2020 Absentee/Early Voting megathread 2

[Florida Man for Crime]

Texas is now at 80% turnout of its total 2016 turnout, still with a week of early voting to go before election day.

Texas appears to be headed towards something like 11-12 million vote turnout (hard to say exactly due to uncertainty on election day turnout).

By comparison, there were a total of only 8 million votes cast in the 2012 election. Assuming that the high turnout does hold up, that means that the electorate will be substantially different from even just 8 years ago, even if not a single voter actually changed their support from Romney to Biden, and even if older voters (the most pro-Romney, most Republican, and most white demographic) were not being gradually replaced in the electorate by more diverse voters over the course of that 8 year period.

Although, as a caveat, we should remember that at least some of the new voters will be rural WWC pro-Trump voters, it is not like every one of the new voters is going to be a Dem. Which is why Trump is even competitive at all.

[Florida Man for Crime]

Black Seniors are turning out hard in the South:

https://twitter.com/tbonier/status/1320701753852047364?s=21

Don't quote TargetSmart.

In reading the replies, Bonier apparently gleaned this info straight from GA state (no, not Georgia State).  Not sure how reliable this is, though.  That's above my head.

https://mobile.twitter.com/tbonier/status/1320544370962112512

The Georgia one could definitely be right, but Texas I don't think has info like this, so I believe that's just speculation.

Correct on both counts. In Georgia voters select their race while registering to vote (for VRA purposes). In Texas, they do not. The GA data comes from that, TX data comes from modeling where they attempt to predict race, which is better than nothing, but not entirely reliable.

[Florida Man for Crime]

Florida Turnout Tracker

Dems proceed beyond 2016 EV Turnout but Republicans catching up.   Just need to keep at same level as Reps in the end.  NPAs will hit their 2016 EV % today

Dems  56.4% (+2.8 )
Reps   53.9%  (+3.7)
NPAs   40.2%  (+3.0)

To me things look on course for a narrow Biden win in Florida. Nevertheless, it does look like the Rs may close/tie the overall Dem advantage by the end of the early vote period. And then there is a possibility that they may pull into an overall advantage with the election day vote.

However, I think the fact that ballots were sent out later in Miami Dade will mitigate somewhat against how good the election day vote will end up being for the GOP. One side effect of that is there will probably be more election day votes left over as some more people who would have sent in their ballots earlier will decide to vote on election day, making the election day vote more Dem. I think the #1 concern has to be late returned ballots getting rejected in Miami-Dade due to the Republican opposition to democracy (deliberately sending out ballots late).

However, I think even with those things, the available data is pointing to a narrow Biden win due to independents/undecideds breaking Dem enough to offset the effects of these immoral and anti-democratic shenanigans. I am thinking though, that GA and perhaps NC may vote to the left of FL because of this.

In 2021, the first order of business for a new Dem trifecta, if there is one, absolutely has to be all manner of procedural reforms, including electoral reform, judicial reform, Senate rules reform, and so on. Policy reform cannot occur without first taking care of procedural reform.

[Florida Man for Crime]

Early voting today has been canceled or delayed in at least 16 Georgia counties due to widespread power outages from Tropical Storm Zeta.

https://www.ajc.com/news/power-outages-delay-early-voting-in-metro-atlanta/G5GD22MWL5G6DMLEKJZC3BYCW4/

The article doesn't have a list of the 16 counties. It mentions a few off hand, but no comprehensive list that I can see. Do we know which ones they are and how badly effected each county is?

[Florida Man for Crime]

Interesting.  The most diverse state in the country.  Clearly minority voters are engaged in a way they weren't in 2016.  

Hey GOP - maybe sitting by idly while your party's standard bearer openly acts like a white supremacist isn't working out in a 2020 election?

In particular, Hawaii has a large Asian-American population. Although there are not that many Indian-Americans specifically, perhaps the prospect of Kamala Harris becoming the first Asian-American VP has at least a small part to do with that also. If so, Asian American turnout may be strong elsewhere, and this could be an indication to expect that.

[Florida Man for Crime]

I saw a chart yesterday that showed the most common name among Harris county early voters was Nguyen.

That is very interesting. There are not many swing states where higher Asian American turnout would make much difference because the Asian-American population is fairly small and somewhat concentrated in safe blue coastal states like CA and NY and VA, but among those Texas is probably the biggest one where a large increase in Asian-American turnout could make a difference, with Harris, Fort Bend, and Collin Counties being the epicenters. It could also make some difference in Georgia.

