Trump vs. The Blue Wave: A Midterm Extravaganza
Still Counting
In the debut episode of 'Still Counting,' hosts Nate Silver, Claire Malone, and Galen Druk dive into the 2024 midterm elections, analyzing competitive Senate
Key takeaways
- Nate Silver's 'Flipper' model shows Democrats with an 83% chance of winning the House, while the Senate forecast remains tightly split at 50-50.
- The shift in polling from registered to likely voters has narrowed the Democratic advantage in the Senate due to reduced enthusiasm gaps.
Main topics
- 2024 midterm election forecasts
- Trump's ongoing influence in the GOP
Notable quotes
"We're going to use [post-Labor Day] as an opportunity to dive in to the midterm." – Galen Druk
Conclusion
The debut episode sets a tone of rigorous analysis blended with cultural insight, highlighting how data and
Transcript preview
Speaker 3 (0:00) I think I'm just like a little bit nervous. That's why I know. I know. It feels weird. It Speaker 1 (0:03) feels weird. Speaker 3 (0:08) Welcome to Still Counting. I'm Galen Druk. I'm Speaker 2 (0:11) Claire Malone. And I'm Nate Silver. And we're talking too fast because it's the first episode and we're a little nervous. Speaker 1 (0:16) I also read my name. I was like, I'm Claire Malone. Speaker 3 (0:21) And this is our first official episode. Yes, we did a freed warmer. Yes, we talked to Pod Save America. But this is it, folks. From now on, on Wednesday evening, you're going to get a Still Counting podcast episode. in your feed. And so for you know, first, I'll say this is what we're going to do on today's podcast, we're going to do a midterms extravaganza, we're even going to kind of play a game. We're also going to talk about sort of the past year plus in Republican voter sentiment. It's been a pretty topsy turvy year for the Trump coalition. We're going to talk about the convention and all that does the anniversary of Charlie Kirk's death. Before we get into everything that we have to cover today, though, we should say, welcome back, everybody. Yeah, to our Past listeners, spread the word. Welcome to the new Speaker 1 (1:07) people. Hopefully, if you have no idea who we are, really, welcome to Still Counting. Speaker 3 (1:12) Well, actually, Claire, who are we? Speaker 1 (1:14) Who are we? I'm Claire Malone, as the script says. We all worked together at a data site called FiveThirtyEight, and I was hired by Nate to be like... the normal magazine reporter. To Speaker 3 (1:28) be normal. Speaker 1 (1:29) I Speaker 3 (1:30) was hired by Need to be Normal. I was hired by Need Speaker 1 (1:31) to be Normal to be like a magazine reporter person. And then I'm sort of a traditional reporter. I am a writer at The New Yorker. I cover media and politics there. And I am currently on book leave for the year, actually going back in like six months. And I'm writing a book about the Maha movement. That's who I am. Speaker 2 (1:49) Nate, who are you? I'm Claire Malone. Yeah, so my trajectory into journalism, if that's what you want to call it, is highly non-traditional. Before I ever got into journalism, I was an online poker player. I was building baseball models. Basically, 2008 comes along. I'm living in Chicago. This campus hipster, as I thought of him, named Barack Obama decides to run for president. It was a very exciting time in politics and also a time when there had been in baseball and in sports a lot of talk about analytics. This was a kind of moneyball revolution, so to speak. I was getting a little burned out with the baseball stuff. The poker was not going super well, to be honest. And so I started trying to build a forecasting model, first for the Democratic primary in 2008 and then general election. And it just kind of improbably took off. But yeah, you know, I... try to look at the world. I sometimes do reporting, but I'm looking at this through a lens. He's Speaker 3 (2:49) written two Speaker 2 (2:49) books. Speaker 3 (2:49) I've Speaker 2 (2:50) written two. For the books, I do actually do Speaker 3 (2:52) real reporting. Speaker 2 (2:52) For the Speaker 3 (2:53) books, I do. Which, for the most recent book, involved hanging out in Las Vegas playing poker for a year. Speaker 2 (2:57) I spent a lot of time in Vegas. It was a nice excuse. Silicon Valley, Miami. But I don't do a lot of insider gossip with, oh, what's the scoop about the internal polls in New Hampshire? I think that is actually a waste of time. You're being spun a lot of time. I think the goal of a model, we talked about this on the Pod Save trailer, it's to force yourself to be more rigorous in analyzing the news. And explaining how you view the world. Speaker 1 (3:26) And now, and Nate had a, he started Five Thread Date, which was at the New York Times, then it moved to ESPN, then they closed it, yada, yada, yada. Now he has a wonderfully- Silver Speaker 2 (3:35) bulletin, which is at natesilver.net. But like, look, we