Loading summary
Transcript46 lines
- [00:00]
A
I think. I don't know, Gail. I think this all reduces to the same point, which is that Graham Platner is a. He's an affluent person's idea of what a working class individual is like. They basically just picked the town drunk in Maine and made him run for Senate.
- [00:27]
B
Hello and welcome to the GD Politics Podcast. I'm Galen Troop. We've gotten some dramatic updates since the last time we discussed the state of the 2026 midterms. The main one is of course, that Graham Platner has dropped out of the Maine Senate race after he was accused of rape accusations that he denies. The main Democratic Party now has until July 27th to replace him on the ballot. They plan on hosting a convention of about 600 delegates made up of local party officials from around the state and the Maine Democratic Party's executive director. Suggested candidates may need to collect signatures from voters in order to be eligible at the convention. Maine, however, isn't the only must win Senate seat for Democrats that's recently been in flux. In Michigan, State Senator Mallory McMorrow dropped out of the Democratic primary this past Sunday, leaving a two person race between establishment backed Haley Stevens and progressive backed Abdul El Sayed. They faced off in a debate this week and will face off on the ballot next month. To talk about these races and much more, I've invited back on our two favorite election forecasting upstarts, Lacha Jane and Zachary Denini. Laksha is with Split Ticket and the Argument. Welcome to the podcast, Laksha.
- [01:35]
A
Hey, thanks for having me.
- [01:37]
B
Pleasure to have you. And also here with us is Zachary, who's with Vote Hub. Welcome back, Zachary.
- [01:42]
C
Thank you very much for having me on again.
- [01:44]
B
Okay, so we are going to dive into the latest drama in just a minute, but I first want to check in on your friend forecasts. So Laksha, when we spoke last month, your Senate forecast suggested that Republicans had a 53% chance of winning the chamber. That's now up to a 61% chance. What happened?
- [02:04]
A
Well, a couple things, Galen. I think number one and the biggest is that Graham Platner's candidacy in Maine has unraveled on a daily basis, going from bad to worse. One of the biggest things is that without Maine, it's really hard for Democrats to actually win the Senate because now you've got to win three of the states that we would call like reach states, right? Like Iowa, Texas, Ohio, Alaska. When the ask was win two of those in Maine and North Carolina. Much easier lane to winning the Senate. Still difficult, but way more plausible without Maine, you can't really do anything. The good news is that with Platner withdrawing as of last night slash this morning, I expect that whoever the replacement nominee is, whether it's Troy Jackson, Shanna Bellows or Nero Shah, is going to be in a much better position to win that race. And so that'll help Democrats a bit. And the other quick thing is that, you know, in places like Ohio, some of the polling that has come out has been curiously bad for Democrats compared to what the other polls used to say. So New York Times had one that had Sherrod Brown down in Ohio, which was completely opposite to what, you know, Fox News and the others showed, which is that Brown had a healthy lead. Because the Times is such a high quality pollster, it drags our average down a little bit. And so the chances of Democrats in some of these other state races have gone down a bit.
- [03:36]
B
Zachary. However, your forecast has not changed at all since we spoke last month, at least as regards the Senate and very little as regards the House, too, which we'll get into. So the Senate forecast last month at Vote Hub said that Republicans had a 55% chance of winning the chamber. It still reads 55% chance for Republicans today. Given what Laksha just said about the changing environment, sort of a deteriorating position for Graham Platner in Maine before he dropped out, as well as some tepid polls for Democrats from the New York Times, why no change?
- [04:08]
C
One reason is just how our forecast is built this far out. Four or five months before Election Day, we find that polls are just not that predictive compared to our fundamental models we make. So because of that, polls right now are only weighted at about a fifth or a third of the total model weight. So even if we get a bad poll in something in some place for Democrats in Ohio, where, you know, from a good pollster where it moves our average a point or two to the right, it's going to be a change you might not even notice just looking at the top line. Overall, I think that our national environments, mine and Laksha's have gotten a little closer in the last month or so. I think Laksha's, I believe, is still going to be higher on Democrats than vote hubs. But at the time when we recorded last month, the public polls and then the arguments LV screen polls were showing a little bit of a bigger difference. And now Democrats have been gaining a little bit in those public polls.
- [05:10]
B
And so where are you guys converging more or less in terms of the
- [05:13]
C
national environment, our predicted national environment right
- [05:16]
A
now is D 7.4 and ours is D plus 7.7. So more or less the same last time.
- [05:24]
C
I think it could have been almost a point different.
- [05:26]
A
That's right.
