2020 Absentee/Early Voting thread

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

I voted early today in Texas. Based on 2016/2018 turnout statistics and the Tarrant County early vote turnout data by hour (showing how many people vote at different times during the day), I guesstimated that probably the least crowded time to vote would be on a Saturday morning very earl in the morning. This also makes sense, because who really wants to get up early on a Saturday, rather than sleep in? I also went to an early voting location that had lower turnout in previous days to minimize the risk that I help to spread the virus.

It is a ridiculous travesty, however, that people are being forced to vote in person rather than allowing everyone to vote by mail. As I signed in to vote, I looked at the poll workers who were signing me in and setting me up to vote. There were 4 of them (and others in different parts of the voting location), all of whom were old, and some of whom clearly had risk conditions such as obesity, and no doubt others had other risk conditions that are not visible that make them more vulnerable to the virus.

As I showed my voter registration card/signed in/etc, I thought to myself that some of these poll workers (and also some of the voters) will end up getting infected and dying in the process of voting/helping people to vote. There is no way to tell which individuals exactly those will be, but there is an absolute certainty that some of them will lay down their lives (or perhaps "only" their health). Hopefully I myself didn't unknowingly infect anyone (or get infected).

The poll workers remind me of British soldiers in the World War 1 Battle of the Somme. Because of poor leadership of General Haig (Republican elected officials), they are told to go over the top and charge enemy machine guns, senselessly. Their deaths will be totally unnecessary and serve no purpose, because everyone could have simply voted by mail. But, they will have died for their country. Due to the negligence, incompetence, and willful bloodlust of their leaders.

The Republican elected officials such as Greg Abbott (and IDK, perhaps there are some Democrats as well doing the same thing) who are refusing in many states/localities to allow universal mail voting are worse than insurgent fighters who have killed American soldiers in places like Iraq/Afghanistan. At least in the case of Taliban etc killing American soldiers, they are killing a foreign occupying army, not killing their own people. These Republican officials, by contrast, deliberately kill their own people, their own countrymen and countrywomen. Rather than protecting the American people, they do the opposite, and knowingly murder their fellow Americans, for no good reason.

I hope that if/when the Democrats are elected and get a Trifecta, one of their first acts is to pass a bill to build a new national monument to the poll workers and voters who needlessly sacrificed their lives to the country in 2020 in order to vote, and also to cover the funeral expenses of all poll workers who die as a result of their service in the 2020 election. Also all poll workers who get infected but survive with any persistent damage to their health should be counted as veterans and receive life-long access to free medical care from the Department of Veterans Affairs, just as do combat veterans from foreign wars. We all owe them a great debt of gratitude.

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

Day 6 of early voting in Texas (In-person + VBM) (October 18)

HARRIS COUNTY (Houston)
2012:   384,618     19.2%
2016:   527,631     23.6%
2018:   468,549     20.0%
2020:   628,708     25.3%

DALLAS COUNTY (Dallas)
2012:   228,376     19.4%
2016:   311,307     24.2%
2018:   307,342     23.0%
2020:   341,607     24.4%


BEXAR COUNTY (San Antonio)
2012:   195,494     21.3%
2016:   255,371     24.4%
2018:   226,845     20.7%
2020:   270,323     22.7%

- SNIP -

This is somewhat misleading, because in previous years early voting in TX began on a monday, but this year it began on a Tuesday. So the first 6 days in previous years includes 5 weekdays and a Saturday, whereas this year it is 4 weekdays, a Saturday, and a Sunday. Since turnout is typically lower on weekends and especially on Sundays (with limited hours on Sundays) that probably makes the comparison understate how relatively high turnout really is this year. For example, Bexar county probably would not be behind 2016 if the comparison were based on same days of the week.

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

Here's another way to look at Texas early vote turnout. It has been compared quite a bit to 2016 turnout, but in some ways could it be more helpful to compare to 2018? After all, 2018 data should show some of the Dem/Beto trend in the suburbs and megacities of TX that we expect to be present in 2020.