got- Flipper is our midterms model. We got an NFL model. Okay, Speaker 3 (3:43) I will do a briefer. Who am I? I was hired by. Nate at 538. Not Speaker 2 (3:48) in the less weirdness initiative. I'll tell you that much. I Speaker 3 (3:51) was hired to be a podcast twink at 538. We were Speaker 1 (3:55) hired at about the same time. Yes, Speaker 3 (3:57) we were. You came Speaker 1 (3:58) from Wisconsin. I Speaker 3 (3:59) came from straight from Wisconsin Public Radio, which at the time I could not have imagined how good of an education Wisconsin would have given me for what was to come in American politics. My Speaker 2 (4:09) earring just fell off. Things Speaker 3 (4:13) are going incredibly well. Anyway, I stuck around FiveThirtyEight for about a decade where I did basically a combination of field reporting analysis, looking at the data, did things like the gerrymandering project. And once FiveThirtyEight shut down, I started GD Politics, which is a podcast that I host on Mondays and Thursdays. talking to a bunch of people who are journalists, authors, sometimes politicians, pollsters who look at data and try to make sense of it all. And I'm going to try to take some of that knowledge that I get from all the people I talk to and bring it to folks here as well. But anyway, enough of that, enough throat clearing. Let's dive into what's on hand today. But first, let me just say, dear listeners, and for our YouTube audience, I'm looking straight to camera for this one. Previous listeners, new listeners alike, spread the word. Subscribe on YouTube, wherever you listen to podcasts. Also rate and review us. Like, I don't know exactly how the algorithm works. That's Nate's job. But apparently on Apple Podcasts, the more you rate and review us, the higher up we go and the more people see us and then we'll maybe get introduced to the podcast. Basically, we're out here trying to rebuild an audience. And it's going to it's like a political campaign. It's going to take all of us. Speaker 1 (5:29) Yes. Speaker 3 (5:30) See, Claire and I would never do that, right? That's why you're a good hire. We're like, do our dirty Speaker 1 (5:34) work, Galen. Speaker 3 (5:36) The podcast twunk now. Okay. We are officially post-Labor Day. I Speaker 1 (5:42) almost wore a white suit coat this morning, which, first of all, wouldn't have looked good. Claire, Speaker 3 (5:46) you're a Condé Nast employee. You must not be wearing white after Labor Day. It's an old school rule. So, yes. We are post-Labor Day, which is officially when we stop wearing white, when we go back to school. It is also historically when people begin to tune into the election cycle in our highly polarized and partisan and engaged environment. I think people pay attention a little bit sooner than that. But nonetheless, we are going to use that as an opportunity to dive in to the midterm. So right now, Nate. Your model, named Flipper, suggests that Democrats have an 83 % chance of winning the House, and the Senate forecast is split exactly 50-50. I mean, did you design the forecast that way to build suspense at the start of election season? I'm kidding. So when you first launched, the Senate favored Democrats. Slightly. Slightly, not overwhelmingly, but I think it peaked at about 58 % less than a month ago in terms of Democrats' likelihood of taking the Senate. What has happened between then and now? So a Speaker 2 (6:44) couple of things have happened. Number one, people get really sensitive about like 50 versus 55 versus 57. It's not a big deal. It's like in an NBA game. It's like one point being scored or two points being scored. You take it from Knicks ahead by a point to the Raptors ahead by a point or whatever else. So when you're near 50-50, small moves affect the probabilities a fair bit. That's getting a little technical. There are a couple of things that have happened. One is that we've started to get more polls among likely voters and not just registered voters. And that matters because turnout is relatively low in midterms, higher than it used to be. But as our model has been adjusting to those. Earlier in the cycle, it felt like there was a bigger, didn't feel like, I guess, I don't know if it has feelings. I guess we use AI, right? I Speaker 3 (7:31) mean, Speaker 2 (7:31) maybe. I used to be an anti-anthropomorphicizer and now, now it's just, you say, oh, the model thinks. Anyway, you know, as we've gotten more generic ballot polls among likely voters, they don't show quite as much of an enthusiasm gap. I mean, I'm being too technical. Speaker 1 (7:50) Now I'm just, by the way, to interject, now I'm just thinking about whether or not you think the model could ever kill you. Per discussions on Twitter Speaker 2 (7:57) about AI. I switched from Claude to ChetchyPT in the middle of Speaker 3 (8:02) Flipper. Speaker 2 (8:02) Because it was trying to kill Speaker 1 (8:03) you? Speaker 