Loading summary
Transcript147 lines
- [00:00]
A
We've all seen the chart showing that more Europeans die from heat related deaths than die from gun related deaths in America. Here's my question du jour. Live in a world, slash, a country where you have lots of guns and basically universal air conditioning or no guns and no air conditioning.
- [00:22]
B
It depends on where I'm geographically located.
- [00:24]
C
Do the police have guns?
- [00:28]
A
Wait, I think in Europe, a lot of the police don't have guns.
- [00:31]
C
Oh, so we're just talking about Europe
- [00:32]
A
versus we're talking about Europe versus America.
- [00:34]
C
Oh, I'd rather live in the U.S. yeah.
- [00:37]
B
Also, I don't want to give up my guns.
- [00:40]
A
That's true. And I don't want to give up my air conditioner. Looks like we've got three Americans on a podcast. Hello, and welcome to the GD Politics Podcast. I'm Galen Drouke and we have a lot to discuss today. Shortly before we begin taping, the Supreme Court released decisions in a couple of the cases we've been watching. In the late arriving ballots case, Watson vs. The RNC, the court ruled that ballots postmarked by Election Day that arrive after Election Day can still be counted in states where that's legal. When it comes to independent agencies, the court ruled that Trump could fire Federal Trade Commissioner Rebecca Slaughter, but ruled that Federal Reserve member Lisa Cook could keep her job while she challenged her firing. This continues a trend that court watchers have suspected would play out that the Federal Reserve has a special designation when it comes to independent agencies. Therefore, the court could expand the president's power over regulators without undoing the independence of the Fed. We'll dig into that and the late arriving ballots case on today's podcast. We're also going to discuss what Americans think about the Trump administration's deal with Iran after all kinds of conflicting messages about what it entails and whether the war is actually even over. And we'll get into some of the latest midterm polling prompted by a listener question about tightening in the generic ballot. And we've got a good data, bad data or not data example based on a Washington Post report suggesting that prediction markets are actually well calibrated. Here with me to discuss it all. We're running it back from Thursday's episode. We've got managing editor of Vote Beat, Nathaniel Rakic. Welcome to the podcast, Nathaniel.
- [02:23]
C
Thank you, Galen. I get first billing this time. Usually Mary gets first billing.
- [02:27]
A
You know, gotta, gotta mix it up sometimes. I also know you've been hard at work this morning. We gotta pay credit where credit's due. Also here with us head of research at 50 Plus One, Mary Radcliffe. Welcome, Mary.
- [02:38]
B
Hey, Galen. How's everything going?
- [02:40]
A
It's going pretty, pretty well. Not too much has changed since last Thursday. I'll also say for anyone who missed it, we had a lively debate about whether the left is in the midst of its own Tea Party on last Thursday's episode between Mary, Nathaniel, and myself. So go check, check that out. But let's begin with what Nathaniel has been up to this morning, which is dutifully writing about the Watson VRNC case at the Supreme Court. So, Nathaniel, what happened?
- [03:07]
C
Yeah, so in a decision that was somewhat of a surprise, although I think you could see the makings of it in some of the tea leaves to be read, the Supreme Court ruled 5 to 4, with Amy Coney Barrett and John Roberts joining the three liberals, that it is totally fine for states to accept ballots after Election Day as long as they are postmarked by Election Day. So this case threatened to basically change the deadline for the receipt of mail ballots in 14 states, plus DSG. This includes states like California that famously accept ballots that arrive after Election Day and therefore kind of slowing down the count. But as a result of this ruling, status quo will reign. So election officials don't have to scramble to educate voters and kind of change all their procedures at the same time. We are going to keep on getting results fairly late, and people like Donald Trump are going to continue to scream about it.
- [03:58]
B
He's already screaming about it.
- [04:00]
A
I'm sure. I have not checked Truth Social this morning, but I'm sure there were a couple areas of disappointment this morning. We won't even get into the Eugene Carroll lawsuit. What was the rationale from Amy Coney Barrett for keeping the status quo?
- [04:14]
C
So basically, the argument that Republicans who were the ones who challenged these laws were making was that Congress has set, by federal statute, Election Day as the Tuesday after the first Monday in November. And therefore, any like, you know, counting or that happens after that is moot, because Election Day is, that is that day. But Amy Coney Barrett in her majority decision said, nope, they're the, you know, the Election day is for like the casting of ballots. It doesn't, it's not about the counting of the ballots. Obviously, there are many steps that happen after the actual voting on Election Day, before results are certified. Right. Ballots are counted, they are canvassed, they are audited double. There are recounts that happen. Obviously, you know, if a candidate were to challenge a, you know, a result that was within like five votes or something in a recount, you can't do that recount on election Day, that's obviously impractical. And so, you know, it was basically a, a case about the meaning of election day and what needs to be done on election Day. And yeah, the kind of more voting rights, voting access folks ended up winning the day.
- [05:14]
A
This ultimately though comes down to statutes. So Congress could change this if it were wants to. It's not a constitutional issue. Do you think that Congress will change the status quo?
- [05:24]
C
No. And you know, Trump did respond to this ruling by calling on America for approximately the 14th billion time to pass the Save America Act. They will not do that because the votes are just not there in the Senate. It would require overriding the filibuster. And that is just, it seems pretty clear that that is not going to happen. You know, Trump has, you know, as we've discussed, like, he has been trying to pull lots of levers. He's been trying to hold other bills like the housing bill hostage to get this done. So it's clear that it means a lot to him. But no, I don't think nothing, I don't think that anything is going to change. Trump has also tried to make this change via executive order last year and that got struck down in court. So, you know, the kind of the, the system is holding in terms of people who might be worried about the impacts on the midterms. You know, I do think it's interesting, you know, Trump has been handed a lot of losses in court over the last like 10 days on election law, had his voter roll, attempts to get voter rolls from states rebuffed by multiple states or multiple judges in states. He has had his second executive order regarding kind of more closely regulating mail ballots that has been blocked in court. His use of the save system to verify or look for non citizens on voting rolls has been blocked in court. So I can imagine that the President is probably pretty angry and this is motivating him. At the same time, I think it is notable that the pieces are coming into place for Trump to be able to blame other actors. Congress for not passing the Save America act, the Supreme Court for not, you know, siding with him on the Watson VRNC case. If and when Democrats end up winning the midterms, he can say, well, there was all this fraud, all these ballots arrived late, you know, mail votes, you know, were continued to be rampant even though I tried to fix it. And, you know, that may be a potent political argument even if legally speaking, it's not having any teeth.
