Nailing the Vote: Jason Corley on 2024 Polling and What’s Next for 2026

After every presidential cycle, the debate over polling accuracy takes center stage. Jason Corley, Founder and CEO of Quantus Insights, joins us to discuss why Jesse Watters said his firm “nailed the popular vote” in 2024, what they got right, and what it means heading into 2026. 

Learn more: Frontlinestrategies.co

Transcript

SPEAKER_00 0:01

This is the Digital Politics Podcast, presented by Frontline Strategies.

SPEAKER_01 0:05

Hello and welcome to the Digital Politics Podcast. I'm your host, Mike Hahn, president of digital at Frontline Strategies. Every month we break down what matters to the GOP digital industry, talk to the people driving the change, and give you the insights you won't find anywhere else. Today's episode is one I'm excited about. In modern campaigns, polling and public opinion research shapes nearly every major decision, from messaging and media to targeting and turnout. As is always the case after a presidential election, there's a lot of debate about which pollsters got it right, which ones missed the mark, and what that means heading into November 2026. So today we're joined by someone who's been right in the middle of that conversation. Our guest is Jason Corley, co-founder and lead pollster of Qantas Insights, the firm that Jesse Waters highlighted on Fox News as one of the only firms to quote, nail the popular vote in the 2024 election. Quantus is a premier provider of polling, election forecasting, economic analysis, and advanced modeling. Their mission to provide political and economic intelligence that influences decision making in elections and public policy. And we're excited to have the opportunity to ask them some questions today. So Jason, welcome to the show.

SPEAKER_00 1:10

I'm glad to be here. Thank you for having me.

SPEAKER_01 1:12

Thanks for doing this the week of the reads. I don't know if you're going or not, but um I'm heading out on a flight after this. So if you are, thank you for fitting this into your schedule. I know many consultants don't like to work in market We don't mind.

SPEAKER_00 1:22

We like working every day, so we like being busy. We're we're the geeks. We we don't mind it at all.

SPEAKER_01 1:27

Data never sleeps. So a prime time slot on Fox News, Jesse Waters saying that you guys nailed the popular vote in 2024. Must have felt pretty nice. So tell me how you guys did it and what others missed in 2024.

SPEAKER_00 1:40

Yeah, so that was a little surprising. It was like winning the lottery, then catching lightning in the bottle all on the same uh on the same day and same 24 hours later. I mean, to tell you the truth, we I was supremely confident to almost to the point to where it was just creepy. I was just calm. Uh flew out to San Diego and set up our little war room and we we watched the returns come in. And it was pretty early in the evening that I knew we were gonna be we were gonna be right because we had models running parallel to all of our polling. And it was just it was just a great uh validation of what we tried to work on and what we tried to pick up on inside the electric. But 24 hours later, yeah, the big Jesse Waters shout out for the national popular vote. But we also did extremely well in the uh Rust Belt states, North Carolina. We missed Virginia, which is that tends to come up a lot for some reason. Amongst some on the on the Twitter world, the X world. But we had Nevada going to Trump uh as early as September. We had North Carolina going to Trump in September, uh and then we got into the field in October for Pennsylvania, uh, Michigan, Wisconsin. We missed Michigan, but it was only like about 1.1 points, and we absolutely nailed Pennsylvania. I think we were off by point zero point three points. And in Wisconsin, we were only off by like 0.5 points. So we we nailed this stuff like Atlas Intel type margins. But going back to how it all started, how did what did we see uh that I guess gave us some additional insight? Was we really concentrated on on voter behavior. So we had three election cycles with one man involved in all three election cycles. So 2016, 2020, and then 2024 coming. To me, it was one of the easiest elections, or only really national election, big time election, but I thought I thought it was pretty easy to nail. There was just so many indicators pointing towards a swing from 2020. We wrote an article about this like in June or July. We predicted about a six-point swing to the right. I think it came out to be like literally 6.1% was actually the actual swing. Um we picked up on that because we again we have economic indicators models, we back test them all the way back to like 1976. And we were nailing all these margins, all these past presidential votes from 1976 onward. So we knew we had something pretty golden. And we used that to sort of help us inform our polling and adjust some of our weighting. But really, it's just trying to define and find the most likely electorate. So the voter file benchmarking helps do that. Demographic modeling is obviously something very important rather than relying pure purely on your panel quotas, et cetera. Try to find out what the electorate is capable of and what it's capable of under certain stress tests. There's really a key insight into how we were able to nail some of these, some of these states. And again, you know, Michigan, Pennsylvania, or uh, excuse me, yeah, Michigan, Pennsylvania, Wisconsin being a trifecta, they tend to vote the same way. There's a lot, there's a good strong delta and relationship between them. There's also an upcoming and ongoing delta between North Carolina and Pennsylvania. So when you start trying to find all these little pieces, this mosaic of uh data points, it really helps you inform how you're what what type of electric you're building. So there was just a lot that goes into it. I could unpack a lot more, but really it's just trying to turn this into data science, you know, as opposed to just sticking out, you know, fire hosing a state and sucking in a bunch of data, a bunch of respondent level data, and just weighting it and slapping it out there. It's not probably the best way to do it, though sometimes you don't really have a choice because you may not have a lot of background information, like say the 25 election. Well, that's a whole new elector and a whole new set of circumstances. A lot of pollsters missed it, including us. I mean, we were we were off in the Virginia race, the governor's race, and we were off in the New Jersey race. I mean, it was just it was a very wild time to be polling because you had a government shutdown, snap benefits, uh, that that that elector was constantly moving and it was moving fast towards the end. So it's not surprising that most of the pollsters just kind of whiffed on Virginia and New Jersey.

