absentee/early vote thread, part 2

[Adam Griffin]

Looks like we're on track for a 59% white electorate in GA this year.

You must be logged in to read this quote.

2016 (2012)

White: 56.7 (59.1)
Black: 30.1 (30.0)
Other: 10.1 (7.7)
Latino: 2.7 (1.7)
Asian: 1.8 (1.4)
Native American: 0.1 (0.1)

Related: Minority voter registration surges in Georgia

[Adam Griffin]

Who are the others?

As I understand it, SoS no longer requires that one specify race on the registration form. However, the form itself has not changed, meaning those options are still on there. I've never been able to find a clear answer as to why, but over the past decade, the number of "Others" has increased substantially, fueled by new registrations. It's not as if the average person registering to vote for the first time knows this nerdy stuff and decides not to specify their race.

There is a tendency in counties with larger Latino populations to have a higher than average share of "Others", but it's not a perfect correlation. It's very difficult for me to believe that a mere 2% of all registrations are Latino in a state where they are 10% of the population. Since Latinos are disproportionately likely to have just registered to vote within the past decade, they're very likely to be under-represented in the formal number due to this phenomenon. I live in the most Latino county in the state, where the Latino population is quite disengaged politically, lower-income, uneducated, heavily undocumented, etc, and the share of registrants who are Latino is around one-third their share of the population (10% of RVs versus 32% of pop); according to this, that same measurement would be one-fifth statewide. Doesn't add up to me. Especially since a large share of Latinos in GA live in counties where they are more likely to be picked up by registration efforts.

Certainly, a good chunk of those people are in reality white, but I'd be willing to say that overall, the "Others" are racially comparable to the makeup of recently registered voters. If that's the case, then I'd say it's probably something like 45% White, 35% Black, 15% Latino and 5% Asian/Other. That'd make the actual electorate 61% White, 34% Black, 5% Latino/Other. It could be less white than that, though: I've seen conflicting data as to the makeup of the Others, so I can't really say for sure. At any rate, tracking with 2012, 2008 and so forth, the white share of the electorate is still dropping at a predictable rate.

[Adam Griffin]

Great registration numbers, but the early voting is almost over and it doesn't look great.

https://docs.google.com/spreadsheets/d/1fz_V3oAUL8XJMEudq5wm5hDT_f554uagt6sIm_sJDro/edit#gid=1996929977
http://www.electproject.org/2014_early_vote
2016 early vote (2014 early vote since I couldn't find 2012)

White 60.8% (58.1%)
Black 27.7% (30.8%)
Hispanic 1.3% (0.5%)
Other 2% (1%)
Unknown 8.1% (9.6%)

I'm still waiting to see what Election Day looks like before rendering any verdicts. 2014 had phenomenal black early voter turnout in GA - to the extent that if it had tracked with election day and early vote in previous years, it would have been something like a 32% black electorate - but those numbers absolutely collapsed in the Election Day vote.

I'm increasingly thinking that early voting numbers are serving as indicators for enthusiasm. We've known for a long time that black voters are nowhere nearly as enthusiastic as they were in 2008 & 2012, but enthusiasm doesn't necessarily matter among reliable voters. I've seen reports that there are a substantial number of likely black voters nationally who have yet to cast ballots - likely being based off of pre-2008 participation. We may just be seeing a return to the mean in terms of voting patterns; that the cannibalization of previous ED vote is reverting back to normal among black voters who a) no longer are voting early because the enthusiasm for Obama is no longer applicable and b) the particulars of this election/the candidates. It really makes no sense for the black share of the electorate to be less than it was in 2010 & 2014, two midterm years where Obama was also not on the ballot.

Also, I'm wondering how they've sourced the 2014 data by race. I had access in '14 to the statewide Democratic voter file, and it showed a substantially larger black electorate in early voting (something like 34% at the end of early voting if I recall correctly).

