Loading summary
Transcript58 lines
- [00:00]
A
Had I known that Robin was going to release her new single at 1:30. Actually, Laksha, you're pretty young. Do you know who Robin is?
- [00:08]
B
I'm 28, dude. I don't know who this is.
- [00:11]
C
Is this a safe space, Galen? Because I was googling rapidly as you were talking about that.
- [00:16]
A
Wait, wait, wait, wait, wait.
- [00:17]
D
You also didn't know who Robin was?
- [00:19]
C
No, but I, I, the song sounded really cool and I don't listen to a lot of cool music, so I thought it was very cutting edge.
- [00:26]
E
Lenny, I, I, I knew who, who, like who that was, but I don't think I could have named a single song.
- [00:34]
A
Okay, wow. This is one of those exper, you know, those, like, quizzes you can take to see how much of a bubble you live in. This is a live version of one of those. Anyway, I was just gonna say, had.
- [00:44]
D
I known that she was releasing her.
- [00:46]
A
New single at 1:30, I would not have scheduled this podcast for two. But we've recorded podcasts in breaking news environments before and we're just gonna do it again today. So strap in, everybody.
- [01:06]
D
Hello and welcome to the GD Politics Podcast. I'm Galen Drook. If you're a regular listener to this.
- [01:12]
A
Podcast, you've already gotten plenty of 2025 election content. We even posted the video of our five hour live stream for paid subscribers at GDPolitics.
- [01:21]
D
And the news cycle is quickly moving.
- [01:24]
A
Beyond last week's elections. There's the end of the shutdown, the.
- [01:27]
D
Jeffrey Epstein emails, new candidates jumping into the 2026 primaries, the release of Robin's first single in seven years.
- [01:34]
A
Yes, that one's for the gays.
- [01:35]
D
And pretty soon, much of our electoral attention will be focused on the midterms. But before we say Goodbye to the 2025 elections, I want to do a postmortem with the advantage of more data.
- [01:48]
A
And less sleep deprivation.
- [01:50]
D
So today we're gonna focus on some of the fundamental questions. In particular, who, where and why, the demographic and geographic breakdown of the vote and the issues driving voters. We'll also compare what we saw to past elections and answer some listener questions about the results. And if everyone behaves at the end, we will get to some of the other things going on in the world, namely those Jeffrey Epstein emails and the raging battles over what kind of candidates the parties should be running if they.
- [02:17]
A
Want to win at the midterms.
- [02:18]
D
Here with me for our 2025 election postmortem is head of political data at the Argument and co founder of Split Ticket, Laksha Jain.
- [02:26]
A
Laksha, welcome to the podcast.
- [02:27]
B
Thanks for having me, Galen.
- [02:29]
D
Also with us is Lynley Sanders, polls and surveys reporter at the Associated Press. Welcome, Lynley.
- [02:34]
C
Thanks. Excited to be here.
- [02:36]
D
And Lenny Brauner, data scientist at the Washington Post. Welcome.
- [02:40]
E
It's great to be back.
- [02:41]
D
I have to say that it was only when I sent you all an.
- [02:43]
A
Email confirming this podcast that I realized I'd inadvertently invited only people with l names. So, Lynley, Lenny, Laksha, a bit of a tongue twister.
- [02:55]
C
Say it five times fast.
- [02:56]
A
Say okay, Lindley, Lenny, Laksha, Lindley. Okay, I don't think I can do it, but here's what we can do. Answer some listener questions. So I'm going to dive right into a question from Rob, one that is charmingly simple and might even be trolling us. But I'm curious what your thoughts are. Laksha, we'll start with you. Rob asks, was it turnout or persuasion?