- [07:17]
A
So you're, are you saying that that's an off ramp for Trump trying to sort of fuck with the midterms, or you're saying be on guard because Trump is going to try to fuck with the midterms.
- [07:28]
B
Both.
- [07:28]
C
Right. I mean, like, it's. Predicting Trump is. Is obviously a futile exercise. Right. He could end up being, like, so angry that none of his things are working, that he'll be more motivated to take one of the actions that he, you know, can theoretically do. Although it would be illegal, like, you know, sending troops to polling places or something like that. But at the same time, he might. Yeah, he might just decide, hey, this benefits me. Like they. He has a very useful foil in Congress and in the Supreme Court right now. And if the goal is to try to overturn the election again after the fact, then, you know, maybe he will stay quiet until then, but we don't really know. I don't want to.
- [08:03]
A
I've been victimized by John Cornyn and Amy Coney Barrett. So if you're mad about the outcome of the midterms, be mad at them.
- [08:09]
C
Right.
- [08:09]
A
Not the war in Iran or anything like that.
- [08:12]
C
Yeah. And like, in fact, the war in Iran is super popular, and I would have won if not for these people.
- [08:16]
A
Got it. Got it.
- [08:17]
B
I just want to throw in also that the Save America Act, I don't think he has the votes in the Senate, even if they did nuke the filibuster. There were four Republicans that voted with Democrats against cloture on the bill. So I just don't think they have the votes. Even if they did nuke the filibuster.
- [08:35]
A
Yeah, yeah. And just to add some clarification to the independent agencies case, for folks who might be confused about why the Fed is different, you don't have to believe or agree with this argument at all. But conservatives argue that these independent agencies or regulators are exercising executive power. So they sue people, they make binding rulings, they bring enforcement actions, and as such, the President has power over that activity. Conservatives argue, however, that the Fed is sui generis. It is, quote, a uniquely structured, quasi private independent entity that follows in the distinct historical tradition of the first and Second Banks of the United States. So that is how the court could weaken the independence of the agencies while preserving the independence of the Fed. Liberals say that this is bullshit and that both the agencies, the independent agencies and the Fed should remain independent.
- [09:26]
C
Well, and it's actually the, you know, in the. The Cook case, which is the, the Fed case. Right. That was actually the liberals who were joining with John Roberts and Brett Kavanaugh to Keep Lisa Cook in her job. So actually, I think the liberals are on, are on that side of it of, you know, they want the Fed to.
- [09:44]
A
Well, they want the Fed to remain.
- [09:46]
C
They want Trump to power these people.
- [09:48]
A
They, they don't buy the argument unique. They say the independent agencies and the Fed are both covered. Well, they say the independent agencies and the Fed are probably both covered by Humphrey's executor, which is no longer, I guess, in existence now, or certainly no longer covers the independent agencies. And that this idea has been sort of craftily assembled so that conservatives on the court could both weaken independent agencies while not weakening an independent Fed because that is a pillar of the, you know, monetary system that we have.
- [10:20]
C
Right. But what I'm saying is that I don't actually. It's not conservative is on the court. It's actually specifically just John Roberts and Brett Kavanaugh. They are the only two people who are consistent.
- [10:28]
A
Right, right, right. Okay. Yes, yes.
- [10:29]
C
And I think that's important because I think they are the two kind of most business friendly, type of like Reagan Republican people on the court. Right. Gorsuch and Barrett, you know, for sometimes they can be idiosyncratic like in the Watson case, but I think they're a little bit more kind of like intellectually and ideologically driven. And then obviously Thomas and Alito are pretty far to the right. So I think that is notable that, like, the camps are pretty clear one way. All one way or all the other way. But it was just those two justices who were kind of splitting the baby here.
- [11:00]
A
If it were up to the other four conservative justices, the Fed would also not be independent.
- [11:04]
C
Right.
- [11:05]
A
So just to round this out, Tuesday is scheduled. Tomorrow is scheduled to be the last day of the Supreme Court's term, meaning we're gonna get decisions in the remaining four cases, including birthright citizenship and bans on trans women in women's sports in the states. So be on the lookout for that.
- [11:21]
C
Perhaps later this week. I'll put in a plug, too. The fourth remaining case is fairly important campaign finance case that could remove limits on party spending in coordination with candidates, which kind of similar to, you know, how, you know, Citizens United and other cases open the door to more money in politics. This could do that, too. So it's actually going to be a pretty big day tomorrow at the court.
- [11:41]
A
It's a big day. Thank you for keeping me honest, Mary. You. Somebody on the podcast mentioned that Trump would talk about how in fact popular the Iran War was and if it weren't for all of this stuff, you know, would have helped at the midterms. So let's address that question because it's been a minute since we've talked about Americans perception of the war and now the peace in quotation marks, which doesn't really seem to be holding. Starting Thursday and into the weekend. The US And Iran exchanged fire near the Strait of Hormuz, though the Trump administration says that continued peace talks are on schedule for Tuesday. We'll see how that pans out. At the moment, Iran has not necessarily said that that's the case. It's been almost two weeks though, since Trump signed the Memorandum of Understanding with Iran at the Pal of Versailles. Meaning, you know, when it, as, as these things go, it's a pretty good time to check in on how Americans are reacting to it all, understanding that this is all still in flux before we do that, and Mary has written about this, so we're going to ask the queen of Polls to sort of bless our understanding here. But before we do that, I'll just cite briefly what was included in the memorandum. So it's a 60 day window extendable by mutual agreement to negotiate a final deal. Immediate permanent end to military operations on all fronts, including Lebanon. That, of course, has already been violated in many ways. The US to fully end its naval blockade of Iran within 30 days and withdraw forces from Iran's proximity within 30 days of a final deal. The Strait of Hormuz reopened toll free, per Trump, though Iran still requires ships to clear its own newly created Persian Gulf Strait authority. U.S. treasury to issue waivers for Iranian crude oil and petroleum exports, plus associated banking, insurance, shipping services. Immediately up upon the memorandum's signing, about $24 billion in frozen Iranian assets to be released, half before the final talks, half tied to implementation. And then Iran pledges, quote, not to procure or develop nuclear weapons. The same language as the 2015 Iran deal, with substance on that deferred to the 60 day talks, such as any binding nuclear verification mechanism. Okay, that's a lot. Mary, what do Americans think about that? I'm sure that pollsters are reading respondents exactly what I just read for listeners.
