Galen’s Cauldron Of Queries
40:40 Share

Okay, I'm looking up the trailer for Hotel Transylvania. Nathaniel, do it one more time. Hello and welcome to the 538th Politics Podcast. Hello. You can't do it, Jalen. You can't do it. Wait, let me try one more time.

Okay, I'm going to have to do the rest in my normal voice because that required a lot of effort. But we're going to keep the Halloween puns to be clear.

As the air gets chillier and the ghost of the October surprise still looms, as we continue to creep towards Election Day, today we're looking into our cauldron of queries, otherwise known as our mailbag, to answer your most haunting questions leading up to the big day. From gushing voters to bone-chilling polling to supernatural phenomenons like voter turnout, we are here to discuss it all.

And joining me to break it down in costume, our senior researcher, Mary Radcliffe. Welcome to the podcast. What's your costume, Mary? Do I just make something up? Because I'm just wearing a T-shirt from a brewery in West Virginia. So you mean to say you mean to say hardworking polling researcher in the days before an election? That's your costume? Yeah.

Also here with us in costume is senior elections analyst Nathaniel Rakich. Welcome to the podcast. What's your costume? Hi, Galen. I am dressed as the scariest thing of all, a 269-269 tie. Oh, God. Folks, he just held up. He is wearing a tie that says 269 twice on it.

My costume is the blue Oxford that I will be wearing for the next two weeks. I just am the Oxford. That's me. That's my costume. You're telling me that I'm the only person who actually put in effort here? Come on, guys. Maybe after the election, we'll host a belated Halloween, belated Galen's birthday party, and I will dress up as something more compelling.

Anyway, before we get to the mailbag, we have a couple of things. Folks might be getting tired of this, but I'm just going to cite where the polling is. So according to our averages, Harris leads nationally by a point and a half in the battleground states in Wisconsin, Michigan, Pennsylvania and Nevada. The polls are tied in North Carolina, Arizona and Georgia. Trump leads by between a point and two.

That is very similar to the last time I cited those numbers. And it is Thursday of Halloween when you're listening to this. Although we're recording a day early, I'll say we expect some of that final polling to come out over the weekend. I'm looking at you, Ann Selzer, in the Iowa poll. In 2020, it came out in the evening of Saturday before the election. TBD on the New York Times final battleground polls. But we will be with you in this feed on Monday to discuss it.

We're also sitting down this morning to a bunch of garbage. We don't need to linger on this, but of course, at a Trump rally at Madison Square Garden on Sunday, a comedian referred to Puerto Rico as a floating island of garbage, along with a bunch of other crass or insensitive things that were said by both the comedian and other folks in attendance. Then on Tuesday, during a campaign call, Biden said,

seemed to refer to Trump supporters as garbage, although it seems to also depend on where you put the apostrophe before the S or whatnot. He suggested his team suggested that he was actually saying the Trump supporters comments were garbage, a.k.a. the comedian's comments about Puerto Rico were garbage.

Nonetheless, this is a bit of what we're talking about after the closing argument speech that Harris gave on the ellipse on Tuesday night, seven days before the election.

You know, we know from past election cycles that information flows matter. In fact, when we did our most valuable data point episode, Patrick Ruffini said that, you know, in the final week of a campaign, things can shift and those sort of not highly engaged voters who are still determining whether they will turn out to vote and who they might vote for can be swayed by information flows during the final clip. I think that was pretty clear in 2016. Does any of this matter or rise to that level?

It could, I think, in a particular handful of states that have significant Hispanic or Puerto Rican populations. I'm thinking of states like Pennsylvania or Nevada or Arizona. The thing that surprised me about the comments at the Trump rally was how much media play they got. You know, the story was leading for a period of time on at least three of the major news sites online.

It seems to have broken through in a way that some previous comments have not. So it's possible perhaps there are some, you know, disenchanted Hispanic or Latino voters that heard those comments and decided, you know what? Fine, I will go vote. Yeah, I mean, speaking of that polling, I was looking at some of the latest Equis research polling from October and Carlos Odio, who's the co-founder of Equis Research, has been on this podcast.

