Why Democrats Beat Historical Trends In 2022
01:01:29 Share

This country was built on a distinctly American work ethic. But today, work is in trouble. We've outsourced most of our manufacturing to other countries, sending away good jobs and diminishing our capability to make things. American Giant is pushing back against that tide. They make the kind of high-quality clothing staples you need for a comfortable, stylish, and active summer wardrobe. Like vintage-style tees, tough tank tops, breathable polos, lightweight sweatshirts, shorts,

and so much more, all made right here in the USA. From growing the cotton to adding the final touches. So when you buy American Giant, you create jobs for seamsters, cutters, and factory workers in towns and cities across the United States. And it's about more than an income. Jobs bring pride.

Purpose. They stitch people together. Support America's workers and get 20% off your first order at American-Giant.com with promo code STAPLE20. That's 20% off your first order at American-Giant.com code STAPLE20. You know, the popular vote's important, guys. It's an interesting measure for us election nerds, and I'm having trouble finding it. We'll have to edit this. Do you know you watch YouTube videos or podcasts and they don't bother to edit long, awkward sections like this? You're like...

I'm devoting my time to you. Okay. Hello and welcome to the Five Foot Eight Politics Podcast. I'm Galen Druk. I'm Nate Silver. And this is Model Talk.

So on Saturday, Nevada was projected for Democrat Catherine Cortez Masto, ensuring that Democrats are going to keep their majority in the Senate, regardless of the outcome in the December 6th runoff in Georgia. As things stand right now at noon on Monday, Republicans are projected to win 211 House seats. Democrats are projected to win 206.

So Republicans need seven more seats and currently lead in 10 in order to take control of the House. Most of the uncalled races are in California, where the tallying is still ongoing. They send mail ballots to every registered voter. You are allowed to send your ballot back as long as it's been postmarked by Election Day. So there are plenty of races there that still only have like 50 or 60 percent of the expected vote tallied.

We're going to talk about all of it. We're going to talk about what it means, how the polls did, and also all of the stuff that is still not projected yet. So let's begin with the House. At this point, would you say that Republicans will win control of the House?

I mean, let's go to prediction markets, which although... Which famously did extremely well this cycle. Not. They did not. Okay. But I think prediction markets are good about counting votes after the fact because it's no longer vibes driven. This is just like nerds who are actually delving into their spreadsheets and reading information from counties and states and being very diligent, right? So like after election vote counting, the prediction markets are pretty good with the caveat that there was still a big position that Trump would

win in 2020, which, but anyway, that's another story, but they have the GOP with the 95% chance of winning the house. I see no real reason to diverge from that. You know, Nate Cohn, the New York times has kind of a breakdown district by district and there are 218, uh,

where you'd say the GOP is favored. And then another several where it's kind of a toss up. So 218 is the exact number you need to take the House. So it's not just that Democrats have to win all these kind of toss up races. They also have to win districts that are already- Take one district that looks pretty good for the GOP and then win all the coin flips. And that gets pretty hard. Is this your way of laundering your soft Nate Silver projection through the betting markets?

No, I don't. You know, I, you know, again, I, I, uh, been on many things, but not politics actually. Oh no, of course. I mean, like, are you in a position where you would normally say, yes, this is a done deal. This race is over. But there are people who have spent a lot more time looking at like, I mean, cause California is tricky because there is like a lot of late counted vote.

it has a shift that tends to be democratic in recent years but the question is how much and it like depends on like the demographics of the district right if it's a district where you have a lot of poorer democrats they may vote late if it's more the upscale high demographic msnbc watching democrats they may vote early so it's district by district but if you kind of extrapolate out

What the shifts in the vote count have been thus far in California and Arizona, another vote by mail state. The math comes up a little short for Democrats. What does a 218 or 219 seat majority mean in practical terms for the Republican Party? I mean, this honestly is partly something where I have to educate myself and we all will get a crash course on like House procedure.

Because we haven't had a House that's quite this close, right? I mean, five seats was very close for Democrats in the past two years. And Pelosi did a good job. But the GOP kind of has a two-headed problem, right? One of which is that you have these Freedom Caucus members who were already a little skeptical of current GOP leadership, and that's a substantial caucus. However, you also have this kind of new cohort of

moderate Republicans that in this like mini micro red wave we had here in New York and New Jersey, you have a group of like six or seven New York and kind of NYC Metro representatives who, let's be honest, are going to have a lot of trouble being reelected in a normal turnout environment. It's not like a micro wave. And so they have a lot of incentive to moderate and compromise. There is a tradition of in New York and New Jersey,

moderate Republicans can do okay for themselves, right? So it's not a hopeless cause. And so you have, you know, so it gets very messy. And also like- Well, because basically every single House member is-

a kingmaker, right? Anyone can jump ship and hold up the entire caucus. And so, I mean, should we expect in a circumstance like that, that these sort of moderate, more marginal district Republicans are the center of power because they're talking to folks across the aisle? Or is it still the freedom caucus? Well, here you get into like, could you have someone other than Nancy Pelosi, like lead the Democratic caucus, right? Someone who is seen as more moderate,

and form a coalition with, right? I mean, I don't know. This is like house of cards. This is where I just don't want to get ahead of myself because it is like ground that in the 14 years I've covered politics, we haven't had a case quite like this. Okay. So let's move on to some more of the uncalled races because we're going to have lots of time, two years to cover the current Congress or the upcoming Congress. Who's going to win the Arizona governor's race, Mr. Silver?

Again, I'm not really adding too much value here apart from like the people who have looked at this in detail think that Katie Hobbs, a Democrat, is the favorite. Again, you go to betting markets, she's around 90, 92 percent or something like that. And I don't have any reason to doubt that. You know, that would be a modest upset relative to polls. I was looking at this earlier. So in Senate governor races, Lake was ahead by 2.4 points in a polling average. So if she loses, that would be the biggest problem.

upset in a Senate or gubernatorial race, which is to say that there weren't very many big upsets in those races. Well, the biggest upset, but not the biggest polling mess.

