We Answer Your Lingering Questions About 2022
01:09:15 Share

You're a podcast listener, and this is a podcast ad. Reach great listeners like yourself with podcast advertising from Lipson Ads. Choose from hundreds of top podcasts offering host endorsements, or run a reproduced ad like this one across thousands of shows to reach your target audience with Lipson Ads. Go to LipsonAds.com now. That's L-I-B-S-Y-N-Ads.com. When are you back in New York? Depends how the tournament goes. Which tournament is this?

It's the win millions, $15 million guaranteed prize pool. You're guaranteed to win $15 million? No. Guaranteed means to the entire player pool. Yeah, if I were guaranteed $15 million, Galen, I would not be doing this podcast right now. If you got $15 million tomorrow, you wouldn't do the podcast anymore? I do podcasts about poker. But you wouldn't do this podcast? Yes, I do podcasts with you.

You would do a podcast? Okay, okay. See, if I won $15 million tomorrow, I think I'd still do... I think I would keep this job. I mean, with, like, some caveats. You know what I mean? There's no f***ing way I would cover the election. There's no way I'd cover an election ever again.

Hello and welcome to the FiveThirtyEight Politics Podcast. I'm Galen Dweck. Now that the 2022 midterms are fully behind us, we're going to open up the mailbag and answer some of your questions about this entire election cycle, the year, the cycle that is to come in 2024. I wouldn't quite call this a model talk. We don't actually have a live elections model, and even our Georgia Senate runoff average has been put away for the time being.

But it is just Nate and me today. So hello, Nate. How's it going? Hey, Galen. How are you? I'm pretty good. I'm pretty good. Speaking of models, I said we don't have an elections model currently running, but how's our World Cup model doing? You know, I was if hypothetically I were making some bets based on the World Cup model, it would have been doing pretty well until Harry Kane missed that second penalty, even though England lost.

Outplayed France. Nothing personal, though. That was a painful match to watch. England definitely outplayed France. For the entire nation of England and for somebody who perhaps, you know, perhaps had a perfectly legal sports wager on England winning. Tragic all around. Sorry to hear about that for you, Nate. Have you been in touch with your buddy Neymar since Brazil lost?

No, I haven't. I felt bad for him, though. If he had actually been like spending last summer preparing for the World Cup and not playing in a poker tournament, the result might have been different. But then you never would have met him. That's true. Yeah. I mean, met is a little generous. So when you're sitting next to somebody at a poker table, to what extent do you get to know them? As like a medium vert, meaning someone in between an extrovert and introvert, I'm happy to carry on the conversation, but I tend not to like...

initiate a conversation. And also like, um, it was a pretty major tournament and he's on the direct right of mine. And in poker you go clockwise, right? So like the player just to your right and just your left is important. So I'm trying to figure out like how to like take his chips. Cause I'm not sure he's very familiar with this particular type of poker that we're playing. So I'm not, you're not trying to get to know the guy. You're just trying to f*** him over. Yeah.

I'm not trying to f*** them over. I'm trying to play this event that I spent a bunch of money to play and maximize my earnings. Let's move on, though. So before we open up the mailbag, we do have a couple topics to get to today. Can you guess the first item on my list? The independent, newly independent senator from Arizona. Is this the topic?

Ding, ding, ding, ding, ding, ding. So Kyrsten Sinema announced on Friday that she is changing her party identity to independent from Democrat. She has made clear that she is not caucusing with Republicans, meaning that Democrats still have a majority in the current Senate. And it appears that she's going to be keeping her committee assignments as well. Why did Sinema do this, Nate?

Since I know that you have a direct line of communication and she's been talking about her thinking throughout this. Well, let me say one thing, which is I think she's bad at politics. So I'm not necessarily looking for like a deep strategic reason. But the basic reason would be that she would expect to lose a Democratic primary. There is some internal polling by other candidates showing her way behind. Maybe she can somehow eke out a second term as an independent. I think the chances go from zero.

low to low or slim to none or none to slim. We can debate what that means, but like, but it does. Why do you, why do you say that though? So if this is, and we can talk about maybe if there are any other reasons to do this, if this is purely an electoral strategy, why do you think her chances of being reelected as an independent are slim to none? Because, um, her popularity is low across the board. Democrats, Republicans, independents have a pretty middling view of her.

Which can we be more specific about that? What is it? She has like 30 percent support amongst kind of all. She's in like the high, the mid to high 30s across the board, I think. It's not like the center is like Joe Biden or something where like everyone is taking their approval rating all the time. But yeah, the polling I see, I think I had it like in the upper 30s and it was like very even among independents, Democrats and Republicans. Yeah.

But I mean, you know, that's a snapshot in time. We don't know what an actual independent candidacy from Kyrsten Sinema would look like. And like she would at least have an impact on a general election. Right. Like how do you see a general election with a Democrat Kyrsten Sinema and a Republican playing out? I think she'd probably get like six percent of the vote and it might not matter that much. I mean, would she spoil the election?

Maybe it depends. I'm not sure which party. Right. If you if you nominated like a Carrie Lake or a Blake Masters and maybe she provide like a off ramp for moderate Republicans. If Democrats dominated like a really left wing candidate, then maybe she could provide an off ramp for moderate Democrats. But again, I think there's no reason to think that this is a case where she's like splitting the Democratic vote. She's not very popular with with Democrats. Right.

And again, you're going to have a whole campaign cycle where she is positioning herself to the center or the right of whichever the Democratic nominee is. Right. So she'll try to draw those distinctions more clearly. I mean, I'm not even sure she's going to run. She might just retire. Right. And get like a CNN gig or something. Well, then why do this? Because she was going to lose a primary challenge anyway, probably. I mean, with with one or two candidates.

caveats to that, right? I'm trying to look up her Biden score, Joe Biden score. Like the reason why I think she's bad at politics is because like she's not actually that much of a centrist, right? She's voted with Biden 93% of the time in this. Yeah, I guess since she's been in, since Biden's been president, right? Which is much more than you'd expect from a Democrat, not a Democrat, but from a regular politician from Arizona. So she's voting with the party, right?

