Almost twenty years ago, political demographer Ruy Teixeira and journalist John Judis wrote a book called The Emerging Democratic Majority. The thesis is that by combining women, racial and ethnic minorities, and knowledge workers, Democrats could create an enduring governing majority. This is exactly what the Democratic coalition in blue places looks like these days. And it bears more than a passing resemblance to the Obama Coalition. Clearly there was something to this.
Some carried this to a further logical conclusion. That Texas would inherently turn blue. If people of color tend to vote Democratic, and Texas was increasingly becoming non-white, then ergo, the second most populous state in the union would join California as part of an Electoral College bulwark, dooming Republicans to decades of needing an invitation to enter the White House. You can see why hopeful Democrats would buy in to this scenario.
In the several elections since the book debuted, Republicans have easily won each of the presidential contests in Texas. The GOP is unbeaten in Senate and governor’s races too. This is despite the increasing percentage of Latino voters and immigrants from blue states. Why isn’t this happening? Or is it, but something that Democrats hoped would take 10 or 20 years will actually require 40 or 60?
2016 Results: Donald Trump 52.23%, Hillary Clinton 43.24%, Gary Johnson 3.16%, Jill Stein 0.80% (R+11)
Sure, this was the first time a Republican failed to win by double digits since 1996, and the smallest GOP advantage compared to the rest of the country since 1992, but it still wasn’t close.
2020 Polls: It’s close. Donald Trump is leading by a weighted average of 1%. AKA well inside the margin of error. Joe Biden could easily win Texas. The nine point gap is being closed in two ways. Biden is running several points ahead of Clinton’s final national result. That’s most of the difference. The rest is Texas being more like a +8 or +9 for Trump instead of +11.
Perhaps the polls are a little bearish on Biden, and he’s actually slightly ahead in Texas right now. Perhaps the state is really R+7 right now. Maybe he does well in the debates, there isn’t a Covid vaccine by November, and the national gap is more like 10%, which would make Biden a winner with just a slight decline in Texas’ redness. Either way, Texas isn’t blue yet. It is just pink enough that a Democrat who does well nationally and is well calibrated to the state can win.
Key Historical Shifts:
Texas first voted in a U.S. presidential election in 1848. Like most slaveholding states, it picked Democrats. In the 1860 election, that meant Southern Democrat John C. Breckinridge over Constitutional Union nominee John Bell. Neither Abraham Lincoln, nor his main competitor Stephen A. Douglas were on the ballot. Then came the war, reconstruction, and a couple of elections (1864, 1868) when Texas didn’t participate in choosing the American president. Beginning with their return to the fold in 1872, Texas has had several distinct phases of presidential preference:
Phase One (1872-1924) Blue AF: Texas voted for the Democrat in each election between 1872 and 1924. The smallest boost any candidate got over his national support was 34 points. The maximum was 79. In 1924, Democrat John W. Davis finished 25 points behind Republican Calvin Coolidge in the national popular vote. But he did 54 points better than Coolidge in Texas. It was essentially impossible for a Republican to win the state.
Phase Two (1928) No Catholics Need Apply: In 1924, the Democrat lost a national landslide while winning Texas by 50+ points. In 1928 the Democrat lost Texas. There’s only one explanation. Democrats nominated a Catholic. Nationally, Al Smith lost badly, but got several points closer to Herbert Hoover than Davis managed against Coolidge. With the normal Texas blue tint, he’d have won the state by 20 or 30 points.
The same thing happened in the ex-Confederate states of Virginia, Florida, and Tennessee. Curiously, deciding to be bigoted against a Catholic in 1928 was a sign the state would not abandon the party for the blatantly racist candidacies of Strom Thurmond in 1948 or George Wallace in 1968 as Alabama, Mississippi, South Carolina, Georgia, and Louisiana did. Nor did they join most of those states in supporting Barry Goldwater and his States’ Rights stance in 1964 after Lyndon Johnson betrayed his native region by pushing the Civil Rights Act through Congress.