But in order to have a chance to win TX, Biden will need that same 2018 trend to not only return, but to be stronger so as to make up the margin by which Cruz beat Beto in 2018.

On the other hand, for Trump to do well, we should expect turnout to be systematically higher among white working class voters in rural counties than was the case in the 2018 midterms (since less educated voters tend to vote less in midterms, Trump's hope is to offset additional Dem gains in the suburbs/cities with higher turnout in among non-college whites than we saw in 2018).

Here's a list I complied of all TX counties sorted by the % of the total 2018 turnout that has already been cast (in the first column). The 2nd column is the margin by which Ted Cruz won the county in 2018.

There is one county (Scurry) which already has 128.4% of 2018's total turnout. That is presumably some sort of fluke. On the low end, Hudspeth county has only 22.0% of 2018's total turnout (also probably a fluke). The statewide average is that votes equal to 63.9% of 2018's total turnout have already been cast.

Another interesting thing to look at is how the % of 2018 vote already cast correlates with the Cruz margin of victory. So far, there is a slight/weak negative correlation between those two things of -0.031

This is a pretty weak correlation, so there is not much of a clear general relationship on the county level between partisanship and turnout as compared to 2018. However, the negative correlation does mean that the distribution of turnout is a bit more tilted towards Beto-friendly counties than was the case in 2018. In other words, at least on the county level, additional urban/suburban/hispanic dem gains seem to be slightly more than off-setting any effect of increased rural/non-college white/GOP turnout so far.

That does support the idea that Biden could be in a position to win Texas. However, it is important to remember that these numbers do not include election day votes and other votes that have not yet been cast, and the election day votes are generally expected to be relatively more favorable for the GOP/Trump.