93% of the time, and yet you deeply alienate that party's entire base by picking stupid fights at stupid times, then you're one of the worst politicians in the country. Mark Kelly, the other Democratic senator in Arizona is, I mean, he probably has like almost the same voting record as Sinema, but like hasn't gone out of his way to like alienate the base of the party. Yeah, so like Mark Kelly, who is not like a capital P progressive, but like a normal mainstream Democrat, but

Voted with Joe Biden 94.8% of the time. Sinema is at, what did we say, 93.2 or something like that? It's kind of amazing that she's become this unpopular with, for the most part, a fairly normal voting record. But the thing is, is that voting record slightly...

misrepresentative of the role that she's played because oftentimes something isn't put to a vote until Kyrsten Sinema and Joe Manchin signed off on it. And so where she might have voted no on certain things, she ended up holding out and then getting her way. I think that's the role that she's played that has pissed Democrats off to some extent.

You want the super hot take? Maybe some of them may have done Democrats a favor. I'm not sure if Democrats had spent a bunch more money and would turn out to be an inflationary environment. I think that it made life more difficult for the party last month.

OK, so lastly, wrapping up on Sinema, is there anything that Democrats could do to punish her? Is that would they actually would that be to their benefit? Would they want to potentially punish her? I have heard some chatter about the response to this being, you know, trying to appoint her to some ambassadorship, leaving a vacancy in the Senate that then Katie Hobbs could appoint a Democrat to.

But where's the fallout here for the actual functioning of the Senate and the Democrats majority? I mean, it makes her so I guess she's still caucusing with Democrats were because even without her, they have a 50-49 majority. It doesn't really matter that much. You know, she might be more likely to vote against the party on key issues if she's trying to differentiate herself, right? In a f***ed up way, right?

If they already have 50 votes, meaning Joe Manchin, the Democrats might not care about that because if anything, they want her to like alienate left-wing voters. So she doesn't cannibalize from, from the democratic candidate in, in next year's election, sorry, in two years from now. Um, so I don't think it's like actually all that consequential really. Now maybe it gives them a little bit less wiggle room. If Joe Manchin says, now I also have to let go and, and, um,

take care of my reelection, right? Joe Manchin would probably have voted against Democrats as often as possible within reason, right? Except maybe he had some infrastructure stuff. So, you know, it probably makes for a moderately more challenging vote on some issues, which only affects things if Joe Manchin is also choosing to be a challenge. But it's really not that much different than like the status quo, except now you need both Manchin and Sinema to oppose something as opposed to just one of them.

And also, of course, Democrats don't control the House. So they're going to be looking at the consequences of passing legislation in the Senate have decreased anyway. I want to talk about the Senate more broadly over the coming two years, but also over the coming decade, which you recently wrote about.

You're a podcast listener, and this is a podcast ad. Reach great listeners like yourself with podcast advertising from Lipson Ads. Choose from hundreds of top podcasts offering host endorsements, or run a reproduced ad like this one across thousands of shows to reach your target audience with Lipson Ads. Go to LipsonAds.com now. That's L-I-B-S-Y-N-Ads.com.

You wrote a piece on the FiveThirtyEight website wondering whether the next two years might be the last time that Democrats hold the Senate for the entire coming decade. And it's something that's been suggested because of Democrats' poor performance in small, rural states that are disproportionately represented in the chamber.

Having done sort of the back of the envelope analysis and the process of writing this piece, how likely do you think it is that these two years are the last time until the 2030s that Democrats control the Senate? Maybe 50-50 or 60-40. That's the last time until the 2030s. The big problem is for Democrats with this next election in 2024 is

And the incumbent Democratic senators in West Virginia, meaning Joe Manchin, Montana, John Tester and Ohio Sherrod Brown. Those are all now quite red states. Those candidates won reelection last in 2018, which was a very good Democratic year. There's a little bit more ticket splitting in presidential years. So, you know, you go into 2024 with them being, I think, toss ups at best. Yeah.

You figure maybe you lose two of those three and then maybe you lose lots of other traditional swing states. Arizona is now more complicated based with cinema, although it was tenuous before, right? Nevada always has close elections. So, you know, if I had to guess how many Democratic Sanders Democrats have after 2024, the answer might be something like 48 as opposed to 40.

And there are also very few Democratic pickup opportunities. They will try again in Texas to meet Ted Cruz. Florida used to be a swing state. Rick Scott is up for reelection. He won by a very, very small margin last time, but Florida has become quite red. You may have Ron DeSantis. And perhaps with either Ron DeSantis or Donald Trump at the top of the ticket, two Floridians might make it hard. Well, you got one real Floridian and one...

I mean, what is more Floridian than a New Yorker who moves to Florida? Fair enough. Miami Beach is the sixth borough. Yeah, I mean, you know, you don't have to tell me. But, okay, so...

Democrats are looking at a situation where this may be the last. They don't control the House anyway, but this may be the last two years where they control the Senate, which means getting nominees through. This has led to some conversation about whether some current liberal Supreme Court justices should go ahead and retire because they could be looking at a decade of not being able to be replaced by Democrats. They're not that old, though. I mean, we're talking about Sonia Sotomayor, Elena Kagan.

Is there precedent for justices stepping down this early in their tenures for this kind of reason? I mean, the age of retirement has gotten later and later over time in a way that kind of tracks with the age of average lifespan. So, you know, the heuristic is probably something like they tend to retire if they think they have about 10 more years of life expectancy left and they want to enjoy their final years not necessarily having...

the job on the court. There's obviously a long history of like strategic partisan retirements. What's different now is that like, I mean, the parties have gotten very effective at not having surprise nominees who turn out to be different than they expected. As you have more partisanship, I mean, and more polarization, the outcome of elections is in some sense,

more predictable, right? We can predictably say that this is a threat to Democrats. We're not just saying, oh, they're randomly going to be 33 Senate races every year. Who knows? There's some predictability. And so, yeah, she has a reasonably strong strategic reason to retire, right? I mean, it really is like if you take someone who's 60, so do Mayer. Like clearly, if you're like moneyballing this, it would clearly be very much in the best interest of

of like liberal causes for her to retire, even though, and be replaced by someone 20 or 25 years younger, right? Even though there is a chance that Democrats can keep the Senate, even though, you know, women live longer than men and healthcare is better than it used to be, right? So she has a decently long like life expectancy, right? Just the chance of whichever party

non-zero chance, right? Small but non-zero. Then, oh, and the other thing too is like, and now you also need not just the Senate, excuse me, not just the presidency, but maybe also the Senate, right? I still don't know what happens the next time you have a opposite party domination where Democrats control the presidency and the GOP controls the Senate or frankly, vice versa. Like, I don't know what happens when that occurs, but it's no longer the case. I have a wild guess.

