2020 Absentee/Early Voting thread

[kph14]

How are these numbers generated, by the way?  Are they demographic/location-based estimates or do people have to state their party when they request a ballot, and then their names are listed and their voting patterns can be looked up?

In Maine voters are registered by party. I suppose in general I think they use all information (party registration, age, ethnicity, county) they can get and model an expected partisanship

[kph14]

One state where I will monitor early votes and probably the only one is Florida. They have a history of mass early and mail votes and Republicans do vote by mail in solid numbers.

Nevada and Arizona are also worth paying attention to since their early/mail vote is usually a very large percentage of the total turnout.

Nevada is all vote-by mail this time

[kph14]

Forbes: When Swing States Will Begin Counting Mail Ballots

Some of their information (such as for GA) was actually incorrect and some (for TX) was incomplete. I've checked as best as possible to verify that each state's status here is in fact accurate.

States can begin counting votes...

22 days before Election Day: Florida

14-15 days before Election Day: Arizona, Colorado, Georgia, North Carolina

12 days before Election Day: Texas (counties with >100k people)

4 days before Election Day: Texas (counties with <100k people)

On Election Day: Iowa, Michigan, Minnesota, Nevada, Pennsylvania, Wisconsin

Unclear: Ohio

DISCLAIMER: Also, here is a separate map that may or may not be up-to-date or directly referencing the counting of ballots. For example, "processing" in some states under normal circumstances simply means opening the outer envelope to simplify the Election Day counting procedure.

Maine has changed their rules by executive order and will allow processing to start on Oct 27:

https://www.centralmaine.com/2020/08/27/mills-signs-executive-order-to-support-absentee-voting-protect-polling-places-from-covid-19/

[kph14]

So with Arizona, Florida, Georgia, and North Carolina counting early we should have an idea of who won on Election Night?

Maybe, maybe not. I'm not sure if any of the non-GA states had these policies as SOP prior to 2020, or if they were implemented precisely because of this year's situation.

In the case of AZ, it has usually taken days to get most/all of the mail ballots counted, so if the procedure hasn't changed since 2018, then there may still be a delay.

With FL, I'm guessing there was at least some semblance of this policy in effect prior, as the state generally gets the vast majority of its mail and non-mail vote alike counted on Election Night.

In GA and as far as I understand, it permits counties to begin 15 days prior but does not require them: with 159 counties, some may (and probably will) choose not to begin counting early, and others (particularly larger counties with notorious difficulties in counting) may still not work through the bulk by Election Night even if they begin 15 days in advance.

But generally, I'd say we'll know a lot about the nature of the evening from TX, GA, NC, FL & AZ much sooner than from WI, PA & MI.

Arizona has changed their laws in response to 2018 and allowed processing 14 days early for the first time in this years primary which I think had no significant counting delays:

https://eu.azcentral.com/story/news/politics/elections/2020/08/04/arizona-primary-election-officials-mail-voting-delay-ballot-counting/5576602002/

[kph14]

So with Arizona, Florida, Georgia, and North Carolina counting early we should have an idea of who won on Election Night?

Maybe, maybe not. I'm not sure if any of the non-GA states had these policies as SOP prior to 2020, or if they were implemented precisely because of this year's situation.

In the case of AZ, it has usually taken days to get most/all of the mail ballots counted, so if the procedure hasn't changed since 2018, then there may still be a delay.

With FL, I'm guessing there was at least some semblance of this policy in effect prior, as the state generally gets the vast majority of its mail and non-mail vote alike counted on Election Night.

In GA and as far as I understand, it permits counties to begin 15 days prior but does not require them: with 159 counties, some may (and probably will) choose not to begin counting early, and others (particularly larger counties with notorious difficulties in counting) may still not work through the bulk by Election Night even if they begin 15 days in advance.

But generally, I'd say we'll know a lot about the nature of the evening from TX, GA, NC, FL & AZ much sooner than from WI, PA & MI.