2 (8:04) The Claude Opuses were kind of weird, right? They would like, they'd be quite stubborn and lazy. What is a Claude Speaker 1 (8:09) Opus? Speaker 2 (8:09) I don't use it. Speaker 3 (8:10) We have the full range of AI use here, which is like, Nate has at least 12 AI employees. I use AI regularly, but like, don't make it code for me. You don't Speaker 2 (8:21) use Speaker 1 (8:21) AI. I Speaker 2 (8:21) do Speaker 1 (8:21) not use AI. I don't know what an Opus Speaker 2 (8:23) is. Here's the summary version, right? As we get closer to the election, I think you see like a little bit of a rebound in Republican enthusiasm. It's having a marginal effect on the numbers and some of these kind of less polled races in the generic ballot. Look, Trump's approval rating is 38 % or so. His disapproval rating is 58, 59%. He's not very popular, but Democrats aren't that popular either. So I think you might be seeing just the slightest bit of. of reversion to the mean. But there's lots of layers in Flipper. There are expert forecasts. You know, is New Hampshire now a state that's in play? How realistically do we take Tallarico's, you know, the model says Tallarico is a slight favorite in Texas. Do we buy that or not? We're Speaker 1 (9:06) going Speaker 2 (9:06) to get into actually Speaker 3 (9:07) all of these states. And we're even going to get into Trump's popularity and his coalition and whatnot. But first we're going to play a little game. I mean, not first. We're going to play a little game in order to talk about all of these states. But it's Speaker 1 (9:17) educational. But Speaker 3 (9:18) it's educational. Worry not. A little game, which is we're going to draw the most competitive Senate races out of a hat. And that's most competitive according to the polling averages. And that means that we don't know the order that we're going to talk about them Speaker 1 (9:36) in. It's the state of the race in Gamified. Yes, Speaker 3 (9:40) exactly. And I will say, I have a GB Politics hat here, which is only because we don't have... You're a shameless shill. Because I'm a shameless shill. And also because we don't have still counting merch yet. But I hear it's coming. I hope it's coming. I think it's coming. I think it's coming. You know, we got to, Nate, you got to stop advertising. I mean, we all love the Knicks. We live in New York City, but you got to stop advertising the Knicks for free on this podcast. We're Speaker 2 (9:59) getting silver bulletin hats. By the way, I'm happy to advertise the Knicks, you know, as long as Dolan doesn't kick me out of. Okay. Speaker 3 (10:04) Claire, you are first. Speaker 1 (10:06) Okay. So Speaker 2 (10:06) pick a race out of the hat. Wait, by Speaker 1 (10:09) the way, last night in preparation for talking about all of the Senate races, I think I watched. ads for every single Senate candidate or most of them just to kind of get a flavor. I was like, I want to see like, did it Speaker 3 (10:20) make you feel more optimistic or more pessimistic about American politics? Well, I Speaker 1 (10:24) have to say it was quite pessimistic. Like that was after watching, like, you know, kind of close to like 40 minutes of just like, you know, don't Speaker 3 (10:32) don't don't like Speaker 1 (10:33) it's, it's, um, Well, none of these people have ever been to Washington, D.C. It was it was basically all about like how much beef costs and isn't it terrible and insurance companies. And that was both parties. It was like, yeah, it was quite striking how public it was. Anyhow. OK, I picked North Carolina, which is a bad pick. It's definitely a bad pick. It's a bad pick because we're pretty sure that the Democrat Roy Cooper is going to win in his race against Republican Michael Watley. Roy Cooper. And that win probability, according to Flipper, trademark Speaker 2 (11:12) Nate Silver. It's pronounced with a dolphin. It's like Flipper. Speaker 1 (11:17) By the way, how did you name it? Speaker 3 (11:21) We've already had a problem with. furries and the FiveThirtyEight Politics Podcast brand. We really don't need to give them any fodder. How Speaker 1 (11:29) did you name it, Speaker 3 (11:29) Flipper? Speaker 2 (11:30) R's are good because R is for regression. All right, I don't even want to know. Speaker 1 (11:34) Okay, but Flipper gives the Democratic win probability at 86.9 percentage points. Let's Speaker 2 (11:42) put it like this. It's not a sure thing that... Cooper wins in North Carolina. But Speaker 1 (11:47) it Speaker 2 (11:47) would Speaker 1 (11:47) be embarrassing if he lost. But if he doesn't, Speaker 2 (11:48) Democrats are fucked anyway. That means the polls are probably way off again. They need four seats. By the way, Galen, you're saying, well, 50-50, is that just a cop-out? It's kind of a coincidence because we think it's going to be, and everyone thinks, it's going to be a very Democratic year. But they have