Poll Me Twice, Shame On Me

25:16Well, this is a question. I mean, how much should al-Sayed, like that vote, right? The Arab American vote. Yeah. that turned against Harris and the Democrats because of Gaza. How much does it make sense strategically if you're in El Saeed's campaign thinking about turnout, right? Okay, like he's not going to win over the establishment part of the party. Like how much can it be a turnout game? Like because I remember, you know, I live in a part of New York that's really South Asian. And on election day here, it was like they did the same thing that Trump did, right? Which is low propensity voters, right?

25:48Like this is your first time voting in an election ever or in a couple cycles. So, like, strategically turning out the Dearborn vote, how much does that make sense? Or how much is that a winnable strategy? If it's 2 % of the Michigan population, I just don't think there's enough of the vote there to make the difference. In a midterm. We're talking about margins here. We're talking about, like, you know, oh, our fundamentals forecast says they should win by three points. Like, you know, one or two points matters in that context. I mean, I'm always someone in general who thinks turnout is a bit overrated relative to persuasion.

26:19but that's less true in a midterm than in a presidential year. Turnout is lower in midterms. You can have these big enthusiasm gaps in the past. They can be as large as like five points where if you poll register, I think, you know, in 2010 and 2014, the Obama coalition, which did so well in presidential years, did not motivate voters very much at all in the midterms. It didn't quite translate. I don't know if it was the young nature of that coalition or what, but like, you know, Obama's approval ratings weren't good to begin with in 2010 and 2014, but that enthusiasm gap really expanded that quite a bit.

26:51And by the way, with Trump's disapproval rating at almost 60 % in some of these polls and a negative 20, you're not going to have Democrats win the House popular vote by 20 points, right? What you might have is a lot of people who are souring on Trump because he's a lame duck, because of Iran, because of tariffs, because of Epstein-Gate, because he's just been in office a long time, they just won't turn out to vote. They'll be watching football on Sunday and not take the time to vote the next Tuesday. And so that's where the lack of enthusiasm could translate. And by the way, people don't realize, they think, well, his approval rating might be 38%, 39%, but they're all really enthusiastic backers.

27:31Like, actually not that either. In our tracking, his very favorable number is down to around 21%. And this is fatigue that you very often get in a second term. And because we're in a modern news cycle, it feels like it turns to be president forever, basically, right? When I read accounts of something Obama did as president, it feels kind of totally foreign. The Mets were in the World Series in 2014. I'm like, Obama was president? That feels like ages ago. Ebola, doc. I mean, that's what I— Ebola, yeah. On sort of where Israel and Gaza fit in, Nate, you may be right that it has more salience among maybe a smaller subset of like hyper engaged people and sort of maybe political elites.

28:15But that can be a vicious or virtuous cycle, depending on how you want to describe it, because that also led to tens of millions of dollars coming into the race to try to turn people against Abdul. And then that becomes a challenge for him to rebuild after a primary in which he was deluged with negative ads. And so, you know, Claire, to your point about, you know, people moderate when they get to the general in a strange way, in part because we're in this sort of post-2020 kind of revision in which people go back and sort of restate or undo things they said in 2017, 2018, 2019, through maybe 2021.

28:58They kind of do their moderation during the primary, right? And so what has Abdul done since? He has not changed his political positions. He really just kind of apologized for a statement that was divisive about an attack on a synagogue, right, in the efforts to try to repair relationships with the Jewish community. But that had come after the primary. And so he's in a difficult position, right, because it's not about moderating, right? There is a kind of specific— And it's not about his issues, right? It's now about a controversy. Right, it's about a controversy. And there are people for whom his remarks after the primary maybe were either kind of either too little, too late, or they had already, there was nothing he could say to undo the fact that they have this fear, some rooted in Islamophobia, some based on just disagreeing with him on his position on Israel.

