This article is part of our 2018 Texas Elections coverage, where you can find the latest in news, analysis, and updates from Texas Monthly. Read More

With rapid fire patter, Congressman Beto O’Rourke of El Paso has criss-crossed Texas—touching all 254 counties as he is apt to remind you—and drawn thousands of progressives to his rallies, filled college auditoriums with students, and raised a prodigious amount of money through the internet. His yard signs seem to be as ubiquitous as mushrooms after a rain. T-shirts. Bumper stickers. The Resistance seems to have embedded its hopes for change into O’Rourke’s Democratic campaign. Some Republicans—unhappy that an uncouth loudmouth from their party is their president—have said they will break tradition to cast a ballot for Beto. The evidence is right there in front of your eyes. A new day is dawning in Texas politics, and Beto O’Rourke will break the drought that has not seen a Texas Democrat win a statewide election during the entire lifetime—or at least since the childhood—of some of his most ardent supporters.

But some of the most recent polls say O’Rourke is going to lose in his challenge to Republican incumbent U.S. Senator Ted Cruz, perhaps badly; that Beto O’Rourke is about to become the latest Democratic disappointment in a long string of disappointments. Is Beto a dead politician walking? Or have the polls seriously missed a rising tide of Democratic turnout?

A Quinnipiac University poll released Thursday found the race had stabilized with Cruz leading O’Rourke 54 percent to 45 percent, the same as a Quinnipiac poll last month. The survey of 730 likely voters had a sampling error of plus or minus 4.4 percentage points, and has pollsters there all but declaring Cruz the winner next month. “Is the Beto bubble bursting or just hissing away with a slow leak? With less than four weeks until Election Day, Congressman Beto O’Rourke has hit a wall and remains the same nine points behind Senator Ted Cruz as he was when Quinnipiac University polled the race last month,” said Peter A. Brown, assistant director of the Quinnipiac University Poll. “The election is far from over, but Senator Cruz would have to suffer a major collapse for him to lose.”

One of the most amazing components of O’Rourke’s high profile challenge is how his image among Texas voters has changed in a negative way over just a few weeks. During the summer, when he was flying high in the polls, anywhere from a third to half of those surveyed did not know enough about O’Rourke to have a positive or negative view of him.  But in the latest Quinnipiac poll, O’Rourke is now viewed favorably by 45 percent of those surveyed and unfavorably by 47 percent. Cruz has a 52-44 percent favorable rating.

A live survey being conducted this week by the New York Times/Siena College had a similar result. O’Rourke is viewed positively by 41 percent of the respondents and negatively by 44 percent. Those are tough numbers for a candidate who is campaigning on the idea of Republicans, Democrats, and independents working together to solve the nation’s problems. In the the survey itself, highly partisan Cruz is leading 52 percent to 43 percent.

In two other prominent surveys this month, Emerson College has Cruz leading 47 percent to 42 percent among likely voters, and CBS New/YouGov has Cruz at 50 percent and O’Rourke at 44 percent. Two surveys in September had O’Rourke slightly ahead or even with Cruz, but none in October have put the race in doubt for Cruz if the traditional turnout ratio occurs. The ABC News FiveThirtyEight political analytics website gives O’Rourke a one-in-four chance of winning.

But are the surveys capturing the Democratic intensity? If a poll models past Texas turnout, then it likely will favor a Republican candidate. The Quinnipiac survey consisted of 23 percent Democratic respondents, 35 percent Republican, and 35 percent independent. When I looked at the responses to the New York Times survey, I found that those talking to the pollsters broke down as 28 percent Democrat, 38 percent Republican, 28 percent independent, and 3 percent other party. Then geographically, the respondents were 20 percent from a major city and 8 percent from South Texas, both Democratic strongholds; while 38 percent were from major suburbs and 21 percent from rural Texas, both Republican strongholds.