SPEAKER_01 5:32

So you guys clearly have a groundwork that was able to give you so much success in 2024. I want to go back in time a bit. Obviously, a co-founder at Qantis. Tell us how the company came to be and how you all started working together.

SPEAKER_00 5:46

So it it really ties back to your very first question. They started off doing modeling, just doing models. We did this as pretty serious amateurs, uh, going back to you know, 2016 got a little better. 2018, we started getting a little better understanding a little more. And then 2020, we started really working in the data. We started having some pretty primitive models looking back on it. Uh, but we learned a lot as we were doing this, and we we all had day jobs. Our background uh, you know, was you know, different walks of life or, you know, cybersecurity or Intel and just some things like that. We pretty crafty guys that knew a little bit about data and and and how to do some uh association using some basic mathematical formulas such as Kravers V or K-means clustering, all those nice buzzwords we like to throw around. But they are very useful. 2022, we got into it pretty heavy, and then we missed 2022. I mean, we we had the Republicans win in the House, but we had them winning by a lot based on just some really strong historical indicators and certainty that inflation was up, gas was up, et cetera, et cetera. We we predicted a much stronger Republican outcome that just did not materialize, it sort of fizzled out there. So we went back and readjusted some things. So in 2024, election rolls around, we get heavy into it, and then we were pretty confident that we had, we learned, we put some a body of work together, a body of knowledge together. And I say we, it was me and a partner that was doing this on the side, and it really was just a Twitter account. But we we prepped and planned months ahead, uh, and then we started unveiling things around December and January of, well, December of 2023, but January of 2024. And then by early, you know, May, early summer, we started rolling out our models and we were hitting some things that we thought we were that we were seeing the most likely electric, seeing what this what the what the outcome was likely to be. I know you'll hear things about realignment. That's probably not the right word, even though we use it a lot. There's only been about five realignments in American history. Six, perhaps, if you count the Democrat or the Southern states breaking from Democrats to Republicans in 1968. Maybe that's the sixth realignment. But the last real true realignment was FDR in the 30s. I mean, we've been operating under that paradigm ever since to a certain degree. But what we really mean by realignment, I think, is what people are coalitions, right? Coalitions are like amoebas. They they they move from a cycle to cycle. They're trying to find security and they're trying to find a home. And that's why these things break apart from one presidential election to the next. Had this huge coalition that people were calling a realignment. It was completely gone by the time 2016 rolls around. So those are the things we we tried to find was what is the coalitions? How is this breaking through? But to get back to the basic point here is we were just steadily pumping out data on Twitter and we came up with a hypothesis. I said, man, I I think we can pull this better than what we're seeing because we we had this. I'm very confident in what we have. Why don't we just throw out some crowdfunding and let's just raise some money? We had a decent little Twitter following and it was uh pretty enthusiastic about our data. Uh so we threw it out there one day. And we, I think our first poll was in Pennsylvania. And we scraped together some money. I put a little money in it too. And we polled Pennsylvania. And what do you know? We had Trump up about 1.7 or 1.8 points, I believe it was. I got a little traction. 538 reached out. We did some vetting with those guys, we sent them some of our methodology, PDF reports, we are, what we're doing, why we're doing it, and they they approved it, got us on 538. And we kept doing this, going back to the well, just staying steady, staying steady, trying to stay up on the election. And then we got notice of some folks who were interested in partnering with us and perhaps even acquiring us. And that's that's where we ended up. Uh we were we were actually pretty much at 1099, but we were fully funded. And then towards the after the November election, we were fully acquired and were absorbed by a digital media company. And that's kind of where we get our funding, and that's how we were able to push these polls out and uh stay active and stay relevant.

SPEAKER_01 9:32

Traditional social media success story. I love that.

SPEAKER_00 9:35

It's it's it's it's the wildest thing.

SPEAKER_01 9:37

Talk to me about the crowdfunding. So it was just people that followed you that see more.

SPEAKER_00 9:41

I mean, we had a really we had a we had a really, like I said, a very passionate following because 2024 was off the off the chain in terms of enthusiasm and engagement. A lot of people really wanted to see change. And we were putting out data. I mean, I you could maybe call it a pacifier. Maybe people liked what they saw, so kind of confirmation bias, but it wasn't. We were we were dead on. We our models were public, our methodology was public. We had Trump winning 312 um electoral college votes, even though our polling showed that Harris was up in Michigan, which turned out not to be true, but our our models said Trump was going to win Michigan, and we didn't trust that. We went with the poll, we went with the people who lied to us, so to speak. Uh but yeah, that's how it happened. It was just a fun hobby, but it was a serious one. I mean, it was serious. It was definitely competing with my day job. I was in cyber and data and consulting and all that wonderful stuff. Um, and I'd much rather do this for a living, I can tell you that. So that that's how that started. And then I I walked away from everything uh probably by August and completely did this full time, and you know, here we are. It's such a cool story.

