President Trump speaks during a rally at the El Paso County Coliseum in El Paso, Tex., Feb. 11. (Eric Gay/AP)

Supporters of President Trump like to point out that Hillary Clinton’s popular vote victory in 2016 really came down to California. She won nationally by 2.9 million votes and California alone by 4.3 million. Drop California into the Pacific Ocean, and Trump has that landslide he kept talking about.

But that coin also flips the other way. But for 807,000 votes in Texas, Clinton would have won there — and the presidency. Big states have a big effect.

Why rehash this fight? In part, certainly, so I can spend my weekend deleting angry emails. But also in part because of new data released by Gallup, showing how respondents in each state felt about Trump’s presidency over the course of 2018.

The good news for Trump is that he had 50 percent approval or higher in 17 states, up from 12 in 2017. (The additions? Kansas, Mississippi, Missouri, South Carolina and Utah.)

The bad news? In 13 states he won in 2016, his approval rating is underwater. And in no state that he won in 2016 is his approval rating worse than in Texas.


(Philip Bump/The Washington Post)

Let’s visualize this another way. Let’s compare Trump’s margin of victory in each state in 2016 with his net approval; that is, the percentage of those who approve of him minus the percentage who disapprove. In 29 states, his approval is “underwater,” meaning that his disapproval is higher than his approval, and his net approval is a negative number. He won nine of those states in 2016.


(Philip Bump/The Washington Post)

That diagonal line shows the trend. We’ve added it to show that there’s a correlation between how he did in 2016 and his net approval now (unsurprisingly) but also because it allows us to visualize how states have shifted.

Notice Alaska and Mississippi, floating way above that diagonal line. Relative to other states, their net approval in 2018 was much better than would be expected, based on their 2016 vote. In Maine and New Hampshire, well below the line, the net approval was worse than you would expect.

Now notice that cluster of states right around the middle. In Ohio, Trump’s net approval is above where the trend would suggest. In Texas, it’s a bit lower.

You might think to yourself, well, sure, but Florida and North Carolina are doing a bit better, relative to the trend, than you might expect, because they’re above the line. And that’s true. But if distance from the line ends up being correlated to the state’s 2020 vote (an “if” the size of Jupiter, mind you), it doesn’t do Trump any good to perform slightly better in North Carolina, Ohio and Florida — states he won — if he loses Texas.

There are more caveats worth applying here. Texas saw an improvement in its view of Trump from 2017, when he polled at 39 percent over the course of the year. (The biggest changes since 2017 were in Alaska, Mississippi, Maine and New Hampshire, states that have come up in this article already.) And Texas also had a statewide race last year in which Sen. Ted Cruz (R) held off Rep. Beto O’Rourke (D), which suggests the state isn’t turning blue anytime soon. (Though O’Rourke did pull within 2.5 points.)

Discussion of Texas going for Clinton in 2016 ended up being a punchline after the election was over, given that the state backed Trump by 9 percentage points. But it was closer than Iowa, a state Barack Obama won by almost 6 points in 2012.

Perhaps it’s not an accident that Trump just endorsed Sen. John Cornyn (R-Tex.) for reelection, or that he just had a campaign rally in the state earlier this month. But, then, that rally was meant to build support for a wall on the border with Mexico, an issue that 45 percent of Texans supported this summer. What’s more, Trump’s declaration of a national emergency to get it built could kick off high-profile eminent domain fights, which likely wouldn’t earn him many fans in the border state.

It’s a state that Trump would be advised not to take for granted.