https://twitter.com/ElectProject/status/1044424766889906176

let the hype begin!

WRONG!

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/1045140690928308225

[Adam Griffin]

What I would be especially interested in is what the equivalent #s were in 2008, when there was such huge early voting for Obama in GA by African Americans. How does that compare (obviously in a Presidential election) to what we see this year regarding GA early voting?

Mail ballots in GA have - by default - historically been more Republican than the electorate, but have been shifting toward the Democrats in the past few years. This probably was also the case with early voting in general before Democrats moved to pursue it here beginning in 2008. Since voting by mail isn't really something that has been pushed aggressively by Democrats (outside of the states where it has been common for some time) until 2016 or so, I feel pretty confident in these percentages being unprecedentedly Democratic compared to past cycles. The only comparison we can't definitely make is to the raw figures (# of votes), but given the big push this year, those numbers are probably going to shatter all records, too.

However, it's worth noting that there was the same vote by mail campaign being pushed in GA in either 2014 or 2016 (can't remember which election) by the Voter Protection group setup by the DPG; the reason I remember this is because the applications/envelopes mailed out don't actually disclose who is sending them, and in both of these elections, I've fielded calls and questions about whether the VBM letters were actually legitimate. I'm not sure why the DPG et al can't just put their name on the letter or envelope to give people more confidence that it's not a GOP ruse or what have you.


EDIT: Comparing figures to the past 3 elections is easy to do since the SoS adopted the new, non-PDF reporting pages.

Mail ballots were actually more Democratic in 2016 than Election Day and early in-person voting (which were almost identical to one another), but you can see going back to 2014 and 2012 that the mail ballots were the most GOP category; a clear trend has been forming since.

[Adam Griffin]

It's worth noting that there does seem to be some volatility in the leans of each group depending on the type of election in GA. For example, the Election Day vote in the gubernatorial primary was the most Democratic category (excluding provisionals), while early in-person voting was the most Republican (though barely more so than mail ballots):

[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

[Adam Griffin]

It's very hard to read anything into this early ballot request spike in MN, because a rise in absentee ballots doesn't necessarily mean higher turnout overall.

Still, I would like to see an increase in turnout in these mid-terms, because turnout in US midterms is always very crappy.

Besides, MN just implemented no-excuse early voting in 2014. Prior to that, you had to deliver a written request for why you should be allowed to vote early by mail or in-person. A huge increase compared to the first time the process is used would be completely normal; certainly doesn't look impressive compared to the infinite percentage growth between 2010-2014!

[Adam Griffin]

The "other" category is disproportionately first-time voters in GA who are Latino and Asian, but of the remaining "others" who don't fall into those groups, the white/black ratio tends to reflect the electorate as a whole (i.e. 2:1 white statewide). This is because, as has been said, race on the form is no longer a requirement. I believe it may also have something to do with more federal registration forms being processed that don't record or forward this information.

I have noticed recently, however, that of those who remain on the rolls for a cycle or two, the percentages identified as "other" seem to decline. This seemingly indicates that the state is collecting information about these individuals from other sources and applying it to their records (not entirely sure about this). The new AVR system via the DMV (that updates or submits voter registrations unless you opt out) may be the primary culprit.

[Adam Griffin]

Has this been posted?

https://sos.iowa.gov/elections/pdf/2018/general/AbsenteeCongressional2018.pdf

In absentee numbers (sent):

Democrats: 83,734
Republicans: 20,564
No Party: 25,979
Libertarian: 756

IIRC, Democrats tend to do well in Iowa with absentee voting but I don't recall it being that big of a gap.

I find this interesting (stats have been updated for today):

IA-01: 60.9% Democratic
IA-02: 66.5% Democratic
IA-03: 65.8% Democratic
IA-04: 59.2% Democratic
IA-Total: 63.3% Democratic

The fact that there's such relative uniformity between the districts (especially compared to IA-04) is impressive. Of course, IA-04 only has 60-75% the # of requested ballots compared to the other 3 districts, so it's not necessarily indicative of anything other than rampant Democratic enthusiasm. IA-03 having almost the same # of ballots requested and being more D than IA-01 is pretty amazing, though.

[Adam Griffin]

And for what it's worth:

https://twitter.com/ElectProject/status/1049707604161773568

[Adam Griffin]

I'm not going to spam the thread, but here are a few updates from the past day or two for Georgia:

According to Georgia Votes, an additional 4,836 valid votes were received compared to yesterday; today's accepted ballot total (returned minus rejected) stands at 45,265 votes. The white share of the vote increased from yesterday, as did every other non-black group. Over 65s have been in relative free-fall the past 2 days, losing almost an entire point of their share of the electorate.

VBM turnout is up 115% compared to this time in 2014.

6,812 new ballots were requested since yesterday, bringing the total of requested ballots to 174,012.