My best guess is this. While the entire former Confederacy voted against Republicans for the first several decades after Reconstruction, party loyalty was less of a thing in the states that went for Hoover in 1928. While they were more likely to opt for a GOP candidate in the era before the South went red, they also didn’t reflexively follow breakaway movements like Thurmond’s and Wallace’s.
NOTE: Texas was still blue in 1928. Smith did 14 points better there than he did nationally. But with a +14 right between a +79 and +58, the intolerance was clear.
Phase Three (1932-1948) Still Very Blue: This stretch encompasses FDR’s 4 wins, plus Harry Truman’s 1948 victory. That last election was D+37, so even in a year where Thurmond (9.1%) provided a segregationist alternative, the Democrat (66%) still got two thirds of the vote. Republican Thomas Dewey (24.3%) couldn’t even reach a quarter. At this point, you would have assumed as long as Democrats avoided nominating a Catholic, they would win Texas in perpetuity. And yet four years later, with a Protestant candidate, they lost.
Phase Four (1952-1968) A Blueish Swing State: Dwight Eisenhower won Texas in both 1952 and 1956. The religious background of his opponent Adlai Stevenson wasn’t an issue. While Eisenhower won by large national margins, his victories were not as large as Hoover in 1928. Texas was slightly more pro-Democrat than the rest of the country, but now +4 instead of +14. Something changed.
In each election between 1952 and 1968 the Texas lean was between D+2 and D+4. Still more blue than red, but close enough that a Republican could win any election in which he was winning by more than a few points nationally. No matter what happens in November, Texas was closer to neutral in this era than it is now. The Democrats needed Texan LBJ on the ballot as VP in 1960, president in 1964, and his VP Hubert Humphrey in 1968 to keep it this close.
While anti-Catholic bias was still a thing in 1960 (Virginia, Florida, and Tennessee all voted for Nixon instead of JFK, as did other normally blue states like Kentucky), it was a lesser issue than 1928. Having Johnson on the ticket was enough to keep the state blue.
When Democrats were winning big nationally in the 1930s and 40s, it wasn’t the time for Texas to start voting Republican. As soon as national conditions changed, Texas was suddenly in play. However, the state didn’t quickly move from blue to red as Virginia and Florida did. The presence of LBJ on the national stage probably had something to do with this. As you’ll see, part of what makes Texas move is the role of Texans in a given party on the national stage. The state might have moved back into deep blueness in 1932 no matter what, but it didn’t hurt that Texan John Nance Garner was FDR’s #2.
Phase Four (1972-1996) Red, But Not *That* Red: Texas officially moved red in 1972, giving Richard Nixon a +10 on top of his national landslide over George McGovern. Jimmy Carter (1976) is the last Democrat to win Texas, and the last (+1) to do better there than overall. So while the era was more hospitable to Republicans, it wasn’t toxic for Democrats. Texan Democrat Lloyd Bentsen made his way on to the 1988 ticket as Michael Dukakis’ understudy, and helped them to a modest D-5 in the state. Dukakis was not the sort of Democrat who does well there, so while it wasn’t enough for a win, it kept him from complete embarrassment.
Bentsen, who also made a perfunctory attempt at the 1976 nomination, was the only Democrat to appear on a national ticket, or make a real try at winning the presidency during the phase. Meanwhile, Republicans were suddenly a big presence. Ex-Democratic Governor and Nixon Treasury Secretary John Connally (the other guy who got shot in JFK’s limo), changed his registration to Republican and contested the 1980 presidential nomination. It didn’t work, but he raised a ton of money.
Transplanted Texan George H.W. Bush ran in 1980, was on the national ticket in 1980, 1984, 1988, and 1992. Republican Senator Phil Gramm (originally Democratic Congressman Gramm) contested the 1996 nomination, like Connally doing better on the fundraising than vote/delegate acquisition side. The center of gravity had clearly changed, but Democrats weren’t gone yet. Ann Richards won the governorship in 1990, and Bill Clinton finished within 5 points (though 13 points worse than his national result) in 1996.
