This morning, a new survey from Public Policy Polling startled national observers with an update on the state of play here in Texas, the most notorious red state in the union. With less than three months to go until the general election, Donald Trump leads Hillary Clinton by just six points, 44-38.

It’s just one poll, of course, and with a 3.2 percent margin of error, the spread may actually be larger. But still the poll confirms some ominous trends for the GOP. Readers may remember that I wrote, back in May, that Clinton had a chance of winning Texas’s 38 electoral votes this year. My reasoning was as follows. First, I thought it was a safe bet that Trump would underperform Mitt Romney, who won 57 percent in Texas in 2012. That is obviously the case, here and elsewhere; interestingly, as Nate Cohn wrote last week, the redder the state, the more visibly Trump underperforms. Second, I thought the Libertarian candidate could pick up a significant number of Republican defectors, between 5-8 percent; Gary Johnson is at 6 percent in the new PPP poll.

And third, I thought Clinton could be expected to at least match Barack Obama’s performance in 2012, when he captured 41 percent of the vote in Texas. I still think that’s the case. In national terms, Clinton is the second-most unpopular presidential candidate in modern history, after Trump, but she has always fared comparatively well among Texas Democrats. She beat Obama in the state’s 2008 primary, and walloped Bernie Sanders by a roughly 2:1 margin in Texas this year. Her support in the new PPP survey is a little lower than I would have expected. I suspect she’ll break 40 percent on election day, particularly if Democrats step up their efforts at voter registration and turnout: as PPP notes, Clinton has a whopping lead among young Texans and Hispanic Texans, two demographic subsets that have, historically, had disproportionately low turnout rates.

That brings us to a key reason Democrats should be contesting Texas this year, which I noted at the end of May. At the time, most observers were skeptical that Clinton would even be within striking distance of victory here. Today, given the latest electoral college projections, it seems unlikely that she needs to worry about Texas, and if the race tightens her campaign might be well-advised to allocate its resources elsewhere. The electoral infrastructure Democrats can build in Texas this year, however, will linger for years to come, even if the Republican Party of Texas’s brand recovers from the damage Trump is currently inflicting on it, with young and Hispanic voters in particular.

Two more small notes, related to the PPP survey. First, as of last week there is an independent conservative candidate, Evan McMullin, in the race; yesterday, his campaign announced that he will be on the ballot in several states, including, probably most crucially, Utah. He is unlikely to be a factor in Texas: the PPP survey found him earning 0 percent support, and the filing deadline for independent candidates was back in May. Still, his campaign has announced that it will sue for ballot access, and Jamie Lovegrove, at the Dallas Morning News, has a fascinating look at the constitutional question at hand.

And then there’s this finding, which PPP notes in its summary of the Texas results:

“We continue to find that Trump voters overwhelmingly buy into his preemptive claims about the election being rigged. Just 19% of Trump voters grant that if Clinton wins the election it will be because she got more votes, while 71% say that it will just be because the election was rigged.”

It was entirely predictable that Trump would make this argument. As I wrote in August, “His ultimate failure will be taken as proof that the game is rigged–against the candidate, but also against people like themselves, his supporters.” Texas Republicans, however, were laying the groundwork for him to make such claims for years prior to his appearance on the national scene—and if a large segment of the Republican base insists that Trump’s loss is due to malfeasance rather than understanding his defeat as the inevitable result of glaring and anomalous weaknesses, the state GOP may regret that in the years to come.