[Adam Griffin]

Great registration numbers, but the early voting is almost over and it doesn't look great.

https://docs.google.com/spreadsheets/d/1fz_V3oAUL8XJMEudq5wm5hDT_f554uagt6sIm_sJDro/edit#gid=1996929977
http://www.electproject.org/2014_early_vote
2016 early vote (2014 early vote since I couldn't find 2012)

White 60.8% (58.1%)
Black 27.7% (30.8%)
Hispanic 1.3% (0.5%)
Other 2% (1%)
Unknown 8.1% (9.6%)

I'm still waiting to see what Election Day looks like before rendering any verdicts. 2014 had phenomenal black early voter turnout in GA - to the extent that if it had tracked with election day and early vote in previous years, it would have been something like a 32% black electorate - but those numbers absolutely collapsed in the Election Day vote.

I'm increasingly thinking that early voting numbers are serving as indicators for enthusiasm. We've known for a long time that black voters are nowhere nearly as enthusiastic as they were in 2008 & 2012, but enthusiasm doesn't necessarily matter among reliable voters. I've seen reports that there are a substantial number of likely black voters nationally who have yet to cast ballots - likely being based off of pre-2008 participation. We may just be seeing a return to the mean in terms of voting patterns; that the cannibalization of previous ED vote is reverting back to normal among black voters who a) no longer are voting early because the enthusiasm for Obama is no longer applicable and b) the particulars of this election/the candidates. It really makes no sense for the black share of the electorate to be less than it was in 2010 & 2014, two midterm years where Obama was also not on the ballot.

Also, I'm wondering how they've sourced the 2014 data by race. I had access in '14 to the statewide Democratic voter file, and it showed a substantially larger black electorate in early voting (something like 34% at the end of early voting if I recall correctly).

I wonder if the GA Democrats, had the early voting electorate at 34% black because of the 'unknown'. That could be plausible.

Nope, because the same method of breakdown is used in the voter file (SoS classifications). There is an alternate way of looking at it (ethnicity), but that's not what I used at the time.

[Adam Griffin]

Michael McDonald has mentioned his voter file is stale, maybe as of September so that could contribute to it, and I think third party vendors have better data on the unknowns (based on modeling).

http://www.cnn.com/2016/11/01/politics/early-voting-update-black-vote-decreasing/
CNN has the Black vote at 31% as of 10/31, I think about 3% higher than Michael's reporting, but that's still trailing previous years.

He had been reporting 28.x% Black, yet it dipped to 27.7% today.

At least in Florida and North Carolina, the Black % is creeping up, but no sign of that in Georgia.

CNN is using Catalist. That is literally the gold standard. They have so much information on every single voter in the country that it's crazy: from your vitals down to what links you click on social media. No offense to Michael, as he does a great job every cycle with collecting this, but if it's Catalist versus a "stale" voter file, then I'd definitely go with Catalist.

Another thing to consider (based on higher overall turnout) is that more white people are simply using early voting this year, which could have an impact on the percentages. That'll leave fewer whites to vote on Election Day, which could result in the racial gap between early vote and Election Day vote being a lot smaller than in previous cycles. If there is no gap, then the electorate would be blacker than it was four years ago (not saying that's going to happen).

[Adam Griffin]

^^^ I haven't looked at the precise numbers yet, but based off of CNN's percentages/early vote totals for GA in 2012/2016, you'd come away with these approximate numbers of black early voters in each election:

2012: 405,000
2016: 465,000

So even though it's a smaller percentage, it'd be an overall larger number of black voters using early voting. It seems that it is just that many (likely older) white people are now using early voting instead. Again, could be totally wrong about the broader implication, but we'll see.

[Adam Griffin]

Assuming I've read McDonald's state-by-state numbers/columns correctly (some of the labeling sounds weird to me, but meh), here's a map showing the most recent early vote in each state as a percentage of the total 2012 vote.

Numbered are the 11 states where the early vote thus far is over 50% of the 2012 total vote.