Yeah. Well, okay. But yeah, but... So it is kind of crazy that this is where these conversations get to surrounding the Supreme Court, but also for anybody who pays attention to this stuff, understandable. There is another question here about...

which is what are the circumstances under which Democrats are able to continue to win a majority in the Senate, right? I mean, in some ways, it's hard to predict the direction that states will go politically. We do have a lot of evidence that they don't do well in small rural states, but is there anything that they could do or that could just happen in the country that would reverse that trend in a way that would make these considerations less for the party?

Yeah, I mean the reason why this goes six years out in this exercise – like that's I think about the longest time horizon that's useful versus a baseline of, oh, it's just kind of random. I mean there are different things that could help Democrats math a little bit. One is that –

There are some states, you know, Alaska elected a Democratic member of the House, Mary Peltola, who I assume will run for Senate at some point. That's how it tends to work in Alaska. You know, even a state like Utah is changing pretty fast. Mike Lee only beat Evan McMullin by 10 or 11 points, right? You know, so those two are interesting, potentially, if you're talking about like a six or eight year time horizon. Yeah.

Republicans, meanwhile, might develop a little bit of a Florida problem of their own. So let me explain. There are certain really big blue states that are vote sinks for Democrats, that California and New York are very far to the left of the US as a whole and also very populous. And so

That gives Democrats lots of votes overall in the race for Congress that only translate into two senators each. Florida is a very large state that's becoming very red. And so that might counteract some of that in terms of how like the median state kind of lines up with the popular vote, right? Yeah, I mean DeSantis won by a much wider margin than Hochul did in the most recent governor's race. Yeah.

And it's up to, you know, Democrats have done well in New England, which are a bunch of fairly small states, very small in some cases. If that ever changed, then they'd be even more screwed in the Senate. But like but again, I think like beyond like these next couple of cycles, there's not a ton of predictability in the Electoral College, for example, in 2012.

Obama did even better at the Electoral College and you guessed based on his share of the popular vote in 2012 and 2008. In 2004, John Kerry was not that far behind from winning Ohio, which would have allowed him, the Democrat, to win the Electoral College but not the popular vote. So these things can shift around a fair bit. Also like one other thing too is like Democrats' coalition is becoming whiter. Yeah.

They have lost ground with Hispanic voters and to a lesser degree, black and Asian-American voters. And so, you know, so that could shift things, right? Like the map that Democrats had last month was full of relatively strong performances in the upper Midwest. So that's that's interesting that would reflect a reversal of of the trends kind of circuit

2012 through 2018 or so. So on a related note, a related note to the Senate, but also potentially having an impact on on every election, Politico published a piece last week by Justin Guest titled How Demographic Shifts Fueled by COVID Delivered Midterm Wins for Democrats. And this was the main crux of the argument. And

Data from the U.S. Postal Service and Census Bureau shows that the pandemic drove urban professionals who were able to work remotely disproportionately Democrats out of coastal progressive cities to seek more space or recreational amenities in the nation's suburbs and Sunbelt. This moved liberals out of electoral districts where Democrats reliably won by large margins into many purple regions that had the potential to swing with just small changes to the map.

And because partisan gerrymanderers carved up new districts before the extent of pandemic mobility data was understood, they could not neutralize the population shifts. And pollsters were none the wiser. I mean, we can leave out that last part for a second, but is this a good or bad use of data? That seems plausible. I would expect the effects to be kind of medium-sized instead of large. I mean, it also is a question of like,

Who is moving out? Right. Like, I don't know that many people who moved to Florida, but like but, you know, the Democrats I know who moved to Florida are not the most liberal Democrats. Right. The people were selecting out. Well, that's a very, very specific kind of move. Right. I mean, there were also people who moved to, for example, California.

Upstate New York from New York City, from Brook, upstate New York from Brooklyn or people who moved to, you know, the Phoenix area for just lots more space and job opportunities or whatever. Like, I think the Florida narrative is a very specific one out of all of the different kinds of pandemic migration we saw.

Yeah, I mean, in theory, a place like New York, where you have a bunch of swing districts upstate and very, very, very blue city would be a good example, except Democrats got like crushed in New York. But in the long run, something like that could help Democrats for sure. Okay, I've got one more data set I want to bring up before we move on to listener questions. So

Today, FiveThirtyEight published its final round of polling with Ipsos, and we've been asking the same 2,000 people about what's motivating them politically throughout the year, mostly before the midterms, but this last wave asked after the midterms. We found some interesting but somewhat predictable things, like actual voters cared significantly more about abortion as an issue than adults in general, which is to say people who didn't vote.

Most Americans, 70%, trust the results of the election, according to this polling. We also asked people if they plan to vote in the 2024 GOP primary and then asked those people who they would like the 2024 nominee to be. Nate, if you haven't already seen this polling, how do you think Trump and DeSantis did? Um...

I mean, most of the polling, I haven't seen this polling. I mean, DeSantis has been holding up pretty well in polling that I've seen recently of the GOP nomination. You want to wager a margin there? DeSantis by seven. DeSantis by seven. OK. 42% said DeSantis. 24% said Trump. 5% said Pence. And everyone else was lower than 5%.

What do you make of that poll? I mean, it's one poll. That's what you make of it, first and foremost. But what's going on here? I think I don't want to sound too like like horse racy or both sides. Right. But like Trump has had like a not very good month. Right. He kind of announced his reelection after having this kind of disastrous midterm and then hasn't really been seen or heard from much except for having dinner with Kanye West and Nick Fuentes, who are.

respectively, an anti-Semite and a white supremacist. And like, that's kind of been the only story really. And like, like, I think, I don't know. I mean, again, I'm looking at the polls and like people seem like, and you also see again, like this is a little bit more subjective anecdotal. It seems like GOP elites, by which they don't necessarily mean like the cabal of like oldie timey people in DC, but just like influential conservative and Republican people are more willing to. Yeah. Like Ben Shapiro.