County   % of 2018 vote already cast   Cruz Margin
SCURRY   128.4%   70.2%
UVALDE   84.3%   10.3%
BRISCOE   83.8%   77.6%
GAINES   83.2%   72.9%
DENTON   77.0%   8.1%
BRAZORIA   76.1%   18.3%
ARANSAS   75.0%   49.3%
COLEMAN   75.0%   77.0%
COLLIN   74.7%   6.1%
GREGG   74.5%   37.4%
WILLIAMSON   74.1%   -2.9%
SAN PATRICIO   73.8%   25.0%
GALVESTON   73.7%   19.9%
CAMERON   73.3%   -25.9%
POLK   73.1%   53.4%
CHILDRESS   72.9%   72.9%
YOAKUM   72.7%   64.3%
GUADALUPE   72.3%   25.4%
CORYELL   72.3%   35.0%
WOOD   72.0%   67.8%
CHAMBERS   72.0%   60.7%
MASON   71.7%   58.9%
BANDERA   71.5%   60.3%
LUBBOCK   71.4%   29.2%
HIDALGO   71.1%   -38.2%
WILBARGER   71.1%   54.2%
COMAL   70.9%   44.3%
HUTCHINSON   70.9%   76.8%
SUTTON   70.7%   55.5%
KENDALL   70.4%   55.3%
NUECES   70.4%   -1.5%
HAYS   70.4%   -15.4%
MADISON   69.6%   58.8%
BAYLOR   69.4%   74.0%
MILAM   68.7%   49.1%
JASPER   68.6%   61.0%
WARD   68.6%   44.4%
GRAYSON   68.5%   47.5%
BASTROP   68.4%   10.9%
REAGAN   68.3%   66.7%
ANDREWS   68.3%   62.0%
LAMPASAS   68.2%   57.1%
ATASCOSA   68.1%   28.0%
RANDALL   68.0%   59.6%
POTTER   67.9%   37.5%
WISE   67.9%   65.4%
MEDINA   67.7%   42.2%
CAMP   67.7%   41.9%
HOWARD   67.6%   53.5%
WASHINGTON   67.5%   50.9%
KERR   67.5%   52.3%
MENARD   67.2%   62.2%
MARION   67.0%   40.9%
EASTLAND   66.7%   73.8%
RAINS   66.6%   68.6%
HARRIS   66.5%   -16.7%
ORANGE   66.5%   61.2%
SOMERVELL   66.3%   65.0%
BLANCO   66.2%   45.0%
MONTAGUE   66.2%   73.9%
SAN JACINTO   66.1%   61.2%
KIMBLE   66.1%   76.3%
STARR   66.0%   -54.1%
SABINE   65.7%   74.6%
CALLAHAN   65.6%   74.9%
BEXAR   65.0%   -19.9%
LEE   64.8%   54.1%
LLANO   64.7%   57.4%
HARDIN   64.5%   73.4%
JEFFERSON   64.5%   -0.5%
PECOS   64.4%   23.4%
BURNET   64.3%   50.9%
LASALLE   64.3%   -9.4%
FORT BEND   63.9%   -12.1%
HARTLEY   63.9%   80.9%
CRANE   63.8%   59.1%
ROCKWALL   63.8%   38.4%
REAL   63.7%   68.2%
ANGELINA   63.5%   45.5%
TRAVIS   63.3%   -49.7%
JEFF DAVIS   63.1%   18.4%
TERRY   63.1%   54.7%
TAYLOR   63.0%   47.5%
EDWARDS   62.7%   60.6%
GONZALES   62.7%   48.9%
STEPHENS   62.7%   77.7%
UPSHUR   62.6%   65.5%
NACOGDOCHES   62.5%   27.9%
WILSON   62.3%   47.7%
EL PASO   62.2%   -49.4%
YOUNG   62.1%   73.7%
CASS   62.0%   59.9%
MITCHELL   61.8%   66.0%
DALLAS   61.6%   -33.0%
REFUGIO   61.6%   31.7%
WALKER   61.4%   30.0%
WHARTON   61.4%   41.0%
GRAY   61.3%   78.5%
FREESTONE   61.2%   60.4%
GILLESPIE   61.1%   58.3%
JACK   61.1%   78.3%
HOCKLEY   60.9%   59.6%
LIMESTONE   60.7%   51.2%
JACKSON   60.6%   65.3%
MCCULLOCH   60.4%   69.1%
ELLIS   60.3%   36.2%
KLEBERG   60.3%   -4.4%
MATAGORDA   60.2%   40.9%
HARRISON   60.0%   44.2%
KINNEY   59.9%   39.1%
REEVES   59.8%   -5.3%
MONTGOMERY   59.8%   45.3%
COMANCHE   59.8%   65.5%
HALE   59.7%   45.9%
AUSTIN   59.6%   58.7%
ERATH   59.6%   60.0%
ARMSTRONG   59.5%   82.9%
SHACKELFORD   59.3%   83.6%
PRESIDIO   59.2%   -47.1%
UPTON   59.2%   66.4%
MCLENNAN   58.9%   23.2%
BREWSTER   58.9%   -6.6%
FRANKLIN   58.9%   67.0%
MARTIN   58.8%   67.8%
TITUS   58.7%   42.7%
