Early & Absentee Voting Megathread - Build the Freiwal

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

I agree that the TX number seems rather low, but I'm not sure you can pin it on either of those things. Being just urban/suburban areas should bias the new/infrequent number upward if Beto's actually bringing in new voters. If the bias is just because of elderly mail voters, you'd expect a similar low number somewhere like ND, but even ND is 13.3%.

This which I posted back 2 or pages prior is the best info on vote history in TX that is available at the moment. Although this is a bit skewed towards mail votes, it is less so than the TargetSmart data is right now:

https://twitter.com/longhornderek/status/1054792918743998465



The TargetSmart data *will* for sure be more informative eventually (probably within a few days and hopefully somewhat by tomorrow), but at the moment it is just not updated enough yet to get a view of voters in general.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Early voting still seems to be at/near/exceeding Presidential levels in heavily Hispanic TX counties. Relatively a bit higher in Cameron and Nueces (and I presume still El Paso), a bit lower in Hidalgo:

https://twitter.com/TheRomanPerez/status/1054933662054129664

https://twitter.com/TheRomanPerez/status/1054952022963863552

https://twitter.com/TheRomanPerez/status/1054931399965319168



Los Hispanos estan votando.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

That's Beet, and I'm Bagel, and for the record, even I think Rosen wins.

The NV Dems would have to be even more incompetent than the FL Dems for Rosen not to win. I assume that won't happen.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Democrats cannibalizing AA Vote here I think. Hillary tried to win FL with the Early Vote in 2016, it failed.

Obama tried the same thing, and in that case it succeeded, twice.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

https://twitter.com/longhornderek/status/1055128208394383361



Note that 28.9% of early voters/mail voters in the 1st day of early voting in the largest TX counties had no primary voting history.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

What would this data imply? I get the primary part, it shows that many non-voters are coming into the election, but what about the ages? What was it like in 2016/2014?

I don't know any specific #s for 2016/2014. Unfortunately TX has pretty bad/limited data.

Generally, it probably suggests what you would think it would - if those voters are younger, then probably the voters without primary voting history skew a bit more Dem. No doubt there are also many Rs among them, though.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

TX Early vote in person data has been updated by the Secretary of State.

https://www.sos.state.tx.us/elections/earlyvoting/2018/oct23.shtml

Here is a comparison of Day 2 to Day 1:



The jumps in Brazos County (College Station) and Jefferson County (Beaumont) are notable. In Brazos, that might be good (or at least good-ish) if that jump is Texas A&M students, otherwise good for Republicans. In Jefferson, there is a large black population but also a large Conservative white population. So could be good or bad depending on who that is that is surging there.

And a comparison of Day 2 in 2018 to Day 2 in 2016:



Harris is lower than 2016 in comparison to other counties because of voter suppression by the Republican County Clerk, who is apparently closing the polls there several hours earlier than everywhere else in TX.

These graphs are in person votes only.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Harris is lower than 2016 in comparison to other counties because of voter suppression by the Republican County Clerk, who is apparently closing the polls there several hours earlier than everywhere else in TX.

Can you provide more details on this?

https://twitter.com/trowaman/status/1055086502047301632

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Here is another graph that someone else made for TX:

https://twitter.com/mcpli/status/1055151219474382849




One thing to note though is that so far today in Tarrant County (Fort Worth), turnout seems to be lower than yesterday so far.

http://access.tarrantcounty.com/en/elections/Upcoming-Election-Information/Live_Turn_Out_Widget.html

For example, 3,894 from 10-11 today, compared to 4,756 yesterday at the same hour yesterday.

So if other counties have that same trend, then turnout may be a bit lower today. and we may start to fall more behind 2016.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

High energy blacks voting in Jefferson County? Betomania!

Hmmm, ok I figured out Jefferson County.

The day 2 # on the Secretary of State website is an error. If you check at the Jefferson County website, you can see that 13,471 is the number of CUMULATIVE in person votes, so that includes the day 1 numbers as well. The day 2 number for Jefferson County SHOULD be 6,374, not 13,471.

https://www.jeffersonelections.com/wp-content/uploads/2018/10/EARLY-VOTING-TURNOUT-11-06-18.pdf

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Tom Bonier of TargetSmart on the early vote data he has (and media not interpreting it very well):

https://twitter.com/tbonier/status/1055166382453768193

https://twitter.com/tbonier/status/1055167322925731840

Keep in mind that the data on TargetSmart right now is still ridiculously incomplete - it only has 200k votes at the moment, 75% of which are still mail votes. Whereas in the top 30 counties alone, 1,187,007 votes have already been cast in TX, along with many more in smaller counties.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

And the big reason Hillary Clinton for starters obtained more Republican Crossover Votes at Precinct Level Results in 2016 is that Maricopa Sheriff Joe Arpaio was on the Ballot and a good chunk of the GOP wanted him out too, not just Democrats. That's the Main Reason AZ was close in the first place.