Phase Five (2000-2012) Red. Very Red.: In the first four contests of the 21st century, Republicans did very well in Texas, averaging a +21 over their national popular vote, and winning each time by double digits. I don’t think the presence of George W. Bush in the first two elections is incidental. He was a very popular governor, and did far better with Latino voters than the average Republican. An important development during this time was having Latinos become a little less Democratic on average as they became an increasingly large part of the electorate.
When only 30 to 35 percent of voters who check Hispanic, non-White on their Census form are choosing Republicans, a state can get blue in a hurry. But if that’s 40 or 45% (and was north of 50% for Bush 43), the process takes a lot longer. Combined with the extreme red tint of rural Texas, as all residual Democratic ties vanished, it made for some very strong GOP math and a lot of 60/40 or 65/35 statewide elections for Rick Perry and John Cornyn.
Phase Six (2016-Present) Hmmm, Less Red. Still Not Blue: This is a different Texas. Trump’s edge here was halved from the Bush 43-McCain-Romney elections. Then there’s the 2018 Senate race between Ted Cruz and Beto O’Rourke. Only 2.6 points separated the two. Sure, Cruz isn’t very popular. Senior Senator Cornyn and Governor Greg Abbott are stronger electoral performers in Texas. Yes, Beto raised a metric shit ton of money, got an insane amount of press, and was a decent candidate too. That was still way closer than any statewide race since the 1990s. A couple suburban congressional districts near Dallas and Houston flipped blue too.
While the ethnic Texas party transition may still be taxiing on the runway, there’s another element that is making a difference. Texas is getting very urban, very quickly. Houston and Dallas are each one of the ten largest metro areas in the country. San Antonio and Austin are getting there. Each of the four is growing rapidly. For all the evidence of women being more Democratic than men, educated voters moving blue while less educated voters are moving red, and people of color voting Democratic far more often than white voters, the single biggest determinant of preference is how densely populated a voter’s community is. No state is becoming more urban more quickly than Texas.
If the correlation between population density and blueness holds, and Texas doesn’t start banning immigrants from the other 49 states, then Texas will be blue one day. It may be quickish, especially if Texas can grow a legit Democratic national candidate in the 2020s. It may be another couple decades, especially if many of the migrants from other states are coming to Texas precisely because it’s not as blue as where they left (i.e. California), but it will happen.
How Biden Can Improve on Hillary’s 2016 Result: In a world where Biden could combine inspiring the Beto Coalition (younger, urban, voters of color) with his evidenced improvement among older voters, white voters, less educated voters, he’d do extremely well in Texas. Biden is likely to run ahead of both Hillary and Beto in the exurbs and rural areas, though not well enough to get even close to Trump.
If he were to pull the same percentages with the Beto Group as O’Rourke and Obama did, taking advantage of demographic percentages far more advantageous than in 2012, you’d see Texas dead neutral if still not blue. Some of those voters will still turn out for Biden because he’s not Trump, and because Texas is looking competitive this time. But they’re not going to have the same enthusiasm they’d have for Bernie Sanders or AOC, both of whom would be poison with the voters Biden is going to do better with. The Democrat who can maximize Texas may not exist yet. It’s definitely not Biden, though he’s a better fit than Hillary was and could well win the state.
How Trump Can Improve on His 2016 Result: He can’t. Unless Biden implodes in the debate and winds up doing worse nationally than Hillary did. There are many scenarios where Trump wins Texas, and a few where he does better than his nine point win in 2016. Improving on +11 in Texas is really hard though. There are a lot of new Texas voters every four years, and these new voters are more likely to pick Biden than the Texans who were already there in 2016.
Forecast: I think Biden will at least make Texas close enough that the networks can’t call the result on election night. I also think he’ll run within 7 or 8 points of his national popular vote number, perhaps even 6. Whether it’s enough to win the state and pad what would likely already be a strong win, no clue. Texas isn’t going to get Joe Biden elected, but it can make a win more resounding.