He's talked about this polling question, which is asking Latino voters who between Trump and Harris cares most about people like me. And when you look at that question in their latest polling, 60 percent of Latino voters say Harris. But the top line in that poll is

was Harris getting only 55% of Latino voters. So I'm looking at maybe that 5% who are inclined to believe that Democrats and Harris care more about people like them, but maybe giving their support to Trump for reasons like the economy or mishandling of immigration or whatever it may be,

who might say, you know what, I have been reminded of this other sentiment that I have, which is there's more antipathy towards people like me on this side of the aisle, and I may end up swinging my vote accordingly. And as you mentioned, there are about half a million Puerto Ricans in Pennsylvania alone. So when we're talking about super small margins, and I don't know if on election day, we'll be talking about super small margins, maybe something like this matters.

Half a million is a lot more than the margin in Pennsylvania in the last few elections. OK, let's get to our mailbag from Eric.

Eric asks, since incumbents typically receive a slight advantage in their polling, when it was Trump versus Biden, it sounded like the common belief was that the two incumbent advantages canceled each other out. Now that Harris will be on the ticket instead of Biden, does that mean Trump now has that incumbency advantage? Or is it still a wash due to Harris being the current VP? Nathaniel, what say you?

I mean, I think it is not at all clear that there was an incumbent advantage. Biden was the incumbent, right? And he was unpopular. Trump has some trappings of incumbency, such as high name recognition, but he was also unpopular. So I don't think incumbency is automatically a benefit when you are unpopular, for example. And so I think actually that one of the things that Harris benefited from, particularly when she jumped back in the race, was that she was the

the farthest from an incumbent of the three, even though she's the sitting vice president, because she was kind of a fresh face and public opinion about whom was still very malleable, as we saw by the fact that her favorable, unfavorable ratings basically reversed themselves. And

She is also younger, which I think was a very important contrast with Biden and also with Trump. So I hesitate to say there's like a non incumbent advantage. I think the sample size is too small. We have like a very long track record of incumbents having an advantage in American elections. So I don't want to throw that out based on one or maybe two elections based on 2020. But yeah.

Yeah, I would say that Harris is probably happy to not be the incumbent right now. Yeah. And I mean, we've seen in the wake of all the inflation and, you know, sort of economic upheaval post-COVID, we've seen across the world voters punishing incumbent parties or incumbent heads of state.

All over the world, people are mad because inflation happened everywhere and it's been really bad. So I think there's actually an argument that incumbency could be a disadvantage in an environment where people are still sort of reeling from all of the economic upheaval in the past couple of years.

All right, next question. This one's anonymous. Ooh, spooky. Could Washington's primary clue us into what will happen in November regarding the general election? And just to clarify here, the reason that Washington's primary can be indicative of how the country is shifting compared to the last election is because they have a top two nonpartisan primary. Everyone competes. So you get to see a sense of

what Republican and Democratic turnout looks like and what those preferences are across a nonpartisan primary ballot. Jeffrey Skelly wrote an article on this. And yeah, historically, it has been a pretty decent indicator, kind of similar to special elections, in that it is an actual vote. People go into the polls.

The Washington primary did suggest a slightly Democratic-leaning, like, popular vote. But obviously, we have Electoral College and all that stuff. I think the Washington primary, if memory serves, was a little bit better for Democrats than the current polling environment looks.

But I think it's just important to note, same with the special elections, which were like basically D plus six this year. All of these indicators, including the polls, are just kind of ballpark indicators. And so if we ended up at like something like a D plus two popular vote, which obviously could go either way in the electoral college, that would just be a couple points away from the Washington primary results. It would hardly be disproving their predictive power. Same with the polls. Same with the special elections to a lesser extent. So, yeah.

So, yes, I think it is part of the magic soup that you can use to predict the election. Sorry, Nathaniel, I think you mean a magic witch's brew. Right. Sorry. Is there anything else in that brew? By the way, we have special elections. We have the Washington primary. We have polling period. I guess we have economic data, which, by the way, also came out today. American GDP growth is just under three percent, which is decent.

Anything else we're putting in that cauldron? There's a lot more economic data than just GDP. Unemployment, inflation. Consumer sentiment is a big factor. Which is like better but not great. Yeah. Some of these broader indicators like the economy, like presidential approval, for example.

They can be useful when you're like a year out from the election. But when you're at this point, a week out, they have such wide error bars that I think I mean, you want to go with polls at this point because it's always possible that the political environment has shifted since the Washington primary, for instance.