Yeah, I mean, Trump kind of looks like a loser and people haven't been willing. A lot of the reason he gained oxygen was with the stop the steal stuff. And like, it seems like collectively Republicans have decided that we don't want to emphasize a stop the steal. I mean, I think Carrie Lake is still going on about the election being stolen in Arizona. And like, she is not getting that much oxygen from like other Republicans or like members of the media. I mean, I have a whole rant about like

the left's extremely inconsistent views about like platforming things, but that's kind of neither here nor there. But like, but like, how does this poll though, this margin that is, you know, 18 points compare with the other polling that we're seeing of this kind? So these polls are sensitive to, uh, what voter universe they define, right? Um,

Is it like the universe of quote unquote likely GOP primary voters or is it universe of like all Republican adults who who are more in the strain of likely? Right. Because we take all all adults. Then we narrow it down to voters. Then we ask, you know, do you plan on voting in the GOP primary? And amongst those people we ask, we ask this question and that's how we got this result. But how does it compare to others? Yeah.

Yeah. In general, when you have these polls of likely voters, the narrower voter universes, DeSantis is often ahead. When you have the polls of like all adults who are GOP-leaning, then Trump is often ahead. But you'd rather – this is not equal. The primary turnout is comparatively low. You'd much rather be ahead in the narrower voter universes. Like so in the Marquette University law school poll, for example –

DeSantis led by 15 among all adults, but 20 among registered voters. So just kind of getting rid of the GOP identifiers who are not registered to vote benefits DeSantis a fair bit. I mean, what I would say is like Trump is now behind in the polls is what I would say. I think we had like this segment like a month ago. It was like, oh, now it's kind of tied. Yeah. I think Trump is behind on average if we had some type of average and also like –

he has less or more name recognition. Usually you're supposed to start out if you're the former president with like an advantage just because you're like the familiar default name. So I think Trump is, um, is not loving, uh, the rise of Ronnie D. When are we going to start? When are we going to have that average? Galen, don't you, man? It's just, it's diving. Okay. Let's answer some listener questions. Um,

First and foremost, from Catherine, what does 538 use for predicting outcomes for runoff elections? How does it differ from other models? This is, of course, a reference to the runoff that was just completed last week in Georgia. We didn't have a forecast for that election. We had an average. But, of course, we did forecast that election during the regular voting. So was our forecast back then a forecast of...

all of the possibilities, including a runoff or not? Yeah, the forecast explicitly considers the possibility of a runoff. It actually simulates out the rest of the election, including if there is like polling of different runoff matchup. So you have had elections in the past where there were like two Republicans against one Democrat and then maybe some minor candidates, right? And so if there was polling of what would a head-to-head race look like

we incorporate that in the model as well. That was kind of a moot point this time because it was just like an independent dropping out, but no, we put a fair amount of work into, into the runoff modeling. We're flying a little bit more by night in Alaska where, where they have a, not what's an instant runoff as opposed to a runoff a month later, like in Georgia. Yeah.

The pre-election forecast explicitly accounts for the possibility of a runoff. Okay. Next, from our friend, the Virginia contrarian, what does the polling error in 2022 tell us about the reasons for polling error in 2020? For example, is 2020 polling error more attributable to COVID and or Trump being on the ballot since the polls did well slash slightly underestimated Democrats in 2022?

Yeah, I mean those hypotheses gain credence, right? I haven't personally done my deep, deep, deep, deep dive into polling accuracy yet. But those hypotheses gain credence, number one, and also like – I mean it may also have been for all the like – I mean people probably know that my view of polling is a little bit more that it's not really a pure science anymore. It's partly like people –

Making polling flavored predictions and it kind of is merging more with other ways of forecasting the election. And so like I just think there's probably like a lot more like kind of consciousness this time around from pollsters that, hey, what we were doing last time around didn't really work. We have to try something different or maybe we're more careful about like which kind of polls we release. But I still want to see that because like a world in which the polls were.

I don't know to what extent that is true. And we should talk about this more in the new year that polling is less a sort of standard scientific method and more pollsters trying to sort of use the art of the craft to poll.

figure out which way the electorate is headed. I mean, that would be disappointing to me. I think, you know, like Nate Cohn at the New York Times is doing a lot of work to try to figure out a way to poll people where you don't have to do that. And I don't know the extent to which other pollsters like the Emersons or Marists or what have you of the world are doing that. But to me, that would be sad, right? Like, I think we want to have a way to ask Americans how they feel about things

and just get an answer and not have to like around. I'm sorry, can you reframe that? I'm a little confused by what? Like you're saying that pollsters are becoming forecasters in a way themselves because they have to try to figure out where the electorate is headed and where they might have some error and make up for it in their numbers instead of just having a standard procedure for

that samples, that uses the scientific method to sample the American public and then just publishes those numbers because those numbers are too shitty to actually publish. That would be, to me, that seems to me like a precarious road to go down. Well, it seemed to be better than 2020. But are we sure that that's actually what's happening? I don't think that pollsters are like, hmm, well, we really screwed the pooch in 2016 and 2020. Now we're just going to throw in the towel and become forecasters ourselves.

I think that – I don't think that's – that's not what I'm – see, that is like a mischaracterization, right? What I'm saying is like – But then I don't really know what you're saying. There is not a reliable gold standard way to do a poll anymore. If you do that gold standard, you're probably going to have some predictable biases. You may on average be biased toward Democrats. I'm not sure. What I'm saying is I think the kind of average pollster – because it is a polling average, right? I think the average pollster is probably saying let's be a little bit more –

practical, pragmatic, empirical about kind of what works and what doesn't. The average includes, I'm sure, pollsters who are like, hey, let's do the same thing we've always been doing. Wait, but what are you talking about when you say, let's be more practical, pragmatic, empirical? Like, in what way? Like, what are you saying they're doing now that they wouldn't have done before? I mean, how they might wait for different variables, right? If you wait, for example, for...