- [13:53]
B
You know what's interesting is, you know, you can at the question wording on some of these polls and see really real differences in support for at least the Memorandum of Understanding. So I want to kind of break this up into little buckets of like, would they support the ceasefire and the deal, the MOU as it stands, do they support the ideas inside of it and how they think this is going to look in the future? Those Sort of the three sort of zones. As far as the mou, I have yet to see a poll with the fact of signing the MOU being underwater. Whether, however a pollster asks it, Americans want this war to stop. And so signing a ceasefire that at least temporarily stops it is generally more favorable than unfavorable. Question wording makes a huge difference. So, for example, YouGov economist had a pretty neutral question wording here, just like, do you support the preliminary Iran deal? Or something like that. They came out with 32% support and 24% opposition. But if you look at some of the earlier polls that were conducted immediately afterwards, they provided a little information about the mou. So there was a poll from Qantas Insights, for example, that mentioned in the question language that the MOU would stop hostilities and reopen shipping through the Strait of Hormuz. And when you, you give that extra information, we've got 56% approval and 13% disapproval. So there's a lot of variation. But no matter how you slice it, I have yet to see a poll with the MOU underwater. The difference often is really in the undecideds, frankly.
- [15:28]
A
Yeah.
- [15:28]
C
And I think that is the key. Right. As we know, Americans are not following every single twist and turn of the negotiations. They know only the broad strokes of foreign policy, if that. And I think that in particular. Right. Mary cited the YouGov poll. The number of undecideds on that was 44%, which is obviously a very large amount plurality. Right, exactly. More people than supported or opposed the deal. And the Qantas one, you know, has still had a decent number of undecideds, although fewer, which makes sense given that they kind of provided a little more information. So I think it is important to note that Americans do not have a firm position on this, and I think that their support for the agreement is probably just born from the fact that the, the one thing that most Americans know about the war is that they don't like it. And so anything that ends it is good. They don't know what's in it, what's in this deal, but they know that it ends the war or, you know, hopes to end the war.
- [16:20]
B
Yeah, I think that's exactly right. And, and when you look at, like, what is actually in it, there are some things that Americans like and some things that they don'. Right. Like limiting Iran's nuclear capacity is by and large quite, quite positive if that is an outcome we're able to achieve here. So in that YouGov economist poll, 69% said they supported that policy. 8% did not. Okay, so would love to get a
- [16:43]
A
member of that 8% on the podcast.
- [16:45]
B
Yeah, so that's, like, enormously popular. Reopening the Straight is very similar numbers. 69% in favor, 6% opposed. But Americans are a little, like, less. Both less sure and less supportive of some of the other stuff in here. So in terms of lifting sanctions on Iranian oil, that only gets 41% support and 27% opposition. Again, you can see those undecideds went up there, and it's much closer. And that reconstruction fund, the alleged $300 billion reconstruction fund is deep underwater. 23% in favor, 46% opposed. There's also in that Qantas Insights poll from last week, there's a piece of the MoU that would require Iran to remove or destroy its stockpile of highly enriched uranium. They're currently discussing method that would require, like, down blending it so it's less highly enriched. 63% in that survey said they would approve making this a condition of any eventual agreement. And this is going to be a really difficult sticking point because it's not clear if it's, like, possible to do this. The highly enriched uranium is, like, under a mountain of rubble from when we struck Iran last year. To have 63% saying they strongly approve of making this a ultimate condition of the agreement, that is like a warning sign for me that, like, whatever we are able to accomplish here is probably not going to satisfy what Americans think they want. They also probably don't care that much about this.
- [18:21]
C
Right, Exactly.
- [18:25]
B
Okay, so the. The thing that I find kind of worrying in the polling right now is how Americans are viewing what could be the future after the negotiation period. Generally speaking, Americans do not think that this is going to work. So Quinnipiac was out with a survey last week where 59% said they are not confident that this will work. And in that 59%, 40 are percent said they are not at all confident, whereas the other 19% are, like, not very confident.
- [18:55]
A
Wait, so not confident that the memorandum will lead to a broader deal, or not confident that the memorandum will actually even result in 30 to 60 days of peace?
- [19:09]
C
Or not confident that, like, Iran will hold up its end of the bargain?
- [19:13]
B
Yeah, so the question was quite vague. It just said, how confident are you that President Trump's deal with Iran will work? Okay, but I mean, listen, I have.
- [19:25]
A
I mean, maybe that's then just a poll on people's views of Trump.
- [19:28]
B
There are other polls, though, that I think suggest that that that's more or less correct. So, for example, immediately, like the day after the MoU was announced, RMG Research and Neapolitan News Service was out in the field. And so this is like in all of the shiny good news headlines, moments of like we got a deal signed. They asked respondents if they thought the war was very likely, somewhat likely, somewhat not likely, or definitely not likely to end in the near future. Right. So is this war gonna end in the near future? That's it. That's the whole question.
- [20:02]
C
Interesting.
- [20:03]
B
Only 47% said yes, 41% said no. I mean, these are slightly different questions than will this work? But like, even in the shiniest, happiest moments in the immediate aftermath From a slightly GOP leaning pollster, there's only a margin of 6% on whether this war is going to end in the near future.
- [20:25]
C
So yeah, that's interesting.
- [20:26]
B
Yeah, Americans are not super confident about this. And the last thing I want to mention is that we see in polls from both Ipsos, Reuters and from YouGov, Americans think this is a better deal for Iran than for America, generally speaking. So I'll just cite the YouGov survey. When asked whether Iran will be better off than it was before the conflict or worse off than it was before the conflict, 36% say Iran will be better off and 18% say it'll be worse off. Everyone else is a not sure or no change. But for America, 31% say it'll be better off, 5 points less than, say the same for Iran, 27% say it will be worse off, 9 points more than say the same for Iran. So generally speaking, people think that Iran won at least the negotiation here, not by a lot. There's a lot of not sures in there. Again, it's foreign policy, but that's where we're at.