29:54So I do think that's like a challenge. And regardless, right, like anytime we're talking about Abdul specifically, even his strengths, you're not talking about Mike Rogers, right? Because I think whatever your position is in the ideological fight, it is pretty clear that if this is an election about what Mike Rogers will do in the Senate and how he will side with Donald Trump, that is the strongest place Democrats could be, which is why on Fox News, it's Assam Piker and Abdul side-by-side day after day after day. Like I have trouble picturing what Mike Rogers looks like, right? Honestly, I think he might too.

30:29Like I don't think Mike Rogers knows is what Mike Rogers looks like. It is, he is, he is, but this is his strength, right? Because Mike Rogers - Sure, no, totally, yeah. I am generic Republican, old school, I'm not a Trumpy person. That should give you some reassurance when in reality, right, the fact that he's going to that midterm convention tells you something about his balance between donors and Trump and what's happening in Michigan. Look, one of my hot takes too is that American voters are actually pretty rational in the aggregate, if not individually. One of my hot takes is that people are just good.

31:05Claire. Innately. Salt of the earth. Salt of the earth, Michigan. That's what I say. I'm always saying that. No, but like, so you see a lot more cross-partisanship in gubernatorial races, right? There's a Republican governor of Vermont. There's a Democratic governor in - Kentucky. Kentucky, for example. Kansas. Kansas. You don't see that as much. Kansas has not voted for a Democrat for Senate since 1932, I think, or something like that. because people correctly infer that maybe I think Mike Rogers is a good guy. He's still going to vote with Trump 90 % of the time, which you could literally name on one hand the number of members of Congress who routinely vote against their party.

31:46Susan Collins, kind of. Jared Golden is retiring. Murkowski has a system she benefits from in Alaska where you have a top four system where you can actually – and ranked choice voting. So people are kind of rational to say, I might be worried about, they'll say it's position on this or that, but he's going to vote against Trump and that's what I care about when Trump has a 37 % approval rating in Michigan or whatever it is. Yeah, I mean, to be clear, I mentioned that I think that he is a sort of below average candidate. That doesn't mean that he's going to lose. As we mentioned, he has a good amount of cushion.

32:21And I think, John, I would be curious for your take on this as well, because we all remember from the 2012 election, or at least all of us millennials in this room remember from the 2012 election, when Romney's advisor was like, oh, you know, we're going to shake him up like an extra sketch for the general and moderate him, right? And it's important to remember that people, first impressions matter to like be your mom about it or whatever. And the first impression that people got of al-Sayed was a fighter and somebody who's fighting the Democratic Party, right? And so I think for a lot of people, that gives them the sense that, oh, we already think of the Democratic Party as quite to the left.

33:00You ask Americans broadly, is the Democratic Party too moderate, too conservative, too left-wing? Americans broadly feel like the Democratic Party is already too left-wing. He's positioning himself to the left of the Democratic Party in a battleground state. So a lot of people, and this is an identity thing too, we know from social science that when voters look at women and in particular candidates of color, they view them automatically as more to the left. They assume that their policy positions are more to the left. He was a lacrosse bro, though. I mean, he comes across as very bro-y. And in the same way that, like, you know, in this ad he has where he's on Senate Island, where he's like, has a football and things like that, right?

33:39And, like, is kind of, like, coming across like a— Or the grandparents' ad. That's fair. And look, John can tell us how effective it is to sort of moderate and reposition yourself and pitch a different, maybe, side of yourself in the general election. It clearly wasn't successful for Mitt Romney. But I think it is something that will be a little bit hard to do because of that primate. It's not just because, you know, AIPAC dumped a mountain of money against him. It's also because voters think the Democratic Party is already too far left. If you position yourself to the left of the Democratic Party, you are even further away from the mainstream.