SPEAKER_01 10:46

2024 in the rear view. We're obviously heading into 2026. We are recording this on March 10th. So we just had our first real stress test for 2026 uh a week ago with the elections in Texas, Arkansas, North Carolina, and of course we have an election tonight, special election in Georgia 14th. Talk to me about some of those races. What are some of the early signs we're seeing heading into 2026? And of course, how has the Trump endorsement loomed over these races? And I know everyone he's endorsed one. How concerned are you as a candidate who isn't Trump endorsed if your election is coming up in the next couple months?

SPEAKER_00 11:23

So I have a a contrarian view on that, but let's unpack it a little bit. Talk about and talk about Texas, we'll get into the endorsement. Because it kind of segues into that because you have a the Paxson Corning situation. I think Texas is at the top of everyone's mind. So I'm okay focusing on that. Go ahead. So we had we had Paxton doing better. I think we had Paxton up like by five points, I think our last poll was. Another one's had him up around four, but I think the average was about four points. We had him doing up slightly better. And that's just that's what our polling showed. It was, you know, it was good polling. Could it be could I have made an argument and say Paxson was really more like 40, 41%? Probably. We might have had, we might have waited a little too much for four-time voters and maybe not enough of those one and two-time voters. Maybe that changed some things. We tried to go back and look at a few issues, but really it was just missing. I think it was just missing the Dallas area, the high population center turnouts in the urban areas that supported corning. Because if you remember, as soon as the polls closed and they started sending updates, Corning shoots out of the gate. And he was just just, I think he won Dallas County by 21 points or something. When I saw that, I knew it was likely to unchange that much. It'd be probably like a 1.1, 1.5 Corning victory because it was going to be too hard for Paxton to overcome that. He just didn't have the votes in other areas. Even if he won Harris County and things, uh, was it Fort Bend, Harris, down in around the Houston area and the South and West Texas, and even some of the East Texas counties that are somewhat populated, Holler, Longview, Nacadochis, et cetera, it probably wasn't going to be enough to overcome that. So Corny did well getting that base, getting that turnout in areas he needed it. He did well in the one thing I noticed, though, where there was really no coalition between any of the candidates. They were just winning spotty areas and then losing this. I'm like, well, if you won that, you probably should have won this, but it's that's not how it happened. It was a very mosaic map when you looked at the color coding. But yeah, so hats off to Corny. He spent$70 million. And I know that last two-week stretch might may have moved the needle in some areas. Maybe Paxson got defined in some areas and then just with a safe bet, Texas being kind of what it is, their electors like that. He outfort story is Cornyn outperformed all of the polling. I mean, there was only, I think, one poll that had Cornyn winning, and that was by 11 points. The rest of it showed Paxon up, you know, like I said, from three to five points. But he outperformed his polling. I think he outperformed it by about four points. And uh looks like Paxton kind of landed where he was supposed to land, and then and then West sort of fizzled out towards the end, and maybe Cornyn picked those votes up. I'm not sure. But that's how that happened in Texas. It was kind of frustrating as a poster to see that, but no knock on Corny. It is what it is. If he can win it, then then good for him. If Paxton can win it, good for him too. So how that works are the dynamics of the Trump endorsement. Listen, we've asked that question a bunch of different ways. If you ask a general question, say you ask the uh respondents, how how influential do you think the Trump endorsement is in this election? More than half will say it's very, it's either somewhat or very, right? You add them two together and you got like 55% of those people say, oh yeah, it's a it's a game changer, man. But then you ask them directly, right after that question, say, how important is the endorsement to your vote? It's it's not important at all. It it's a factor, but they if you ask them, give them another option saying, hey, I I'll decide the candidate I want to vote for on the merits in which I decide that, whatever, whatever your response is. They that always outweighs the Trump endorsement factor. It is certainly at play, but it's not the key driver. And we've done this in about four or five polls in a row now, and it's never, it never outdoes the individual level responding. You see that in Texas Senate race. Let Lowe, congressman from CE5, which is in northeast Louisiana around Monroe. She was endorsed by Trump. And you have Fleming, John Fleming, Dr. Fleming, who was a former congressman at CE4, which is on the west side of Louisiana. And he was also, he's also the state treasurer, and I think he worked in the Trump administration as well. And then you're running against Cassidy, obviously, who's the incumbent. We had Fleming up in our last poll, regardless of the Trump endorsement. And there was other polls, JMC Enterprises. He also showed Fleming up in that race. We just had Fleming up a lot in that race, actually. Head to head, he's crushing both both candidates. Um, and we'll go back and pull it again. Maybe we just got a hot sample for Fleming. I don't know. But I do think the Trump endorsement helped Letlow, because she's, I think she's a relatively unknown candidate, certainly nationally and probably somewhat statewide, but Trump's endorsement helped her probably get over that hump. And then, of course, the governor, Governor Landry uh Landry in Louisiana is also back in Letlow. So she's got some, she's got some institutional power that's helping her. But I don't know if it's the driving force in this elector, because when we pulled it, 60 something percent of those Republican primary voters and an independent leaning Republican voters said their number one objective is to get Cassidy out of the Senate. I mean, that's their motivating factor. It supersedes all other metrics in that sample. When you do clustering and you start doing association with, you know, Kravers V, what how important is this question to that question? And you start building these relationships, the driving factor, the number one sentiment in that election is anti-Cassidy. It's the anti-Cassidy vote in that election. Sort of a long way around your very simple question, but there's a lot of dynamics that's usually more candidate-driven as opposed to, you know, an endorsement for Trump. While important, it's not necessarily the number one motivating factor in how people are going to choose these, uh, especially in your primary elections, how they're going to choose their their representative or their senator, et cetera.