Speaking of indicators of how things might go in this election, our next question is this. Say you had to make a bet on the winner of the presidential race, but you're given a crystal ball that you can look into that would let you know the winner in just one state and the winner's margin of victory in that one state.

Which would you pick to give yourself the most certainty possible on the winner and exclude Pennsylvania? Because that's the most obvious answer. Well, that's not fair. I think I'm going to go with Wisconsin on this one. It's the whitest of the northern battleground states. And so what I would be looking for is Pennsylvania.

especially thinking about margin of victory, any indication that what we've been seeing in polls among white voters is actually coming to fruition. In polls, we are seeing white voters move several points toward Democrats.

If that were to be true, if white voters moved several points towards Democrats, even if the Democratic ticket lost significant support among non-white voters, it would probably still be enough to push them over the edge in all three of those northern battlegrounds. So I'm looking at Wisconsin and I'm thinking about how the margin there might translate into support among white voters specifically.

Nathaniel, you got a state? I'm going to go with North Carolina. According to our forecast, it's the second most likely tipping point state after Pennsylvania. Pennsylvania is by far the likeliest. So, you know, darn you for not letting us choose that one. Wisconsin and Michigan, like obviously it's good for one side if they win that one.

state, but there's still like about a 25 percent chance that the loser of that state could win the presidency, according to our like what if interactive. But North Carolina, if Trump wins, he wins 80 percent of the time. If Harris wins there, she wins 90 percent of the time. In addition, I hear what Mary's saying about

white voters. But I also think there's a lot of question marks about voters of color. And Wisconsin has very few of those and North Carolina has more of them. And so I think we'd get just like a better cross-section of the country that way. It'd be also particularly useful, I think, if we could see like county to county data in the state. But maybe that's asking too much of the crystal ball.

All right, next question from Lizzie. If 19% of the general electorate in 2024 will not have voted in the 2020 election, what are the different subcategories that make up this group of new voters? We're talking about churn in the electorate here. Mary, do you have thoughts?

First of all, you always get a bunch of new voters that are generally younger than the population as a whole. In part, young voters, especially voters that are voting for the first time, are really excited to vote. So you'll get a bunch of 18 year olds who are like, yeah. So not that many. Right. Overall, those young people are voting like at super low rates. Yeah. I mean, if every single one of the newly eligible voters is.

This cycle voted every single one of the eight million newly eligible voters. It's like five percent maybe of the total turnout. Right. So that's not anywhere near that 19 percent that Lizzie's talking about. So to get into this, I actually went to the most recent New York Times Siena poll because they include a crosstab of non-voters in 2020, but that are categorized as likely voters for 2024. Right.

So I thought that would be an interesting way to sort of get a sense of what these voters look like. They tend to be less college educated. They're much less white. They're significantly less partisan, right? Less likely to identify with a political party strongly or to identify as liberal or conservative.

Interestingly, they're much more likely than both Biden and Trump voters from 2020 to say that democracy in America is not under threat. So these may be voters that are not, you know, when people say this is the most important election of our lives, these voters go, is it? They're not as concerned about that sort of thing. And they're much more likely to get their news from social media. So this is really a description of sort of a

Less engaged voter that's maybe paying a little less attention, less likely to have college education, lower income, more likely to pay attention to social media. That's this sort of churn in voter for 2024.

And on net, who does that group support? Trump, right? Yeah, but not by a lot. I think it's a five point margin, if I'm remembering it correctly. Right. Historically, that group of voters, the low propensity voter, has been a benefit to Democrats, has chosen the Democratic candidate over the Republican candidate. And it seems like the dynamics have shifted as the coalitions have changed along educational lines and, of course, racial lines as well. Our next question is one that I love.

From Andy, politicians from both parties have historically high unfavorable ratings, and you've spent a fair amount of time talking about double haters, potential voters who have an unfavorable view of both candidates. Is there some tiny fraction of the voting population that has a favorable view of both candidates? Who are these double lovers, and who are they voting for?

That's a great question. I mean, yeah, there's a small sliver of those people, but I have no idea who they're voting for because they're such a small sliver. I haven't seen any breakdowns of their support. But yeah, I would be curious to talk to one of these people. If you're one of these people, get in touch, please. I'll say that looking over the 2016 exit polls, I found that 2% of voters in the exit polls, which of course have their own caveats about accuracy, said that they had a favorable view of both Clinton and Trump.