For example, Ron DeSantis was leading by 12 points in the polling average. He won by 20 points. Yeah. Or, you know, what's another example? I think Wisconsin may have been another situation. Or New Hampshire, Shaheen won. Shaheen. Yeah. It's not Shaheen. It's Maggie Hassan. Maggie Hassan. So there were bigger polling misses statewide. In fact, Pennsylvania was even bigger than the Katie Hobbs, Carrie Lake miss would be. Right. But in terms of calling winners, the polls did...

very well in House and Senate races or Senate and gubernatorial races. And in fact, that one was like actually like Leaning Lake. Leaning Lake. But not by much.

I think there was also, we're going to get into all of this, but there was a lot of narrative building around Carrie Lake's charisma and her potential position as a running mate for Donald Trump in 2024. And so the vibe started to also take root in how- Yeah, and Katie Hobbs hadn't debated her. And yeah, it's not great, but she was ahead in the polling by a bit and not just like the

GOP flavored polling, right? So you're saying Cary Lake was also ahead in the institutional, with the institutional pollsters, a divide that we're about to dig into a whole lot more. I mean, you know, data orbital, research code, data for progress, which is a democratic leading firm, Emerson College, they all had Lake ahead. Marist College had Hobbs ahead. They may look smart, but it's like hard to find. There are not very many polls apart from Marist College. I'm scrolling way back here.

Susquehanna back in mid-October. So really only one of the final polls out of 10 or 12 had Hobbs actually ahead. It's from Marist College, so pat yourself on the back, Marist College. I will say that, like, this looks like a little hurdy. Everyone has Lake up by, like, one to four points. Mm-hmm.

Maybe people, if they actually had Hobbs ahead in their numbers, were a little... Well, that's what I was saying. The narrative kind of took hold. Little scaredy cats. Okay. So Alaska Senate, Lisa Murkowski or Kelly Chewbacca? I haven't looked at it. Haven't looked at it. Okay. Well, folks, I think the conventional wisdom is that Lisa Murkowski is going to be able to pull this out. Okay. If she is trailing by only 1.4 points in the initial round of voting, then I would favor her. Yes. Yes.

Okay, so last uncalled race, and then we're gonna dig into the drama of it all. Which isn't exactly an uncalled race, but the Georgia runoff. So I think there are somewhat competing narratives, which is that historically, Republicans do better in runoffs in Georgia because the electorate tends to be older, whiter, more likely to have a college degree, which historically favors Republicans. However, in 2021,

That was not the case. In fact, Democratic turnout stayed high, particularly in Black communities, and rural turnout dropped off. And so we saw that John Ossoff and Raphael Warnock were able to win in a runoff environment. Should we expect the runoff environment in December, considering that this is not going to determine control of the Senate overall, that it will be more like sort of historical Georgia runoffs or the 2021 example?

I'm not sure how robust that history is given that Georgia is an entirely different state right now. I mean, there are a couple of reasons why you'd rather be Warnock. Um,

One is he did win by like 0.9 or one point on the initial round of voting. So if you take the remaining votes and split them 50-50 to libertarian vote, then he would win. I'm not saying it's the right way to do it, but that's actually not irrelevant that he came fairly close to winning the first time. Although my thinking would be that a runoff election is more about turnout than persuasion. Maybe, but Democrats, I think, are happy to play the turnout game in Georgia, particularly when the GOP is going through a period of

Whereas for Democrats, it feels like, I mean, don't get me wrong. Like, it's nice to be in the opposition, right? But like the GOP is like really kind of knocked off its heels, I think. And like, and you might have Trump make an announcement that which, you know, is going to motivate Democrats. Oh, interesting. You think that Trump announcing might help Herschel Walker? No, no, no, no, no. I think Herschel Walker might mistakenly think it would help him or hurt him. And you do have 2.1% of the vote or something went to the libertarian candidate vote.

If you look at the exit polling of voters who voted for that candidate, it's a very small sample. However, in that small sample, a majority of them thought that Walker's views were too extreme, but not Warnock. So they may be Republicans who just couldn't stomach voting for Walker. And like if Walker was necessary to like.

have a 50-50 Senate, then you might think that they'd be more inclined to vote for him. But now that it's like, we lost the Senate anyway, right? Are they really going to go to the polls and like... Well, right. My thinking is that they just won't turn out. Like, there's a lot of people who turned out to vote for Kemp because they like Brian Kemp.

Obviously, Brian Kemp ran seven points ahead of Herschel Walker. So there were a lot of people who support a Republican for governor who said, OK, but I'm not supporting Herschel Walker. And I would imagine that those people just don't turn out. I think a lot of them

won't turn out. And then of those that do, I'm not sure we count on necessarily that many of them going for Walker, right? It's like, well, you, you know, I mean, now that you lost anyway, so I'm going to, I mean, yeah, it's hard to see them kind of turning out for either candidate. Do you think candidate quality becomes more of an issue in a runoff election like this where Kemp can't kind of carry Walker to some extent? With Kemp kind of having been

From the standpoint of Democrats...

Democrats, what difference? So like, my thinking is that it's people who actually like Raphael Warnock in large part and want to see him reelected are the people who are most enthusiastic to turn out because this isn't going to determine control of the Senate. But from the perspective of how 51 seats may make a difference from having 50 seats, what is it from Democrats perspectives? Like, is there anything that they can do with 51 seats that they couldn't already do considering that they probably won't control the House?

Sure. So there are some things about like committee membership, you have an actual majority and not a 50-50 majority broken by the tiebreaker, then committee memberships change. But more importantly, from a forecasting point of view, if you look ahead to 2024, it's a really brutal map for Democrats. Let me bring that map up. You got Manchin, you got Tester. You got Manchin, Tester, and Sherrod Brown, right? So those are three states where it's pretty challenging for Democrats to hold on.