I’m not too worried about Wisconsin. The process is so decentralized (each municipality/town counts their own votes and releases them all at once). Only place I’m worried could take time is Milwaukee as they’ve had some issues as recent as the August primary.

Maybe someone from Wisconsin can enlighten us here: but I think in the August primary absentee ballots and election day votes were reported together as parts of each precinct. Meaning that no election day vote in a precinct was reported unless the absentee vote in that precinct was also ready. Which in my opinion is a very good system. I might be wrong though.

[kph14]

The explanation was just posted a couple posts above. Stop worrying.

https://twitter.com/gercohen/status/1303389733050109953

[kph14]

New data from North Carolina

Total Ballots returned: 1,373 (0.2% of requested ballots)
Total Ballots accepted: 1,287 (93.7% of returned ballots)

Acceptance rate by party identification:
DEM: 96.3%
REP: 88.4%
UNA: 94.0%

Accepted ballots by party identification:

DEM: 791 (61.5%)
UNA: 363 (28.2%)
REP: 129 (10.0%)

They have added a new category "Witness info incomplete". 17 ballots so far (15 DEM, 2 UNA). Mostly older African-American voters seem to have issues here. The Democratic party is suing to allow voter to cure those ballots (https://www.democracydocket.com/wp-content/uploads/sites/45/2020/09/2020.09.08-NC-Cure-Guidance-Complaint.pdf), like they can cure a missing a signature of themselves. According to the lawsuit though those voters will get send a new one. 

[kph14]

Ballot requests in North Carolina

Total 710,798 (+ 21,818 in a day)

DEM: 368,171 (51.8%) (-0.3%)
UNA: 223,278 (31.4%) (+0.1%)
REP: 116,852 (16.4%) (+0.1%)

Source: https://dl.ncsbe.gov/?prefix=ENRS/2020_11_03/

[kph14]

As I satiate my need to put election data on maps, here's a map of Democratic ballot requests as a percent of Hillary's final vote count in 2016. Doesn't seem to tell us too much at this point except that the heavily black rural areas in the northeast seem to be behind in requests and college areas are ahead.



Lightest color = 0-10% of Clinton's final vote share
Middle color = 10-20% of Clinton's final vote share
Darker color = 20-30% of Clinton's final vote share

I would expect most black voters (especially from rural areas) to vote early in person.

[kph14]

Michigan has over 2.2 million requests for absentee ballots, which is around 46% of 2016's total electorate

https://twitter.com/ElectProject/status/1304133829968277505

[kph14]

When do we start getting consistent daily updates from states with modeled/registered party IDs? I know they don't mean too much, except maybe this year when most voters are likely to vote using VBM, but at least it gives me something to look forward to in this subforum.

As with many aspects of North Carolina voting and registration, the data on who has returned an absentee ballot is publicly available.

https://www.ncsbe.gov/results-data/absentee-data

Current Accepted Ballots:
Democrat — 13,075 (57.6%)
Unaffiliated — 6,557 (28.9%)
Republican — 3,018 (13.3%)
Other — 59 (0.3%)

Do with that what you will. It's only 22,709 votes. Or 0.5% of the total votes cast in North Carolina in 2016.

We are now up to 45,141 accepted ballots. There are already the first voters who cured their ballots with an affidavit and also so the first voters who sent in their second ballot correctly after their first was rejected because of missing witness information.

When voters got a second ballot after their first got rejected, both ballots remain in the list. Therefore data about the rejection rate will be unreliable unless it is corrected for this.

[kph14]

when do we usually get a lot of votes coming in per day for florida? i am excited to see how they trend.

You can find that data here:
https://countyballotfiles.elections.myflorida.com/FVRSCountyBallotReports/AbsenteeEarlyVotingReports/PublicStats

So far 67 voters have returned their ballots.