- [05:26]
C
And we both moved in.
- [05:28]
B
And I'll say that after we recorded our podcast last month, we heard from a lot of folks who thought it was really interesting to hear both of your takes and how you crafted your forecasts differently and why that resulted in different things. And so I am curious, Lacha, given that for your response to the non movement in Vote Hubs model and sort of de emphasizing polls at this point in time, I imagine there are also arguments for emphasizing polls more. We have become more polarized, our elections have become more predictable further out in reaction to that. But what's, what's your response to Zach?
- [06:01]
A
First of all, Zach's point is very relevant, but it's not like we wait polling that heavily right now either. Polling is about 35% of a forecast at max right now. So in places where we have a lot of polling, like Maine, it's 35%. In places where you don't have a lot of polling, like Vermont or not even Vermont, like Alaska maybe. Yeah. Or. Or like California's 22nd with David Valadao. Right. Like pulling is going to be worth much less in those places because we don't have that many surveys. So it's not just because of polling volume. It's also because, primarily speaking, our model just doesn't see as many paths to victory for Democrats if you close off Maine. And Zach's main probability has not gone down as much as ours has, I think. And that's part of it. The other part is that in Ohio, right. Like we thought the fundamentals were better for Democrats than Vote Hub did. And when the polling comes out that diverges from that a little bit, it forces the model to calibrate a bit more in favor of the polling and be like, hey, what's going on here?
- [07:14]
C
I think another interesting thing on the main point is so we're running a forecast update literally right now. I'm sure we're recording before people are seeing this. So it will be out by the time when people see it where we're removing Platner from our forecast. And what that means is we're removing all of the Platner Collins polls. So today, Thursday, July 9, Dems are going to jump, I don't know, maybe 5 to 7%, I'm guessing, in our main Senate prediction.
- [07:41]
B
And how does that affect your forecast overall? Does it go up from 55% to closer to Laksha's at 61%, I think
- [07:49]
C
it will tick up maybe a point or so.
- [07:51]
B
Okay.
- [07:51]
C
Just because Maine, Maine is still just one state and it's not the tipping point state. Right. It's a necessity for Democrats to win, but especially without Platner now, you know, if they lose it, they're. They're not really going to have a path.
- [08:02]
B
Lacha, have you removed Platner from your model yet?
- [08:06]
A
We will remove him when he sends the official documents of withdrawal to the Secretary of State. Until then, he is formally the nominee. He has released a video saying he will drop out. We are de facto treating him as if he's dropped out in, like, discourse and columns and whatnot. And I think that's correct. But until he sends the documents saying I am no longer the nominee, by policy, we don't remove that because, I mean, truth be told, it's a very volatile situation. I don't think these are especially stable human beings and like, goodness knows what can happen. I've seen extraordinary rows. I'm not saying he's going to do that, but I am saying we're not going to remove him until he's formally off the ballot.
- [08:45]
B
Okay, so we've got some slightly different takes on forecasting just in that as well. Another way to underline how there are different ways to do these things. Let's talk about the House for a second before we really dig into Graham Platner's withdrawal. On the House side Last month, Laksha, you gave Democrats a 90% chance of winning the chamber, at least when we spoke. And that was before you had officially released your forecast model. Today, that's down to 83%. Zachary, a month ago, you gave Dems a 72% chance. Today that's at 73%. So also little changed on the House side over at Vote Hub. I'm curious, Laksha, why we saw that movement. Of course, the House is not dependent on Maine in the same way the Senate is or Ohio either. What's going on when it comes to the national environment that's sort of shifting the odds down for the House?
- [09:36]
A
Well, I'll say two things. One is we had a couple of final methodological tweaks and bugs caught in final qaing before we released. Right. So that took us down from like 93 to 87 when we launched. It was just like a couple of minor things that we fixed, like correlated issues here and there, you know, polling averages, candidate quality, imputing names and whatnot. Like, very minor things. That all in all, just added up. I'll give you an example. Right. Like in doing name matching for candidates and pulling against WAR databases, we had to make sure that we didn't pull the wrong. De La Cruz is war because there's a lot of candidates.
- [10:13]
B
Juarez wins above replacement, which is exactly. Of the candidate running.
- [10:18]
A
Right. Like, we used to do a last name match. Then we realized De La Cruz and Ted Cruz both would match on the last word in the name. And so we had to do a fix for that. Right. So those are all.
- [10:31]
B
Did you catch that or did Claude catch that?
- [10:34]
A
I caught that.