[Florida Man for Crime]

I saw a chart yesterday that showed the most common name among Harris county early voters was Nguyen.

That is very interesting. There are not many swing states where higher Asian American turnout would make much difference because the Asian-American population is fairly small and somewhat concentrated in safe blue coastal states like CA and NY and VA, but among those Texas is probably the biggest one where a large increase in Asian-American turnout could make a difference, with Harris, Fort Bend, and Collin Counties being the epicenters. It could also make some difference in Georgia.

Nguyen is a Vietnamese last name. Vietnamese-Americans are the most pro-Trump subgroup of Asian-Americans. There is a generational divide between older Vietnamese-Americans who favor Trump and the GOP, meanwhile younger Vietnamese-Americans favor Biden and the Dems.

Many older Vietnamese-Americans believe that the Democrats are "communists/socialists" and that Democrats "suck up to China". There is a historical anger among many Vietnamese people towards China because China occupied Vietnam 4 separate times in the past. Many Vietnamese people like Trump's "tough" rhetoric on China. Also, there is a decent amount of Vietnamese-Americans who are Catholic and thus they are heavily anti-abortion and anti-LGBT.

It's true that Vietnamese-Americans historically voted Republican, but over the past decade or so they have been trending more and more Dem, especially in the Trump era. Just look at precinct results in places like Little Saigon in Orange County CA over time and especially in 2016/2018 and you will see a dramatic change. As with other groups, 2nd generation/younger Vietnamese-Americans are more likely to be Dem (and also more likely to be citizens/eligible to vote).

In the case of Harris County TX in particular, there is plenty of reason to be skeptical that the Vietnamese-American vote is going to vote for Trump based on previous elections. For example, see:

https://en.wikipedia.org/wiki/Hubert_Vo

No doubt Trump will win some votes among Vietnamese-Americans, and I would also bet he will do better among them than among other Asian-American subgroups, but I strongly doubt he is going to win them, especially in urban counties like Harris.

[Florida Man for Crime]

Why are Democrats freaking out so much about turnout in Miami Dade specifically?  The turnout doesn't look way out of whack.  Though the Republican turnout looks a bit elevated there.

Because it is a few days before an election, and at this point it is a second nature unconscious tic to always worry about Miami-Dade County, regardless of the circumstances.

Either "Miami-Dade County will be bad" or "Miami-Dade County will be good, but not good enough." At least it isn't any worse than worrying about the I-4 corridor.

[Florida Man for Crime]

https://twitter.com/mcimaps/status/1322006684680785920

I'm gonna go ahead and make an observation on this Florida Hispanic turnout freakout.

What is being compared here is the overall turnout rate of Dem Hispanics in 2020 to Republican Hispanics in 2020.

However, this is not really the comparison we ought to be making. This would only be an appropriate comparison to make if Hispanic Ds and Hispanic Rs were demographically similar and if, other things equal, we would normally expect them to actually have equal turnout rates.

It may be (and I strongly suspect probably is) the case that demographic differences between Dem Hispanics and GOP Hispanics mean that we should expect GOP Hispanics to have higher turnout and that we should be surprised if that were NOT the case.

What would those demographic differences be? Well, :

1) Hispanic Republicans tend to be older than Hispanic Democrats, for similar generational reasons to why White Republicans tend to be older than White Democrats. Thus, other things equal we should expect higher percentage turnout from Hispanic Rs than Hispanic Ds.
2) Hispanic Republicans are disproportionately Cubans. Unlike other Hispanic groups, Cubans automatically are eligible for US citizenship and thus older Cubans who were not born in the US are eligible to vote, whereas older non-Cubans who were not born in the US are less likely to be citizens and less likely eligible to vote. There is the exception of Puerto Ricans, who are also automatically citizens, but Puerto Ricans are not the main non-Cuban group in Miami-Dade (Central Americans from Nicaragua and Honduras are the main non-Cuban group, with lots of other national origins spread across Central America/South America). Anyway, the fact that a disproportionate share of Cuban olds, as compared to Nicaraguan/Honduran/other Hispanic olds, are eligible to vote biases the age distribution of Hispanic Rs as compared to Hispanic Ds in Miami-Dade upwards even more. And that also biases the expected turnout of Hispanic Rs upwards as compared to Hispanic Ds, since we all know that olds vote more than youngs.

It may well be, and probably was, the case that in 2016 and in previous years (including 2008 and 2012, and we know how things turned out then) GOP Hispanics also had higher turnout than Dem Hispanics for those simple and pretty obvious demographic reasons.