SPEAKER_01 16:48

It's interesting. I think a lot of consultants will be surprised to hear that. Some may actually even be delighted if they're that's what we're finding. I don't want to poop poo-poo it too much now.

SPEAKER_00 16:55

So it's not important, but it it is it's just not what we're picking up on. We're not seeing that is they're not going to outsource their vote. They're just not. That's the whole point in voting. People are a little independent-minded when it comes to that type of type of scenario.

SPEAKER_01 17:08

Trevor Burrus, Jr.: What does this mean for the runoff? It sounds like the president's gonna get involved in one way or another, either Cornyn or Paxton. You had an incredibly close primary. Have you guys done any forecasting for what it's gonna look like here in a couple months?

SPEAKER_00 17:21

No, we're gonna break that apart. We have until what? March, April, May? May before that runoff? Yeah. We we we thought about jumping in the field, but I think it's too soon right now. We're gonna wait. We make we just wait till April, and we will have a full we won't touch a poll again until we completely model Texas, find all of the strength areas for Cornyn versus Paxton. And then again, we're gonna do some machine learning that helps us find these patterns and we're we can use Vodafile data to get the turnout propensity. Meaning, is it gonna be comprised of 80% of four-time voters, or is it gonna be more like 65% of four-time voters? We'll figure this out using Vodafile. We'll go back several elections if we have to. We'll run this all through the hopper, shrink a little magic dust on it, and we'll find out what what we're looking at before we poll it again. But what does it mean exactly? We had a poll out, I believe yesterday, showing Paxton up over Cornyn, but it was a Democrat poll. Not that not that that necessarily matters. Maybe they want to see Paxton win, because they think there's a there's some supposedly there's some nugget of wisdom floating around that Paxton would be the easier candidate to face off in the general election. I think that's fool's gold. I don't think that's uh I don't even think it's relevant, to be honest with you. I think the voters are gonna support who they want to get into the Senate, and then Texas being what it is, which is a Republican stronghold, and was a it was a Republican fortress in 2020, or it's been a fortress for years, but 2020 2020 and 2018 seems to get people excited that our Democrats are trying to put the state, but the the data and the trends just simply don't show that. It is definitely a Republican-leaning state. Even if you had the circle, the semicircle around Harris Counties and Houston, let's say it turned out and it broke the Democrats by 56%, and you got Dallas with a favorable turnout, and they they they broke, they broke the Democrats by 55% or so-and-so. That does cut your margins extremely close, but you're still probably talking about a 280,000 vote padding left over for Republicans to win it. You get a Beto Award versus Ted Cruz election outcome. You get a close race, but I I don't think there's enough votes there uh to flip it. So I don't know if it's any different to having Paxton or Cornyn on as your general election Republican candidate. I think either one of them is going to win the general election. And it doesn't matter if Cornyn wins it by seven and Paxton wins it by five. You you still have a Republican senator. And if the Republican voters want that senator in in the office, then they'll get their senator. So I don't know. I have a weird way of looking at that type of. I understand the scheming and the strategy behind it all. I'm just not really sold on it yet. Maybe they uh these Democrat polsters wanted to push that out, or maybe that's the real deal. I I don't know. But it's not surprising me that Paxton wins in a head heads-up contest. I sort of regret not doing a heads-up contest when we polled it, but I didn't really want to want to pollute the sample. I just wanted to get a good read on a three-way race between West, Cornyn, and Paxton and leave it at that because I knew it was going to go to a runoff. And I said, we'll figure out the runoff when it gets here. But um I think Paxton's probably gonna be a strong candidate in this primary election. And then if Corning outperforms the polls again, then we'll we might have a little surprise.

SPEAKER_01 20:30

I think he was kind of a few weeks ago seen as progressive, but also somewhat moderate compared to Crockett. But man, as the stuff is coming out on this guy, he may actually be worse than Crockett.

SPEAKER_00 20:41

You know, I I don't know why that set-aside label only exists for Democrats, moderate. I I don't understand that. Only a Democrat can be moderate, it seems like. I really don't, but no, I there's a lot of APO research going to come out about that gentleman. He's not harmless. Uh the some of the stuff that he's I've seen, that's just it's gonna be ammo. And I I just don't think the Republicans should fear too much about running against that guy. I that's just my my opinion on it. But uh we'll see.