- [21:27]
A
I think a lot of listeners will be interested in this on face value, how Americans are reacting to the war with Iran, the quote unquote, peace with Iran, et cetera. I think a lot of listeners are probably also just curious how this is shaping Americans perspective of the Trump administration, what it all means for the midterms. So I'm going to move us in that direction towards the national environment with a question from Andre who says, can you please talk about the recent tightening in the generic ballot polls? Is it real? Also, do we see any other metrics supporting this tightening? Which is, I think a question about is Trump or Republicans becoming a little more popular than they were during the depths of this war? And I will say we can get into generic ballot polling, presidential approval polling. At the very least, oil prices have been falling. So maybe if we think that there's a really close relationship between oil and gas prices and presidential approval or approval of Republicans, we could be on the lookout for some movement there. What are you all seeing in the data?
- [22:34]
B
Well, when it comes to presidential approval, at our 50 plus 1 polling average, we had Trump at his lowest at negative 24 percentage points underwater. He's up to negative 22 percentage points underwater right now. So there is a two point shift since the lowest lows, which were like during the depths of the conflict. That's not like, and I'll say that
- [22:57]
A
is backed up, I would say, by some other aggregators that don't have. Exactly. So like Silver Bulletin, it moved from net negative 20 to net negative 18. The New York Times moved from net negative 22 to negative 20. So I think we are seeing like a shift of 2 percentage points on net across the different aggregators, even if the top line is a little bit different. So yeah, that does suggest maybe a little bit of movement.
- [23:22]
B
Yes, yes, it absolutely does suggest a little bit of movement in the President's favor. I don't know that there's like a really big difference between a negative 22 point polling average and a negative 20 or negative 24 point polling average. When it comes to how people view the precedent, all of those versions seem quite bad. So like, like it may be, Yes, a, like a little bit of support coming back to, to Trump in the wake of the mou. Whether it stays there or not. I, I don't know. As, as to the generic ballot, I don't actually think that there's been any movement that's significant enough to discuss at least the 50 plus one polling average. We did at one point have Democrats ahead by seven. They're now ahead by six. They dropped to five at some point and moved back up to six. Right. Like these things are jiggly and noisy. We are still in a moment of the campaign where many pollsters have not turned on their likely voter models. We are still in a moment of the campaign where we have quite, quite a few undecideds in the polling still. This is going to continue to be jiggly and noisy for a little while. So, you know, like voter opinion really starts baking in. Like we can say something more in like mid September. Right. Until then, jiggles of a point or two here and there. I would just set them aside.
- [24:42]
A
Side jiggles of a point or two. The Mary Ratcliffe story. No, not a Good name for a memoir. Yeah, I mean, and I think I've looked across Silver Bulletin, ddhq, et cetera, and we have, we have seen movement. I mean, Andre is seeing something real. And I think they specifically cited the movement of a point in the Silver Bulletin average from a 7 point advantage for Democrats to a 6 point advantage for Democrats. So at least what they're noticing is that, that we're not seeing a continued trend in favor of Democrats at the moment. Nathaniel, do you have thoughts on all of this?
- [25:21]
C
Yeah, I mean, I agree with Mary. All these different aggregators use slightly different methodologies and if there is a genuine change in public opinion, it's going to be obvious. It's going to show up in everybody's aggregation. And if there's like a little bit of noise in one person's average, but it's fairly steady in another person's average, it's probably just kind of like a, a mirage of the data. And I completely second everything that Mary said about like it's early. Don't get too worked up. I think what people are really the real question that Andre and people who are looking at these averages are wondering is I would assume, like, which party's going to win the midterms or like, you know, is my party going to win the midterms? And like whenever, if you're a Democrat and you see you're standing go down by a point, what you're really wondering is like, oh, is that going to go down to like, you know, even. And then are we going to be screwed or you know, vice versa if you're, if you're a Republican. And I think that it's just, it's too early to tell. I completely agree that like negative 20 and negative 22 approval is basically the same thing. You know, plus 6 and plus 7 for Democrats is that's, those are both very good years for Democrats. And just bear in mind folks, the average error for national polls is like four points. And so we are not talking about huge movement here. Like the, the polls can on, on a lot, you know, go into election day with Democrats being up by eight points and they might only win by four and that would be completely normal. So just don't obsess too much about those small movements and those decimals.
- [26:50]
B
And given that the listener wrote in and they used the word tightening, I'm going to make what is perhaps a wrong assumption that you are a leaning Democrat voter looking for some cope if you want some. We did recent analysis at 50 plus 1 looking at the generic ballot. In each of the last three midterms, we found that where Democrats are now is better than where they were at the same point in the cycle in 2018, and that the percentage of Democrats planning to vote for a Democrat this year is higher than we have seen among any party voting for their own party in the last three midterms. Democrats are so on board, Republicans are less on board, but also generally on the high side for wanting to vote for their own party relative to other midterms. But if that helps.
- [27:40]
C
That's interesting. That's actually, that's all a little narrative busting too, because we've been talking about how dissatisfied Democrats are with their own party, but that's very interesting.
- [27:48]
A
Well, there's been a lot of agreement here. So let's bring in something a little bit spicier, which is the New York Times latest poll out of of Maine. Monday morning, they released a survey of the Susan Collins Graham Platner race, showing Graham platner leading Collins 49 to 47, which is within the poll's margin of error, and suggested some troubling signs for Platner underneath the hood. So they asked a couple questions about views of the candidates overall. One was, is this candidate too extreme? 47% said that Platner is too extreme. Only 34% said Collins is too extreme. They asked, does this candidate have good character? 44% said that Platner did. 66% said that Collins did. Does this candidate have the right kind of moral values? 45% said Platner did. 61% said that Collins did. Now, first of all, people may think, well, that's a contradiction. Why does the poll sort of overall, the top line numbers favor Graham Platner if, if people think he's too extreme and doesn't have good character? Well, let me remind you of Donald Trump, somebody that people broadly viewed as being extreme and not having good character, but was still able to get enough votes to win elections because that is not the only thing that people vote on, and we've talked about this on the podcast already, that basically Graham Platner's job will be to nationalize this race and make it look like a referendum on Trump. And Susan Collins job will be to remind people that, like, hey, I brought in all this money for Maine. I'm, you know, you know me, you can trust me, have again, good moral character, et cetera. But, you know, New York Times two point race for one of the top tier races that we're going to be watching this fall. What are Your takeaways from these results.
