Midterm Election

I have thought long and hard about not writing this post. Yet, I have decided that I should, otherwise, there will no record of the anxiety I am going through, and the total amount of time I have spent looking at the polling data in Texas. Many of you know that I like to plot data and study. Election polling dataset analysis is a quintessential application of data visualization, statistical modeling, and graphical communication. Because poll data combines binary/multinomial response outcomes, temporal tracking, spatial distribution, demographic stratifications, and quantifiable margins of error, it presents a rich playground for data science and graphic design.

Voter registration in Texas reached a record 19,041,684 registered voters. Of those 19M voters, as of this writing 6.2M have already voted. For those who perhaps are not politically motivated, you know that this year Texas is a swing state. It is therefore critical to understand how Texas votes. Now, I will try my best to keep this post away from political opinion and focus just on the data. Because to me ultimately that is the goal of this post.

Blue Spine of Texas

I do not know how many of you know this, but democrats in the State of Texas have long tried of find a footing in the state and win a statewide race. Every election in recent past we have tried and lost in the end. The closest we have been in a statewide race in recent memory was the Beto O'Rourke - Ted Cruz senate race in 2018. I remember that race well and we lost with 2.7 points.

If you are a democrat, this is the best map of blue in Texas you have seen in recent times. On this map, the blue urban counties along the I-35 corridor, from north of DFW to Austin to San Antonio to the Northern Rio Grande Valley is known as the Blue Spine of Texas. Whoever wins this spine, wins Texas. It matters not the most populated county in Texas, the Harris County, Houston, where I live. That is always blue, but it can't pull the state numbers. That is how the math works.

Early Votes

If you look at the 2018 Senate race, a little more that 8.2M people voted. At 5-days out right now, we are at 6.2M. We are tracking lower than 2020 and 2024, both were presidential elections, which typically show higher turnouts. However, we are tracking higher that 2022, which was a midterm Governor race, O'Rourke - Abbott, which Beto lost by almost 11 points.

So it is fairly safe to say, this year we will definitely turn out more than 2022 midterm and likely more than 2018 8.2M number. The question in my mind is, is the higher turnout going to be good enough to pull the votes along the Blue Spine. As of now Talarico is polling consistent 4+ point higher than Paxton, but the trouble is in Texas, democrats typically poll about 4.2 points higher, so the lead is still within the margin of error. I am also deeply invested in the Governor race where I prefer Abbott to be gone forever, but that is even a tighter race where Gina Hinojosa and Greg Abbott is polling even to Abbott with a point lead.

I don't know what will happen on Nov 03, but I indeed am anxious, and consuming as much data as I possibly can!