SPEAKER_01 21:21

I think the consultants in Michigan, North Carolina, and Georgia will thank you for that because we we've certainly spent enough in Texas at this point.

SPEAKER_00 21:29

Yeah, no kidding. No kidding.

SPEAKER_01 21:30

Let's talk a little bit more broadly about 20. What are some of the the good, the bad, the ugly for Republicans that you're seeing in the numbers?

SPEAKER_00 21:37

I always t I always tell folks, you know, we we poll we poll pretty often. Again, we try to stay relevant and we we will we want the data. We want to collect the data. We want we want to keep the data so we can use it for other things when we when it gets to crunch time. Having said that, though, I don't trust any of the polls right now. I don't even trust our own because it's just not a good time to poll, especially for Republican voters. It's just not a good We see it. We see the response rates. Why is that? They're just not engaged. It's I don't know why it's like that. I truly don't. It was like that in 2024 up until the polling got easy late September. And certainly by October, I started getting full healthy samples, just almost from the vine. And I I didn't get that during the summer and late summer. It was uh a lot of response bias from the part of Harris. And it really started making me question things. But then in October, everything just got really easy. We started getting really good balanced samples. We started getting, you know, what we thought would be white males, right? Republican men or just Republican men, period. You know, 18% of our sample was Republican men. Our last national sample, October 27th, I think was when we polled the national poll in 2024. 18% of our sample was Republican men. That's dead on. That's exactly what you want. So it helps make, it puts things a little more in perspective. Start getting your most likely electorate. So, but going back to what we're seeing here, the good, good, bad, and the ugly, yeah, I I think the Republic the Democrats are up, no doubt. They're up four, five, maybe six points because that's what they're supposed to be at. And I'll elaborate on that in a. We're also not getting a lot of high engagement Republican voters who are taking polls. And then when you ask the ones who are taking your poll, they're running about 11 points less than Democrats when it comes to uh certain to vote or are very enthusiastic to vote. They're always at a deficit of about 10 points. That will probably change in September. Now, will it be enough to overcome this Democratic lead? I I don't know. History says it won't. But the bad is it's just this is the midterm effect. This is what happens. It's a course correction. Uh there's a whipsaw, it changes every presidential cycle to the next, and it changes by about seven points. So what do I mean by that? The election, the electorate swings about seven points in the opposite direction of the preceding presidential election. So if Trump won the margin by 1.5 points and we add about a six or seven point swing to that, you would expect the Democrats to have a five-point lead in his generic congressional ballot, which is about where they're sitting. They're sitting at about four or five points. Well, as a matter of fact, I did some work on it last night and updated my Twitter with some of this stuff. Took all of polling for 2026, went through some smoothing lowest, et cetera, and find uh and found what you know what we're probably looking at. And it's about a 3.8 to a 4 point, you know, whatever, 8 uh spread. And that's about where it's falling RCP and DDHQ average. And we're finding that too, uh forced leaners. So we asked from straight up just for the firm voters. I think we had it 45.6 to 40.8. And then when you ask the leaners, it gets up to like 47.6 to 42. So it's about a six-point lead with leaners. This is where our last uh three polls in a row is showed a plus six uh Democrat lead. And it hasn't changed this entire year since January. The bad though, if they don't, if they don't turn out, and this electorate sort of morphs into highly engaged Democrat, more educated voters, bourbon kind of goes away from them. Independents swing wildly. You expect the the independents to vote for the party out of power by plus 11 to plus 18. That's enough to get you over the 50-point mark, maybe even 51, 52 percent, depending on how your suburbans vote, how your other women vote. If your electorate is more, say, 45% or 44% educated to 56, 58 percent non-educated, non-college, uh, then you you you got yourself a recipe for a good a good four and a half, five point uh victory for for the Democrat voters. Or the Democrat Party, I should say.

SPEAKER_01 25:24

Well, how do Republicans reverse the trends? I mean, what are some of the issues that you're seeing or the big things that voters are talking about? I know we keep hearing the A word thrown around, affordability, right? Yeah. So uh what are some of the other things?