- [03:20]
B
The thing is, turnout is persuasion in a lot of ways. Think about it this way. Non voters actually care about a lot of the same things as swing voters do. We know from the 24 election survey data and the post election studies that had everyone turned out, Trump would have actually won by more because non voters straight up leaned towards him by more than the people who showed up to vote did. It's not really true that like Democrats had amazing turnout in 2024 either. It's just that they lost even more ground among the people who stayed home. All that is to say that I think The Democrats in 25, they got good turnout for sure. And then also we know from the exits it's like 7% of Trump voters in 2024 voted for the Democratic candidate in 2025 in Virginia and New Jersey. And we also know that among the people who didn't cast a ballot, Sheryl and Spanberger won those handily as well.
- [04:18]
A
I'll follow up with Khalil's question here, which jumps off of what you were just saying. He asks when looking at the Hispanic vote moving back towards Democrats and we talked about this on election night, we did see the Hispanic vote moving towards Democrats decisively, particularly in New Jersey. There's a larger Latino Hispanic population in New Jersey than in Virginia. It makes up about 20% of New Jerseyans.
- [04:42]
D
Khalil says how much of that is.
- [04:44]
A
Actual Trump 2024 to Democrat 2025 voters versus the existing trend of say, more educated Hispanic voters leaning Democrat and being more likely to vote in off years.
- [04:58]
B
I look at the exits and I can see that like almost 20% of Trump 2024 Hispanic voters outright voted for Mikey Sherrill. And these are the types of numbers where you have to look at it and think like there is a substantial persuasion component. Now there it is true that some of the low propensity Hispanics that voted for Trump sat out 2025. This is true. And these people don't tend to vote as much like low propensity voters don't tend to vote as much in off year elections. But like there was a substantial amount of true persuasion going on. Now I think something that I haven't seen talked about as much though that I should bring up is that I'm pretty sure a lot of those 18% voted for Biden in 2020. We don't have evidence like hard evidence in terms of like voter survey data, but what we can see is that there's a lot of reversion going on with the margins that Mikey Sheryl got. She won by a similar amount to Joe Biden and her margins in those heavily Hispanic towns also came very close to his. So I think like a lot of this persuasion is Biden 2020 to Trump 2024 to Mikey Sheryl, 25.
- [06:06]
A
So it's something of a last in, first out dynamic in terms of swinging electorates kind of.
- [06:13]
B
And that's to say like that that realignment that we saw in 24 could continue. Maybe this was just down ballot reversion, but like it could also just be that it was a, it was a blip on the radar and then Hispanics swing back because we are seeing Hispanics just like wildly swing against Trump in polls. And it's also true that the Hispanic reversion here was massive beyond what I just expect, a normal down ballot lag.
- [06:35]
E
We do know from political science research generally that it is voters that, I mean also this generally makes sense. Voters that change their vote, that swing are more likely to swing again.
- [06:44]
D
Right?
- [06:44]
E
There are this idea that swing voters exist is real. So it would make sense from what Lexa just said that it would be most likely the voters that did swing from Joe Biden in 2020 to Donald Trump in 2024 to be the ones that are most likely to actually swing back, rather than this being a new set of voters entirely or ones that had been long term Republican Hispanic voters.
- [07:03]
A
Lindley, you at the AP along with the networks worked with SSRS to do the voter poll, which is something akin to an exit poll, except, you know, because a lot of people are no longer voting on election Day. They're voting in advance or they're Voting by mail. It also incorporates some traditional polling methods to reach out to people, you know, who say they voted but may not have been at the polling location on election Day. In any case, it tracks a whole bunch of demographic information about what the electorate looked like and what motivated them. So I'm curious, were there particular surprises in that that get at this question.
- [07:38]
D
Of turnout versus persuasion, or maybe get.
- [07:40]
A
At other trends that, you know, we should be focusing on when assessing why people voted the way that they did in 2025?