And there aren't too many good pickup opportunities, maybe hardly any really, right? You know, Florida and Texas are the closest thing, but like those are not states where Democrats have done well in recent elections. They did poorly this year, right? And so you kind of are hoping to catch lightning in a bottle. Maybe there's some Rick Scott, you know, he's, you could understand him being one of the easier comers to defeat, but Florida is like a red state now. It's, it's,

It's pretty tough. You know, Utah, I don't know. You know, Romney is a very strong candidate if he runs again. So they're likely looking at losing the Senate majority in 2024. But you can imagine that like if it's 51-

Then Tester and Brown, who have survived some tough elections before, although not really, haven't really been in really tough circumstance. That's part of the problem for them. Oh, they can trade off being the vote that doesn't agree with the rest of the Democrats to build their bone a few days. Montana is getting a little bluer. And you can imagine...

Brown Holding on I mean he was reluctant last time mentions the hard one but like but one thing is like if you have 51 The mansion can vote go vote against everything that Biden would yeah would do right? Um

So, you know, in the long run, too, I mean, there's questions about like what happens in 2026, right? Right. Which is basically what you're saying is here. Like, yes, one Senate seat right now doesn't determine control of the majority. But Senate seats are held for six years and down the line. In 2026, you have a seat in Maine. Susan Collins might have trouble again. I think Alaska is actually an interesting state.

Going forward. All right, we're getting too far ahead of ourselves. Let's talk about why what happened last Tuesday night happened

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.

We're in that time of year, Nate, where the market for takes could not be hotter. And lots of people have incentives for describing the outcome of the election in different ways. People on different parts of the political spectrum have their own motivations. Analysts may have their own motivations as well. I want to try and see if we can get to the bottom of this, knowing that one, we don't have full data. We're not going to have full turnout data until December, that not every ballot has been counted. So there's

places like particularly in Southern and Central California where we just don't have a full picture yet. But if we can start to piece things together with, again, the asterisk here that it will take some time. So how would you describe what happened on Tuesday night? Why it happened? Let me try to give the one sentence version. Republicans underperformed historic precedent for how the party does at midterms based on nominating candidates

Extreme candidates based on the Dobbs decision, based on how many inexperienced candidates and underperforming in key swing state and swing district races. Here's how I think about it.

is that if you look at the landscape before the Dobbs decision, Republicans were overperforming their baseline by about two points. If you look at last November's election, they were overperforming by even more than that in special elections. And then last November, I'm talking about regular elections. After the Dobbs decision, Democrats began overperforming the baseline by nine points. All of a sudden, it was like a 2018 blue wave style environment over the summer.

There were things that brought that back down to earth, time, arguments about inflation and crime, etc. But that ultimately without the Dobbs decision, we probably would have been looking at an environment that was more like last November than this November.

So Republicans were able, with time and sort of campaigning on those issues, to pull roughly, roughly even. But it was ultimately the fact that they nominated bad candidates in a lot of competitive Senate races that they lost. And House races. And House races that they lost. And so Dobbs sort of changed a lot of the playing field from where we were headed last November. But then it was the poor quality candidates, you know, extremist candidates, election denier candidates that were largely backed by Trump.

that lost on the Senate majority. They could have still won a Senate majority after the Dobbs decision had they had better candidates. I think so. Although, again, the least worst GOP candidate was maybe Lex Alt in Nevada, and he lost too. So Democrats had... Yeah, I mean, I think it's a basic story, but it's kind of more... I mean, looking back to last year in Virginia and New Jersey, I think there's a little bit of an untold story here, which is that there was some...

return to normal after COVID that there was less of a year ago, right? So there was a lot of talk about shutdowns and some of the conversations in 2020 about defund the police and policy in schools, right? Like, you know, Democrats low key, I think kind of moderated a lot between 2021 and 2022. And some of their weakest issues, I think kind of came off the radar a little bit.

But, you know, but yeah, I mean, because that was a very bad election for Democrats. But but no, I mean, I think without Dobbs and the GOP probably has a normal ish midterm, maybe not a landslide, but like a normal ish midterm. I mean, looking at the position that the country is in. So 40 year high inflation, the incumbent Democratic president has a 42 percent approval rating on average.

Violent crime has risen over the past couple of years, particularly murder rates. Yeah, with some caveats. I mean, with some caveats, illegal immigration has we've seen record record levels of border crossing. So all of those things you would think Republicans can message well on Americans register in polls as unhappy with those things favoring Republicans on the economy, on immigration, on crime. And yet this happens.

And he's like, is that how powerful of an issue sort of like Dobbs and the Supreme Court is? Like how powerful just candidates are? Dobbs, the Supreme Court, Trump, stop the steal, kind of all the above. I mean, the thing is, there's always been like, I mean, Dobbs is the single most important thing, but they all complement one another. I mean, again, like the basics foundations of midterm election politics.

results is that people want to check the party in power, right? When you vote in 2020, you don't know what the outcome is going to be. It turns out that you get a Democratic sweep, although by very narrow margins, right? So you want to then, when you have an opportunity to course correct, to check the party in power. But Republicans have demonstrated time and time again how much they exercise power even when nominally out of power. Through the Supreme Court, through the threat to overturn

election results through very aggressive actions in the states where they govern you know so so this was more kind of a you know voters kind of felt like actually power is pretty evenly distributed between the parties now and they kind of reaffirm the status quo when actually incumbents did pretty well to me that's part of the story to what do you ascribe to

the pretty stark geographic differences that we saw. So to describe some of them, and we have described them over the past week on this podcast already, but Republicans doing quite well in Florida and New York, TBD how Republicans will do in Southern California, but Democrats doing quite well in Pennsylvania, Ohio, Michigan. Yeah. I have a pet theory about New York. I

which is that it reminds me of Beto O'Rourke in Texas in 2018, where he lost, but like only lost by two some points to Ted Cruz, right? And was like the first Democrat in a while who had had a chance there. I think actually in these red and blue states, there are sometimes more opposite colored voters than you would think, but they kind of aren't very active in politics most of the time because they never win anything, right? So the fact that Lee Zeldin was kind of given a chance, I guess, credibly enough to win, right?