Party if you wait by past vote right in theory waiting by past vote is Problematic in part people don't always recall their past vote correctly, right waiting by party ID is problematic because Party ID is an attitude and not like a permanent demographic variable So techniques like those might might be more common than they were before So things that we're familiar with you're not talking about like a

sort of actively putting their finger on the scale saying, let's mess with these numbers to make sure Republicans do a little bit better because we think that we have been under. That probably happens too. That probably happens too. Or like what they publish and what they don't. So whenever higher rated pollsters like Monmouth University did a poll of Pennsylvania,

in the closing days of the campaign, right, where they framed things in such a way that it was like not actually a head-to-head poll between Emmett Oz and John Fetterman, even though you could infer from it, because they were just scared. They were scared. They didn't want to be criticized again. They didn't want their poll showing up in the 538 average. If they hadn't been scared, their poll would have been almost exactly perfect. So that's a clear example of like a pollster with reputational risk chickening out, right? That's lame.

Okay, I think we will talk about this more in the new year. When are we going to do our new pollster ratings? Galen. You're like, stop it. Galen is. But come on. Come on, man. Come on, man. I'm just asking questions. I'm just asking questions. I've got 13 more days to go into Christmas. I'm here in Las Vegas for five of them. I'm not sure we're going to get pollster ratings before January. First quarter of next year?

Oh, yeah, sure. Oh, OK. Fantastic. I mean, that's, you know, that's soon enough for me. OK, next question. Guillermo asks, did the McCaskill strategy of picking their rivals backfire in any place for the Democrats? I know the answer to this question. It's no, I don't think it did backfire anywhere. Yeah.

It may be. What was the strategy? I'm sorry. So what Guillermo was talking about is that Democrats during the Republican primary spent money to boost the more extreme candidate who they thought was less electable in place, like, for example, Don Bolduc or, for example, the governor's race in Pennsylvania with my my election brain. I've already lost it. I'll blame it on covid. But who was the Republican candidate for governor in Pennsylvania?

Uh, Mastriano, Doug Mastriano. Yeah. So like Doug Mastriano, Don Bolduc, some Republican candidates in House districts like Peter Myers, previous district, you know, his opponent, they helped boost. Did that strategy of sort of boosting the more extreme candidate, even when that extreme candidate was sort of counter to Democrats own purported ideals of supporting democracy, etc., etc.,

Did it backfire anyway? And I would say no. We know electorally at this point that it didn't. From a message perspective, you know, ousting people like Peter Meyer from Congress after he voted to impeach Donald Trump and was a sort of powerful messenger amongst Republicans for these Democratic ideals, I don't know if in the long term strategically that was a great idea, but it didn't electorally in the near term backfire. Yeah. Yeah.

But, you know, maybe it damaged the brand in some intangible way. But at least...

The first part of it, I mean, the idea is that it helps you win elections and they win those elections. But it's, I mean, it's sort of funny. In a world where Republicans are going to have a majority in the House, Peter Meyer is more helpful in office to Democrats than that extra Democratic vote in the minority, I would argue. Right? Like having a Republican in the slim Republican majority who is going to sort of like make anti-Trump arguments and things like that

Getting a lot of press attention. I I agree So I think it did backfire in that sense, but not in the electoral. It's not it's not easily measurable It's not easy, but the strategy looks better than it would have if Don Bull was the senator from head one anyway Yeah, we got lots of questions. So I want to get to a few more. I

You're a podcast listener, and this is a podcast ad. Reach great listeners like yourself with podcast advertising from Lipson Ads. Choose from hundreds of top podcasts offering host endorsements, or run a reproduced ad like this one across thousands of shows to reach your target audience with Lipson Ads. Go to LipsonAds.com now. That's L-I-B-S-Y-N-Ads.com.

You're a podcast listener, and this is a podcast ad. Reach great listeners like yourself with podcast advertising from Lipson Ads. Choose from hundreds of top podcasts offering host endorsements, or run a reproduced ad like this one across thousands of shows to reach your target audience with Lipson Ads. Go to LipsonAds.com now. That's L-I-B-S-Y-N-Ads.com.

All right, continuing along with our listener questions, Joe asks, have you ever looked at what day the model is most correct across the years? Like, is the day before the election the most accurate week, month? You talk a lot about the hurting that happens at the end. So I wondered if an earlier endpoint is more accurate. I think this year it was. I think like this year, actually, like three weeks before or like five or six weeks before was like more accurate in the election day model, at least in terms of like

the overall calibration, right? Maybe individual races, you have more information. But no, in the long run, I mean, election day should always be the most accurate forecast. This probably matters less for races for Congress than like other types of races because they're like a bit more slow moving. It's not like there's like one dramatic scandal that can affect every race. It might affect one race, but not every race. Yeah, maybe pollsters are getting more hurdy. I don't know, right? Yeah, I don't think there's a lot to be added

certainly in like polling the past week, which is just everyone kind of covering their asses for the most part. Maria asks, is there any way to tell how redistricting may have impacted the results? I think this is something of a debate over whether redistricting, because the House majority ended up being so small, could have sort of flipped it towards Democrats. I mean, certainly Republicans did not do particularly well in terms of converting votes to seats, right? They are going to win the popular vote by, um,

two or three percentage points, I think, for the House. And they just barely won a majority, suggesting that if you had a neutral popular vote nationally, the Democrats would probably have won some of these close races and actually would have kept their majority. How much of that is about new districts versus kind of subtle demographic shifts in the coalition is not quite as clear to me. Again, if you have like, if you have voters of color

in cities that are very, very, very blue actually becoming more Republican in some cases and trading those with like suburban voters. That's a very good trade for Democrats electorally. And that's sort of the trade they seem to have had in the past election cycle or two. So yeah. But what would the election have looked like with 2020 districts instead of 2022 districts? I haven't looked at that myself.

I think Nathaniel came to the conclusion, Nathaniel Rakich came to the conclusion that it would not have changed the outcome in terms of party control of the House. I think one thing that may have changed the outcome, though, is if you had had Democrats want to hire a share of court rulings in different states, right, like in New York, that may have let them keep the House. But despite the fact that, like, the GOP controlled the redistricting process in more places, I don't think you can say the map is, like,

So empirically, I'm not sure you could say the map is like biased against Democrats. It might even have a slight bias against Republicans just because of the way now that voters have chosen to distribute themselves between districts. I think you mentioned earlier where like there was also like allocations were done based on a census that did not assume there had been much movement due to COVID when there was a decent amount of it. So there were a lot of little things that wound up helping Democrats. Yeah.