- [29:42]
B
This is going to be a close one.
- [29:44]
C
Yeah.
- [29:46]
B
I mean, this is one poll. We'll put it on our pile of polls, but it jives with, with the data we have to this point, which is. This is going to be a close one. I don't know. I don't know if there's anything more to say.
- [29:59]
C
Yeah, I mean, you know, I will. I think that's absolutely right. I think this is, like you said, Galen, it's a push and a pull between the partisanship of Maine and the national forces at play versus the individual candidates factors. You know, Collins is not particularly beloved in Maine anymore, but she does kind of score higher on those, you know, like, you know, trust and, you know, and moral code questions.
- [30:23]
B
Yeah. Just to throw in, Depole had her favorability underwater by two points and Platner's favorability underwater by five points.
- [30:30]
C
Right. So this is another kind of like double haters race and not to bring back that, you know, that term from 2024. But. Yeah. And, you know, it does make me think, you know, I don't think that Graham Platner's troubles are over. I think there are probably going to be more stories that drop between now and election day. There is still the outside possibility that he, you know, there'll be something so bad out there that he'll drop out of the race and Democrats will appoint a new candidate. I, I tend to think, given these numbers, that a kind of generic Democrat, somebody without baggage, you know, somebody like Jenna Bellows, the Secretary of state, or Troy Jackson, like, would be able to beat her by a fairly healthy margin. But we don't live in that world right now. So, yeah, with the, with the, you know, kind of field as we see it and the data as we have it, I agree with Mary that this is a toss up.
- [31:20]
A
I do think of this poll as a little bit of a wake up call for Democrats amidst the delirium of primary season, because folks have been looking at a lot of races where it's Democrats versus Democrats and drawing conclusions about sometimes, you know, not responsibly so, but drawing conclusions sometimes about the direction of the country or the parties or whatever. And this is a poll that targets everybody in the state. Democrats, it's likely voters. So it could be Democrats, Republicans and Independents, people from a state that is not super ideological. Yes, it leans left, but it is not Massachusetts or Alabama. And they, you can see in this poll that probably some of the policy positions and values that we were talking about on last week's podcast as being, you know, championed by the Tea Party of the left or the burgeoning Tea Party of the left or the something entirely different of the left. Probably the, the whole, the country as a whole is not on board.
- [32:21]
C
Sure. I also think that Graham Platner has his own issues that are not ideological. Right. All of his scandals. And that I think is certainly weighing into it. But. But no, I don't disagree with anything you said.
- [32:32]
B
Yeah, I think this race is going to be probably not the one that I would want to look at to declare the like progressive left as not being able to appeal to the electorate at large because of Platner's specific personal issues. Right. Like this is not, not an ideological fight in the same way that you might see in other districts around the country. Just because like it's comp. It's like it's all complicated by the fact that there's like candidate quality issues in some ways.
- [33:04]
A
As elections watchers and politics watchers, what we really want is for Abdul Al Sayed to win the primary in Michigan. And then you have.
- [33:12]
C
That would be an interesting one. Yeah.
- [33:14]
A
A top tier battleground state. Like top, top tier battleground state state. A an out and out progressive, unabashed progressive winning there and going toe to toe with a viable Republican.
- [33:27]
B
Yeah. Like a normie Republican.
- [33:29]
A
Normie Republican candidate. And then. And then we've got really got a test on our hands, I think. You know, I'm sure there will be some people listening who don't want that to happen. But from a like, let's really test some of these theories, especially in a super blue environment or presumably super blue environment. That would be really, really interesting.
- [33:48]
B
Yeah, agreed. That raises a weird mess though, by the way. Michigan.
- [33:53]
A
You mean the primary or the general? We'll talk.
- [33:56]
B
We'll.
- [33:56]
A
Don't worry Mary. We're going to talk about. Mary is currently a Pennsylvanian, but she is ancestrally a Michigander.
- [34:03]
B
To be clear, our former colleague Cooper Burton is also a Michigander and we text about the Michigan Senate primary like constantly.
- [34:12]
A
GD Politics is powered by you, the listeners. If you enjoy the show, the data driven analysis, the genuine curiosity and lack of partisan bs, and yes, sometimes the silliness too. Please consider becoming a paid subscriber@gdpolitics.com paid subscribers help make the podcast possible and they also get twice the number of episodes and access to the paid subscriber chat where you can send me whatever questions you've got got. Independent political media only works if the people who value it support it so if GD Politics is part of your week, head to GDPolitics.com and become a paid subscriber. That's GDPolitics.com I so appreciate it. Thank you. There's one more thing to note about this Times poll, which is they start to make some methodological changes that they're that preview greater methodological changes coming at the Times. This is pretty wonky, but we do have a former math teacher on the call. So this is as good a time as any to try to understand some of those changes, which are, they say just to kind of contextualize all of this. They say that they've seen a trend where gold standard pollsters are the most rigorous in trying to reach everybody oftentimes, but they sometimes aren't using the more advanced sort of waiting mechanisms or methodological mechanisms that people who actually work in politics are using. So they say that their goal is to use the absolute best practices in terms of trying to reach voters. But because that's not enough, they're also going to use the more wonky, highfalutin politics operator mechanisms for like deciphering those numbers once they come in.
- [36:00]
B
Yeah, so. So there's two big changes here. One of them is actually not that big of a change. I think it's a slight tweak to an existing method which is to use a new variable called they're calling the support score in their weighting. This is in place of 2024 recalled vote. So when you ask people who did you vote for in the last election? Studies routinely show that people lie or forget. Get if the politician they voted for has become unpopular, they may manifest a memory of voting for somebody else or
- [36:37]
A
not voting at all.