I think it kind of brought out Republicans in New York and Republicans have won governorships in New York. And we've had Republican mayors of New York City until fairly recently. So it's like a latent Republican vote in New York, more upstate than the city, but like, you know, also Long Island and stuff like that. Every county in New York shifted red compared with 2020. Yeah. So that vote was activated in New York is not quite as blue as it seems when GOP voters have a reason to come out. Right. But also Democrats maybe weren't, you know, I know a friend, this is like,

like in Tom Friedman territory, right? But like, well, I was in this cab and no, but like a friend's wife was like, who's a Democrat was like, you know, kind of taking care of her two kids. And at six o'clock, she's like, you know what? I actually think

based on the coverage or meeting, this election might be close. I have to go out and vote for Kathy Hochul, even though it's kind of been different before, right? I mean, people, I think, woke up very late, Democrats, to the fact that New York was kind of having this red shift. Well, I think there's another story going on here too, which is that the strongest issues for Democrats in this cycle didn't resonate as much in a state like New York. So abortion rights are already codified in New York, and Lee Zeldin said that he didn't plan to change abortion law in the state. I don't know to what extent

folks were paying attention to that because Kathy Hochul was still running ads about abortion. The other thing is sort of election denialism, extremism, etc. Now, Lee Zeldin is not, you know, a Pataki-style Republican. He tried to moderate his...

Trump alignment did kind of make him a weird candidate there. But nonetheless, New York isn't a swing state. And so there wasn't the same concern about certifying New York state elections, et cetera, et cetera. And then the things that did get a lot of attention are like crime in the New York metro area. And it's a high cost, high tax state. So maybe inflation comes more to the fore as well. So I think there's like an issue story being told there in addition to your like Republican voters being activated thing.

You know, living in New York, I don't know what to think about the crime. I mean, now I get to, like, self-selecting who kind of, which friends I tend to hang out with. But you did see, like,

Republicans did a lot better with Asian American voters in New York City. In Southern Brooklyn, yeah. Yeah. And so that starts to add up and that can very plausibly be a crime-related story. Ultimately, Kathy Hochul won by six points in a state that Biden won by 23 points. Do you think that New York State is more of an example of what this midterms might have looked like without Dobbs?

Um, without Dobbs and also without like, I mean, Schumer went by 13, right? And there's still votes being counted. It was because like, I think like, it's like not like she is like that disliked. It just kind of people are indifferent to her. I mean, this is kind of the scenario you get where A, Dobbs less a factor and B, there's like a big enthusiasm gap. I can imagine like an election where Dobbs doesn't happen, but there isn't that much of an enthusiasm gap because Democrats are still concerned about January 6th and Trump and so forth.

But yeah, this is like this kind of like you had an example of the downside case for Democrats here in New York. And then we've already talked about this on the podcast, but the upside case is like Michigan, where abortion becomes very salient. You see it in the exit polls. You see it on the referendum. You see Gretchen Whitmer running on the issue. I mean, obviously, she's quite a moderate governor and runs on things like infrastructure and roads and bipartisanship and whatever, but also ran on the issue of abortion. Yeah.

Michigan is an upside case, Pennsylvania, and you actually have some state legislators that might flip here too. The fact that Democrats kind of swept the gubernatorial races in Pennsylvania, Michigan, Wisconsin, and Minnesota. I mean, it was a really bad election for election deniers. Yeah, every single election denier in a competitive secretary of state race lost.

And we were able to call project those earlier on than a lot of the other races. Honestly, the biggest story of the election might be that the chances of a constitutional crisis in 2024 have, in my view, gone down a lot. They're still much higher than the average person on the street might think. Right. I guess I don't want to come with like a quantitative estimate. But, you know, if it goes from like some.

low double-digit probability to a single-digit probability or something like that. My guess is if you looked at trading markets at somewhere where it would be directionally, then that seems like pretty important. Well, speaking of 2024, all the reporting suggests that Trump plans to announce his candidacy for president Tuesday night from Mar-a-Lago.

There's a lot that we can say about this, and I'm sure a lot that we will say about it in the coming months and potentially years. But given everything we know right now, given everything we've talked about right now, what does that mean for American politics? I mean, there was actually a poll that came out. You go, I think, actually had DeSantis ahead. I mean, look, nationally, I think it was the first one, right, that showed DeSantis nationally ahead. Yeah. Because he was already ahead in some of the early state polling. Look, I think I mostly...

I mean, Trump had this kind of halo of invincibility. It's kind of funny that this kind of gets burst in the same week that Sam McMinfreed's FTX operation blows up and that Elon Musk is doing a very good job of running Twitter. Wait, you mean the emperor has no voice? The halo of, yeah, it's kind of the theme. No, look.

Based on the conservative outlets that I read, it seems that like the anti-Trump sentiment is penetrating much deeper than it has at any previous point. Right. Pretty conservative Trumpy outlets are kind of saying, let's take a careful look at this. Right. And that's, I know, a subjective observation, right.

You know, I mean, but look, it couldn't have gone better for DeSantis relative to Trump. He nominated several candidates. They're going to lose New Hampshire, Arizona. I mean, all these candidates like underperformed. Meanwhile, DeSantis wins by six gajillion points in Florida and generally other Republican governors in that mold. Right. Where, yes, you're an ideologue, but also you're.

You have a pragmatic streak, and I'm sure we'll get hit mail for saying that. But like, you know, Greg Abbott and Brian Kemp did very well as well. Right. You have the Glenn Youngkin model in Virginia where they were not Senate governor, but the GOP had a good night in in other races on the ballot in Virginia. So like DeSantis has like a much better electability argument at this point. I think accurately so. I mean, the Trump era has been kind of because here was the thing, right?