Next question. Spondina asks MVP of the election season. If it is even possible to single out one candidate, if not, which voting block or demographic has been the most influential in this election? So MVP of the election season, who overperformed the baseline of their district or state the most in 2022? Uh, I mean, I, you know, correction Whitmer, um,

Josh Shapiro, Ron DeSantis, those seem like pretty good answers potentially. So Gretchen Whitmer won by about 10 points in Michigan. Ron DeSantis won by almost 20 points in Florida. I don't remember the margin in Pennsylvania, but I'm pretty sure it was double digits. Does a 10 point margin in Michigan, does it trump or come close to a 20 point margin in Florida? Yeah.

Like who had a better overperformance, DeSantis or Whitmer? I guess you'd say 20 is bigger than 10, right? I mean, I'm not quite sure what the baseline is in each state. Marco Rubio also won by a lot against a pretty good Democratic candidate in Val Demings. So, you know, YOLO Florida seems to love Republicans now, or I think actually there's evidence that like Democrats are just like not turning out, right? They've kind of given up on

on competitive races in Florida. Wait, so that brings us to the next question from Sreenu, which is, so has Democratic overperformance nothing to do with low Republican turnout, as Nate Cohn and others already pointed out? Did Dems still have a lower turnout but are saved by some Republicans and independents voting for them? Okay, this is a somewhat convoluted way of saying,

According to the election analysis after the fact from Nate Cohn and others, they found that there was still some drop off in terms of Democratic turnout in this midterm cycle. But what really saved them in these competitive races is that Republicans were turning out at high numbers. And then a lot of Republicans or even sort of Republican leaning independents were voting for Democrats in key races because of poor candidate quality. And that's sort of what made the difference in a lot of these races.

Yeah, I basically buy that theory. I think it's pretty consistent across...

different indicators that the electorate was not particularly democratic, although it varied a lot state by state in case where you maybe had an issue like abortion more directly affecting voters. It may have mattered. Um, meanwhile, in states like New York and Florida, you had a very red turnout. And so that helped propel the many GOP waves in those two states. Um, also states maybe where people didn't feel like the Republican candidate at the top of the ticket was radical. Right. Um,

But no, I think the basic story of the election is that Democratic turnout was decent to okay. Republican turnout was better than decent, but not great. And independents really sided with Democrats. Jim asks, how did young voters play a role in this election? Was participation up? Did they make the difference in close races? I haven't looked at that in detail. I mean, clearly, like, you know, you had like crazy, like,

The vote on some college campuses was like 97 to 3 for Democrats, which even for college campuses is a pretty big margin. Well, you say that. I have looked into this a little bit. So the turnout was not as high as in 2018, which was a record high turnout for young people, people under the age of 30. It was –

still high, but not quite as high. And record turnout for young people in a midterm is like 30 something percent, right? It's not super high.

Here's what I also see from the AP in terms of their vote choice in this election. Voters under 30 went 53% for Democratic House candidates compared with only 41% for GOP candidates nationwide this cycle. That's down compared with 2020 when such voters supported Biden over Trump, 61% to 36%. That's comparing a midterm to a general election, so maybe not really fair. But I

I think that I don't really understand what the desire is, but I have get the sense from different segments that I've seen or stories that I've seen written or whatever, that there's this like strong desire to say that like, oh, this one portion of the electorate like did it for this party or that party. And that like young voters are a piece of this argument where people want to say, well, young voters, young voters, young voters are like, you know, winning it for Democrats.

Young voters don't vote that much. They do prefer Democrats when they vote, historically by very wide margins, but electoral outcomes don't happen like that. And also, even if they did, even if you were going to say, oh, but there was an overperformance amongst this group, that group, whatever, it's not going to happen.

In this election cycle, there wasn't some crazy high turnout amongst young voters and there wasn't some crazy large margin by which they won, Democrats won them. So if you do want to make that argument, young people aren't the voters with which you want to make that argument. I say that as an almost young voter. It's a little bit of the how do you do fellow kids meme. What's that? You know, like the how do you do fellow kids meme, right? Which is Steve Buscemi, I think. That's what it's like. If you're like a,

54-year-old resistance Dem, right? You're going to get all these likes if you're like, it was the youth...

The youth of America propelled. Well, the young people are saying it too. Like David Hogg from the very sad Parkland shooting. But like he was like, oh, young. I saw his commentary saying like young people did it for Democrats or whatever. It's just not how this stuff works. And all of these people were making these arguments before we even had the data, actually. So people just want to people just want to seem smart. They don't want to be smart. Saying like it was the youth of America. Right. That seems smart at a cocktail party.

Would you rather seem smart and be stupid or seem dumb and be smart? I just care about being right, man. I don't give a shit what stupid people think. Would you rather be hot or smart? Hmm. Toss up. Toss up. Andrew asks,

There didn't seem to be much of a punishment for Democrats in regards to extensive COVID restrictions. Is this because they started backing off them around March of 2022? If they hadn't, would that have led to a worse result for them? Yeah, I mean, so 2021, Democrats had this very bad interim election in Virginia and New Jersey. And you can make a decent argument that those were caused in part by COVID and some educational policy issues that were tied in.

directly and indirectly to COVID. I mean, by the time elections were held, there hasn't been a lockdown anywhere in the United States, probably some weird liberal arts campuses for like a year and a half, right? The government got rid of the mask mandate. If you ask people what's the most important issue, like only 1% say COVID, even though you still have quite a bit of disease and a fair amount of death still. But no, I think the combination of the Omicron-related variants being

less deadly people being burned out on lockdowns and Democrats kind of giving up on like trying to enforce for better or worse any COVID safety measures. I mean, it was kind of like kind of off the radar as an issue by 2022. I mean, I do think it has some like local effects in like Florida maybe, right? The gambit of being anti-COVID restrictions, I think, A, created kind of a brand for Ron DeSantis and like also like literally encouraged people

people who were against COVID restrictions to like move to Florida. But no, it kind of became, I mean, I know, I mean, I also think that like a lot of, I think a lot of stuff is downstream from COVID, right? Like, you know, I think basically everything that happened in like 2020 and 2021 from like crypto to like stop the steal to the George Floyd protests, which in some cases, you

um, it was spilled over. I think it's all related to COVID, this hugely traumatic thing that everyone experienced. Um, but it wasn't like you were really like seeing like

specific COVID policies being litigated that much. Okay, so I have one final question on the 2022 cycle. And then we do have a few questions casting it forward a bit. A listener writes, Nate, are you going to apologize to Galen for getting mad at him for asking whether this was a unique midterm election? Let me add some context here. So I described the 2022 cycle as unique.