Trump. And at the time, I tweeted something that I thought was funny, but I was like, oh, this is Kanye West. Because in the run up to that election, he basically said that he liked or respected both Clinton and Trump. And I was like, okay, you know, I guess there are some people out there who see positive things or positive potentialities in both candidates.

Although this seems kind of silly to talk about today, it wouldn't have been silly to talk about in 2008 or 2000 or 1988 when both candidates were very clearly above water on favorability. And so you would have had as opposed to like one fifth of the American public having an unfavorable view of

both candidates, it would be more like 20% of Americans have a favorable view of both candidates. And these are the halcyon days of us being like, oh, you know, the like highly attentive voter who's really just like paying attention to what the candidates say. And, you know, oh, but they said this on that policy and this on that policy. And I trust them on sort of like lining up their different positions and then swinging between the two candidates accordingly.

I think that image of a swing voter is now a little outdated in this new era of unpopular politicians. But once upon a time, that may not have been such a ridiculous caricature. Yeah, polarization rules everything around me. It sure does. All right. The next question is a meaty one, but we're going to dive in nonetheless.

FiveThirtyEight has frequently labeled this race a toss-up or 50/50, despite Harris sometimes being ahead and Trump sometimes being ahead. How can each candidate still win? For example, does Kamala have more upside with more likely voters while Trump is depending on low propensity voters? Another way to ask this is, if Kamala or Trump were to win, how could it happen? How would the turnout look different from the current polls?

So if the polls are exactly accurate, right, on average, we have a super tight race. But we should expect, if we have an average error, that there's something of a clearer winner. And perhaps there's a good chance that one candidate sweeps all the battleground states. In either of those cases, say, you know, Trump plus three in the battlegrounds or Harris plus three in the battlegrounds, what happened? What did the sort of like turnout or persuasion look like?

I mean, I think based on the demographics we see in polls right now, if we are looking at a Harris victory in all the battlegrounds, I think what we are probably seeing is, you

A high turnout and by high, I mean high relative to historical trends in urban areas and suburban areas in particular. I think we are seeing movement among white non-college educated women toward Democrats, which is something that's been hinted at in some of the polling.

And I think we are seeing less attrition for the Democratic Party among voters of color. That's like the the pieces, right?

that I would expect to see if we see a Democratic sweep in the swing states. If we see a Republican sweep in the swing states, I think we are looking at sort of a relatively low turnout environment in urban areas, Republicans maintaining their advantage among white voters that they've seen in previous cycles. And we're looking at low turnout or significant gains for the Republican Party among voters of color.

Whoa. Okay. I would have thought something different, but I'll let Nathaniel respond first. I don't know. I struggle with questions like this. I think we don't know is the answer, right? Like there are many ways that Harris could beat her polls. There are many ways that Trump could beat her polls. No, seriously. Well, no, no, but wait, let me, let me say my piece. Like, you know, you could see, it could just be that like the

the polls have uniformly missed. And like every demographic group is just three points bluer, three points redder than we thought. It could be because there is racial depolarization or the urban-rural divide is getting wider and certain some groups go in one direction and some groups go in another direction.

Yeah.

You know, that wouldn't suggest to me any major shifts. I think actually probably if it turns out to be Trump plus three, that would presumably, you know, Trump might win the popular vote. And that would be, I think, interesting in terms of like coalitional analysis. But if Harris ends up beating her polls, her map could look exactly the same as Biden's. And we'll be like, well, that was much ado about nothing. So I don't know. I don't know.

I think we need to wait and see the actual patterns of results before answering that question. Yeah, I think that's a pretty fair point, right? Like I'm thinking about what we see in polls now and what each candidate would need to beat that. But it's totally fair to just be like, actually, what if the polls are just off by three points in one or the other direction? But I think the turnout piece is notable, Mary, given what we've already said about low and high propensity voters, right? If Democrats could run a repeat of

2022, they might choose to do that and keep it to a relatively low, highly engaged, more educated electorate.

if Republicans could get sort of the turnout that we had in 2020. Obviously, Trump lost in 2020, but that might be a more advantageous environment for them because what we're seeing, right? And I looked at a bunch of this polling back when we did a podcast about Black voters and shifts among Black voters, which is that most of these shifts towards Trump come amongst lower propensity Black voters. The most reliable Black voters are almost as Democratic as they have ever been.