- [10:35]
B
Oh, okay. All right. Human in the loop. Human in the loop.
- [10:38]
A
Yeah. No, I.
- [10:39]
C
Back in 2024, we had Sam Brown and Sherrod Brown. We had two S. Browns even in Senate.
- [10:44]
A
Yeah. No. So that's the type of thing we fixed. And then after that, the national environment has tightened by about, like, half a percentage point, and that's resulted in a drop from 87 to 83.
- [10:54]
B
Okay, got it. And, Zachary, on your end, little change. What's going on there?
- [10:59]
C
In the last month, there was a little bit of change. I think it got to a low of about 70% after we talked last time, Democrats continued to get some pretty poor national polling. You know, since. Since then, it's recovered a good amount, which is, you know, often expected. Or we try to make sure our forecast doesn't overreact to, you know, just swings that we think could be basically random or, you know, different pollsters releasing. So that's why it only went down from 72 to 70, and now it's back up to 73. So it's very, very subtle changes. But overall, yeah, I think it kind of comes back to the fact that. That our, you know, general midterm fundamentals was a bit lower on Democrats a month ago, and we just haven't seen any polling that's really out of line with our. What, you know, our prediction was on the environment a month ago. So our projected national environment just hasn't changed very much.
- [11:54]
B
Laksha, you mentioned how the New York Times Senate polls affected your Senate forecast. For folks who haven't seen those, we have talked about them on the podcast, although I think it was on a paid episode, potentially. Democrats were only leading in two of the battleground Senate seats. Those being North Carolina in Maine tied in Texas, and then Republicans leading in all the rest. So Alaska, Iowa, Texas, Ohio. Did those Senate polls affect your House forecasts at all? Did it suggest that maybe, you know, because the New York Times is also going to great lengths to reach certain kinds of voters that maybe your average either generic ballot poll isn't reaching or even a House poll might not be reaching. Did those polls affect your broader understanding of the national environment and therefore the House forecast as well?
- [12:44]
A
They didn't because in general, we don't really like to have too much of an influence between the Senate and the House polls.
- [12:52]
B
Like we.
- [12:54]
A
We found that that is actually pretty noisy historically. Right. Like, as Zach can also tell you, candidate quality matters a lot in the Senate. And I don'. Want a model where like Talarico doing really well because Paxton is just a really bad candidate ends up dragging every single Democrat up by the amount that like Talarico does or whatever. Now, there are ways to play with the interaction factors so that like a good poll for Talarico helps Texas House Dems, but on the net it's pretty cloudy as to how much actually helps and we have enough polling to make it not that big of a difference. So we kept them separate. It's a design choice. I wouldn't fault people for making them correlated a bit more. I know Nate does that. I don't. And I've found that it doesn't really make a difference in the final forecast.
- [13:41]
C
Yeah, 2022 was a really strong data point supporting what Laksha did here. And that's also why we don't do this as much. You saw people like Kathy Castor in Florida even in 2022, you know, outrun. Was it Val Demings on the Senate ticket by I think was eight or nine points in Arizona. You saw, you know, Kerry Lake really struggle and Republicans do fine in the House ballot there.
- [14:06]
B
Final question before we get into Grant partner specifically, which is we're in July, which is typically a lull in the election cycle. You know, we still have some primaries upcoming in August. Likely voter screens don't really come into play until around Labor Day. While we are in this moment of pause or historic moment of pause, there's still a lot in flux which we're going to talk about in Maine and Michigan. Are there certain things you're looking for to get a read on the national environment going forward?
- [14:39]
C
One thing I think is always interesting to monitor right now, right now would be the Iran situation. I think there were two things at the start of the year that drove polling starting to move away from Republicans nationally.
- [14:51]
B
That's the end of today's Preview. Head to GDPolitics.com to become a paid subscriber and catch the full episode. Laksha, Zach and I spoke for an hour and discussed the factors that will shape the midterms between now and the fall. We talked about about how Platner dropping out would reconfigure the forecasts. We dove into who might be the strongest candidate to replace him on the ballot. We also asked what lessons there are to be learned for the Democratic Party, both the progressives and the establishment, in what happened in Maine. And we got into the Michigan Senate race and what the median Democratic voter really believes. It was a great chat. Like I said, head over to GDPolitics.com to catch the whole thing. For eight bucks a month or 80 bucks a year, you'll get twice the number of episodes. And we'll help ensure we can keep making an independent podcast again. Again, that's GDPolitics.
- [15:38]
C
Com.
- [15:39]
B
Hope to see you there.