Because it was an environment in which this was like weeks or months before the election. It was an environment in which the president was clearly unpopular, right? 42% approval rating, significantly lower than Bush in 2002 or Clinton in 1998, the asterisk elections where the party actually gained at the midterm. So it was clear that the sitting president was

President Biden was unpopular, but his party was still doing pretty well in the polling, nonetheless. And I suggested that this was a unique circumstance because of that. And you did not like my use of the adjective unique. I don't know if you remember this. But anyway, are you going to apologize? You know, I don't remember, but I'm sorry, Galen, for everything I've ever said and done. Oh, oh, my God. Just one blanket, one blanket apology. Yeah.

To you, not to everyone. Oh, okay, okay, okay. Oh, my God. I mean, I feel so special. Would you now condone the use of the adjective unique to describe the 2022 midterms? I mean, every election is some special snowflake, as are you. As are you. No, nothing, nothing. It was a weird election. It was a weird election. Okay. I'll give you that. I gotcha. I gotcha. I gotcha on the record. That's a feather in my cap.

Moving on just briefly to 2024, because we're going to have two years to talk about this, you know, when it actually in 2023 and in 2024 itself. So Chris asks, are we really starting with just six swing states for the 2024 presidential? And I think the six that they're talking about are Wisconsin, Michigan, Pennsylvania, Georgia, Arizona, Nevada. I mean, you know, North Carolina could...

I, I'm all for broadening out the map. I would say, say, I mean, you know, who, who knows, but like include Minnesota in that those elections have been pretty close, even though Democrats have won reliably. Why not still include Florida? We don't have enough data to say that like those margins can't be close. I would say, you know, what about New Mexico? Florida went Republic? Maybe, maybe. What about Alaska? Maybe Alaska. Oregon. Oregon.

I would say keep an open mind until you have evidence that tells you you shouldn't. Yeah, I think that's correct. I think that's correct. And like, remember, we had like the same GOP nominee two times in a row. Maybe were they including New Hampshire, but not Nevada? New Hampshire is definitely a competitor. I don't know. Maybe they weren't including New Hampshire. Yeah, it's just, yeah. Yeah, I mean, New Hampshire is like only like four electoral votes, right? Nebraska, Nebraska's first congressional district. Second congressional district.

I mean, that is quite blue at this point. Yeah. So maybe it maybe that's not competitive. OK, next question. How would changing the order of the Democratic primaries, if it happened, be reflected in the primary model? Would any early state factor be split among extra states moving up to the front? Have you thought about this?

I've thought about having to think about it, but I haven't processed it yet. I mean, you know, before it was safe to assume, I mean, our assumption was, and this is all like a little speculative, right? That Iowa has a bigger impact and that diminishes over the course of the first four states. I don't know what assumption we would make now. I mean, my guess is we kind of say that like the first states matter a little bit more, but it's spread out a little bit more evenly potentially. But the fact is that it's all this kind of, you know,

coordination mechanism among parties and the media, right? If everyone agrees it's really important to win state X, then state X takes on greater importance. I mean, I think you can argue that people already were becoming a little bit more skeptical of Iowa. They kind of knew that Iowa was like non-representative. Well, and not announcing a winner also sort of made the impact of Iowa a

Not possible because we didn't know who the winner was. Okay, so we'll talk about that more next year. Next question. There has been an amazing amount of advancement in AI this year, especially with interpreting information from huge and varied databases. If similar progress continues, would FiveThirtyEight continue researching if any of this tech could be applicable to modeling? We're going to automate you out of a job, Nate. So I think these models are not very good at like

mathematical reasoning. Maybe there'll be separate models someday that are. But they're good at like reading all the texts produced on the internet and kind of like giving you a coherent human-like pastiche mix-up of those texts. That's a very different skill set than like designing a model from first principles, especially in cases where like where precision is important, right? Like AI technologies are good at giving you like an approximately correct

in more and more cases, right? It's not going to give you the most uncannily correct answer that requires like true expertise. And in fact, these AIs like kind of like lie and gaslight a lot to make themselves seem more confident and more correct than they are. Again, this is all drawn from

records of human discourse and communication, right? So humans are kind of lie, cheat and steal and are political about their answers, um, and are inconsistent and are overconfident, um, and are predictable biases along demographic and other lines. Um, so all that's replicated by the, by the AI. So if, um, this artificial intelligence were to listen to every podcast we've ever published, um,

How do you... Right, the AI could create like a fairly... How would that go? AI could create like a... Maybe the AI could replace the podcast. Yeah, automate me out of my job. You get AI Nate. Yeah, podcast shit. Oh my God, we're gonna... Let's have AI Nate on the podcast.

sometime when you're not available. We're just going to plug in all the questions and see how it responds. You guys can have AI Galen sometime when I'm not in the country. I really don't want to be confronted with that. Thank you very much. Next question. Yeah, but this stuff is pretty... By the way, we've had this AI development and these claims that...

there's now, is it cold fusion or I don't know if it's cold or whatever, but like, like the tech press should be covering this. This AI is like a hundred times more important to the future than anything involving Twitter. Um,

So if you're like a tech reporter, cover AI. You don't have to tell me. Honestly, Twitter is unusable now that the only discourse on Twitter is about like the ownership of Twitter. I mean, I already don't like Twitter, but it's now literally unreadable. And like, I'm not saying that it's because like Elon Musk changed my feed or whatever. I'm just saying because the people that I usually follow are now all just talking about Twitter itself.