34:12Look, we're a representative democracy. I think flip-flopping is fine if you're saying, I'm trying to meet the voters halfway to where they're at, right? I've never - Oh, wait. Yeah. Flip-flopping should be normalized. But that's like a normative take. That's not how American politics necessarily works. What's the point of a mind if you can't change it? I've gotten in years-long arguments with internet leftists about, I think the data says that you're more to the center that helps on average. I've just found the whole El-Sayed discourse a bit weird because he has gained ground in the polls. Immediately after the primary, Rogers was ahead by one or two points in the polling average, and now El-Sayed's up by— by three or four.

34:53I mean, a lot of it is— And that reflects, like, consolidation of the Democratic Party, and there was a lot of nationalism between them. Oh, I totally agree. But a lot of it is just— it is literally private party politics, right? And the fact that, like, I mean, I said at the top, like, this is the race where the drama is as a member of the media who formerly covered, like, campaigns in that way. Like, obviously, there's something going on here, right? Like, we gravitate towards— like, the New York Times Politics Assignment Desk and the Wall Street Journal to a lesser extent. Like, that really does drive a lot of the way that race is covered and the way— And we lost—scrutiny on El Sayed increased when the main race became less interesting because Graham Plattner dropped out.

35:36Oh, my gosh. It has actually been so funny to me. I've been thinking about that, that we're all like, all right, who's this guy in Maine? And then, like, Susan Collins did have the really hilarious— Oh, I saw that. I'm not going to say anything about the fact that the mother— of his children as his second cousin. That was so fucking funny. Call her Cicero. Oh my gosh, so funny. Where was I going with this? Oh yeah, but now we're just like, okay, so now we're doing Michigan. Wait, wait, wait. But it's like seeking. Back this truck right the fuck up. This is a fantastic exercise for us to learn something about the American voter.

36:18Testing a sort of a left activist in a battleground state is a really interesting thing for Americans to find out about whether or not El-Sayed can win. And look, it's going to be a blue wave environment. So like, I don't think that most voters are going to like drill down on, did he underperform? Did he not? Whatever. They're just going to be like, if he won, he won. And like the left is ascendant and they can win in battleground states. And if he doesn't win, then I think, you know, there's going to be some recriminations with the Democratic Party, whatever. But, like, this is not just, like, ooh, you know, like, if we can't talk about, like, rape allegations, we have to talk about the fact that there's, like, you know, an Arab-American Muslim guy running in Michigan.

36:59Like, this is a genuinely interesting race.

37:08Pod Save America is brought to you by CookUnity. I love CookUnity. I had it today, John. I had it today. I will have it tonight. I had it last night. I had the chicken meatballs with quinoa, which were very good. And I think tonight I might have the crab cakes. Wow. I eat it. So we're just, look, look at those two different ways you can use it. Cause for me, I don't have it at night. Like this morning, I went to the gym. I raced home. I had pad crab pow with pork. That's what I had for breakfast. That's right. That's right. Wow. And last night for dinner, I had chicken chicken masala after I did love it or leave it.

37:39I went home and had that. So it's cook unity at night, cook unity in the morning. You're talking to some cook unity boys right here. We're the cook unity boys. CookUnity is the first chef-led meal delivery service that makes your meals in small batches inside local micro kitchens. Do you think the kitchens are small too? They're little, it's Oompa Loompas. Across the US. Not factories, so every dish arrives with the kind of freshness you'd get at the restaurant itself. Meals from Michelin-starred chefs, James Beard winners, and Food Network stars. Go to cookunity.com slash crooked or use code crooked before checkout for 50 % off your first week.

38:10CookUnity works with some of the best chefs in the country to bring creative, delicious meals to your door. hundreds of dishes to choose from, menus updated constantly, over 10 dietary preferences, high protein, low sodium, GLP-1, pescatarian, gluten-free, all the good ones, over 25 global cuisines, Mediterranean, Korean, Thai, Mexican, Italian, and more. Taste what happens when real award-winning chefs make fresh, small-batched meals just for you. Taste what happens. Go to cookunity.com slash crooked or enter code crooked before checkout to get 50 % off your first order. That's 50 % off your first order by using code CROOKED or going to cookunity.com slash CROOKED.