SCHLEICHER   58.6%   55.3%
PALO PINTO   58.4%   60.6%
GOLIAD   58.2%   52.4%
BOWIE   58.1%   43.0%
WALLER   58.0%   23.1%
RED RIVER   57.9%   55.5%
STERLING   57.8%   81.9%
GRIMES   57.4%   51.8%
TARRANT   57.3%   -0.7%
ROBERTSON   57.1%   37.5%
BEE   57.0%   21.2%
FOARD   57.0%   47.8%
MOORE   57.0%   60.6%
KAUFMAN   57.0%   36.8%
SWISHER   56.8%   54.9%
LIBERTY   56.7%   56.5%
LAMB   56.6%   59.1%
ECTOR   56.5%   38.6%
DAWSON   56.5%   45.8%
CALHOUN   56.5%   38.0%
HAMILTON   56.4%   68.7%
HOPKINS   56.4%   56.7%
GARZA   56.4%   67.2%
DEWITT   56.4%   62.7%
COLORADO   56.3%   51.8%
ARCHER   56.3%   78.6%
TRINITY   56.2%   60.8%
VAL VERDE   56.0%   -5.3%
COKE   55.9%   78.1%
DEAF SMITH   55.9%   42.7%
MILLS   55.8%   76.6%
VICTORIA   55.7%   40.3%
WHEELER   55.6%   85.9%
SMITH   55.6%   39.5%
BELL   55.5%   10.4%
KENT   55.5%   72.4%
TYLER   55.4%   66.3%
FLOYD   55.2%   48.9%
OCHILTREE   54.9%   79.9%
COTTLE   54.8%   64.8%
FRIO   54.7%   -10.3%
CLAY   54.7%   73.8%
HENDERSON   54.2%   58.4%
DONLEY   54.2%   74.2%
SAN AUGUSTINE   54.0%   50.8%
ROBERTS   53.9%   91.7%
ZAVALA   53.8%   -59.1%
LOVING   53.7%   75.9%
MIDLAND   53.3%   53.9%
PARKER   53.2%   62.6%
FAYETTE   53.0%   57.5%
PANOLA   53.0%   63.1%
HOUSTON   52.8%   51.4%
COLLINGSWORTH   52.7%   75.1%
BURLESON   52.7%   55.7%
WEBB   52.5%   -43.2%
CALDWELL   52.5%   8.0%
RUSK   52.3%   55.2%
HALL   52.2%   66.5%
JOHNSON   52.2%   51.8%
LIVE OAK   52.2%   66.5%
COCHRAN   52.0%   58.5%
FISHER   52.0%   53.5%
LAVACA   51.8%   73.3%
MORRIS   51.7%   40.0%
CROCKETT   51.7%   46.1%
HARDEMAN   51.2%   67.8%
CROSBY   51.1%   38.0%
SAN SABA   50.9%   75.6%
DALLAM   50.9%   74.5%
STONEWALL   50.8%   62.7%
WINKLER   50.8%   55.1%
HEMPHILL   50.8%   76.5%
NOLAN   49.5%   53.8%
HASKELL   49.5%   63.3%
BROOKS   49.5%   -43.3%
CHEROKEE   49.3%   56.4%
JONES   49.1%   65.9%
CARSON   49.0%   79.2%
NAVARRO   48.5%   44.9%
BAILEY   48.4%   49.4%
RUNNELS   48.3%   76.0%
LEON   48.3%   73.7%
BRAZOS   48.2%   12.6%
MCMULLEN   48.1%   80.5%
COOKE   48.1%   64.2%
KARNES   48.1%   41.0%
FANNIN   48.1%   60.1%
LAMAR   48.0%   54.2%
OLDHAM   47.9%   79.7%
SHELBY   47.8%   59.3%
JIM HOGG   47.8%   -44.0%
BORDEN   47.5%   86.9%
CONCHO   47.3%   65.3%
DICKENS   47.2%   69.2%
WICHITA   46.9%   40.3%
ZAPATA   46.9%   -25.7%
ANDERSON   46.5%   54.5%
GLASSCOCK   46.2%   85.9%
JIM WELLS   45.8%   -8.2%
NEWTON   45.8%   57.0%
DIMMIT   45.3%   -41.5%
KING   45.0%   90.1%
SHERMAN   44.9%   73.6%
MOTLEY   44.2%   84.1%
IRION   44.0%   73.2%
BOSQUE   44.0%   60.7%
MAVERICK   43.8%   -44.3%
VAN ZANDT   43.0%   70.0%
DELTA   42.9%   62.5%
DUVAL   42.7%   -34.9%
HUNT   42.6%   49.0%
KNOX   42.6%   57.3%
WILLACY   42.6%   -28.8%
TERRELL   42.6%   39.4%
HANSFORD   41.7%   82.7%
THROCKMORTON   41.5%   76.8%
HOOD   41.0%   61.5%
BROWN   39.8%   71.9%
KENEDY   39.4%   12.8%
LYNN   39.4%   61.6%
HILL   39.0%   56.7%
TOM GREEN   37.2%   43.2%
FALLS   35.7%   37.8%
CULBERSON   34.7%   -27.1%
PARMER   33.0%   63.3%
LIPSCOMB   31.2%   77.4%
CASTRO   31.1%   50.8%
HUDSPETH   22.0%   10.9%