You could always say, oh man, Trump hasn't performed really that well, right? He wins the champion nomination. That's impressive. But then 2016, you lose a popular vote to Hillary Clinton, right? You have this miraculous electoral college map. Maybe it's skill, maybe it's luck, but like you pull it out very narrowly in a few states. You have a really bad midterm in 2018. You lose reelection as incumbent in 2020, which you're not supposed to do as an incumbent, right? And then you have like

This, you beat the bed in 2022 based on candidates that you handpicked. I mean, this year, I guess we would have always said that, oh, Trump's electability argument is pretty bad, right? He's always been mediocre and unpopular. But this year kind of proves that to people. This year, you cannot spin your way into saying it was a good election.

for Republicans, right? To lose, they'll, you know, to let Democrats keep the Senate, a 50-50 Senate, maybe probably even lose a seat with 8.X percent inflation, right? And Biden had the same approval rating that Trump was at when Republicans lost a whole bunch of House seats in 2018. Yeah, although they gained Senate seats in 2018, which is complicated. But this is, no one's even trying to make excuses. And the fact that, like, you know, that you go, probably proved to be a mild outlier. But the fact that, like,

you know, we have God. What's God doing? We have like two years for the next election, although kind of, you know, a year and a few months before the Iowa caucuses. The fact that DeSantis is within squinting distance of Trump at this point is not a good sign for Trump, who is the former president. Yeah. All right. Well, we may have to have an emergency podcast on Wednesday pending that news. But now it is time for me to pivot to the forecast.

Okay, so the classic and light versions of the model showed the Senate a dead heat, a 50-50 proposition. The deluxe version showed it a 60-40 proposition. Meanwhile, the light version showed Republicans with a 75% chance of taking the House, and the deluxe version showed Republicans with an 84% chance. Colin, listener, asks, should there only be a light forecast going forward? I think that's not a good argument. I think the argument is about

versus deluxe, right? You know, you gain a lot by accounting for fundamentals, especially in states where there's not much polling, right? And kind of the classic forecast is such that like it differs to the polling in states where there is a lot of polling, but like it has better heuristics for states and districts where there's not a lot. I mean, we'll look at it, right? But like, but classic, I think is applying some pretty common sense kind of general empirical corrections, whereas deluxe is looking at the expert ratings and like,

The vibes. The vibes. The vibes took us down the wrong path. Yeah, I don't know. I don't know. I really want to know. I mean, I want to examine these on a case-by-case basis. But the other problem with, like, which we use, Cook, Sabato, and Inside Elections, the problem is all these people, like,

talk to one another and see the 538 forecast. So it's like in a world where you're isolated from it, but it gets a little weird and recursive where we publish a forecast that can affect the forecast. Yeah, yeah, yeah. And I think folks have sort of been saying this because we used to default to the classic version, right? We did. And kind of my view was that in a world where people are busy and where there's a lot of stress around elections, right?

If you come to 538, I want to kind of show you the number that I would bet on, basically. Because historically, the deluxe model would do better. A little bit better. A little bit better. But like, you know, because I would take our forecast and hedge it a little bit because that's how a good bettor acts, right? They won't say, oh, well, the market's entirely wrong and every edge that I have here must be right. You'd say, okay, you know, if we're hedging, then you would hedge a little bit toward the convention of wisdom. Not a lot, but a little bit. And it's kind of what deluxe does. Yeah.

if I were putting my own money on the line, right? I would have thought that this 41% price was a pretty good price, which was more optimistic for Democrats than the markets were actually. But I would have thought that was a pretty fair price if I were setting a betting line, right? And wanted to be accurate and be right half the time. But can we say at this point which version was the most accurate? So from a macro point of view, I think light would be the most accurate. From a micro point of view, you know,

Are there races where the expert forecasters took us in the right direction? I don't know, actually. I think, I mean, but this is a live question, right? The question is like, you know, should the deluxe forecast go the way of the now cast? And say farewell? And say farewell. Are we going to bury the deluxe model? Or maybe we kind of say go back to before. Maybe we can go back and say, okay, classic is like,

Our default version and our official version, right? Deluxe and light are for fun and games, right? But like, yeah. But we have to...

We'll have to look at that. Okay. One more hot topic. I mean, not one more. We have more hot topics, but another hot topic. So we talked a bit about partisan pollsters flooding the zone before the election and showing a better environment for Republicans while institutional pollsters showed a better environment for Democrats. Can we say at this point that that is what happened? The institutional pollsters had a more accurate portrait of the national mood, public opinion, et cetera, et cetera. And the Trafalgar's of the world,

skewed the averages essentially in Republicans' favor? I want to be, so I want to like, this is where I would like actually want to go through and look at every race and every poll, right? I mean, you know, in the most competitive races, the polls generally, well, they had a mild anti-democratic bias. And yes, if you had not had, I mean, there's a lot to unpack here, right? One is like different polling averages come up with rather different

And we should say that when a Republican, for example, partisan pollster publishes a poll and they are demarcated as a partisan pollster in our pollster ratings, then we obviously adjust that based on their historical track record. Were there a lot... Okay, so the narrative, Republican pollsters were trying to flood the zone. Let's kind of break that down piece by piece, right? Number one, Republican pollsters, some of them are not per se partisan pollsters.

Trying to flood the zone. I don't know what their motive is. I think a lot of them are true believers and they thought they were like publishing numbers that were going to look smart. So I'm not sure what the objective is, right? Number three, like I think the 538 polling averages have more robust defense mechanisms against this than you might think, like the house effects adjustment. And although there are some pollsters that actually had a fairly high rating, like Trafalgar, some of the newer fly-by-night ones don't have a very high rating and don't influence the model much. Oh.

I don't mean like flood the zone as it's some sort of conspiracy to trick people. People do mean that way. And I understand that there were some folks on MSNBC describing that to be the situation, but more. OK, these pollster, these sort of up and coming partisan pollsters do think they sort of have this new way of doing polling that's more accurate, that captures Republican support where other institutional polls don't. We had this conversation. We knew that institutional pollsters did well in 2018, despite it all.