- [36:38]
B
Or not voting at all. If they voted for someone who lost, they may manifest the memory of having voted for the winner. I always liked that guy. And so rather than relying exclusively on 2024 recalled vote, the Times is using this new variable, which it includes recalled vote as part of how they calculate it, but it's actually an estimate of how likely that voter is to have supported Trump, Harris or not voted based on their demographic characteristics and partisan characteristics and how they answer the 2024 recall vote. So they're basically accounting for the fact that you might be wrong about how you voted, whether that's intentional or not. The second change is how they are actually computing weights. And I'm going to apologize, folks, because I'm going to get so nerdy for a second. The normal way that we typically calculate weights in a survey is what's Called raking or iterative proportional fitting, if you would prefer. I will call it raking. Here's what we do. For all those characteristics that you want to weight by, they'll be like race, gender, age, education, whatever it is, whatever the characteristics are. You set a goal for how much of your survey should match each characteristic. I want if you're in Maine or whatever, maybe you want it to be 80% white and 15% black. Whatever the characteristics of Maine are.
- [37:53]
C
If Maine is 15% black, I don't
- [37:56]
B
know, I'm making stuff up over here.
- [37:58]
C
If you've ever been to Maine, you know that it's not 15% black.
- [38:01]
B
I've never been to Maine. Not convinced it's a real place you should go. So you set the rule like the targets, the goals for each of those groups. Then you go and you ask all your questions, you get your responses and you don't have 80% white like you wanted. Maybe you only have 65% white, white, whatever. So you go through each of those targets and you like adjust how much you weight each respondent based on each of those target characteristics. And then you get a new weight and you look at it and say, is it representative now? And if it's not, you do that again and you keep going, making slight adjustments at every stage until you get a final weight that's, that's close enough to the target distribution. Now this can guarantee you get enough white people and it can guarantee you get enough young people and you get enough women and you get enough college educated people. But what it does not do is guarantee you get enough young, white, college educated women when you put those buckets together. This method is not going to ensure that those individual parts of the electorate are well represented. And we can't really set those as targets in our variable set because the populations are too small. The sample sizes will be way too small. So we won't have have enough young, white, college educated women in the sample probably, or at least some of those buckets. We won't have enough people for this to be reliable. So what the Times is doing is using this very new methodology which is called energy shifting.
- [39:33]
C
Energy energy balancing.
- [39:35]
A
What is it called?
- [39:37]
B
Energy balancing. I found this described in a paper, I found this described in a paper from 2024. So it appears to be quite new as a technique. If you are a real sicko out there and you want to Google this, you will have to add a math E word to your Google search. You can't just go, go for energy balancing because you will get like some chakra stuff, like I, I. If you throw in the word distribution, that should work. Basically, the idea is this, since we can't wait every little bitty tiny bit bucket in the population, what we'll do is we'll do the iterative proportional fitting, we'll do that raking process, we get a distribution, and then we're going to use like a fancy math technique for looking at how far is this distribution overall, the whole thing, from the actual distribution we want overall. So rather than looking at individual respondents, this is a holistic measure. We're saying you have a distribution of shift, shape, goal, and you have a distribution of shape, what we've done so far. And if they're too far apart from each other, we will move what we've done so far closer to what the actual goal is in a holistic way rather than on the individual demographic buckets. This is very interesting and there's lots of integrals, so it's really fun. For anyone who used to teach calculus, this is fascinating and it is, like I said, very emergent technique. Right. Very new.
- [41:06]
C
Yeah.
- [41:07]
A
And it pushes us closer to this idea that we've been talking about for a while, which is that increasingly pollsters are modelers. Right. Because in order to do this, you have to project what you think the electorate is going to look like, and then you massage your number. I mean, I don't want to say massage in a way that sounds flipping or whatever, but you kind of massage scientifically. Scientifically massage. You know, this is more of a sports medicine massage than a, I don't know, Reiki massage. But ultimately you're still massaging even though
- [41:34]
B
it is energy balance.
- [41:36]
A
Exactly. Ultimately, you're sort of massaging what you have to look something more like what you expect, which I totally understand why pollsters have been freaked out by the direction of the industry and Mrs. And repeated Mrs. In the same direction. And I understand the motivation to do this. There's like, a lot riding on getting this stuff right. It is also a little bit worrying because it blends. And this was already happening, so, like, I'm not trying to, like, be a fearmonger, but it blends together what we see out in the world with sort of what we expect the world to look like and makes it harder to see the difference between the two.
- [42:22]
B
Yeah, I mean, I think that's a fair point. Although, I mean, a likely voter model has always relied on pollsters doing some elect modeling of the electorate. Unless the pollster is using the, like the laziest Definition of a likely voter, which is like I asked you if you're going to vote and you said yes, which some pollsters do, right? Like, I, I don't know.
- [42:43]
A
And Salt are famously dead.
- [42:45]
B
Yes, Marquette Law School does, right? So like, okay, fine, that's a great pollster, right. Just to be clear. And maybe that's accurate.
- [42:56]
A
Like, okay, is this good data, bad data or not data? Nathaniel, I see you've been, you've been. I see, I see you're ready to go. Go.
- [43:06]
C
No, that's the key question, right? I mean, I think so for me. You know, I'll let Mary speak for herself. Obviously she's much more knowledgeable on this stuff than I am. Like, this is certainly above my pay grade, right? As Mary said, this is some real cutting edge shit. And like, I think I am inclined to say good data because I think it is good when pollsters experiment, try to innovate, try to find solutions to these problems that are plaguing the industry. I think the folks who are running the Times Sienna Poll are very smart and I think I trust them when they say that, you know, they think this is going to make their surveys better. And in fact, we should note that they, you know, they didn't just be like, oh, like, you know, this might work, let's try this next time. They like back tested it rigorously on their past polls and found that it would have made them slightly better. Not by a huge amount, but like, kind of like this is not going to be like some big, huge change that is forever going to like fix the polling industry. But they, they did see modest improvements when they looked and when they applied this method to previous polls. So I am inclined to think it's a good thing that they are iterating or are innovating and I'm eager to see how it works out.
- [44:13]
B
Yeah, I tend to agree with that analysis. The article that they wrote about this indicated that it improved their 2024 surveys by about one or two points. As Nathaniel said, that's not like earth shatteringly better, but I would love data to be one or two points more accurate than, than it is.