What the polls are picking up when they show Trump getting, you know, between 15 and 20 percent of black support are black voters who say they are not sure whether or not they are going to show up to vote on Election Day, which is why we are also seeing an advantage for Harris amongst likely voters when the polls use a likely voter screen versus registered voters overall. And so, again,

I do think that it's not just getting the, well, this does ultimately affect the margins, but it's not just getting sort of like, if everyone voted, where does every demographic fall? It's figuring out who's going to vote. Is it going to be these highly engaged, high propensity voters who seem to advantage Harris? Or is this going to be more of a 2020 situation where we're approaching 70% turnout and a lot of not very engaged people who are not happy about the economy

are voting basically against the Democratic incumbents of the past four years. We'll find out. We will find out. We will find out. Okay, now we each have a take that we have to defend come election night. This is an ad for BetterHelp Online Therapy. What's something you've always wanted to learn? An instrument? A new language? How to not burn pasta? Whether it's big or small, imagine how great it would feel to finally do it.

As kids, we were always learning and growing. But as adults, sometimes we lose that curiosity that made life feel so exciting. It's easy to think that we are who we are and we can't change. But BetterHelp is here to tell you that it's never too late to be who you want to be. Therapy can help you reconnect with your sense of wonder. Because your back-to-school era can come at any age.

BetterHelp makes it easy to get started with affordable phone, video, or live chat sessions you can do from anywhere, and the option to message your therapist between sessions if anything comes up. Rediscover possibility with BetterHelp. Visit BetterHelp.com today to get 10% off your first month. That's BetterHelp, H-E-L-P, dot com.

Hey podcast listeners, tired of ads barging into your favorite news podcast? Good news! With Amazon Music, you have access to the largest catalog of ad-free top podcasts included with your Prime membership. Stay up to date on everything newsworthy by downloading the Amazon Music app for free or go to amazon.com slash adfrenews. That's amazon.com slash adfrenews to catch up on the latest episodes without the ads.

All right, next question, also from anonymous, also kind of spooky.

Over one month ago, I saw reports that voter registration of young Black women increased 179% and Latina women almost 150%. And those were not the only groups with which we were seeing an increase. Does this constitute an expected large increase in young voter turnout? Aren't newly registered voters highly likely to vote? How effective is a potential shift in young voters factored into the polls?

We did some data research here, the producers did, to try to figure out what our anonymous caller was talking about. And it appears to be a CBS report from August of 2024, which relies on data from TargetSmart, a Democratic-aligned data firm. And it showed how much higher voter registration was in the month plus following Joe Biden's withdrawal from the race on July 21st, compared with the same month

in 2020. So in other words, how did the first weeks of Harris being the de facto nominee compare with the same period in 2020? So how would you respond to our anonymous caller? I guess I would just say that that probably just tells us about enthusiasm for Kamala Harris over Biden. And I am not sure that it

necessarily means anything broader than that, similar to how she raised, what, like $100 million in the first week or whatever. It was like 36 hours. Yeah. Yeah. Those people who are going to rush to register at that point, sure, they're going to be

motivated. But in terms of absolute numbers, I don't know how big they're going to be. And also, you don't know how many of those people would have registered later on anyway when somebody was knocking on their door on September 15th and they're like, OK, I guess I should do this now. And maybe they were less excited about doing it in a scenario where Biden stuck around. But at the end of the day, they still do it. And obviously, a vote that is cast unenthusiastically counts just as much as a vote that's cast enthusiastically. So this to me just kind of

smacks of trying to read beyond the polls using other data, which I understand the temptation of. But at the end of the day, the polls and some of the other kind of factors that we talked about that go into our models and our analysis are the things you want to pay attention to, not

little cherry picked pieces of data like that. Yeah. And on the topic of looking at the whole as opposed to the parts overall, when it comes to voter registration in America, Democrats advantage on registration has been eroding.

nationally and in the battlegrounds too, specifically, Mary, if we look at Pennsylvania, for example, a few years ago, Democrats had a more than 600,000 person registration advantage. By October of this year, that was cut in half or more than cut in half to just 300,000. So if you're looking at who has been registering over the past several years, on net, Republicans have been winning that

battle. Yeah. And I would also add that voter registration tends to be a lagging indicator. That is to say that people will vote Republican or Democratic

and then they'll change their registration later. And in a state like Pennsylvania, I think that reflects a lot of those blue-collar Pennsylvanians who have been registered Democrats for a long time because maybe they belong to a union or something like that, but maybe voted for Trump in 2016 and are voting for him again this year and are just now getting around to changing their registration. All right, here's a fun one. I would love to hear a discussion about how weather on Election Day affects the results! Exclamation point, exclamation point.