- [07:48]
C
I do agree. Claude showed that it was and probably always is, persuading voters to turn out. In this case this year, we saw it was over the economy. If you look at Virginia, where Spamberger spent a lot of her time talking about the cost of living and federal layoffs that had impacted Virginia voters, the voter poll showed that about half of Virginia voters said that the economy was the top issue facing the state. I think about 6 in 10 said that their family had been affected by the federal government cuts and spammer very handily won those federal worker households that included government contractors. So I'm sort of straddling the lane on this question, and maybe Lenny has a bolder take on it, but I think it's persuasion that leads to turnout, what we saw last week.
- [08:34]
A
So I'm getting a little bit of it's both. But I'll say that Nate Cohn tried to break down mathematically for the New York Times, how much of it was persuasion versus how much of it was was turnout by looking at, you know, what percentage of voters in that exit poll said that they voted for Trump but then voted for a Democrat this time around. I'm curious if you guys had thoughts on that breakdown, accounting for the fact.
- [09:00]
B
That turnout is one vote and margin change and persuasion is two. Right. Because you're going from Republican to Democrat. I would have said it's like, you know, 2/3 persuasion, 1/3 turnout, or like in Virginia, slightly less, like 60% persuasion, 40% turnout. I've seen some smart people say half and half, but I haven't dove into the results at like a specific precinct level enough to say it that conclusively.
- [09:24]
E
I totally agree with that characterization and I think that very much is similar to what NACON shows in the New York Times. So NACON actually was able to take a look at voter file data. So this is individual level data, so we know exactly who turned out. So a few counties and New Jersey had already made this data available, which usually takes actually months to come out. But this data was made available a lot quicker. And so you can take a look at sort of the exact differences in turnout by party registration or by model party registration, if you want to include people that you know are not registered with either party. And he showed that democrats had a 19 point turnout advantage by party registration. And this is up from around 16 percentage points in 2024. The additional modeling that the New York Times did, which I assume includes sort of non partisan registrants, the difference goes up to 3.5%. But obviously the total shift between 2024 and 2025 was closer to 9 points, which is, I think where these 2/3, 1/3 weighting comes from. So from this 9 points, 3.5 percentage points can be explained by turnout and the rest would have to be explained by persuasion. Mind you, I'm talking about New Jersey specifically here where we have this data.
- [10:33]
A
So that's the fine grained detail of the turnout versus persuasion math. But I think something that you all were alluding to before is that regardless of how you allot Democrats margin in this case, that in many ways the issues that will motivate somebody to turn out are the issues that will motivate somebody to switch their vote. So it could be that you're very upset with the economy or the state of the economy. It could be that you're very upset with the Trump administration. Both of those can end up having the same effect on a voter, getting you off the couch or making you change parties. I'm curious, Lynley, when it comes to motivation, you said the economy, can we break down a little bit more of sort of the issues that were driving people to, well, turn out or switch parties this year?
- [11:16]
C
Yeah, absolutely. So the economy and pocketbook issues were pretty clearly seen as the top issues in the major races in Virginia, New Jersey and New York City. Trump may have motivated some voters even though he wasn't on the ballot. We saw that about half of California voters on Prop 50 and about 4 in 10 New Jersey and Virginia voters said that they were voting to oppose Trump. Voters were much less likely to say that they were casting a ballot with the idea of supporting the President in mind. And there was a lot of discontent in some of the elections. We saw that more than half the voters in New Jersey and Virginia said that they were angry or dissatisfied with how things were going in the country. And it's, I mean, hard to not argue that that has an impact on how people end up voting and what they end up caring about.
- [12:04]
A
Did we ultimately see like, compared to similar elections like 2021 or 2017 higher turnout, or is this on par with what we've been seeing in recent elections?