- [44:28]
C
Yeah.
- [44:29]
A
Can I bring up something slightly more esoteric that I think will ultimately indict some of the work that I've done on this podcast, maybe some of the work that we've all done, which they say that in doing this they're incorporating more of the ways that campaigns classify voters, that this is kind of some of the methodology that campaigns have been using.
- [44:49]
C
I think that's a Reference to the support score. Right. That's something that campaigns have used for a long time to model. Like, you know, oh, this person who lives at, you know, 1, 2, 3, fake stories street. Like, you know, should I go knock on their door and get them to vote? Like, but like, based on the demographics, I think they're likely to be a supporter of mine. The energy balancing stuff seems totally new as far as I know. I don't think campaigns are, there's no reason for campaigns to use that stuff.
- [45:09]
B
No, I have seen no reference to this.
- [45:12]
C
Yeah.
- [45:12]
B
At least none that I recall until this New York Times article.
- [45:17]
A
And I do wonder, like, I, this isn't, this doesn't mean it's bad for the New York Times to necessarily do this. But increasingly endorses this world that campaigns have already endorsed. And we in the polling observer world and election watcher world have also endorsed. We just increasingly talk about the world as these predetermined sets of groups. You know, like the whole, oh, well, we need to make sure we have the right proportion of young, white, college educated women, not just blah, blah, blah, blah, blah. And that is, it is representative of increasingly the ways that campaigns think they need to talk to voters as, as more groups than a whole. And I think that it makes it harder sometimes to create space for our society, to find common bonds or to talk to our country as a whole. And that over time it becomes a feedback loop that if we're always talking to you as your subgroup or trying to find the subgroup that you belong to in order to categorize you in our political analysis or appeal to you as part of a political campaign, you sort of increasingly see yourself as that group. And it makes it harder. Like, I do think that politicians almost always are lying when they say they don't have a pollster. They don't look at the polls. But there is something powerful about being able to sense or talk about something that can appeal to everybody. And, you know, I think the best politicians do do that, despite probably still having data teams that are doing this stuff behind the scenes. I mean, I think Barack Obama is a good example of somebody who really talked to Americans as a who group, even though he did have one of the most, like he had the most advanced data team in presidential political campaigns at the time. So I just worry sometimes when we get too carried away with who we are as groups and not who we are as a whole. And here I'm getting more esoteric. I'm not just talking about accuracy, like it could jeopardize Accuracy when we do see, right, shifts in the electorate that where groups who have historically done one thing, a realignment, if you will, like, it can jeopardize accuracy, but it also, I think jeopardizes harmony.
- [47:29]
B
Yeah, I don't know, I, I, I'm gonna have to disagree at least somewhat when it comes to how we think about polls, right. Like this is not micro targeted advertising. We're not trying to send a message to anyone. In fact, we're doing the opposite thing when we're doing polls. We're trying to understand how the electorate is right. So, so like we knew in advance of the 2024 election and found ourselves this to be confirmed that there was going to be changes in certain demographic subgroups, that they're not some sort of monolithic thing that's always going to be the same and have the same opinions. We learned from the pre election polling using scientific methodology to ensure that we represent a, a sample of Americans that looks like the electorate, that the Hispanic vote was going to shift towards Republicans in 2024. And then that happened and now we're learning in polls because we are doing, doing really good science to reach voters and understand their opinions that some of that those gains are going to be clawed back. Right. Like this is a different thing than how you message to these voters. This is about understanding what the electorate is. So I don't know, I think actually like if you don't try to make sure that your sample meets these demographic quotas in your survey, you're gonna have a worse understanding of America.
- [48:47]
A
In fact, this kind of data can put a check on certain stereotyping and the like.
- [48:51]
B
Yeah, exactly.
- [48:52]
A
I think that, and so I think that Go for it. New York Times do this. I think it's valuable. I think the fact for me that when I read that like we're incorporating the way that campaigns classify voters, it reminded me of something that makes me sometimes uncomfortable about American politics and maybe sometimes makes me uncomfortable as even, even an analyst of American politics that we are. And look, the vast majority of Americans are not listening to podcasts, this podcast or podcasts like it at all. And so they're not thinking about people as their census driven category or whatever, you know, but to the extent that that kind of analysis does permeate all of political analysis and it increasingly has, which like people are like, wait, then why the fuck are you even doing this podcast? That's like so much of what you do. I want, I want you to know that I am a human and I do have my doubts, even though I general believe in the work that we're doing. Like there are moments when I step back and think, is this always productive? Are there times when it becomes counterproductive?
- [49:54]
C
Yeah. And I think that that kind of philosophical and thoughtful mindset is a good one to have. I do tend to agree with Mary. I think that, you know, it is good analysis to do things this way. And I don't think that what's dividing the country is the New York Times polling methodology. But, but I do, I absolutely take your point, Caitlin, that, you know, kind of like if politicians talk to Americans like they are segments, then we may segment ourselves and that may not be productive. And to some extent I think that also goes, is true for the media as well. Although I do think that nobody trusts the media anymore. So I think that our, our influence is often overstated. But I do take your point, Kaylin, and I, I think it's good for anyone who speaks in that kind of segmented of a way to keep in mind.
- [50:46]
B
Yeah. I mean, this is why despite some others in my life at various points having wanted it, this is why I don't publish crosstab level polling averages. For example, I might publish an article analyzing crosstabs in polls, but I would never want there to be a polling average out there because the individual jiggles, as we were discussing earlier, are even less meaningful at that level. And there's, it's just too easy for people to overthink. Like, oh, what if the black vote moved a point?