And so there was a question of, are these new pollsters who really have opaque methods for conducting these polls kind of screwing around with the averages? It seems like at this point, at least to some extent, they did. And the institutional pollsters, like the New York Times, did really, really well. Well, this is where I want to see if the institutional pollsters had...

had a de-bias or were totally exactly on, right? Yeah, yeah, yeah. And I don't know that, or even in our bias, I don't know that yet, right? Right, so they might have had a slight sort of Democratic bias that was overdone by the partisan pollsters and it netted to a slight Republican bias overall. Yeah, we have, I guess, we will have to- I mean, and then you have- We'll get back to you, listeners. You have some states where like Trafalgar had like a 15 or 30 point error, right? Like in these non-competitive races. And so-

I think the polling in the non-competitive races may weigh the accuracy down a little bit. One funny irony that people pointed out to me is that Data for Progress, which is a Democratic-aligned pollster, had pretty bearish results for Democrats. So, you know, so they're not quite traditional, but they're not GOP-leaning either, so I don't know if they're trying to flood the zone too. So the question I'm getting at is like, did the GOP polls, and again, they're not formally GOP polls, a lot of them, did they kind of

quote, two wrongs make a right, because even though they weren't very good, they made the polling average more accurate because otherwise they would have been too democratic, right? Like a big theme is like people I think were double counting in some ways. If you already have a lot of GOP aligned pollsters, that already makes it much less likely that you'll have a anti-GOP polling bias. It's a different mix of pollsters. So you shouldn't then assume that

There's some further bias on top of that that becomes infinitely progressive. Which is basically the light version. Which is the light version, yeah. The light version was showing just the polls and no assumption that the polls would overrate Democrats. Right. If you just looked at the polls, then you kind of get to 50-50 Senate and a gain of 15 GOP House seats, and they'll probably wind up being 51-49 Senate and a gain of...

geez, eight or nine GOP House seats. So it's actually not too far off. It's like pretty good. But you're looking at like in the aggregate, a mild, pretty good accuracy in the aggregate, a mild anti-democratic bias in the aggregate,

But with some pollsters, Trafalgar, that really had much more severe problems. So basically what we're telling listeners is we're going to come back to this question once we've teased out how individual pollsters did. And if you take out the Trafalgar's of the world, did institutional pollsters still overestimate Democrats or were they spot on? And we are going to get back to you listeners. One more weedsy polling question before we get to listener questions.

Our generic ballot average showed Republicans leading by 1.2 points on Election Day. Our forecast suggested that the House popular vote would look more like four points, the margin. The House popular vote margin would look more like four points because of uncontested races and likely voter adjustments. How did those two numbers stand up? So it looks like we'll end up with a House popular vote that's like R plus two once California is counted. So what's that saying is like the deluxe forecast

did have a little bit of an anti-democratic bias. Well, within the robust part of the confidence interval, but yeah. I mean, what it's kind of saying is the light forecast is pretty unbiased and deluxe, you wind up having a little bit of a pro-GOP bias. So again, I expect that when we evaluate this, that light will call more races correctly than deluxe. So all told, we got

Polls underestimating Republicans in 2016, polls being pretty spot on in 2018, underestimating Republicans in 2020, underestimating Democrats a bit, right? I think that's fair to say in 2022. When you add in the 30-point miss by Trafalgar and Vermont and stuff like that. Although, again, you have some states like Florida where DeSantis did a lot better than these polls. Right, an eight-point polling miss. So what's the message here?

Don't try to predict the direction of polling bias. In part because people are very aware of these conversations. Like it wouldn't shock me if like I've said before, I think polls are becoming more like models, right? It wouldn't shock me if like this was the year when the institution of pollsters said we can't take

a purist approach right we can't just assume that we're doing polling the same way that we did in 1995 and that we're going to get um accurate results they kind of acknowledge that like doesn't mean we're going to like make it up out of thin air or be as flaky as some of these pollsters that are very non-transparent are but like but we have to like dig in and re-examine our methods and so like and i you know i suspect there was also a little bit more

hurting potentially? I mean, I don't know. Yeah, I mean, I think one thing that's going to help us in this process is that Nate Cohn, friend of the pod, and the New York Times in general seems genuinely curious about the possibility

project of getting an accurate read on public opinion. And they're doing experimentation where they're sending out sort of letters to... In Wisconsin, they're doing this experiment where they send out letters to respondents with $5 in it. And if they respond to the poll, they get $20. And it turns out the response rate is like 30%, whereas the response rate to regular phone polls is like less than 1%. And they're really trying to figure out what is going wrong, why people aren't responding, who isn't responding. And so I think

At least there are people out there who are trying to get to the bottom of this and not turn... I think it would be weird if we got to this place where public opinion polling becomes this sort of like...

Well, this is my own me as a pollster. This is my opinion about what public opinion is. It really should, at the end of the day, still try to stick to scientific principles where you're sampling a public and not just being like, well, I don't know. I don't want to it up and overestimate Democrats again. So here's my opinion about the likely voter model.

What I found interesting from that is that it turns out if you have higher participation, you actually have a lot more moderates. Yeah. Okay. That was interesting too. Because they are not, maybe Democratic partisans are more inclined to answer a poll than Republican partisans, but like the cross pressured, sick of all the ads I see on TV voter, but you hear something about abortion and that makes you turn out or you think Blake Masters is just weirdly creepy and

Okay. And so you turn out this, I saw that in the New York Times experiment and I was like, this validates my hobby horse for the past like a year or two, which is that the way that we describe the public, which is closely divided and somewhat extremely partisan is just not right.