Proof? Anything at all?

This seems like CNN pundit-level analysis where they just randomly pick some reason to attribute to why an election happened the way it did, yet it usually ends up being wrong.

The problem with 2016's analysis is that Maricopa County's swing to Clinton was hardly unique. In many other states, in similar counties (Orange County in CA, Dallas and Houston suburbs in TX, Atlanta suburbs in GA, etc), Dems made similar sorts of gains. And Joe Arapaio had nothing to do with any of that.

One can always come up with a just-so story to explain away any inconvenient fact, but if you have to start coming up with lots of special stories for explaining results everywhere, that is usually a sign that it is just a story you are telling yourself. Similarly, in 2010, Dems told themselves lots of such stories about how things were not going to be that bad. In each race, there had to be a different story, because if you looked at the overall picture and weren't pre-committed to trying to explain everything away, it didn't appear good.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

If you look at precinct results in Arizona, either in 2016 or the recent special election, it is pretty clear that Clinton/Tipirmeni obtained majorities among unaffiliated voter and also got a decent sized chunk of GOP registered voters in some suburban areas.

You can't compare a Special Election vs a General Election.

Can you come up with any reason Republicans would vote for the Dem in large numbers in a special election, but not a general election a few months later? If this were AL-Sen, I would grant the point that it was weird, but Lesko was generic R and Tiperneni was "some lady" D, it was as close to a generic ballot as you can get.

The real problem here is it is just another just-so story. 2016 could maybe make a reasonable case about any one thing. Maybe there is a problem here with comparing a special election vs a general.

But when one has to tell multiple different such stories about multiple different things, the probability of all of them being correct eventually becomes vanishingly small, after you have told enough such stories.

There is probably some different just-so story for why Scott Walker or DeSantis is going to win.

And it is possible that any one of these various stories could be right. But all of them?

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

TX early vote data is updated for day 3. The graphs below are for in person voting only.


First, here is Day 3 in 2018 as compared to yesterday (Day 2 in 2018):



There was a decrease in all counties, but it seems to have been a bigger decrease in some counties than others. The biggest drops seem to have been in the Dallas area - Dallas County, Tarrant, and Collin. Much less of a drop in Denton County (University of North Texas students still voting?). Proportionally, there also seem to have been a bit larger than average drops (relative to the overall size of the county) in Williamson (North Austin suburbs), Lubbock (Texas Tech), and Ellis (R South Dallas exurbs). Other counties like Travis (Austin) and Bexar (San Antonio) seem to be holding up a bit better so far. Turnout didn't drop much in Harris (Houston). That is probably, I would guess, because of the voter suppression (polls are only open until 4:30 each day) - probably people who would otherwise have voted on Day 1 or Day 2 but didn't because the hours were too short are voting on Day 3 instead.



Second, here is Day 3 2018 as compared to Day 3 2016:



Turnout was lower than 2016 basically everywhere - except for Nueces (Corpus Christi). Something or other interesting must be going ton there to get there to be such high turnout, but no idea what that is. Relatively, it seems like Bexar, Travis, El Paso, and Nueces held up a bit better than other counties, relatively. That suggests good Hispanic turnout, if true. On the other hand, Cameron and Hidalgo (South Texas) look a bit worse, relatively in comparison to those.

This is probably an early indicator that while turnout will still be insanely high for a midterm, we are not going to match 2016.



Thirdly, while turnout does look like it will end up lower than 2016, it is still completely insane as compared to 2014, which is really the fairer comparison:



Here is the % increase in each county's Day 3 turnout as compared to 2014. El Paso really stands out as being MUCH higher than 2014. Other counties with notably larger than average increases relative to 2014 are Denton, Williamson, Nueces, Travis, and Fort Bend:

Harris   204%
Dallas   215%
Tarrant   176%
Bexar   146%
Travis   250%
Collin      233%
Denton   251%
El Paso   343%
Fort Bend   248%
Hidalgo   124%
Montgomery   185%
Williamson   250%
Galveston   208%
Nueces   251%
Cameron   194%

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Of course Rosen and Sisolak aren't doomed or anything, but this doesn't justify the level of confidence some posters are having about this race.

Please tell me NV Dems are not going to blow this. I should be able to be supremely confident about NV. This is not 2010 or 2014, the national environment is good and demographics have continued to shift Dem more quickly in NV than anywhere else. I don't want to have to worry about NV, and shouldn't have to. WTF is wrong with them if this is even a question?

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

-snip-

Not to be that one guy, but it was raining pretty hard across most of the State yesterday, we got 2 inches in Tarrant!

Ah, ok. Thanks. I am not in TX now so I didn't know that (I did actually look up "weather yesterday Dallas" on google, but it didn't say it had rained). So the rain may well explain it, in particular why turnout seemed to drop more in the DFW area.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

-snip-

Not to be that one guy, but it was raining pretty hard across most of the State yesterday, we got 2 inches in Tarrant!