SPEAKER_00 25:37

Well, I mean, you you definitely I mean it is, it's cost of living and economic stability. So they're overwhelmingly focused on that. When when you stack the when you stack the issues, you give them ten issues, you give them five issues, you do research on previous polling and you find, hey, what do these issues look like? You go five go find five or six polls that are somewhat recent, see what that questions they are the issues they raise, see what the results are, try to try to get an average, see what it looks like, and then put cheap put together a good survey instrument to measure that same sentiment. And it always comes out to cost of living and economic stability. And so the argument with that is well, look at the GDP. It was up to, you know, whatever it was, 4.8 and 5.3 and and whatever. Now it's dropped down this, I think Q1's probably gonna be a little less than I don't even know if Q4 uh when it's released. I can't recall exactly, but it it is definitely showing improvement. GDP is accelerating. Looks like wages are up 0.4 month over month. So we've had rage wage increases as well. And that also helps with real disposable income, et cetera. These are all good predicted variables in a in anyone's economic in an economic forecasting model. ISM index shows some really good strength. It was pretty flat during Obama's or Biden's years, like I think towards the end of Q4 when Harris was running, it was around 47.4 or 48 tops. That's a contraction area. Anything over 50% is growth. And Trump's been getting that, it seems like what, two or three, three months in a row now? These are all good, strong indicators. But here is the real problem to all of this is yeah, inflation's down. I don't understand why people are upset. This doesn't make sense. This is counterintuitive. Well, inflation is down and wages are up, and GDP is moving in the positive direction. ISM index is showing some strength and some resiliency. But the prices from pre-COVID to post-COVID are still high. They're higher than. They're higher than 2019. They're higher than 2018. So you compounded a decade's worth of inflation in a matter of three and a half years. It's not easy to get over. The lagging, lagging part of this, the memory of having prices lower is still very much on the minds of voters, especially when it comes to household items or or just household concern about how far your dollar stretches. They're they're mad. They're mad at this. Not necessarily mad at Trump, I don't think, but they're also mad at not necessarily mad at Biden personally. It's just one of those dynamics to where things aren't right. You said you were going to fix it, and the prices are still higher. It's a it's a weird, it's a weird thing. It takes about 16 months for economic recovery to actually show up in sentiment analysis because the memory of higher prices stays in people's minds. That's it's documented. There's a lot of research about that. I don't know if it should stretch out quite to 16 months, but that seems to where it falls, tends to be where it falls. So getting back to that though, nailing this messaging on the cost of living and economic stability, I don't know if you can. I don't know until there's some real tangible changes and prices go down, which they're not going to go down. Corporations, supply chains, they've already baked those prices into their operational models. It's not going back. It's just going to be a new norm. And hopefully a productivity wage increases sort of compensates for that. But I don't know how long that would take. It's it's a big, it's a big deal to lift an economy and certainly turn it around. But we'll see. Yeah. And then really it's just the incumbency factor of getting in the White House and having a trifecta. Voters tend to punish that. I mean, it's it's happened in 18 of the last 20 midterm cycles. So it's a 90 90% chance that it'll happen this cycle. And there's really not a lot you can do to buck those types of trends. It's baked in, it's behavioral, and it's measurable.

SPEAKER_01 29:09

Voters saying about the foreign policy agenda of the administration seems like, you know, feels like it's all over the place these days. I'm curious what people are saying.

SPEAKER_00 29:17

If you asked a blanket foreign policy question, Trump's underwater, maybe 10, maybe 10 points, something like that. I have to go back and look at the aggregates. We've had it, we've pulled it several times. And if you have specific policy issues like Ukraine-Russia, Iran, Gaza, et cetera, you you start you get a little more mixed results because people may focus more, they may have a little slightly stronger opinion on a specific foreign policy issue. And you might get some different readings on that. But I think what the well is poisoned in that regard. I just don't think voters trust any president or any Congress or any or any Senate when it comes to foreign policy and foreign policy entanglements. And Democrat voters were extremely supportive of Ukraine, even up to the point of war, which was just mind-blowing to me. But yeah, when during the by during the Biden years, they were they were for it to the point of war. But then when the Republicans in office, they automatically recoil against that. And then the Republicans sort of switched up, and now they're way more supportive of foreign policy than they were under Biden's administration. So again, this is behavioral, this was bias from from voters. It's the partisan bias of the electorate showing through here. But we polled the Iranian and uh U.S. involvement, epic fear. We polled that last week. I think we released Tuesday or Wednesday. We had it um 46 to 50 or 44 to 50. So 44 approved or 46 approved to 50 disapproved. I think it was 46-50. So we had it close, closer. But then other polls showed it similar. They they were showing in the 40s and the 50s we had we had a decent little chunk, then you just didn't know enough. They said didn't know enough, I'm you know, not sure. And then moved it to black, you know, maybe 6% of electors there. So again, the issue, it's issue driven. So if you ask the blanket foreign policy, you may you may get a little maybe not so quite clear of a reading, but if you put a specific issue to the test, you tend to get a little more of a clear, a clear reading about foreign policy. So if you take all that together, you're still gonna probably have negative numbers. But I don't know if it's as bad as some of the polls have shown. I just think America has war fatigue. I think they have Middle East fatigue. We're gonna get out of the Middle East. We're never gonna stop having influence in the Middle East. As you can see from the oil markets and the oil price, the barrels, the price of oil skyrocketed over the weekend last week. Then of course they come back down yesterday, Monday, which was kind of bizarre to see that happen. It's the largest reduction in price in the, I think, in history. There's no way to not be involved in the Middle East. It's never gonna we're never gonna lose our presence there. It's all these proxy wars and and things that sort of complicate stuff. China supporting Iran in a lot of ways to weaponry and intelligence, so is so is Russia. They do that to occupy the U.S. instead of directly confronting the U.S. on certain issues. They prop Iran up and make Iran a little stronger than probably what it would be on its own, which occupies the U.S. because of the strategic interest that the U.S. has in the Middle East. And then Iran by proxy will support Hezbollah to interfere with Israel and and then vis-a-vis the U.S. and then Gaza and Hamas and Gaza, et cetera, et cetera. So it's this big chessboard. And if Trump doing this, this uh maybe a master stroke or maybe it's hubris, I don't know. But if he can topple Iran, which I think that's what he wants to do, we'll see how bad he wants to do it, you sort of remove that piece from China. You remove a plea that China and Russia can use to leverage against U.S. interests in the Middle East, and then you deny China a lot of its oil. We've already denied it, Venezuela oil. And if you deny China, the Iranian oil, it sort of sort of isolates China in a bit and puts them in a more of a, I would say, a strategic pinch because they have big designs and big ideas and big ambitions in their region as well. So it's just a big piece, a big puzzle that a lot of people may not quite follow. But when you start putting it all together, it becomes a game, the great game, I guess you could call it. And there's there's a lot of statesmanship and one-ups. But yeah, it's it's a wild thing to pull, but of course you can't pull all of this. You can't write a freaking essay and ask them to now that you know all of this, what do you think about that? It doesn't work that way. I might have got us off on a tangent there. No, it's great.