I still think that 80% of Americans are too busy to feel as strongly as the vast majority of people who are loud about politics. And we can go back to this interview that I did earlier this year with some researchers who tried to show that it's 15 to 20% of partisan Americans who define our political. And this is a bias of, because if you're in the media, you encounter partisans, right? You encounter people who work for campaigns who like, Politicians themselves. You may be,

In some sense. Yeah. But you kind of people who are strong partisans. Right. And like the most vocal voters that you hear from are strong partisans. You also got a lot of centrists, but even the centrists are like centrists in like almost a partisan way. Right. I think the media has made a lot of overcorrections and it used to be kind of way back when I was first getting into following politics, like in 2004 before I even covered it really like, um,

There was lots of talk about the importance of independence and the swing voter, and it kind of really overcorrected toward like, oh, it's the base, it's the base, it's the base. It's kind of ironic if like the errors in the polls kind of correlate with that. Okay, so I have occupied a lot of this podcast with my own thoughts and questions. Let's get to some of our listener questions.

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, we're going to try to do some rapid fire listener questions here. Maria asks, is there any way to tell how redistricting may have impacted the results of this midterm? I'm sure some smart person will go back after the fact and look at this. But like, it's very possible that like if Democrats had not had some kind of bad breaks is a term to use some court decisions late in the term that

their redistricting gains, they would probably have won the house, right? In New York alone, the Democratic gerrymandered map was struck down. That may make a difference of several seats.

Also Florida. I mean, Rhonda Santos sort of against the will of the Republican state legislature forced through this even more partisan map that if Republicans only win the majority by like two seats will have been in part because of that. Yeah. I mean, maybe another win. So it may be true that like Democrats did better than they would have with the 2010 map, but worse than they would have with, uh,

You know, with the version where they didn't kind of unilaterally disarm in some states. Well, right. California and New York. California has independent redistricting and New York had a map drawn by the courts after their map was thrown out. So, yeah, I mean, if California and New York could be gerrymandered in Democrats favor, they would be the majority in the House. Probably. Yeah. Next question from Abe. Will anyone take Trafalgar seriously again?

The guy hasn't been heard from. I keep waiting for... I guess other people aren't following this FTX Sam Bankman-Fried story as much, but his Twitter feed is mysteriously sending out. Let me see. Is SBF tweeting about Trafalgar? He is not. His Twitter feed, though, let me check this out. He's just sending out these mysterious letters. So he tweets last night just the word one and then the word what, right? And then he tweets after that two H and then three A and then four P, right?

I think he's just trying to name Elon Musk's next child. Maybe, or like maybe he's sending secret, but like that's kind of imagine what the Trafalgar bolster is doing. He's like in some like, I don't know, man, uh, in some country where is there like an extradition agreement? No, but like, it's, you know, it's like, I mean, come on, like you can't just disappear like that. That's kind of lame. Um,

That's kind of lame. It says the men who stuck around to keep fighting all these years. So Trafalgar, come on. Make your case for yourself. Tell us what happened. No, I think... Because like he also like, you know, Republican investments, they're going to lose...

Arizona, Nevada, maybe the runoff in Georgia by like one point each. Pennsylvania wasn't that close actually, right? So if you lose all those by one point, then you wonder, okay, what the fuck were we doing thinking we were going to beat like Patty Murray in Washington or Michael Bennett, which weren't close at all, right? And so, you know, Democrats overperformed in the swing districts and states. And if that's true,

then GOP like unrealistic red wave scenarios may have led them astray. Okay. Next question on this topic. Keegan asks, what seats are you surprised flipped? Should we talk about Washington's third district? So that was a district where the forecast gave the Democrat a 2% chance of winning election. And that Democrat won.

Let's talk about this for my forecasting. That's Marie Glucencamp Perez, the Democrat, who won that. I mean, let me go to our- The biggest upset of the 2022 election. Yeah, so this is a district that has an 11-point Republican partisan lean. And so in 11-point, so a year where the national popular vote favors the GOP by a point or two, a district that's 11 points more Republican than the country as a whole in a midterm year. There's a reason why-

the model would not expect that to be a competitive race, especially because like the Pacific Northwest overall, like Democrats kind of had like a rough year in Oregon. So that's like a, that's a really weird one, but 2% is not zero. And the reason why, like, I can say that with like a straight face is because like, we literally run projections for 435 districts. My guess is there probably were 50 or so where you wind up in the bucket where it's like,

less than 5%, but greater than zero, right? You know, one or two or three or 4%. And we'll probably have no one out of 50 where you have an upset. And that's exactly how the model is supposed to work. You know, why that particular district flip that much,

I think we'll have to look at it. It's just a little weird. I mean, it does come down to some of the specifics of the candidates who are running, which I am not familiar enough with to describe. But maybe the next time Nathaniel Rakesh is on the podcast, he will dig into it. Magnum asks, I believe the model thinks results in states are correlated, but it looks like election outcomes were very regionalized with Florida and New York going very well for Republicans. Will the model treat states as a bit more independent in the future or was this more of a fluke?

No, I think it's actually about what the model would assume as far as correlations. I mean, the model, like to use some big round numbers, like presidential elections, like 60% of the error is correlated. And like in midterms, it's like 35% or something. You know, one more particular thing is that like, so the model kind of assumes there can be variations based on regions or demographic groups. You know, maybe you want to have also like state by state variations.

where because of top-of-ticket effects, right? Yeah, we talked about this before. So I don't know if it says, like, oh, New York, one thing happens, but New Jersey, something totally different happens, right? I'm not sure if it deals with that particularly well, but people, I think, over-indexed. Obviously, people kind of erred in part in 2016 by not realizing that Michigan, Pennsylvania, and Wisconsin are not independent states really in the presidential election. It's all kind of the same coin flip, basically. There was maybe an over-correction where, like, yeah, but midterms can always be somewhat...

The fact that this was more than usual, you know, look, I mean, in some ways it's like an argument for like, this is kind of why you look at the longer period of history and not just the most recent couple of elections. But, but yeah. Okay. Jig asks, why did Dems perform so well in competitive races, but collapse in safe districts? Also explain Florida.