to win four seats, two of which are in quite red states. North Carolina is not a... quite red state. It's a slightly red state. Like we have in your handwriting is actually better than I thought, Gailen. This is Speaker 1 (12:15) Saul. Speaker 2 (12:16) This is Saul. This is producer Saul. Speaker 1 (12:18) We're introducing producer Saul who has in Nate's reading pretty good handwriting. And I would agree with that. Speaker 2 (12:25) That's the only job that might survive the AI apocalypse is like good handwriting. It's rare. What you were going to say is that Trump won North Speaker 3 (12:32) Carolina by three points in 2024, which is a slightly red state. And Harris Speaker 2 (12:35) only, he only won the whole. thing by one and a half point, right? So North Carolina actually moved toward the tipping point. It was one of the few positives out of Kamala Harris's night in November 2024 was that North Carolina actually became more purple. And Cooper's a strong candidate. He's the former governor. If he weren't to win, then it means the polls were way off again. Right. Speaker 3 (13:00) Yeah. I mean, in terms of overall trends to look at, I think an interesting one here is that we've all been talking about the anti-establishment sentiment amongst the electorate this go round. Roy Cooper is like a two term governor. He has a good Speaker 1 (13:14) candidate, like well known in the state. But Speaker 3 (13:17) he is I mean, he's not really the establishment because he's running against the Democratic Party in part. And he's like, you know, I don't belong to anybody. I'm my own man. He's very much been a feature of state politics and not national politics. And so I think. The benefit of being a governor. People Speaker 1 (13:31) would suggest Speaker 3 (13:31) that he is a good candidate to run in North Carolina, but he's not running, you know, the Republican Michael Watley is not really a populist outsider who can say the establishment has sort of taken the country in the wrong direction, whatever. He's like the former chair of the Republican Party in North Carolina. So I think whereas maybe being so entrenched in politics could have been a vulnerability, the contours of this race are not going to let us see that. Great. So Speaker 1 (13:56) takeaway is panic at the disco for Democrats if. Speaker 2 (13:58) Roy Cooper loses. Speaker 1 (14:04) Still Counting is brought to you by Sundays. We are what we eat, and so are our dogs. A long, healthy act of life starts with clean, real food nutrition every day. Sundays for Dogs is different. It's human grade, real food you'd feel good eating yourself. Sunday's recipes contain 80 % plus all natural meat. That's about twice the meat content of some leading frozen brands. Plus ingredients like kale, ginger, and blueberries. No fillers, no chemicals, just simple, complete nutrition. Doesn't even look or smell like dog food. Looks like high quality jerky. Do you eat jerky? Speaker 3 (14:45) they're really selling us on like actually eating our dog food here, which I mean, if that's a measure of, you know, the quality, I think. Speaker 1 (14:53) Should we, should we say here that none of us own dogs, but Speaker 2 (14:58) we know dogs. We've met dogs. We Speaker 1 (14:59) have met dogs. We have met dogs. I Speaker 2 (15:02) have a stuffed animal or two. Speaker 3 (15:06) Wait, Nate, do the, do the stats live for us. According to my search right now, 45 % of American households have dogs. What's the likelihood that you would get three separate households together and none of us would have dogs? Speaker 2 (15:22) One in... Six or one in seven, roughly. But that assumes things are uncorrelated. I mean, I would guess the dog ownership rates lower in New York. Speaker 1 (15:32) Yeah. Well, can I can I tell you something? This Sunday's for Dogs is that founded and formulated, created by Dr. Dr. Tori Waxman, who I'm betting owns a dog to meet her high standards as both a veterinarian and a dog parent. Yes, she is a dog parent. Do Speaker 3 (15:47) you guys want to know the best part? Yes. It's effortless. Speaker 1 (15:50) No fridge, no freezer, no prep, no mess. Just scoop and serve. Over 100,000 dogs are now eating sundaes. And many dog parents notice the same things. Shinier coats, better digestion, more energy, and finally getting excited at mealtime. Even the pickiest eaters. Speaker 3 (16:08) And you know those dogs can be, some of them are picky eaters. I used to have a roommate who had a dog. A dog you've Speaker 1 (16:16) met? Speaker 3 (16:16) A dog I've met. Shout out Memphis. A Bernie's Mountain Dog. Super picky. But I bet Memphis would have eaten Sundays. Speaker 1 (16:25) Well. You too can make the switch to Sundays. Go right now to Sundays for dogs.com slash count and get 50 % off your first order. Or you can use code count at checkout. That's 50 % off your first order at Sundays for dogs.com slash count