SPEAKER_01 33:18

I mean, it's gonna play an important role in in the midterms, whether people like it or not. So really good insight there. Speaking of uh making you predict the future, let's talk about the polling industry over the next 10 years. Last question here. Where do you see the industry and do you think AI will have a big role in shaping it?

SPEAKER_00 33:35

Yeah, I mean, I think I think AI is gonna have a big role in shaping everything, everything we do, from finance to industrial capacity to medical research to I don't know where it's going. I mean, you get some of these guys that are on the bleeding edge of this technology, these these drivers, these innovators that are that have been, you know, proselytizing and stuff, they're saying it's gonna change dramatically in the next two years. There's gonna be a lot of people out of work. Maybe, maybe we get a universal income from that, because if GDP goes to 10% or 6% and these people are out of work, what are you gonna do with a bunch of unemployed people? You know, you can't make them AI agents, so you gotta figure something out. So yeah, they just want to impact everything. It's all gonna be downstream from this innovation that we are that we're kind of unlocking here. I mean, we're seeing we're seeing the AI play out in this war uh to such a degree that uh it's never happened before. And one day we'll read a book about it or a paper or a research paper about it. How much how much AI had an effect on planning this war? Because something, something pretty unique is happening over there in terms of the speed and capacity in which they're delivering this these payloads and uh knowing where things are at. Not to say that the U.S. intelligence infrastructure could do that on its own, but something's going on with AI and this war, and they did it with Venezuela as well, and I really can't wait to for that to come out. But I think polling is going to answer your question specific specifically as it relates to polling, it's gonna look very different in the in the next years. I don't think it'll take 10 years. I think by 2028 we'll probably see something, a bleeding edge when it comes to there'll there will be some firm that comes up with some polling apparatus, some AI-focused polling, and we'll see what the industry, if it accepts it or not. I think one day it probably will. There's no other way around it. But response rates to traditional surveys, they're they're declining every year. So the industry is going to have to evolve beyond just calling people, sending online panels, texting, all the wonderful things that we try to use to get those, to reach those most likely voters. But I think what we'll see is probably a hybrid approach where we blend survey research, behavioral data, voter files, and certainly the machine learning aspect of AI, how it can take disparate amounts of data and draw correlations and find trends that just a human can't find in spreadsheets. So I think it'll it'll play a big role in how we analyze patterns fast with tech bias and samples and simulate different turnout scenarios, et cetera. Hard to do with spreadsheets. You can do some basic arcs, a basic Python, probably get pretty close, but you're never gonna be as quick as machine level and machine speed. So you're still gonna need good data, you're still gonna need good methodology and good research practices. But I think analyzing information, fix, fixing flawed data sets, I believe AI is probably gonna play a huge role in that. Will it replace posters? I don't think it'll replace them, just like it won't replace the human in the loop when it comes to anything, nuclear research, for example. I don't think they're gonna outsource AI to some nuclear facility just because it can process metrics and machine level speed versus a human. There's always gonna be a human in the loop. There has to be. And it'll be more like using AI as a tool, serving to build models and understand voters clearly. So that I think that's where AI is going to go with with polling.

SPEAKER_01 36:42

Pull in the chest, especially for pollsters, like you said, from the the data learning perspective. So it'll be fun to see in 2028. It's crazy.