Is that a clear trend that we've seen, that Democrats did better in the competitive districts than they did in the safer districts? I think that's true for sure. Yeah. If you lose a popular vote for the House by a couple of points, and in the aggregate, there is moderately favorable mass of GOP-based army districting, and you almost tie the House, you kind of almost have to have overperformed in swing districts. What's the explanation for that? One explanation could be that actually the 66%

gazillion fundraising emails that come from Democrats actually helps. I'd like to see how the fundraising was in those districts. One is that if you have a return to normalcy election, maybe having some of the incumbents in those districts did well. And also Democratic incumbents in the House survived a fairly bad cycle in 2020. So you have survivorship bias where incumbents that survived 2020, not terribly for Democrats, but they might be stronger than average candidates. But also voters are intelligent.

I'm aware of like which races are close or not. And that might affect who I vote for, whether I get cute with a protest vote or not, you know, or whether I turn out or not. Right. So it's pretty rational behavior on the part of voters. And in some ways, you know, you can argue that like, OK, it's an interesting paradigm. Like, let's say let's actually only look at what happened where the vote mattered. And if that's the case, maybe if you had had national stakes, the Democrats would have done better in the popular vote. Although the flip side of that is then maybe you actually have to retire these electoral college candidates.

arguments, right? The fact is that in the most important votes, Trump won in 2016, right? So, you know, and maybe Democrats had inflated margins because Republicans in New York, for example, didn't turn out for Trump in 2016 when they would have if he had had a chance to win the electoral college. Asim asks, can you talk about how the lack of district level polling this year impacted the accuracy of the House model? So one thing I have not looked at yet is...

Yeah. What happens? How did the individual House district polls do where we had them? Right. Because the light model had Democrats with a 25 percent chance of winning the House. Right. You know, if that were the model we've been highlighting, like 25 versus like 16 or whatever is I think in terms of like how we would.

tell the narrative of the race I think fairly different right like 16 is like this can happen it's not actually that crazy but like it's unlikely whereas 25 is like actually this is even the house is kind of comparative it's a little different um no I have not looked at that and um and yeah I'd want to see that maybe if you had had um like those surely the last set of New York Times polls for Democrats in the house were good yeah and they kind of pretended they were good for Republicans which was wacky but yeah maybe if you'd had robust house polling then then you would have had

an even narrower GP majority protected and like a credible chance of a Democratic House. Okay, Lex asks, how does the turnout breakdown compare to previous midterms? And how accurate was the translation from registered voters to likely voters this year? According to Dave Wasserman of the Cook Political Report, so far 101 million votes have been counted. And of course, there are

More to go my comparison in 2018 where we can look at Wikipedia US House election there were a hundred and twelve million or so So my guess is that turnout will wind up being raw turnout will wind up being a little bit lower Than 2018 and if you adjust for population growth, it's a bigger gap So it'll be turn out that is like still pretty high for a midterm but but as a percentage of registered voters

I think we'll be down a bit from 2018. And then the translation from registered voters to likely voters this year, I think it was pretty accurate. That's a different question. I mean, they didn't show it when polls were all over the map, right? They didn't show a huge gap. It will probably work out that like if there was like a one or two point shift on average and if polls had a Democratic or Republican bias on average, right, it will turn out probably that like the registered voter polls will do better in the aggregate. But, you know, we can look at a case by case basis. But yeah, I mean, you know, the one thing about that is like

You know, likely voter models are where pollsters can sometimes put a little bit more of a finger on the scale. On the other hand, like you don't necessarily want to assume that like each registered voter is equally likely to turn out. And so, yeah, that's something to look at. All right. We're going to have a lot of time to continue answering questions about this election and to continue to dig into all of the data. So the last question we have here is from Thomas.

5e lives in Miami and the model in Manhattan. Is that the reason why only Florida and New York experienced a red wave? It sure is suspicious. I don't think 5e lives in Miami. Miami is not hospitable to foxes. Wait, where does Miami... The model lives in Miami. Oh, the model lives in Miami. And 5e lives in Vermont. 5e lives in Vermont. I guess Thomas got this a little mixed up. But the model still lives in Miami.

How did Vermont, I guess Vermont reelected Phil Scott, the Republican governor. So still, you know, still applies. What's going on? What is going on with the red wave induced by the model and 5e? I mean, Florida is like, I think it's become kind of self-consciously DeSantis, I don't know, MAGA or whatever. It's become like kind of self-consciously.

conservative in maybe ways that Florida wasn't before to a point where I think it may affect who migrates to Florida. We had this episode in the podcast a while ago. Polls sometimes ask, are you a recent transplant? And then they ask for your preference according to whether or not you're a recent transplant. Maybe we can track some of those polls down for Florida. So one thing I would want to see is a breakdown of how U.S. citizens who were not born

In the U.S. voted in this election because both Florida and New York index very high on that number. And they have more in common in some ways than you might think. Right. If recent or not so recent, if they've been become citizens. Right. But if like if non-native born voters have shifted GOP, Asian-Americans, Hispanic-Americans, etc. Americans. Right. That would have some effects in Florida. Right.

I mean, ironically, like we have seen a reduction in racial polarization. Well, as education has become more of an indicator. Yeah. So one big one big question is like, let's say that Florida has shifted several points to the GOP side. Like actually now that actually helps to even out the Electoral College popular vote gap. Right. It's a high population state. Right.

if the GOP wins Florida by 10 points in a neutral year, and then maybe California and New York shift a little bit more red also, or although I know California will, right. But like New York does at least, right. Like, and then meanwhile, like Democrats numbers hold up really well in Pennsylvania, Wisconsin, Michigan, you know, that starts to reverse a fair amount of the electoral college edge that we saw in 2016 and 2020. So from that standpoint, like,

Democrats had a very good election. Efficient election. Yeah. And if Georgia and Arizona, although Brian Kemp did very well, but if Arizona and Georgia are kind of at the national median instead of less red than the national median, I mean, that has implications for next 2024 for sure.

All right. Well, we are going to leave things there for now and we'll leave 2024 to Wednesday morning. But thanks, Nate. Thank you. My name is Galen Druk. Tony Chow is in the control room. Chadwick Matlin is our editorial director and Emily Vanesky is our intern. 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.

Ooh!