In other words, a great deal (if not all) of this turnout differential may be explained not by partisanship, but by age. And to the degree it is explained by age, it is not unexpected. It is not breaking news that olds vote at higher rates than youngs, and that is already taken into account in polling and campaigns' voter modeling.

So the more appropriate comparison to look at would be to compare Dem Hispanics in 2016 to Dem Hispanics in 2020, and GOP Hispanics in 2016 to GOP Hispanics in 2020.

The more telling questions are not if GOP Hispanic turnout is higher than Dem Hispanic turnout, but are whether Dem Hispanic turnout now is higher or lower than in 2016 (and other previous elections) and likewise if GOP Hispanic turnout now is higher or lower than in 2016 (and other previous elections).

If someone with access to the relevant voter file data feels like telling us that, then we would actually know something interesting and relevant. But absent that, this freakout is little more than a bunch of uninformed hot air.

[Florida Man for Crime]

These numbers in Texas are just insane. I seriously doubt all these people are coming out for Trump lol.

One thing to keep in mind is that in some previous years Texas early vote turnout seemed really high at the time, election day turnout ended up being a bit more underwhelming than expected at the time (or at least than hoped).

Of course, "really high at the time" and the numbers we are seeing now are in somewhat of a different league. Now, the numbers are way higher than anything seen previously, and it is hard to know how much of that portends high election day turnout, and how much of it portends a lot of cannibalized votes.

Honestly, I am not entirely sure whether additional very high election day turnout would be good or bad for Dems at this point. Normally you would say that higher turnout is good for Dems, but given that polling is showing that Biden has a pretty clear and consistent lead among those who already voted, if election day turnout doesn't end up quite as large as some of the predictions, that may not be the worst thing in the world, as at least it means fewer votes with which Biden's early vote margin can potentially be cut down.

[Florida Man for Crime]

This may sound crazy but... is it time to send Hillary Clinton to Miami??

It crossed my mind, but Bill Clinton would be the better option.

Get Bill Clinton back together with his old HHS secretary Donna Shalala and ONLY good things can happen in Miami, amirite? Shalala-Clinton will electrify the Dem Miami-Dade Hispanic vote.

[Florida Man for Crime]

This chart single handily eased my mind on the matter. Those 3-3's are gonna vote! Either this weekend or Trump's attack on the mail scared them into voting on Election Day. Look how good Democrats are doing at getting infrequent voters to vote! Plus those NPA's 0-3's are amazing too.

Not sure if you are interpreting the table correctly, note that it is Hispanics in Dade who have NOT voted. For context we need (but don't have) a similar chart for Dade Hispanics who HAVE voted. And more importantly, we need a comparison to 2016 which is totally absent. If a similar pattern was present in previous elections, which makes sense based on the likely age demographics, it is not that big of a deal.

[Florida Man for Crime]

They said its a point lower than in 2016? That needs to get fixed before Election Day, but I don't think it would be impossible.

Am I crazy or where there similar concerns in 2016 until the last weekend? I could have sworn there was.

You are correct, there were.

And that is precisely why what is needed is a comparison of the turnout data to 2016 (and 2012 and 2008). Not just looking at it in the abstract without any context from previous years.

But instead, all we have is just 2020 Miami-Dade data presented with no comparison to past trends, and some breathless politico articles which are based on some quotes the author selected from various consultants in order to try and get clicks and ad revenue.

[Florida Man for Crime]

Here's an update to the Derek Ryan reports on Texas early voting, which include primary voting history and some basic demographic data.

https://mcusercontent.com/d3064a2fadaf6089dc58a8393/files/c2de2855-be79-4d06-95de-f085578dc256/Statewide_Report_Day_16.pdf

Honestly, there are some good things and some bad things for each side here, depending on how you want to interpret it.

Texas early vote by primary voting history:

First, here is turnout so far by primary voting history, as compared to 2016/2018:



At the start of the early voting period, there was a big surge of Dem voters, and for a time the Dem primary voter share of the electorate was higher than in 2018. That trend has since receded as the electorate has broadened out, and GOP primary voters and voters without primary history have gone up.

On Day 9 of early voting, the Dem primary voter share of early voters was 26.1%. By Day 14, it dropped to 24.0%. Now it is down to 22.9% (Day 16). The R primary voter share also has decreased since Day 9, when it was 31.3%. On Day 14 it fell to 29.8%, and now is down to 29.1%. So it has fallen, but not by as much as the Dem primary vote share.