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

One other thing to note is it seems like, among the counties that voted Dem/for Beto but have relatively low turnout compared to 2018, it seems like rural Hispanic counties dominate those.

Some examples of that are Maverick County (which Beto won by 44.3% but which has only 43.8% of 2018's turnout), Jim Hogg county (which gave Beto a 44% margin but where there is only 47.8% of 2018's turnout so far), and Brooks County (49.5% of 2018 turnout, gave Beto a 43.3% margin).

So, while turnout does seem to be relatively good in some urban Hispanic areas such as Nueces County (Corpus Christi), El Paso County (El Paso), and Cameron County (Brownsville), so far early vote turnout in rural and smaller-town Hispanic areas seems relatively weaker.

It is also that weakness in those rural Dem Hispanic counties that makes the negative correlation between turnout so far as a share of the 2018 total turnout and Cruz margin not be stronger. If it were not for these sorts of counties weighing things down with weak turnout, so far, then there would likely be a pretty clear positive correlation between Beto vote and high turnout relative to 2018 on the county level.

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

So far the discussion of TX early voting data has been lacking partisan data on turnout by party, because Texas does not have party registration (unlike states like Florida).

However, TX does have primary voting history and some other demographic data, and that can be seen in the voter file when you look at which individual voters have voted so far.

Here is a pretty good report on that data for TX early voters so far. It also happens to be from a Republican consultant/strategist, so it definitely ought not to be skewed/selectively interpreted in favor of the Dems:

https://mcusercontent.com/d3064a2fadaf6089dc58a8393/files/862e2ef3-a87a-4640-8919-0137e43dc5b6/Statewide_Report_Day_8.04.pdf

So far, 31.2% of voters have Republican primary voting history but not any Dem primary voting history, and 27.5% have Dem primary voting history but no GOP primary voting history. 3.0% have mixed voting history but their last primary was a Dem primary, and 0.9% have mixed primary history but their last primary was an R primary.

That compares to 28.7% Republican (0.9% mixed with last R) and 25.1% Dem (1.5% mixed with last D) in 2018 early voting.

In 2016, it was 30.1% R (1% mixed last R) to 20.3% D (0.9% last D).

So it is clearly a very substantial improvement for Dems so far relative to 2016. It also appears to be an improvement over 2018 for Dems, if you add up the R + mixed R and compare to D + mixed D, it is 32.1% R to 30.5% D in 2020 so far (a margin of 1.6% for the Rs), compared to 29.6% R to 26.6% D in 2018. If people generally voted similarly to in the 2018 Senate race, with a shift in the partisan composition of the electorate of about that much, that would take TX squarely into the pure tossup zone.


Voters with Democratic primary voting history are consistently turning out to vote at higher rates than voters with Republican primary voting history. For example, 65.2% of people who voted in the last 4 R primaries have voted, compared to 76.35 of people who voted in the last 4 Dem primaries. 48.7% of people who voted in 1 of the last 4 D primaries have voted, compared to 42.2% of people who voted in 1 of the last 4 R primaries. Dem's also hold similar turnout advantages among 2 of 4 and 3 of 4 primary voters.


One interesting thing, however, is that voters so far are older than in 2018. Only 10.5% of early voters so far are age 18-29, compared to 12.5% in 2018 and 13.4% in 2016. Whereas 21.5% so far are age 70+, compared to 17% in 2018 and 15.1% in 2016.