[kph14]

The Pennsylvania secrecy ballot stuff is worrying.  The letter says 6% of ballots were rejected in previous elections for this reason alone, and that a higher share will be rejected this time around since there are more first-time voters.

If half of PA votes by mail and the VBM share is 70% D, 30% R, and 10% of ballots are rejected, that is a 2% swing in Trump's favor.  That could easily be enough to decide the election.

Hopefully the Biden campaign is ready to absolutely spam the airwaves with instructions on how to vote properly in Pennsylvania.  Get some hashtags going on social media as well.  Get celebrities to show the kiddies how to vote.  Send out physical media and put up billboards in the cities.

And make sure to not tell the rurals... they voted for these Republicans who actively want to disenfranchise voters, so it's only fair for them to be the first group disenfranchised.

I think we should keep some facts in mind before jumping to the conclusion that this is yet another nefarious Republican plot to steal the election:

• The PA Supreme Court is 5-2 Dem and upheld the secrecy envelope requirement.
• The secrecy envelope is included in the mail ballot materials (see https://www.phillymag.com/news/2020/05/14/pennsylvania-mail-in-ballot-instructions/)
• Biden/the DNC/Mike Bloomberg/whoever is free to run ads reminding people that this is a requirement. Lisa Murkowski was able to get a plurality of Alaskan voters to write her in and spell her name correctly. It can't be that hard to get people to put a mail-in ballot in two envelopes.

Why is there no fat "Put your ballot in here before you return it" on the secrecy envelope? Pennsylvania is setting up its voters to fail here. It just says "Official Election Ballot"

[kph14]

The Pennsylvania secrecy ballot stuff is worrying.  The letter says 6% of ballots were rejected in previous elections for this reason alone, and that a higher share will be rejected this time around since there are more first-time voters.

If half of PA votes by mail and the VBM share is 70% D, 30% R, and 10% of ballots are rejected, that is a 2% swing in Trump's favor.  That could easily be enough to decide the election.

Hopefully the Biden campaign is ready to absolutely spam the airwaves with instructions on how to vote properly in Pennsylvania.  Get some hashtags going on social media as well.  Get celebrities to show the kiddies how to vote.  Send out physical media and put up billboards in the cities.

And make sure to not tell the rurals... they voted for these Republicans who actively want to disenfranchise voters, so it's only fair for them to be the first group disenfranchised.

I think we should keep some facts in mind before jumping to the conclusion that this is yet another nefarious Republican plot to steal the election:

• The PA Supreme Court is 5-2 Dem and upheld the secrecy envelope requirement.
• The secrecy envelope is included in the mail ballot materials (see https://www.phillymag.com/news/2020/05/14/pennsylvania-mail-in-ballot-instructions/)
• Biden/the DNC/Mike Bloomberg/whoever is free to run ads reminding people that this is a requirement. Lisa Murkowski was able to get a plurality of Alaskan voters to write her in and spell her name correctly. It can't be that hard to get people to put a mail-in ballot in two envelopes.

Why is there no fat "Put your ballot in here before you return it" on the secrecy envelope? Pennsylvania is setting up its voters to fail here. It just says "Official Election Ballot"

Dude, come on. It's common sense here. Why would they send you the ballot with an envelope that says Official Election Ballot  if they didn't want you to put it in there?

We have the same secrecy envelope system in Germany. The blue envelope, which is the secrecy envelope, has instructions on it. 5%-10% are not using it correctly, that does not sound like common sense.

[kph14]

The Iowa SOS has started to publish their absentee numbers today.

https://sos.iowa.gov/elections/pdf/2020/general/AbsenteeCongressional2020.pdf

545k requested ballots (roughly 35% of total 2016 turnout)
357k already sent

Numbers are surely high this early on but look still very managble for IA's system. In 2018, IA sent out 576k absentee ballots

[kph14]

Tbf, Fall’s Church is the exact type of place that would have insanely high turnout this year.

Falls Church always has very high turnout (for American standards).