The major group that has increased in the later part of early voting has been voters with no vote history whatsoever; brand new voters (those are definitely not cannibalizing anything). On Day 9, new voters were 12.7% of early voters. By Day 14 that was up to 14.9%. And now it is up to 16%. Voters with GE vote history but no primary vote history also have increased over the early vote period, but not as much, from 26.3% on Day 9 to 28.1% on Day 14, to 28.9% by the end of Day 16 (yesterday).

To some extent, it is to be expected/inevitable that the higher turnout goes, the more the Dem primary voter share will go down, because you simply run out of Dem primary voters from previous elections. Part of what it means for a state like TX to swing/trend Dem is that you are getting new Dem voters, not ones that previously voted in Dem primaries or previously voted for Dem candidates.

The part that is potentially troubling from a Dem perspective is that the R primary voter share of the electorate is basically the same as in 2016; no decline at this point. As a caveat though, R primary vote history is not the same thing as being a Republican. I personally know partisan Dems who basically always have voted in R primaries in Texas in the past, because they have effectively decided the General Election. Even the Democratic Senate nominee MJ Hegar has R primary voting history as recently as 2012 (she voted for Romney)!

By this point near the end of the early voting period, on both sides voters with previous primary voting history seem pretty close to tapped out/"cannibalized." If there is high turnout on the last day of early voting and on election day, it is not going to be primarily composed of people who have voted in primaries, it will be made up of new voters and people who may have voted in a General election or 2 in the past but never in primaries.

90.3% of people who voted in 4 of 4 D primaries have already voted and 87.9% of those with 4 of 4 R primaries have. When you take into account people moving to a different address/county/state and who have died, that means that pretty much every last partisan super-voter on each side has already voted. Most, though a bit less, of those who voted in 1 of the last 4 primaries have also voted (71.2% of Ds, 68.4% of Rs), and people who voted in 2 or 3 primaries are in between the 1 of 4s and the 4 of 4s. So there is a little bit more that can be potentially squeezed out of those groups, but probably not that much more.

Another thing this makes clear is there is not some huge untapped group of election day voters with Republican primary vote history. It is possible that there may be a significant group of R-leaning voters waiting to vote on election day, but if so, they are not the ones that vote in R primaries (and that includes people who voted in the 2016 Republican primary for Trump). This makes me skeptical of just how much Trump can plausibly be expected to win the election day vote by. Since polls consistently seem to show that Biden is leading among early voters even in polls that have Trump ahead overall, the fact that at least the hard core partisan R vote seems to be mostly tapped out seems like a good sign for Biden's chances.


Texas early vote by age:



At the start of early voting, young voter turnout (18-29) was comparatively low. Since then, it has caught up, and is now exceeding the 2018 share of young voters, and has matched 2016 (with early voting still not done). Interestingly, despite that, the age 70+ vote share is still 17.6%, which is higher than 2016 and also even higher than 2018 (despite 2018 being a midterm).

The fact that the Senior vote share is higher is no doubt partly a result of the general aging population. It might also be an indicator of some vote cannibalization, with lots of Seniors voting early. So the Senior vote share may be pretty low on election day. If so, that is another reason to be skeptical that Trump will carry the election day vote - at least in Texas - by some massive unheard of margin. Probably a lot of this is Seniors voting early by mail to reduce their risk of contracting the virus in the pandemic...

Over the course of the early vote period, the 18-29 share has gone steadily up, while the Age 70+ share steadily declined. At the end of Day 9 it was 11.2% 18-29, up to 12.7% on Day 14, and 13.4% on Day 16. The age 30-49 share has also been going up, while the Age 50-69 share and the age 70+ shares have been declining as more and more votes come in.


Texas early vote by gender:

Currently 52.1% are female, 43.5% male, 4.4% unknown gender. It is very clear that more women have been voting than men. There has been a slight decline in the female share, however, from day 9 when it was 52.4% female, 43.4% male, and 4.2% unknown. But that is a very small change compared to the other demographic & vote history changes we have seen.


Overall, the more people vote and the more time that early voting has gone on:

a) the younger the electorate gets
b) the more the Dem primary voter share goes down
c) the more the GOP primary voter share goes down (to a lesser degree)
d) the more the share of brand new voters with no history at all goes up
e) the more the share of voters with GE vote history but no primary vote history goes up (to a lesser degree)
f) the electorate becomes slightly more male (though not by much, and there still seem to be clearly more women voting than men)

In terms of demographics, I would expect this general trend to continue on election day, though it is possible the election day voters may vote for Trump to a higher degree than those demographics would predict.