- [51:16]
A
Like, stop a little Monday morning existential crisis is always healthy. Right? Okay, we have one last actually good data, bad data or not data before we close. And it's been on the back burner for a couple episodes now, so I want to make sure we actually get to it. And this example comes from the Washington Post. I think we can do this relatively quick, but I'm reading from the Post. A few days before the first round of Los Angeles's mayoral primary, Kalshee and Polymarket gave reality TV star Spencer Pratt a 75% chance of advancing to the general election. Kalshee published campaign style updates on his odds, and polymarket hosted multiple active markets on whether Pratt would finish first or second in the city's nonpartisan ranked choice primary. On Kelshi alone, the mayoral race market surpassed $40 million in bets before voting ended. And then Pratt ended ended up badly trailing Nithya Rahman, a left wing rival, for the second slot in the general election. That pattern, a prediction market favorite ending up as a loser, has occurred several times this primary season. Candidates favored by the markets, for example, also lost in the Georgia gubernatorial primary, where a candidate endorsed by President Donald Trump lost to a healthcare executive, and in Kentucky, where longtime Representative Thomas Massie lost to a Trump endorsed primary primary challenger. But people who rely on prediction markets to help them understand politics, however, aren't panicked because these losses aren't a failure. They're a sign the markets are doing their jobs. For each of those high profile misses in midterm primaries this year, a Post analysis found there were about three other elections where a favorite with a 75% chance of victory did win. The Post examined 268 candidates who received a 70 to 80% chance of winning on either count Kalshi or Polymarket on any day in the two months prior to their primary. I have read a very long opening to a Washington Post article that all comes together to say we checked whether or not the prediction markets were well calibrated and we determined that they were. Is this good data, bad data, or not data?
- [53:30]
C
I First of all, I think this is exactly what you want to do, right? Like you, you know, we have talked about the problems of prediction markets on this podcast before. But you know, we want to you want to look at the data and you want to see how often they are correct and it is the correct approach to be like something that prediction market gets give a 30% chance of happening to should happen about three and 10 times, right? And you know, that is what the definition of well calibrated is. I would love for predict prediction markets to be well calibrated and they found that they were according to this methodology. I'm going to have to give it a bad data though, because of something you said. Ghillen the post examined 268 candidates who received a 70 to 80% chance of winning on any day in the two months prior to their primary.
- [54:11]
B
It's like you're reading my mind, Nathaniel.
- [54:14]
C
This is not helpful. And in fact one of the critiques I think that we've had of prediction markets is that they are way too noisy. They're very reactive to new polls or you know, a Trump truth or something like that. They jump around a lot. And so if you're looking at a race where one candidate went from a 30% chance to a 90% chance and then finished at a, you know, 80% chance, like that goes through the 70 to 80% range but like that's not that helpful. And so what I would really love the Post to do is to go back and do this analysis with people, candidates who had a 70 to 80% chance of winning, like on the morning of their primary. And if that ends up being well calibrated, then I will have more respect for prediction markets. But until then I find it very, I find this analysis to be incorporated incomplete. And I also, I just don't see the utility in using prediction markets particularly far out because of how noisy they are.
- [55:07]
B
Yeah, to, to expand on that a little bit, the way that we calculated how well calibrated our election forecasts were at 5:38 treated every day of the forecast model in the two months prior to the election as a separate prediction. So we had however many races times 60 days, many predictions. Right. Because we did have a separate number forecast for each race on every day. And so the model would be well calibrated if in general, all of those forecasts were typically well calibrated. Right. So that means like if a candidate moves over time from like 40 to 45% like that should, that's going to count a bunch of different times to understand the calibration. But here they're not taking that into account at all. And I think Nathaniel's point about the noise in the prediction practice is exactly right. Yeah, at one point it was 70%, but then at another point it was 30% and that 30% prediction was still valid. Like you just happened to cherry pick the day you want to look at, but that 30% prediction was still a prediction just the same as the 70% was. So I think this is close to being good data. It's adjacent to good data.
- [56:18]
C
Frustratingly close.
- [56:22]
A
Nathaniel said straight up bad data. And it's rare that we get a full blown call on this segment. So Mary, am I also hearing that
- [56:29]
C
as a really ornery son of a
- [56:31]
A
bitch, as a bad data?
- [56:33]
B
I, I actually, I, I, I'm closer to not data because like, and here, this is why. But this is why. But this is why. Because this is like the basic idea of this methodology is on the right track. If they did one more step, I would be calling it good data.
- [56:50]
C
Right.
- [56:52]
B
So I don't want to say it's bad.
- [56:53]
A
Incomplete data.
- [56:55]
B
Incomplete data. Yeah.
- [56:57]
A
I'm just giving you an out. So what we want to know is for every single day that that market was live, how did its prediction do? And if it ends up being well calibrated, fantastic. It is good data on the part of the Washington Post and, and it is good data on the part of these prediction markets.
- [57:15]
B
But until then, yeah, we need more.
- [57:18]
C
Yeah, I suspect, you know, and somebody should do this. You know, maybe the Post can do it, but maybe, you know, you guys at 50 plus one or whoever. But I suspect that on the day of the election, prediction markets probably are pretty well calibrated because I think that the people who are trading on these markets are fairly savvy consumers of data. I would expect that like a month before. I bet there's, like, very little. I bet it's very badly calibrated.
- [57:41]
A
Well, the good news is we're gonna get a lot more data very soon, Nathaniel. And I'm sure the Washington Post will listen to this podcast and apply these lessons going forward. Jk JK all right, that is it for today. Thank you so much for joining me. Any other existential crises that anyone wants to air before we get off the call?
- [58:01]
C
You know, as a country, I am really nervous about getting over the hump of being 250 years old. You know, there's just something different between 249 and 250. I just feel really old right now,
- [58:11]
B
right now experiencing some sort of like, midlife crisis.
- [58:16]
C
It's a quarter millennium crisis.
- [58:19]
A
Oh, my God. Yeah, it's really. This is all.
- [58:21]
C
This is all just America's quarter millennium crisis, folks. Don't worry, we're going to get through it.
- [58:25]
A
That must be what I'm feeling this morning, honestly. Nathaniel, thank you for putting words to my feeling. All right with that, Mary. Nathaniel, thank you so much for joining me.
- [58:33]
B
Thanks, Galen.
- [58:34]
C
Thanks, guys.
- [58:36]
A
My name is Galen Drooch. Remember to become a subscriber to this podcast@gdpr politics.com and wherever you get your podcasts. Paid subscribers get about twice the number of episodes. You can also join in our paid subscriber chat and pass along questions for us to discuss on the show. And you ensure that we can keep making a podcast that prioritizes curiosity, rigor, and a sense of humor. Also, be a friend of the POD and go give us a five star rating wherever you listen to podcasts, maybe even tell a friend about us. Thanks for listening and we will see you soon.