SPEAKER_00 36:50

I think so. Yeah, I think we'll see something. There's already some research flirting with it right now. Uh there was a research I think out of the Netherlands. We were actually in that. That's how I found it. It was c I think it's called Possum, and they basically used all the polling and all the data from 2024, and they tried to find polling errors, ISO data using open source available data from the polling industry. And they were trying to find benchmarks, meaning which pollster could you consider a benchmark of 2024, meaning could you build your model around these list of polls, et cetera? And it was a pretty good uh research paper, but again, that's just another level of AI and how it can be used. I don't know if humans are able to I don't know if they could do that. I don't know if a staff of people, I don't know if a staff of researchers and research assistants could do that. So anyway, very interesting. I I think it's got some I think it's got some use, but we'll see how far it goes in the polling industry. I want to move into our rapid fire up here. So best state to poll the worst state to poll. That's a good one. Best state would be, man, I don't know, gun to head. I got two. So Florida, probably because it's large and diverse, but it's surprisingly stable. The turnout patterns are fairly predictable. I I think it's an easy poll, or at least an easy state to model. I mean, we modeled it back in 2024 without polling it and nailed, nailed it. Like we had Trump at 56 and Harris and Biden, we we swapped out Harris and Biden, but we still kept the same model like at 42. So I think we had about plus 14, that's exactly where it fell. Just using voter file data and turning out models. But I think Pennsylvania as well, because it has a large white voting block. That's white college and white non-college, but it also has that strong urban pull in Philadelphia. So it's diverse in those regards, but also surprisingly stable as well. Interesting. It's the worst state. I just polled it. Maine, small population, quirky regional differences, a lot of independent voters. Publicans are habitually underpolled, especially non-college, conservative men, et cetera. It's hard to get those guys to answer your polls. It's notoriously difficult to model. It's just, it's hard. You look at 2020, was losing by eight points in the polls, and I think she ended up winning by nine. I mean, that's a egregious polling error. I'm seeing the same thing now in that state. Yeah, I polled it yet. Yes, I just released yesterday and I had uh Platinum up over her like 49 to 42. She was beating uh Mills, but when Platner, when I introduced Platinum to the sample, and the left-leaning independence, and it's hard to trust. You just don't think you're you just don't think you have enough of the right ideology in the sample, and it won't go off the rails. Const is probably a close second, but it's also tricky for for similar reasons.

SPEAKER_01 39:23

Interesting. Interesting, cool.

SPEAKER_00 39:24

What's something that campaigns should stop pulling? Stop caring about. Man. Probably those tiny message testing questions that ask voters like to rate two different slogans. I'm just not convinced that voters are motivated and thinking slogans, but if I was in the inner workings of a campaign, I guess I could see how it matters. Maybe for a tiny sliver of voters, maybe say you're you're you got a two-week sprint and the campaign is closed and you have five million dollars left and then you're just gonna blast ads on these slogans. Maybe I could see how if it moved the needle a little bit, then it would then it's probably worth it. It's probably worth every bit of that. But I I just think overall, most of your, especially in a general election, I mean 90% of your electorate is pretty much accounted for between two candidates one way or the other. Hardly more than that as the time goes by, and then you're all fighting over that three or five percent that's left over. So I don't know. Maybe maybe those guys are smarter than I am, because I mean I I give them credit. I I don't I don't work on the internet workers of a campaign. So maybe there's some use to that. Maybe they have some metrics to say I'm wrong. But anyway, but as a polster, I I don't like I don't particularly like uh I don't particularly like seeing that. Good to know.

SPEAKER_01 40:27

Last question. You're the first Louisiana native on the show. So when you head back home, where and what is the best restaurant in the state?

SPEAKER_00 40:35

But the case is a good thing. Depends on where I'm depends on where I go. Uh if I'm going to New Orleans, I don't know, I'm less like a tourist here because I'm not from New Orleans. I've been to Commander's Palace in New Orleans a couple of times. I've been to the oysters right there on the right on the right on the bay, like they go and get the oysters out and bring them to you, it feels like. But uh it's it's a it's classic Louisiana cuisine and it's probably one of the best dining institutions, I think, in the country. But if you're up, if you're up north around Streetport Boser, Fat Calf uh Brazierie in Streetport is it's some good food. Honestly, anywhere in between your mom and pop's, it's all cuisine, good food. I mean, even gas station food in in Louisiana. Oh, yeah. Because they it's not like grabbing something out of a cooler. There's people that cook that stuff and bring it to the gas station, and then they get their money. But you're you're talking moodan balls and meat pies and you know, there it's different in Louisiana. There's a different couple of different regions that go on and how they what they're known for in their terms of their food. You can get the straight up traditional southern food, which is where I'm from in the state. I'm more of the traditional southern side, but then you get the Cajun and Creole, but then sometimes that line moves from that demarcation, and it's just a fusion of them all. So, yeah. Short answer though, three port, fat calf, New Orleans, you gotta go to Commander's Palace, but you wouldn't do wrong by finding a mom and pop shop somewhere.

SPEAKER_01 41:50

Listening who's working on the center rates, you are now hooked up. Thank you for doing this. Tell the people there they can follow you, follow what Quantus is doing.

SPEAKER_00 41:58

Sure. You can find us on Twitter at Quantus Insights. Our website is quantusinsights.org. Um that's where we usually release our polling. Well, we always release a polling there, excuse me, in some of our research papers. But then we have a substack, Quantus Insights, quantus. That's where we do a little more of our free freelance, free willing writing. We kind of, you know, we might do some drop some research papers there. We might, we might write something quick that happened about, yada yada. That's just more of like a almost an open diary that we we keep for our for election analysis. But yeah, follow us on Twitter. We uh like engaging over there, we like sharing information. We also have an interactive polling site, which you'll find on our quantasinsights.org. We don't we got away from static spreadsheets. We're trying to make it a little more interactive, keep the quantus brand alive, or you can go in and see the the polling and and then um interact with the data itself.

SPEAKER_01 42:44

Thank you for listening to the Digital Politics Podcast. If you like what you heard, please share with your friends and be sure to like, follow, subscribe to the show. Be back soon for episode six.