For comparison, here are previous reports from earlier that were posted in this thread by me and ON_Progressive:

https://mcusercontent.com/d3064a2fadaf6089dc58a8393/files/9411ad67-cc0e-4028-b4c6-5472ef16256a/Statewide_Report_Day_14.pdf

https://mcusercontent.com/d3064a2fadaf6089dc58a8393/files/c269aebb-c027-490d-ae9b-487678a34e83/Statewide_Report_Day_9.03.pdf

[Florida Man for Crime]

Delays in delivery of the mail-in ballots is a very clever ploy.

In states where Trump has the lead, these late mail-in ballots will be contested and thrown out by the Supreme Court.

In states where Biden has the lead, Trump would not contest these late ballots because they would give him a chance of a victory.

You say clever ploy, I say criminal action.

If Biden wins, Dejoy needs to be prosecuted for vote tampering.

Indeed, and if for some reason it is not possible to prosecute him (and anyone else involved in this crime against democracy) for that, then prosecute him for Income Tax Evasion or whatever else is possible instead, Al Capone style. There should be 0 tolerance for this, and reforms need to be passed to ensure that nothing like this can ever happen again.

[Florida Man for Crime]

Travis County just hit 550,000 with 1.5 more hours of early voting left. Looking forward to updates from other counties but I’m particularly proud of my new home. Shattering turnout expectations and breaking records with each new vote cast.

wow, it was 468,000 in 2016!

If you apply the margins from the last election that should net Biden at least an additional 40k votes.

Except Biden will definitely get a higher percentage margin than Clinton got. It will be more similar to the Beto-Cruz margin, and likely higher than the Beto-Cruz margin.

Realistically, if you take Harris and Travis counties alone, between the two of them their additional turnout/Dem support are very likely to sufficient on their own to erase the entire statewide 2018 Cruz margin of victory.

However, that does not necessarily guarantee the state to Biden because it is plausible that Trump will gain enough offsetting votes elsewhere (with turnout in rural and suburban/exurban safe Republican counties likely being higher than in 2018, and giving Trump a chance to get better margins out of some of those than Cruz could get).

But I really don't see any realistic argument at this point for Texas being anything less than a tossup for Biden.


- edit - I should really amend this. I said that Harris + Travis alone are "very likely" to be sufficient on their own to erase the entire 2018 Cruz margin of victory. But really that is just excessive caution talking. It is really hard to see how that could possibly not be the case. Not so much "very likely" as "near certainty."

[Florida Man for Crime]

https://twitter.com/Redistrict/status/1322550386373414913

This is a very positive sign for Biden's chances in Texas.

Why?

Not so much because of the impact of these particular votes, but because it suggests that the typical pattern of Hispanic turnout from previous elections is fairly likely to hold at least to a significant degree this year as well. That pattern is that Hispanic voters tend to turn out late, and to vote on election day. So a big turnout on the last day of early voting in these heavily Hispanic border counties suggests that the early vote in those counties is not merely cannibalization of the election day vote, and we should expect relatively high turnout to continue on election day among Hispanic voters. If that is the case among Hispanic voters in border counties, it is probably also the case to some degree or another in other less-Hispanic urban counties that nonetheless have large Hispanic populations like Harris, Dallas, and Bexar.

[Florida Man for Crime]

Why aren’t you posting about the numbers anymore Buzz? You were so confident a few days ago? Why did you stop all of a sudden? Whatever could it be?

To be fair, Buzz has been one of the more consistently sensible voices of the right here. 

Yes, the left-leaning posters should remember that they will likely be in the same position as Buzz 2 years from now.

[Florida Man for Crime]

For all the talk about cannibalization in Florida the Republicans still have something like twice as many super voters left to vote as the Democrats (~300k vs ~150k). If anything its the STTP that's been cannibalized by mail in votes and reduced church attendance.

If Democrats bring their lead back up to around 120k in time for election day then the Democratic lean of the independent early voters could give Biden the edge, but otherwise Trump is in the drivers seat in Florida. But for the millionth time, Florida is lean Trump under normal conditions and he'll still probably lose even if he wins it.

But turnout is so high at this point that a non-trivial # of the super-voters that still have not voted yet will be deceased people who have not been taken off the roles yet. Those are more likely to be older people, and I would think that R super voters tend to be older than D super voters. So even if the #s you quote are right, they may be slightly deceiving for that reason.