Previous work has shown rain could be beneficial to Republicans, but will the realignment, what we were actually just talking about, affect this trope? I'm seeing a 50% chance of rain in Philly on Tuesday and not sure what to think. Boy, Andre is the name of this caller. You are in it. You are looking up weather patterns in the most populous city in the likeliest tipping point state to

to try to get a read on this, man, we need to hire you as a sleuth. Does the weather matter? A tiny bit. The New York Times had a piece out recently looking at this, and they found that it can have very marginal impacts, especially if you have an all-day kind of dreary day.

Now, your 50% chance of rain probably doesn't imply an all-day kind of dreary day, the kind where you want to just sit on the couch with like a cup of tea. Could it have an impact? It's possible. I would suspect any impact would be very, very small, unless we're having a torrent. Right, exactly. I think, obviously, the severity of the weather matters. I think that the...

enthusiasm around the election matters. I think this year there is a sense that a lot is on the line. And so it'll take more to get people to stay home. And there's more ways than ever to vote not on Election Day. Right. If people plan ahead and look at the weather forecast, which they may not be doing. But yeah, at the end of the day, I think, you know, they drizzle

isn't going to keep a lot of people from voting. I do think it's worth noting that, you know, that past research that has found that it can help Republicans that may be out of date in a world in which Republicans are the ones who use Election Day voting now predominantly, although there are signs that that is shifting back a little bit and that we won't see as dramatic of a split as we saw in 2020, where basically like everyone voting on Election Day was Republican and everyone voting by mail was a Democrat. And this year, it looks like both methods will be a little more mixed.

This is a pretty tricky question. I'm curious what you'll have to say.

People have talked a lot about how pollsters have changed their methodologies this year and also in 2020 to try to account for undercounting Trump voters in 16 and 20. I'm curious, what would the polls look like today without those adjustments? This may be an impossible question to answer depending on those changes, but if it's a matter of things like changing weighting and assumptions, how do those differ? What would 2024 poll data look like if analyzed by 2020 methods and vice versa?

Mary, do you have thoughts on this? I mean, I would agree with our questioner. This may be an impossible question to answer. We don't always have all the ingredients under the hood for various pollsters. So, for example, one thing that pollsters may have changed between cycles is how they model the likely electorate, how they think about who's likely to vote. And this is going to be influenced by lots of things such as like the past history of different demographic groups of voters, things like that. But we don't usually have access to that information.

over here at 538. So we know there are shifts happening under the hood, and we expect there to be, and we think there should be. And sometimes pollsters will tell us about that, and sometimes they won't. So I'm not sure that we have a good answer to this question. Yeah, Mary, I think those are all really good points. And the reality is that even according to—this is

probably not very satisfactory. But even according to the American Association for Public Opinion Research, we don't know exactly why the polls were wrong in 2020. And so we can't say exactly what kind of shifts would change the outcome. You know, like sometimes waiting by education does make a big difference, but also sometimes it doesn't. And Ann Seltzer, everyone's favorite pollster, oftentimes doesn't wait by education. So, you know, it's really hard to say.

Broadly, we'd probably still be undercounting voters without a college degree, which could, you know, undercount Trump's support. But even that sometimes is questionable because there are other ways to wait that can, you know, make up for things like that. All right. Next question. How does favorability affect the likely outcome of the election or does it?

So to be specific here...

Harris has a net negative a point and a half favorability rating. Trump's is net negative nine advances somewhere similar. And Walls is positive like two and a half. So that's everyone's favorability of everyone who's running for president or vice president. And it is clear that overall, the Democratic ticket seems to have a better favorability rating overall.