There are different ways you could interpret this. One is that young voters are not voting and that seniors are turning out. In TX, if that is the case, that would definitely be good for Trump (even if Trump is doing a bit worse with Seniors nationally, if he wins TX it will absolutely be with the support of older voters bringing it home for him, because younger voters are much more diverse and more Dem). Another way to interpret that, however, is that we are only part way through the early voting period in 2020 so far, and the votes that we have so far are still disproportionately skewed towards mail ballots. In TX, only Seniors are automatically eligible to vote by mail, so it is unsurprising that they should be overrepresented when the data does not include all in-person votes yet, since there has been more time for mail ballots to be returned so far than for people to vote in person. My educated guess is that the latter is mostly the correct way to interpret this. I would be extremely shocked if the electorate doesn't get younger as more and more votes come in.

One implication of that is that the later early vote and the election day vote in TX may end up being less tilted towards the GOP than in some other states where more younger people are eligible for vote by mail. And another way to interpret this is also that if the partisan composition of voters is as favorable as it is to Dems (more favorable so far than 2018) even with the electorate being skewed towards older voters, then once the Senior vote share starts to go down towards what it was in 2018/2016, things will be even more favorable for Biden/Dems than they already are.


Another thing to look at is the gender of early voters. That also looks good for Biden; 52.1% so far are women, and only 43.1% men (4.8% unknown). This is more skewed towards women than overall voter registration in TX, which is 50.9% women, 45.3% men, and 3.8% unknown. A more female electorate is definitely good for Dems.


Voters with Democratic primary voting history also seem to be out-voting people with Republican primary voting history not just in the largest urban/suburban/Beto-trending counties, but also in basically all the significantly sized counties where data is included in the report.

In Harris County (Houston, 60.5% of D primary voters have voted, compared to 52.9% of R primary voters. Dems have similar sorts of advantages across the big urban counties such as Dallas, Tarrant, Bexar, and Travis.

Importantly, since people were wondering about whether high turnout in places like Denton County and Collin County is good for Ds or for Rs, the Dem turnout advantage is also present in those counties. i.e.:

Collin: 66.3% D - 59.3% R
Denton: 65.8% D - 60.6% R
Williamson: 66.7% D - 57.0% R
Fort Bend: 53.5% D - 42.5% R
Galveston: 64.7% D - 59.7% R
Brazoria: 67.1% D - 60.3% R
It is even the case in GOP juggernaut Montgomery County: 54.0% D - 47.5% R

This is all very consistent with what we would expect to see if these suburban counties continue to trend Dem as they did in 2016 and 2018. And regardless of whether Biden wins, it is very bad for GOP chances to hold the State House and a bunch of those competitive suburban congressional seats... Honestly right now I would probably rate the TX State House at tilt/lean Dem rather than tossup, unless something unexpected changes.

Importantly, there is also a similar sort of Dem advantage in most of the smaller city counties outside of the major megacity metros (note however these are all in central or east TX). For example:

Bell (Killeen/Fort Hood Area): 56.0% D - 37.7% R
Jefferson (Beaumont area): 59.1% D - 54.8% R
McLennan (Waco area): 60.5% D - 47.8% R
Smith (Tyler area): 63.0% D - 47.1% R
Brazos (College Station area): 46.7% D - 39.2% R
Comal (New Braunfels/San Antonio exurbs): 61.9% D - 53.7% R



In Fort Worth Exurb/rural counties there is also a Dem advantage, though a small one. For example:

Johnson (South Fort Worth exurbs/semi-rural): 48.2% D - 43.0% R
Parker (West Fort Worth exurbs/Weatherford/semi-rural bleeding towards West TX): 47.7% D - 42.8% R

These areas are the deeply R of the deeply R. The fact that there is anything remotely positive at all to say about these areas for Dems is a very positive thing for Dems and very bad for Trump/Rs.

