I don’t know whether Joe Biden will win Texas, but Joe Biden can win Texas. At the front end, let me emphasize here that this is not an attempt to forecast Texas. I do not know, and the little Excel spreadsheet model that I built requires one to make a ton of assumptions about things that are yet to happen. In fact, Texas is interesting precisely because much of the information we have about 2020 is cancelled out by what we don’t know. What we do know is that Texas will have millions more new voters than it had in either 2016, when Trump first won election, or 2018, when Ted Cruz narrowly edged out Beto O’Rourke to win re-election to the Senate.
But the nature of new voters is that it is hard to know what they will do. Forecasting models give Trump the edge to hold on in Texas, but one need not go far to find polls showing Biden leading the state–a PPP poll today showed Biden leading 50-48. And Trump won the state by nine points in 2016–similar to the margins by which he won Iowa and Ohio. Since those states are also competitive, it’s probably no surprise that Texas is as well. But Texas is Texas. Texas is supposed to be different.
What we do know is that approximately 9.7 million ballots have been cast in Texas so far, and we know where those ballots have been cast. That is approximately 700,000 more votes than were cast in all of the 2016 election, and some are predicting a turnout of 12 million votes–which would mean approximately 2.3 million more voters before Election Day ends. About 60% of the 9.7 million ballots so far have been cast in Texas’ 10 largest counties, which when combined lean toward Democratic candidates:
Harris County – City of Houston. Once an urban GOP stronghold during the Reagan/Bush years, now the GOP is collapsing and Democrats are ascendant. Every successful Biden map in Texas requires massive Harris County turnout.
Dallas County – City of Dallas. The urban center of Dallas is itself heavily Democratic, while its suburbs–some of which are listed below–have been historically ruby red Republican.
Tarrant County – City of Fort Worth. More Republican than its neighbor, but has also been sliding more Democratic over the past decade.
Bexar County – City of San Antonio. Historically Democratic, but new growth has kept the county relatively competitive.
Travis County – City of Austin. The freewheeling id of Texas Democratic politics. Democrats are likely to win 75%-80% of the vote here, though its suburbs have been heavily Republicans.
Collin County — Dallas suburbs, including the city of Plano. Once as conservative as territory could be, an influx of new corporate interests has diversified the county. Underlying district-level polls suggest even the Republican congressional incumbent may be endangered here.
Denton County — Collin’s neighbor, and the same is happening here as in Collin.
El Paso County — Another historically Democratic county, here the issue is always turning out voters in a corner of the state many hours from the other urban centers.
Fort Bend County — High-growth suburban Houston, where Harris’ slide toward Democrats is also manifesting.
Hidalgo County — City of McAllen and the Rio Grande Valley. Like El Paso, Democrats need to make sure to turn out their voters here, which is why Kamala Harris made this a stop on her trip to Texas on Friday.
Then there are 200+ some odd other counties in Texas, but what is interesting about those is that even as O’Rourke was closing Trump’s 9 percentage point gap to 2 in 2018, he was only winning 33% of the vote in those 200+ some odd other counties–2.7% more than Clinton had won in 2016. Take a look.
So, in 2016, Trump won by approximately 9 points after carrying the larger counties by 13 points but swamping Clinton in the rest of Texas by nearly 40 points. Two years later, O’Rourke sliced Ted Cruz’s margin to 2 1/2 points by winning the largest counties by 20 points, and narrowing Cruz’s lead in the rest of the state to 33 points.
The good news for Biden is that with 4,000,000 new votes to play with, it is far easier to imagine him making up that remaining 200,000 vote margin–assuming he holds on to all of O’Rourke’s voters. In any event, there’s enough information that one can make a series of calculations as a baseline for when the votes are counted on Tuesday.
|2020 Early Voting – conservative assumptions|
|Other Texas Counties||2,685,479||1,263,755||3,949,234||68.00%||32.00%|
|Assume 12,000,000 turnout scenario|
|Top-10 Votes remaining||456,421||688,926||1,145,347||39.85%||60.15%|
|Other Votes remaining||778,836||366,511||1,145,347||68.00%||32.00%|
In this model, I assume that Biden wins the historically Democratic counties (plus Harris and Fort Bend) by large margins, but that Trump ekes out majorities–albeit reduced ones–in the Dallas suburbs. The model also assumes that even if Trump isn’t able to match his 40-point margin in the smaller counties–he outruns O’Rourke and wins 68% of the vote in those counties (including 68% of the vote remaining). This model also assumes that larger counties have been somewhat cannibalizing their election day vote, and that the remaining 2.3 million votes are split equally between the Top 10 counties and the rest.1To date, just short of 60% of Texas votes have been cast in the Top 10 counties, itself an uptick from 2016, where those counties had about 56% of the vote. Trump wins here by approximately 4 points–unimpressive, but a solid win.1The model assumes the 12,000,000 voter turnout some have been predicting. No one knows what turnout will actually be.
|2020 Early Voting – moderate assumptions|
|Other Texas Counties||2,606,494||1,342,740||3,949,234||66.00%||34.00%|
|Top-10 Votes remaining||520,217||854,200||1,374,416||37.85%||62.15%|
|Other Votes remaining||604,743||311,534||916,278||66.00%||34.00%|
Now it gets more interesting. Here, I assume that most of the new Harris County vote is turning out for Biden, that he breaks through and wins by modest margins in the Dallas suburbs, including Fort Worth, and that O’Rourke’s rural voting initiatives are able to help Biden reach the same 34% among the smaller counties that O’Rourke did. Further, I assume that the remaining ballots to be cast before election day are allocated in the same ratio as they have been cast to date. Accepting those assumptions, Biden manages to eke out a narrow win.
|2020 Early Voting – aggressive assumptions|
|Top 10 Counties||2,115,488||3,644,584||5,760,072||36.73%||63.27%|
|Other Texas Counties||2,527,510||1,421,724||3,949,234||64.00%||36.00%|
|Top-10 Votes remaining||504,823||869,593||1,374,416||36.73%||63.27%|
|Other Votes remaining||586,418||329,860||916,278||64.00%||36.00%|
Meanwhile, this is the one readers of this site are probably most interested in: the jailbreak scenario. Here, Biden runs up the score in Harris County, sets modern records for Bexar and Fort Bend, and the great realignment in the Dallas suburbs comes to pass. Further, Democrats make up another two points of ground in the smaller counties, capturing 36% instead of Clinton’s 30%. This is a lot to ask, but it does lead to a 4.5 point Biden victory.
Again, none of these scenarios are likely to hit the mark. I don’t know what turnout will be, much less who will actual vote. Texas doesn’t register voters by party, so they can potentially be matched to various primaries, but doing so will be imprecise, and millions of voters have no Texas voting history at all. The point is not to forecast what will happen in Texas, but to demonstrate what may happen on election day, and how you can fit the vote count as it comes in so that you can recognize the trendline before all the votes are counted. That said, I do think the results will end up somewhere within the spread described above, and that this is as good a start as any for when the votes actually come in.