- [12:16]
E
It depends a little bit on what state we're looking at. I'll start with New York, because it's the clearest example here. There was, you know, know, very high turnout. It's the highest turnout since at least 2001. But I think even going back further for a New York City mayoral in New Jersey there, there was higher turnout, too. So, like, if we just look at ballots cast, which is, like, an imperfect measure because usually you want to look at the exact turnout because population changes, too, over time, then the ballots cast was 23% higher than in 2021. Just sort of a. As a baseline here in New Jersey in the past usually has around 2.2 million cast ballots for governor. In 2021, it was 2.6 million cast ballots, and this time around, it was 3.2 million cast ballots. So, like, quite significantly higher than in 2001, which was already a relatively high turnout election compared to sort of the past in New Jersey. In Virginia, if I'm looking at the data correctly, turnout was actually, you know, down a tiny bit compared to the last governor election there. That being said, we are generally in a high turnout environment. I mean, this has been true, basically, since Donald Trump came down the elevator in 2015. Elections have seen higher turnout. So that 2001 governor election in Virginia did have higher turnout than Virginia governor elections in the past. But compared to that election, we had slightly lower turnout. But it's, like, still even also for Virginia, where turnout went down a little bit. Generally, we're seeing high turnout sort of across the board.
- [13:35]
B
Yeah. And I want to add one thing to that. So, Len, you brought up, like, turnout being high, but I think something no one talks about is that when turnout is high, usually it's, like, high across the board. Like, Republicans turn out more and Democrats turn out more. And we are talking about, like, maybe, like, Democrats turning out at a rate of, like, I'm just gonna make up numbers. You're like, 79% Republicans turning out a rate of 76% versus like, 60% and 59% or whatever. But what I think is really interesting here about New York is that, like, Mamdani's margins against Cuomo were actually pretty tight compared to the polling. He only ended up winning by, like, less than 10%. But the electorate he got was perhaps the most young electorate that I have ever seen in, like, any of these elections. It was actually crazy. Like, these are the types of things that, like Bernie Sanders and his supporters thought they would get and they didn't. And what was interesting to me is that like Zoran got that kind of electorate. And now look, I have actually, on a quick note, I have said I don't think Zoran is a national model for Dems because it turns out his national numbers are pretty bad for himself. But I do think there is something for Democrats to look at and be like, how did he get turnout that good? It didn't necessarily help Zoran win by more because people were persuaded against him, but it had implications for down ballot races too. Right. Like the rising tide lifts all boats thing. It did have implications for all the down ballot races in New York in margin. So, like, there is something to look at there in terms of, like, how did he get an electorate like that? You can hold two things to be true at once. And that's kind of what was interesting for me in looking at the New York turnout numbers.
- [15:17]
C
Specifically, the voter poll is not something that measures turnout, but I did find it very interesting. Just like you're saying about how well Mamdani did with younger voters. Close to eight in 10 voters under 30 voted for him, while older voters were going more for Cuomo. And it was such a clear thing along age. It really was a generational race. And then Mamdani, who was gonna be the first Muslim mayor, had the support of about nine in 10 Muslim voters in the city. And we saw about six in 10 Jewish voters went for Cuomo. So all the media narratives that we were seeing there really did play out in how people went out and voted as well.
- [15:56]
A
Yeah, we've talked a little bit about Hispanic voters, which was one of the big trends coming out of 2024. Another big talking point coming out of 2024 was the degree to which President Trump won young voters. Was there as much snapback amongst young voters as there was amongst Hispanic voters?
- [16:15]
B
It was is crazy. The.
- [16:17]
D
All right, that's the end of today's preview. Head over to GDPolitics.com to become a paid subscriber and listen to the full episode. We chatted for about 50 minutes and covered a lot of ground. We talked about what the 2025 results portend for the 2026 midterms based on historical trends. We also had takes on the realignment debate, the gender divide, and folks did, in fact, behave. So we also got to Epstein File polling and more. Like I said, head over to GDPolitics.com to become a paid subscriber and catch the whole thing. Paid subscribers get about twice the number of episodes can join the paid subscriber chat and most importantly, keep this podcast going. When you become a subscriber, you can connect your account to wherever you listen to podcasts so you'll never miss an episode. There's a link in the show notes explaining how. Again, head over to gdpolitics.
- [17:07]
A
Com. See you there.