The turnout in these sorts of counties, along with the previous category of smaller city counties in central/east TX, is also suggestive of how things may be going in smaller rural TX counties. They suggest that insofar as there are Democrats in rural TX, they are voting, and the white rural non-college vote (which was already pretty much maxed out at very close to 95-100% Republican) seems like it is quite unlikely to offset the very clear urban/suburban trend we are seeing in the data. In other words, I would expect to see a similar sort of thing as in 2018 in these sorts of counties. Yes, Biden should get demolished similarly to Hillary/Beto, but I would not expect any sort of sharp Republican swing to offset the Dem swing elsewhere, and some smaller city & rural counties may even swing very slightly Dem.



There seem to be only 2 sets of counties where anything is possibly the slightest bit amiss for Dems. The first of these are West TX counties:

Lubbock (Lubbock area): 59.8% D - 59.9% R
Randall (Amarillo area): 57.0% D - 56.4% R


The second set that looks less than completely 100% ideal for Dems are South TX Hispanic counties.

Hidalgo (Edinburg/McAllen) - 48.9% D - 51.7% R
Cameron (Brownsville) - 51.6% D - 54.6% R
El Paso (not South TX and Beto country, looking more similar to the non-Hispanic urban/suburban counties): 55.6% D - 47.9% R
Webb (Larado): 33.4% D - 34.7% R

These counties are clear outliers (with El Paso being a clear outlier amongst the outliers). One thing to keep in mind, though, is that there are very few Republican primary voters in these counties in the first place. If you are a Republican primary voter in these counties, there is something that is making you be a committed voter and go against the grain. For example, even with Rs having a turnout edge in Cameron County, they are only 12.8% of voters there (compared to 43.2% D primary voters). And in all the other counties, it is lower than that 12.8% R, down to 4.8% R in Webb County.

It has often been the case in many of these counties that the Dem primary has been effectively the general election, and it has not been at all unusual for Dem primary turnout to actually be higher than general election turnout.

Nevertheless, the bottom line with regards to the South TX counties is there are a lot of Hispanic voters in South TX who have voted in previous Dem primary voters but who have not yet voted in the 2020 General election. This is pretty much the one and only thing that could be very favorable for Dems that could theoretically happen that is not clearly visible as already happening.


So overall, the early vote demographics look pretty favorable for Biden here and seem quite consistent with the polls that have been showing TX as a true tossup. I would not say they are favorable enough to guarantee a win or for TX to be lean or even tilt D, but given the comparison to 2018/2016 and how we know people voted in 2018 both of those years, it does appear that TX is truly competitive and Biden has a very real shot to win it.


The one area where there is clear room for improvement, though, is Democratic turnout in heavily Hispanic counties in South TX. Anything that can be done to increase that ought to be done ASAP, that could make all the difference in the world and could easily be the difference between Trump winning TX by a fraction of a point and Biden winning it by a fraction of a point.

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

Amazing post. Now, the important question: who will win Kenedy County (a strongly neoliberal rural Romney-Clinton-Cruz county )?

Thanks, glad you liked it. I know you are probably not expecting a serious answer to the question, but actually I can give a somewhat informed one.

A few pages back I have another post comparing turnout in all counties so far to the total 2018 turnout. https://talkelections.org/FORUM/index.php?topic=392429.msg7674306#msg7674306

I won't update that in full now so as to not spam with a large list of numbers (though I can re-post updated data if anyone wants or can PM anyone who wants it updated data, it is just a matter of copy/paste). The current data is basically similar to then.

Anyway, in 2018 there were 180 votes in Kenedy County, and Cruz won 100 votes to 77.

So far in 2020, 80 votes have been cast in Kenedy County, putting it at only 44.4% of the 2018 vote being cast so far. That compares to the statewide average of 70.3% of the 2018 vote total already being cast. At least so far, turnout in Kenedy County is abysmal compared to other places. There are only 8 counties with lower turnout compared to 2018 than Kenedy County (of those, the only one with any remotely significant county is Tom Green county, in the heavily R San Angelo area).