Code:

White	20143	44.5% (+0.3)
Black	18789	41.5% (-0.5)
Latino	1079          2.4% (+0.1)
Asian	1381	         3.1%  (+0.2)
Other	3873	         8.6%  (+0.1)

Female	26498	58.5% (+0.1)
Male		18023 	39.8% (-0.2)
Unknown	744	        1.6%   (0.0)

18-29	 2928	6.5%   (+0.1)
30-39	 2424	5.4%   (+0.2)
40-49	 3407	7.5%   (0.0)
50-64       10294	22.7%  (+0.1)
65+         25536        56.4%  (-0.5)
Unknown	  676        1.5%    (0.0)

Looking at the 5 big metro counties, the white share of the electorate in VBM seems to be higher there than it is statewide; inversely, the non-white share of VBMs is lower there:

"The Big 5" (Fulton/Dekalb/Clayton/Gwinnett/Cobb) VBM Returns
44.4% White

GA Outside "The Big 5" VBM Returns
44.1% White

And here are the 2017 Census estimates:

"The Big 5" (Fulton/Dekalb/Clayton/Gwinnett/Cobb) 2017 Pop
37.2% White

GA Outside "The Big 5" 2017 Pop
61.5% White

Think about that for a second.

I'm not sure what percentage of requested mail ballots usually get returned (and I'm not sure where I might find those stats for previous years), but, there were 107k returned mail ballots in 2014. In 2018 and as of right now, 159k ballots have been requested.

I just got home and saw that the Abrams campaign sent me a reminder to return my ballot. Already turned in but wow at the earnestness of getting each and every vote. She rocks!

I got one today too (and I've also returned my ballot already - later than I planned, but alas):

 

[Adam Griffin]

Iowa is looking very good so far. It's still early, though, but it looks like we're at least going to see a good improvement on the 2014 numbers.

Keep in mind that EV in IA skews Democratic and looked deceptively good for Democrats in 2016 as well, though. I don’t think we can draw too many conclusions from this.

I'm not sure about 2016, but:

And for what it's worth:

https://twitter.com/ElectProject/status/1049707604161773568

[Adam Griffin]

According to Georgia Votes, an additional 76,929 valid votes were cast yesterday; today's accepted ballot total stands at 129,517 votes: 5.08% of total 2014 turnout. More than three times the number of votes were cast on the first day of in-person early voting this year than were cast on the same day in 2014.

The white share of the electorate obviously surged today, with the electorate that voted today pretty closely resembling that of 2014 and 2016's electorates overall (60% white, 30% black). The female share of the electorate fell precipitously, as did the 65+ share (mostly due to 50-64 year-olds voting). In short, the electorate got considerably whiter, more male and younger (at least when comparing extremes).

1,377 new mail ballots applications were received today, bringing the total of requested mail ballots to 187,101.

Early vote totals are 181% higher than at this point in 2014 (up from 127% yesterday).

Code:

White	69548	53.7% (+9.3)
Black	44491	34.4% (-6.9)
Latino	2134          1.6% (-0.9)
Asian	2347	         1.8%  (-1.3)
Other	10997        8.4%  (-0.2)

Female	69829	53.9% (-4.6)
Male		57177 	44.1% (+4.3)
Unknown	2511	        1.9%   (+0.2)

18-29	 6936	5.4%   (-1.3)
30-39	 7243	5.6%   (+0.1)
40-49	11727	9.1%   (+1.5)
50-64       36077	27.9%  (+4.8)
65+          65158       50.3%  (-5.2)
Unknown	  2376        1.8%   (+0.3)

[Adam Griffin]

Just as a quasi-comparison: in 2014, when there were approximately 210,000 early votes banked in GA (which was on October 22), the electorate was 66% white, 29% black and 5% other.

As of last night, approximately 210,000 early votes had been cast: the current electorate is 57% white, 32% black and 10% other.

Ultimately in 2014, the electorate became less white as early voting progressed and settled at 61% white, 33% black and 6% other; Nathan Deal won this group by less than 1 point.

By the way, this is a great resource for anybody who's not aware of it: http://georgiavotes.com/

[Adam Griffin]

Kansas mailed ballots (as of 8:00 AM on 10/17):

128,220 Sent (compared to 133,822 in 2016; 89,285 in 2014)

GOP - 51,558 (40.21%)
DEM - 51,326 (40.02%)
UNA - 24,665 (19.24%)
LIB - 670 (0.53%)

[Adam Griffin]

^^^ FWIW, there are only 423 returned ballots in KS thus far: all UOCAVA ballots. Their breakdown:

DEM - 225 (53.2%)
GOP - 115 (17.2%)
UNA - 79 (18.7%)
LIB - 4 (0.9%)

I wonder how many of those are actually uniformed ballots...

[Adam Griffin]

Two observations:

#1: combined requested and returned ballots are already at 55% of 2014's turnout. Additionally, look at these stats compared to 2016:

https://twitter.com/ElectProject/status/1053111387084546050

#2:



Not crazy about those figures in #2. The thing is, if this election plays out like 2014, then beginning next Saturday, the winds should begin to shift back in the Democrats' favor. If this plays out like 2016, however, then it'll only get whiter from here. Thing is, we can't be for sure which is which (especially given point #1 - turnout is approaching presidential levels in early voting, which makes any reversion that much more difficult).