Got to talk about something else. I don't really care about Twitter. Isn't this boring at some point? It's so boring. I'm so bored. We have to talk about something else. Believe me, if you are someone I follow, I thoroughly have registered your take on Elon, right? It's fucking boring. One way or another, I don't care. I don't care. Okay, anyway. Raid says, I have never been called by a pollster. What are the odds of that happening at least once for an American during their lifetime? I like this question.

This is the kind of question that I should have a precise answer to, but like... Yeah, come on. Are you even Nate Silver? Don't. I mean, let's say... Okay, let's do a little... Let's bring up the Google calculator. We'll do a little estimation as you might... Oh, my God. Wait. Can we start a new segment called Math Live where I ask you statistical questions? Live math. Live math. And I ask you statistical questions and you have to sort of like work your way through them live on the air. Let's say there are 500 polls conducted in election cycle...

that interview 1,000 people on average each. That's 500,000. But some of those aren't calling people. They're actually just interviewing people online. So let's say half those are phone polls, 250,000. So that's like not actually that many, man. Like 250,000 times like, let's say there's an election every two years. Let's say you're, I know how old our correspondent is, right? Let's say you're 50 years old and therefore when you're 18, what is that? That's 32 times 16, right?

Yeah, so it's pretty unlikely that you'll be polled, actually, by a horse race poll. I think the chance for one in... Wow, yeah, it's quite low, if you think about it, man. Wait, what is it? This doesn't have to be your final, final answer. I know this was back of the envelope, but what does it look like to you? Am I doing that? Let me double check that, right? So again, let's talk about assumptions here. So let's say 500 polls, of which half are... Well, I guess it's been more than half historically. Let's say...

75% are done by phone and they talk to 1,000 people each. So every cycle, 375,000 people are polled, right? Let's say that you have a- Wait, hold on. 1,000 people respond, but they contact a lot more people who don't actually take the poll. So are we including here people who just have been called by a pollster, period, or people who actually respond to a poll?

Oh, yeah. I mean, you've probably been called by a pollster. Yeah. But like, I mean, it's like and there are also polls that aren't election polls, right? There are marketing polls. There are. I've been polled. I was polled once as a famous pollster.

Turned 18. We know because they asked you if you used drugs and you lied to them and said no. This has been on the podcast before. No, man. I've never used drugs. What was the number you got, though? The likelihood that you would actually respond to a live caller poll in your lifetime? If you're like in your 50s, it's probably like

For a horse race election poll, are you voting for John Kerry or George Bush? Like, I think around like 5%. Okay. Well, I've never heard from a horse race poll. If you live in like... If you live in New Hampshire, though, or something, much higher. Okay. Now that the 2022 elections are over, Devin asks, what will 5E be up to until the 2024 models launch? He's got to make a primary model. 5E...

Can I be honest? 5e had a lot of his wealth in FTT and other exotic crypto tokens. And so 5e is looking at various options for declaring personal bankruptcy. That honestly... And it's kind of taking up a lot of... I feel so bad for 5e. I do. You know that 5e is the one that lives in Vermont. I would think the model would have been the one to invest in all of these exotic assets. Yeah.

Down in South so the model invested the model to escape to the Bahamas No, the model invested in Bitcoin like 2011, right? So it's way down versus the peak but like it's doing pretty well The model finally convinced me five years so eager to please right? He's like I don't want to miss out on this, right? And so five you like bought crypto like right at the peak, right? He thought like the Larry David commercial was funny and stuff like that like Kim Kardashian's doing it I gotta do it too

If I have these timing was pretty bad. Final question of this episode of not model talk. We're going to have to come up with a name for our chats when we don't have a live elections model up is from Kashif aisle or window. Oh, window. I like seeing the, uh, pretty shit going on outside. So this is the, by the way, airlines, what's this airlines always have in the windows closed. Okay. Wait, we're not, this is not, this is not your space for, uh, airline grievances. Uh,

The context of this question is that you on Twitter were like, oh, the way that the hack when you're traveling as a duo is that you always book the window and the aisle seat and then just hope that no one sits between you. But if they do, they'll obviously switch with one of the people because they'd rather be in the aisle or window seat and you can just sit together anyway. And then I said, no, no, no, no, no.

The thing you definitely do is you both sit in aisle seats on either side of the aisle when you're traveling as a couple or two friends or whatever. No fly list. No fly list for you, Galen. You can't do that. Absolutely. And you're like polluting all the... The best move. You're not going to talk to somebody the entire way. If you do, you do it in like hushed tones. But you each get sort of space to get up, walk around, go to the bathroom. You can talk to each other if you want to. It's definitely the move.

straddle the aisle. I have a pretty good bladder. If you like looking at, if you like looking at, especially like a cross country flight, man, they're like LA to New York flight or vice versa. You don't pee once between LA and New York? Depends on if I'm drinking on the flight. The answer, your neat silver, the answer is yes, obviously. It's a good, I have a good bladder. No, I, no, I can, I can hold it, man. I'm telling you. Okay, folks, we would like you to vote.

Okay, no. Straddle the aisle. Do not vote about my butter. No, we're not voting on your platter. Jesus Christ, I'm not a psychopath. We're voting on whether when you're traveling in a pair of two, do you straddle the aisle? Do you get the aisle and window seat? Or do you get two seats right next to each other, aisle, middle, or window, middle?

We would like to hear from you. This is not close. It is not close because it's aisle aisle. Occupy. No, if you occupy the aisle and the window, you often get a free seat because no one wants the middle between two people that don't know. And if you get aisle aisle, you always get a free seat because the aisle is open.

The aisle is people lining up for the bathroom and like beverage carts and like, I have a wide stance. So like, I like, we'll like clip. Are you manspreading an economy? A little bit, a little bit. So like, I don't like a clip by the beverage cart, right? Yeah. Okay. Well, we're going to have to leave it to the listeners because I think we've made our, we've made both of our arguments. We're going to leave it there for now.

Good luck at your tournament in Las Vegas. And I will talk to you next week. Thank you. My name is Galen Druk. Tony Chow is in the control room and Chadwick Matlin is our editorial director. You can get in touch by emailing us at podcasts at 538.com. You can also let us know, aisle, window, window.

middle aisle, middle aisle, aisle on Twitter, or you can also let us know any questions or comments on Twitter. 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'll see you soon.