Low turnout in rural Hispanic counties tends to be bad for Dems... So unless turnout starts going up, tbh I would not be surprised if Trump were to win Kenedy County. I actually would not be at all surprised if Biden won TX as a whole while losing Kenedy County, lol (but also would not be surprised if it ended up the other way around)!

There may be some reasons why turnout is so low there so far. It is a very small county and it probably has only 1 early vote location, so there may be more people than in other places waiting to vote on election day or something. I would say that Kenedy County, given the turnout so far, is one place that Dems should hope that there is more turnout on election day (and in the remainder of the early vote period) bringing it up closer to the statewide average.

So I would say that Kenedy County and TX as a whole are both tossups, albeit for entirely different reasons

Now, mind you I am not currently looking at anything beyond this publicly available data from the TX secretary of state (i.e. individual voter data), so it is hard to say for sure. But turnout so far in Kenedy County looks extremely low, and in general low turnout in Hispanic counties is not good for Dems.

Also, btw, the county-level correlation between Cruz margin and current turnout as a share of 2018 turnout is now slightly more negative than before in that previous post, down to -0.042, which is another modestly good thing for the Dems.

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

Here is an update on the demographic/primary voting history data I posted yesterday for TX early voting.

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

Yesterday's report for comparison is here - https://mcusercontent.com/d3064a2fadaf6089dc58a8393/files/862e2ef3-a87a-4640-8919-0137e43dc5b6/Statewide_Report_Day_8.04.pdf


First of all, there is something weird going on in the report. The Day 8 report had, for example, 76.3% of voters who voted in 4 of the past 4 Dem primaries had already voted. But in the Day 9 report, it now says that only 74% of 4 of 4 Dem primary voters have already voted.

That should not be happening, votes do not get subtracted over time, they should only be being added. So there is probably some sort of error/mistake in one of the reports (could be either one). Maybe the guy that made the report found some sort of minor methodological mistake/calculation mistake and fixed it. Or possibly the voter file he was using might have just happened to be updated yesterday or something. In any case, this does make direct comparison of the Day 8 and Day 9 statistics a bit suspect, but we may as well look at them anyway and just keep this caveat in mind.


Perhaps partly as a result of this methodological issue/inconsistency, the share of 2020 early voters with Republican primary vote history went up slightly from 31.2% to 31.3% (but voters with mixed primary vote history who last voted in an R primary went down from 0.9% to 0.8%). However, the share with Dem primary vote history went down from 27.5% to 26.1% (and the share with mixed history but last voted in a Dem primary went down from 3.0% to 2.9%).

So with this new data, the partisan composition of the electorate looks a bit worse for Dems than yesterday. This is clouded, though, by questions about how the % of dems that voted went down.

If this change reflects an actual change in more Rs voting at this point in the early vote period, that would obviously be good for Trump. If not, and if it mostly relates to the methodology issue, then it is more ambiguous, as one would expect more voters without primary voting history on either side to vote as time goes on and the electorate broadens (and brand new voters, with no vote history at all, went up from 11.7% to 12.7%, which is already higher than 2018, but still lower than 2016).


Another thing we can look at is age... On that point, there is a pretty sharp shift towards the electorate getting younger for just a single day's additional votes being added. The share of voters age 18-29 went up from 10.5% to 11.7%


In terms of gender, BOTH the male and the female vote share went up slightly by 0.3% (this is because voters with an unknown gender went down from 4.8% to 4.2%). I would guess this might suggest that there was some sort of one-off update to the voter file and now have more information/better guesses about gender of voters.


In sum, there are multiple ways to interpret the latest data from TX especially given uncertainty about what seems like an inconsistency/methodological change. It may be good for Dems that apparently more younger voters and more brand new voters (whom one would think would tend to lean D except for non-college whites in a state with TX's demographics) are now turning out to vote. But if there was an initial wave of Dem voting which is now dying down and more Rs are starting to vote, that would be good for GOP hopes of holding on to TX.

To figure out which of those is more likely the case, we probably need to wait for tomorrow's report.