Why isn't that more evident in, say, national head-to-head polling? I mean, I think it comes down to like what we said before, partisanship, right, is that, you know, people don't have to –

be in love with the candidate that of the party that they're supporting. And but they think that, you know, the other candidate is worse, for example. I do think that favorability helps. Right. So I wrote an article a long time ago when Biden still was in the race about the double haters and what it meant that both candidates were were disliked. And when obviously when that happens,

It's a question mark because in most cases, when one candidate is liked and the other candidate is disliked, the candidate that is liked ends up winning. But even though Harris is significantly more favorably viewed than Trump, she is still underwater. There are still a decent number of double hitters and those people could break any one of which ways. And some of those people might feel like Biden has really struggled and democratic policies are really letting them down, for example.

Or, you know, there may be sexism and racism at play. You never know. But it's not just as simple as looking at favorability ratings, especially now that partisanship has taken such a strong hold right back in the day. Favorability rating may have mattered more because people really did assess the candidates based on the candidates themselves and personal qualities. And that just isn't really how people make up their minds to vote anymore.

Yeah, to put a finer point on this, in ABC's latest polling with Ipsos, fully 21% of Trump's supporters said they had an unfavorable view of him, which sort of gets at what has been perhaps a strength for Trump in his political career, which is getting people who generally don't like him to still vote for him. And we did a whole...

Audio documentary back in the spring about the beer question and whether it was sort of more important that somebody wanted to have a beer or a glass of wine with you or if they thought you were a strong leader or competent or whatnot. And it turns out that those latter questions end up being more important.

All right, we're going to end on this question, which is fun and nerdy, and it comes from Steven. He's referring to a recent poll from the Cooperative Election Study that had a sample of 48,000 likely voters. Steven says, I think I recall the FiveThirtyEight podcast saying the response rate to polls was around 1%.

So they reached out to 4.8 million people. I need to know more about this method. Smiley face. So what can we tell Stephen about how the cooperative election study with YouGov got access to nearly 50,000 likely voters? First of all, YouGov's panel is

is huge. Globally, YouGov has 24 million people in their panel. So you can expect that in America, which is a significant amount of their surveys, they've got probably at least 10 million people in their panels. Huge, huge, huge, huge, huge panel. Wow.

This has taken over a much longer period of time than a typical survey. The CES, the Cooperative Election Study, is its own sort of special beast that's intended for research purposes. After the election, they make the data available to researchers so people can really dig into what people said before and how that corresponded to their behavior afterwards.

So they're actually doing sort of a special design here. They have a 25-day field period compared to what's typically like three days for most polls. So because they have a much longer period of time, they can reach out to a lot more voters. And the recruitment for the study is not the typical call on the phone, send a text message, and so on. They're sending email recruitments for the most part to members of the panel. Email recruitments

tend to have a slightly higher response rate than phone or text recruitment. So that helps with this response rate issue. And they've designed the sample carefully so they can follow up with multiple emails to the same respondents, which will also choose those response rates. So did they probably reach out to 4.8 million people? I don't think so. Is it possible they reached out to 2 million people? Sure.

All right, so there we have it. Our final, final question on a lighter note was, what is your favorite drink on...

election night. And we're going to assume that this is election night after we have recorded the election night late night podcast after we've shut down the live log or at least paused it for the time being. What are we drinking on election night, Mary? Manhattan. Manhattan. All right. Yeah. And you will be in Manhattan on election night drinking that Manhattan. I'm very excited for you to be here in person. Nathaniel, what about you? I

I'm partial to tiki drinks myself. So, you know, give me a good Mai Tai or painkiller or something. Yeah. Okay. All right. Nathaniel with the tiki drinks. For my part, I'm going to be an image of John Boehner's book cover on election night where I believe he's drinking a glass of red wine and smoking a cigarette. That'll be me after we record our election night podcast. With that, thank you, Nathaniel and Mary.

Thanks, Galen. Thanks, Galen. My name is Galen Druk. Our producers are Shane McKeon and Cameron Chertavian, and our intern is Jayla Everett. You can get in touch by emailing us at podcasts at 538.com. You can also, of course, tweet at us with any questions or comments. If you're a fan of the show, leave us a rating or review in the Apple Podcast Store or tell someone about us. Thanks for listening, and we will see you soon. ABC Election Night. When we fight, we win.

We will soon be a great nation again. Who will be your next president? Election Night. This is where America turns for answers. ABC Election Night Live.