[kph14]

how long til Virginia surpasses North Carolina in total early votes?  The numbers are insane.

Very soon. NC has been adding around 20k a day while VA is close to 50k a day. NC has no in-person until the middle of October

[kph14]

Are there any websites similar to http://georgiavotes.com/ but for other states?

There is https://carolinaelections.com/votetracker/gen2020

[kph14]

Incredible how people are trying to read anything out of those early numbers ...

That’s like trying to predict how the winter will go based on the first snowfall of the season.

Early voting numbers are in general pretty useless for predictions in higher turnout elections like the presidential election. That's even true close to election day. For low turnout elections (e.g. Wisconsin's spring election) they were useful to gauge enthusiasm

[kph14]

welp that didn't take long, Virginia has already surpassed NC for the most early votes at 261,000.  LOL, people in Virginia really really despise Trump.

Do we have any idea the party breakdown of these voters and regions?

This is a good site for early/mail vote in Virginia:

https://www.vpap.org/elections/early-voting/

It seems the more Democratic areas of the state have higher early voter turnout than the Republocan areas, which is expected.

Thank you!! super helpful - i see they are showing the district breakdown by district - they aren't showing the party breakdown anywhere?

They can't. Virginia has no registration by party

[kph14]

People should keep in mind that 2016 non-voter just means that the person in question did not vote in that particular state. There is almost no way of knowing whether that person voted in a different state

[kph14]

For Virginia, part of the shift is just the sheer changes in population.  Loudoun county has added 43,000 voters since 2016 from 232,000 to 275,000.  Meanwhile, take three coal counties from SW Virginia, Buchanan, Dickenson, and Lee they've gone from a combined 42,000 voters in 2016 to 40,000 in the latest numbers.  So, Loudoun has literally added the voting equivalent of those three counties in just 4 years.  An extreme example, but very illustrative of Va.

Yes, Northern Virginia (depending how you define it) used to be like 30% of the statewide total, it's more like 35% nowadays.  Democrats winning NOVA + Richmond area is probably enough to win statewide now even if they lose the Virginia Beach area + rural Virginia.

The Hampton Roads region as a whole seems to lean slightly Democratic but I feel like it will be the region that gives Democrats the most headaches under any Democratic president. Will not even be close to being a solid GOP region but its relatively large population is what would cause some handwringing in close state wide races.

In the Hampton Roads region Democratic support depends very much on minority turnout which tends to plummet in non-presidential elections

[kph14]

The highest turnout in NOVA appears to be from the cities:

https://wtop.com/virginia/2020/09/northern-virginia-voters-pack-recently-opened-early-voting-sites/

"When it comes to the percentage of ballots cast, as of Friday, the City of Falls Church leads the pack with more than 17% of its more than 10,000 registered voters showing up to vote early.

The City of Fairfax is next with more than 12% turnout so far in a city of just over 17,000 registered voters.

More than 5,400 voters have voted in Alexandria, which is more than 5% of its more than 101,000 registered voters."

To be honest, it is not physically possible for just one location in Fairfax county to be facilitate 10% turnout. It's much easier in smaller jurisdiction where daily capacity is much higher relative to the bigger ones.

[kph14]

Virginia:360,050
Wisconsin:308,555
North Carolina:275,144
Minnesota:75,511
Georgia:40,285
South Dakota:36,060
New Jersey:34,107
Florida: 33,423
Michigan:28,003
Illinois:26,423
South Carolina:2,692
Montana:708
Iowa:587
Ohio:153
https://twitter.com/Elections2020Us

Leading based on share of 2016 total turnout

WI: 10.4%
SD: 9.7%
VA: 9.0%
NC: 5.8%
MN: 2.6%

[kph14]

New Hampshire has started today to publish its absentee ballot request numbers (hopefully they add returned ballots later):
https://sos.nh.gov/elections/information/notices/absentee-ballots-requested-for-general-election/

It currently stands at 148,630 (20% of 2016's total turnout)