Keep in mind, President Trump received 52 percent of the Texas vote; Democrat Hillary Clinton, 43 percent, with the rest spread among other candidates. High-profile Democratic candidates have not broken 44 percent of the vote in more than 20 years. So are the polling models based on this traditional turnout in Texas? Or do they take into account that this is supposed to be the year of the “blue wave” in which there is a palpable backlash against President Trump that has resulted in the largest engagement of the Democratic electorate nationally since 2006, when the party took control of the U.S. House for four years?

Age also is a factor—73 percent of the respondents to the New York Times survey were 45 years old or older. That age group is most likely to vote and vote Republican. O’Rourke’s support is embedded in youth. Nationally, a new survey found Democrats holding a 30-point lead among millennials in congressional races, but that doesn’t mean they will vote. As Democratic political consultant Keir Murray of Houston noted on Twitter on Wednesday, O’Rourke needs a vote from those who seldom do. Early voting begins October 22.

Rice University political scientist Mark Jones said all polls have to model an expected turnout. “If it misses badly in terms of turnout model, all bets are off,” Jones told me. “So especially if Beto is able to drive unprecedented turnout among Anglo college students and other millennials, then the poll numbers may be off.” Jones said O’Rourke would have to just about triple the typical youth vote to make that happen, though.

The polls themselves have the potential to dampen O’Rourke’s voter turnout efforts, he said. O’Rourke this week is on a statewide tour of college campuses, receiving large crowds. “One of the ways to give them enthusiasm and get them to the polls is to convince them that Beto has a realistic chance of victory. But the more we see poll after poll that says, ‘he’s close but no cigar,’ that could lead to those voters effectively staying home because Beto is the only thing driving most of them to turn out to vote this cycle,” Jones said.

Southern Methodist University political scientist Cal Jillson told me he believes O’Rourke’s negatives were driven up after his first debate with Cruz by the question of whether O’Rourke tried to leave the scene of a drunk driving accident in 1998. O’Rourke had admitted the drunk driving arrest and apologized for it, but the question came from a police report that said a witness had to stop O’Rourke from leaving the scene. O’Rourke has adamantly denied that he tried to flee, but a Washington Post fact check said he lied, and Republicans started using it to counter questions that Democrats were raising about whether Supreme Court nominee Brett Kavanaugh had sexually assaulted a woman while he was drunk in high school.

Jillson also believes O’Rourke’s hopes rely heavily on a youth vote that finds it easy to answer a cell phone for pollsters and difficult to go vote. “If your electoral strategy depends on things happening that don’t normally happen, you’re probably in for a rough ride,” Jillson said. “With Beto, that’s the idea of that collage tour: to go around and try to increase young people’s voting, increase single women voting, Hispanic voting. All of those things are a pretty heavy lift.”

Although O’Rourke has raised more money, Jillson said Cruz is benefitting from heavy spending on television by Governor Greg Abbott and Lieutenant Governor Dan Patrick, who are not facing serious challenges. They create “background messaging” that fires up the Republican base. “Ted Cruz is in the driver’s seat, and the polling has settled back to him being somewhere in the mid- to upper single digits in the lead, which is what you would expect of an only marginally attractive Republican candidate in a good Democratic year.”

Abbott is up 20 percentage points over his Democratic opponent, former Dallas County Sheriff Lupe Valdez, according the Quinnipiac poll. Jillson said Abbott is better positioned because of “a difference in personality.”

The standard line for political candidates who are trailing in the polls is: the only poll that matters is the one that occurs on Election Day. That survey of the ballot box usually is a foregone conclusion. I’m not so sure this year. There were about two million Democrats who turned out for Hillary Clinton in 2016 who did not vote for gubernatorial nominee Wendy Davis in 2014, and Clinton outpaced gubernatorial nominee Bill White by 1.7 million votes in his 2010 midterm election. If O’Rourke can inspire a substantial portion of that drop-off vote to turn out, then he has a chance. Plus, there were 475,000 people who registered to vote after the March primaries. Texas does not register by party, but I have to believe much of that registration was by Democrats.

Both O’Rourke and Cruz have been driving their supporters on the idea that the election will be decided by turnout. If the polls are correct, Cruz already has a substantial advantage and the likelihood is that he will receive another six years in the U.S. Senate.