And instead of a raw vote-to-raw vote comparison, here's a day-to-day comparison:



https://twitter.com/gtryan/status/1053281487628525568

[Adam Griffin]

And just a quick pull from McDonald's spreadsheet:

Code:

STATE	% OF 2014 ADVANCE
CO		0.80%
KS		0.80%
AK		1.20%
WV		2.30%
MD		6.40%
CA		8.80%
OH		9.00%
ME		11.00%
AZ		13.50%
NM		13.50%
ID		13.80%
SC		14.00%
MI		20.00%
IL		22.30%
NC		23.40%
FL		24.70%
IA		24.70%
MT		27.90%
WI		29.80%
NE		33.60%
TN		34.40%
ND		36.60%
GA		40.10%
WY		46.50%
VA		47.00%
MN		67.70%

[Adam Griffin]

And just a quick pull from McDonald's spreadsheet:

Code:

STATE	% OF 2014 ADVANCE
CO		0.80%
KS		0.80%
AK		1.20%
WV		2.30%
MD		6.40%
CA		8.80%
OH		9.00%
ME		11.00%
AZ		13.50%
NM		13.50%
ID		13.80%
SC		14.00%
MI		20.00%
IL		22.30%
NC		23.40%
FL		24.70%
IA		24.70%
MT		27.90%
WI		29.80%
NE		33.60%
TN		34.40%
ND		36.60%
GA		40.10%
WY		46.50%
VA		47.00%
MN		67.70%

I think I read somewhere that Minnesota is a head of 2016 as well.

I believe this is the first year that MN offered real no-excuse early voting, which is why its numbers are so large (I believe prior to this year, you had to send a letter to the board of elections detailing a justifiable reason to vote early). I know they didn't have it in 2014; changes in the law are skewing the above 2014 metric at minimum.

[Adam Griffin]

I think I read somewhere that Minnesota is a head of 2016 as well.

I believe this is the first year that MN offered real no-excuse early voting, which is why its numbers are so large (I believe prior to this year, you had to send a letter to the board of elections detailing a justifiable reason to vote early). I know they didn't have it in 2014.

I think 2016 was the first year it was available.

Looks like you're right about 2016! And technically, it looks like they did have it in 2014, but I'm not sure if the no-excuse portion was still intact; it may have just been that nobody was accustomed to the new law then and therefore didn't use it.

[Adam Griffin]

As of 10/19:

Code:

STATE	% of 2014 Advance  	% of 2016 Total
AK		1.30%			0.50%
KS		1.30%			0.30%
CO		1.70%			1.30%
WV		2.60%			0.40%
MD		6.90%			0.90%
OH		9.00%			1.40%
CA		11.50%			3.70%
NM		16.20%			3.60%
AZ		18.90%			11.10%
ME		22.40%			4.00%
NC		23.40%			5.80%
IL		24.40%			3.40%
FL		24.80%			8.30%
SC		27.60%			2.10%
MT		27.90%			12.50%
ID		32.60%			6.10%
WI		33.40%			4.20%
IA		33.60%			10.10%
NE		33.60%			4.70%
MI		36.50%			6.20%
ND		36.60%			9.50%
TN		44.30%			11.20%
WY		46.50%			6.70%
VA		47.00%			2.40%
GA		50.60%			11.60%
MN		67.70%			5.40%
IN		84.40%			7.00%

[Adam Griffin]

Not sure if it's been pointed out yet, but TargetSmart isn't some no-name firm: it's used by many, including the Democratic National Committee, and is actually the firm that's providing state and county parties with turnout and support modeling solutions in their voter file this year (as a baseline; I'm sure they have others). For example, I can see TargetSmart models for my county via our voter file.

However, I'd be willing to bet $10 that the way NBC conducted their analysis was that they took TargetSmart's modelling projections for these states and split them down the middle (which I similarly did for my county at the beginning of early voting). These models work on percentage chance of doing something; a 0-100% chance of being a Democrat, a 0-100% chance of turning out, etc.

It's important to note that classifying everybody who has a <50% chance of voting Democratic as a "Republican" and everybody with a >50% chance as a "Democrat" is inherently flawed: especially if independents are leaning notably in one direction or another this cycle. This would definitely make the baseline scores lean more Republican in their projections than they would be in reality.

[Adam Griffin]

I mean, the likelihood (for example) that Abrams is doing 9-10 points worse than Carter among a similar demographic EV group in GA should speak volumes to the accuracy of these NBC projections.

[Adam Griffin]

Not sure if it's been posted yet, but TargetSmart has uploaded EV data by state, age and other factors and is displaying its models (which again, don't be like NBC and confuse them with vote share). I thought there were maps on there too, based on his tweet image, but I haven't found them yet:

https://targetearly.targetsmart.com/

[Adam Griffin]

Democrats have been consistently winning 60-75% of independents in high-turnout special elections all across this country for the past 2 years. I'm not sure why everybody is obsessing over simple D-R turnout figures; it is not going to paint a full picture. I'm glad we don't have them in my state.