Ah, ok. Thanks. I am not in TX now so I didn't know that (I did actually look up "weather yesterday Dallas" on google, but it didn't say it had rained). So the rain may well explain it, in particular why turnout seemed to drop more in the DFW area.

http://access.tarrantcounty.com/en/elections/Upcoming-Election-Information/Live_Turn_Out_Widget.html Tarrant seems to be recovering today

A little bit now in mid-day/early afternoon, but it was lower again in the morning.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

This thread is EVERYTHING that is wrong with atlas. We spent weeks before early voting opened warning each other how stupid it is to make predictions off of EV and yet that’s exactly what we’re doing here. Shut the computer off and read a book.

It is not stupid to make predictions off of EV. It is stupid to make stupid predictions based off of EV.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Right, the EV is certainly not at the point where Rosen is guaranteed to win, but from the hysterical reactions of some of the posters here you would think Heller already had it in the bag. And yes, the early reactions were also hysterical. I guess that's the one common thread here.

To me, it seems perfectly reasonable to by hysterical if Rosen is not guaranteed to win. Rosen should be guaranteed to win by all rights. It is not about Heller having it in the bag. It is about Rosen not having it in the bag, which is really pathetic. Don't succumb to the soft bigotry of low expectations.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

I can't find the North Carolina chart I saw a couple hours ago. It was really good. I'll try to keep looking in a couple hours.

The problem with NC is that there are no major statewide races driving turnout. Other than Supreme Court, the races that matter are Congressional Districts and maybe some state legislative districts. So just looking at statewide #s doesn't tell very much.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

https://twitter.com/tbonier/status/1055500285337092097

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Just got my ballot!

What are you waiting for? Fill it out and send it in!

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

TargetSmart has updated early vote data. There is better/more complete data for a lot of states now, but I will focus on TX:

https://targetearly.targetsmart.com/index.html

The TX data looks pretty bad to me for Beto.



The big problem is not per se just that there are more modeled Rs than Ds - you would expect that regardless in TX. The problem, however, is that there are not that many Infrequent Voters or Never voted, and that there is no Dem lead with them (there is a slight lead with "Never voted," but that group is so small that it makes little difference). Even if you counted every last one of the "Modeled Unclear" as being Dem, there would still be only a slim Dem lead with Infrequent Voters/Never Voted.

Most of the people who have been early voting, even after we now have more than a million votes in, are Super Voters or Frequent Voters, which suggests both parties are mostly getting their usual voters out, but they are just voting earlier than usual. The electorate is not really being expanded all that much, and not in a way that would clearly favor Dems. Modeled Party is not Party, and is not the same as candidate support, but it is also not all that different.

Of course, the election day voters may well be and probably will be more Dem, but if there were a sufficient surge in Dem turnout to put Beto over the top, it would be more noticeable in the early voting data as well by now.

While this is not perfectly representative and is not complete, it is bad enough to make me comfortable removing the "probably" from my name and going back to just Cruz Will Win. I might revisit this again if there is some sort of crazy unexpected surge of unlikely voters in the last few days of early voting, but there is no reason to expect there will be if a surge of unlikely voters is not already more recognizable.

But barring that sort of unexpected change, stick a fork in Beto. There was some reason for hope at the start of early voting when we saw the huge turnout on the first day of - but IMO that chance is pretty much gone at this point. Too many Rs are voting as well as Ds, and not enough new Ds are voting. The turnout will still be historically high for a midterm, and Beto should do well for a Dem in TX in a midterm year. But nonetheless, I do not see it being enough.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

Do everyone a favor and ignore Targetsmart. We know from actual voting history of early voters thus far that there are actually a lot of infrequent voters.

The Targetsmart data is actual voting history data. It is individual-level voter file data.

[Former Dean Phillips Supporters for Haley (I guess???!?) 👁️]

What are their methodological definitions of the different types of voters?

They are buckets that voters are categorized into based on their vote history. I don't know the exact criteria for the buckets, but it is a safe bet that the definitions are reasonable. This is a widely used, well-known voter data firm.

You must be logged in to read this quote.

One way you can see (more or less) what the categorizations are is by looking at the % of registered voters that fall into each category. In TX, it is 16.51% Super Voter, 32.74% Frequent Voter You can then compare that to historical turnout as a share of registered voters. There was 38% turnout in 2010, 34% in 2014, and 59% in 2016.

So more or less normal midterm turnout seems like it would be most Super-Voters + 1/2 or 2/3 or so Frequent voters + smaller #s of infrequents and new voters. Clearly we are going to be well over that, but the problem is there are lots of Rs turning around in the lower-rung vote history categories, regardless of whether you want to say that lower rung is "frequent" or "infrequent."