Code:

NV	75.61%
AZ	69.18%
TN	68.16%
FL	67.64%
NC	64.36%
GA	60.86%
CO	60.45%
OR	59.35%
MT	56.35%
TX	56.28%
WA	54.21%
--------------
UT	47.21%
AR	46.89%
NM	45.93%
IA	37.68%
MD	36.16%
ND	34.82%
DC	34.50%
KS	32.43%
CA	32.22%
ME	30.51%
WV	27.43%
ID	26.85%
LA	26.31%
WI	25.32%
IN	24.31%
IL	24.04%
MA	22.14%
SC	21.71%
SD	21.61%
NE	21.58%
HI	21.03%
--------------
OH	18.90%
MI	18.10%
AK	17.41%
WY	16.33%
VT	15.13%
OK	14.77%
MN	14.22%
VA	12.52%
DE	5.41%
NJ	4.73%
KY	4.08%
RI	3.78%
NH	3.64%
MO	2.01%
MS	1.08%
AL	0.12%
--------------
CT	0.00%
NY	0.00%
PA	0.00%

[Adam Griffin]

As of today's early voting figures. Florida crosses the 70% threshold and NC comes very close, too.

'16 EV as a % of '12 Total:

Code:

NV 75.87%
FL 72.61%
AZ 69.18%
NC 68.96%
TN 68.16%
GA 61.00%
CO 60.45%
OR 59.35%
MT 56.35%
TX 56.28%
WA 54.21%
-----------
UT 47.21%
AR 46.89%
NM 45.93%
IA 37.68%
MD 36.16%
ND 34.82%
DC 34.50%
KS 32.43%
CA 32.22%
ME 30.51%
WV 27.43%
ID 26.85%
LA 26.31%
WI 25.32%
IN 24.31%
IL 24.04%
MA 22.14%
SC 21.71%
SD 21.61%
NE 21.58%
HI 21.03%
-----------
OH 18.90%
MI 18.10%
AK 17.41%
WY 16.33%
VT 15.13%
OK 14.77%
MN 14.22%
VA 12.52%
DE 5.41%
NJ 4.73%
KY 4.08%
RI 3.78%
NH 3.64%
MO 2.01%
MS 1.08%
AL 0.12%
-----------
CT 0.00%
NY 0.00%
PA 0.00% 

[Adam Griffin]

Florida is over. Nevada is over. It's over. Even if you concern-troll over PA, all Clinton needs now is NH and she can afford to lose PA, MI, OH, NC, IA, AZ, ME-2 and NE-2.

[Adam Griffin]

Miami-Dade: 53k voters today

[Adam Griffin]

Just so we have the figures readily available to compare when yesterday's in-person partisan breakdowns are finally published:

Miami-Dade (in-person through Saturday)
D: 194,777 (46.1%)
R: 113,386 (26.8%)
NPA: 108,797 (25.7%)
Other: 5,783 (1.4%)

Miami-Dade (mail through Sunday)
D: 116,819 (41.0%)
R: 97,921 (34.3%)
NPA: 66,757 (23.4%)
Other: 3,640 (1.3%)

[Adam Griffin]

Miami-Dade In-Person (Sunday)
D: 23,847 (44.9%)
R: 11,916 (22.4%)
NPA: 16,584 (31.2%)
Other: 748 (1.5%)

With Sunday's vote now in, Republicans sink to third place in Miami-Dade in in-person voting, behind NPA.

Miami-Dade (in-person through Sunday)
D: 218,624 (46.0%)
NPA: 125,381 (26.3%)
R: 125,302 (26.3%)
Other: 6,531 (1.4%)

Miami-Dade (mail through Sunday)
D: 116,819 (41.0%)
R: 97,921 (34.3%)
NPA: 66,757 (23.4%)
Other: 3,640 (1.3%)

Miami-Dade Total (through Sunday)
D: 335,443 (44.1%)
R: 223,223 (29.3%)
NPA: 192,138 (25.2%)
Other: 10,171 (1.4%)

Miami-Dade early vote as a share of 2012 total vote: 86.6%


Florida early vote as a share of 2012 total vote: 75.6%