You may have heard this somewhere before: In the run up to Election Day, a high-profile Democratic challenger in a closely watched race appears to be pulling ahead in the polls, a race that could flip a seat. But then comes Election Day and a lot of long faces among expectant Democrats.
Then comes the inevitable follow-up question: “Wait a minute. What about those polls?”
Alexandra Samuels of Texas Monthly puts the matter more pointedly in a new piece asking whether polls saying the Lone Star State could turn blue can be trusted. She joined Texas Standard to discuss. Listen to the interview above or read the transcript below.
This transcript has been edited lightly for clarity:
Texas Standard: We’re about to enter the homestretch here – under 50 days before Election Day. Let me ask you right off the top: Anyone this election season saying Texas could go blue? I’ve seen some polling out there that suggests perhaps a purplish hue, but what are you hearing?
Alexandra Samuels: So I don’t know if “blue” is the exact phrasing I’m hearing, but there’s a lot of chatter that the state could be very competitive.
Democratic vice presidential nominee Tim Walz was at a Dallas fundraiser last week where he proclaimed that the U.S. Senate race is both within reach and winnable for Democrats. And then at the DNC last month, too, many Texans there expressed a similar optimism that the state could turn blue in a few years.
But you know, kind of to your point, this line of thinking isn’t anything different than what we’ve heard before. Every two years or so, Texas Dems say that their next big win is just around the corner. But victory so far has been pretty elusive.
Well, I guess you’d have to say “our next victory is around the corner” or else you wouldn’t have much turnout. If you’re a Republican, the line is going to be, “Well, we’ve got everything to lose here, and we can’t take anything for granted.” On the Democratic side, it’s “This could be the year. All we need is an improved turnout.” And that’s been the story for many election cycles, it seems.
That’s very true.
I think one reason why Republicans express some nervousness and why Democrats could express some excitement is because every year these statewide races do get closer and closer, particularly at the presidential level. We haven’t exactly seen that trend for U.S. Senate races and other statewide races. But at the presidential level, former President Trump only won the state by about five percentage points in 2020.
So, I think when we see those margins narrowing, that gives Democrats the feeling of “Next year could be the year.” And that’s why we see that excitement come every election cycle.
» RELATED: Rapid population growth is changing the voter map in Texas
And to your point, we’re not talking there about pre-election polls. We’re talking about a narrowing at the ballot box, and so there’s something consistently happening there over the last several election cycles. But it’s been decades since any Democrat won a statewide office in a Texas election. I think you have to go back to the 1990s.
I think about 2018, it was Ted Cruz running for Senate race. And that seems to be a sort of fulcrum here – something that has become a closely watched thing over the last couple of election cycles. O’Rourke was running against Cruz, and I remember seeing black Beto signs in people’s yards. It did seem like we had reached a kind of tipping point. The polls were showing that Beto was moving up close, and it seemed as if Cruz was running a little bit scared. What happened ultimately there in that race, do you remember?
So that was a year of palpable Democratic excitement. But I remember in August of that year, Texas Lyceum released a poll suggesting that the race between Cruz and O’Rourke was an essential tie. I think they had it within two percentage points.
Of course, Cruz won his reelection bid by about three points. And that’s a very small margin for a Republican running statewide, but that’s just one example of the polls showing one result and then the end election result being fairly different from what the polls predicted.
I want to turn to your article. It’s a fascinating read, and you actually put together what you call “guidelines on how to temper or stoke your own hopes or fears” of a blue tide rising in Texas. You mind if we go through a few of these recommendations?
Not at all.
Number one: Don’t get too excited about outliers. What do you mean, “outliers”?
From now until Election Day, we’re going to get a bunch of new polls. Some are going to look better for Democrats and some are going to look better for Republicans.
In August, we had a University of Houston/Texas Southern University poll that showed Allred was only behind Cruz by two percentage points. So essentially, what they were saying was that the race could be a toss up, but what we advise in the story is to avoid getting too excited by individual polls.
Instead, it is more helpful to use polling averages and to look at trends over time. Since the March primary elections, Cruz has actually led Allred by an average of nearly seven percentage points. So in this case, the two polls showing the race being an essential tie would be considered an outlier.
So you would want to look to polling averages like those put together by FiveThirtyEight?
Exactly, yes.
Here’s tip number two: Pay attention to possible methodological issues. What exactly are you referring to there?
Well, we’re never going to see a survey that polls all of Texas’ 30 million residents, let alone our voting population. What pollsters like to do instead is to select the sample of Texans that are representative of the state, but not all polls will do this super carefully.
So, if you’re reading a poll and you have time to look at its methodology, a good tip is to check how many people were surveyed. The gold standard for a statewide poll is between 1,000 to 1,400 people. And if you have extra time, you can also look at whether the demographics are a reflection of Texas’ electorate in terms of race and education.
Very interesting. So, for instance, you look for voters who might not have a college degree and how well-represented are they in this survey, that sort of thing?
Exactly, yes.
» GET MORE NEWS FROM AROUND THE STATE: Sign up for Texas Standard’s weekly newsletters
Let’s go on to suggestion number three: Note what types of voters are being polled. What do you mean, “what types”?
So related to my last point, you’ll see that some pollsters interview “likely voters,” while others speak to registered voters or Texas adults. And in election polls in particular, likely voter surveys tend to be the most accurate, and that’s because higher percentages of likely voters cast ballots in presidential elections.
So especially now after Labor Day, we really want to be looking at surveys who interview likely voters versus registered voters.
Roger that. Now, you have another point here. Number four: Consider the source. What do you mean by that? I know that there are some news organizations out there that are doing their own polling. There are some familiar names in the polling game. How do you know what sources to trust?
So that’s why I really appreciate FiveThirtyEight’s polling average landing page for 2024 in the Senate race, because what they’ll have in there is it’ll show who the sponsor of the poll is, and you can see whether it’s a Democratic sponsor or a Republican sponsor.
One trend that I’ve picked up is that the polls that were sponsored by Democratic groups give Cruz maybe like a two-point edge or a four-point edge. And the polls that are sponsored by a Republican group give Cruz a ten-point margin. And so when we look at these more nonpartisan polls, we just want to be a little bit more skeptical because often they will paint a rosier picture for whatever candidate they’re ultimately backing.
How is it that the polls get such radically different numbers, unless you’re tweaking the data? Maybe that is what’s happening on the back end – you’re building in some presumptions about what the data you collected suggests.
I think part of it comes down to who is being interviewed. And again, whether it is representative of Texas’ population, how many people are being surveyed, when you’re in the field… Some polls were conducted while President Biden was still the nominee for 2024, and some were conducted after Kamala Harris took over.
So you have to look at a lot of these factors that will tie into it. But since there are so many polls, again, that’s why it’s good to look at the average versus looking at individual samples and individual surveys.
Alex, let me ask you a little bit behind-the-curtain question, if I can. Do you trust the polls?
I do. And that’s because polling is still the most useful tool we have to gauge where the electorate is.
You know, polls are not meant to be predictive. They are snapshots of where the electorate is at any given moment in time. And I think polling is most useful when we’re honest about what it can and cannot tell us.
So obviously, prognosticators took a lot of heat in 2016 when a lot of them didn’t correctly predict that Trump would beat Clinton. But pollsters are always refining their methods, and the industry looks vastly different now than it did in 2016 or even in 2020. And I think looking at polling in the aggregate or as an average is still very useful.