39:10to understand or maybe even dare to predict the future of American politics. New episodes every Wednesday starting September 9th. Subscribe to Still Counting on youtube.com slash at Still Counting Podcast or wherever you get your podcasts. Hello, appliance fans. I'm here at Grand Appliance where Bill just bought a premium Bosch Benchmark dishwasher. Talk about your experience, Bill. My salesperson was an appliance genius. She told me about their free install offer on Benchmark Appliances, and it was a done deal for me. Wow, free install from the pros at Grand? That's a no-brainer. Agreed. They make it so easy.

39:49There you have it. Grand for the win. Check out GrandAppliance.com for full details.

39:59Yes, whatever happens, I think people are going to try to find ways to confirm their priors, right? Because let's say, look, I desperately hope that Abdul wins, but let's say Abdul wins and he underperforms compared to other results. People will say, oh, that is evidence that the left isn't as strong electorally. But I think rightly people will then say, but also he was dealing with tens of millions of dollars dumped on his head in a primary that kind of put him back, right? Like people will find a way to sift through the data. But I think part of the reason they don't believe the polls is because instinctively, for reasons based on, I think, some bias, but also just evidence that like, oh, they just assume that a left candidate is going to do worse.

40:42They see the fighting over anti-Semitism. They see like Fox News focusing on all of this. The polls were wrong in the primary, right? They gave them too big of a lead. And so maybe they're wrong again. and it's either tighter or we're behind, even if the model, the flipper model, has him at 60 % to 70%. Nate, is it going to be Lucy in the football with the polls? Or do these polls correct for what they got wrong in the primary? I'd say a couple of things. One is that there is a long, long history, including in Michigan particularly, of primary polls being way off. Bernie Sanders was down 20 points in the polls, in the primaries, and was it 2016?

41:22It was the biggest forecasting error in the history of our polling database, basically. So if you can have a 20-point error, you had, by the way, a 20-point error in Wisconsin in the Hong race. So polling is very, very difficult in general, but much more so in the primaries. In general elections, you have anchors. You know roughly which people are going to turn out and what the demographic makeup will be in a state like Michigan where it's quite regional. you don't really have that in the primary where they're sometimes very high turnout, sometimes very low. Young voters can now be overrepresented in polls because they're online.

41:58That used to be the opposite problem. They wouldn't answer the phone, right? There might be issues with that potentially where, hey, youthful enthusiasm is always a hell of a thing to have. But there's a bias sometimes in which voters poll survey. But the thing is, I think all the indicators are kind of surprisingly rational in this election. If you look at the history of midterms, if you look at special election results, you can talk about Republicans' big corporate money raise, but Democrats are raising a lot more individual contributions on a race-by-race basis. You can look at the aggregate amount of turnout in the primaries for Democrats and Republicans.

42:35You can look at states that have had nonpartisan primaries like Alaska and Washington State, for example. I don't know. You can look at generic ballot, look at Trump approval. So both the polling and the non-polling indicators align pretty well. And the models do say that El-Sayed's likely to underperform. And so he'll probably win by two or three points. But there's a big margin of error on that, in part because polling is inherently uncertain. And he has weaknesses and strengths, right? He's not just a Francesca Hong who seems totally disinterested in reforming her past views and doesn't come across as a very, frankly, charismatic politician.

43:16I just want to – can I put a couple of numbers, lest it sound like I'm fully talking out of my ass? Like when I'm talking about like above average candidate, below average candidate, I'm referring to work that friend of the pod, Lakshad Jane, has done over at Split Ticket looking at do candidates over or underperform a generic Democrat in any given race. And so you can look like who's the most underperforming Democrat in America. It's Ilhan Omar. She underperforms the generic Democrat by about 13 percentage points. Then I think it's Pramila Jayapal and then Summer Lee or something like that. And then after that, it's Elizabeth Warren.

43:49And, you know, Elizabeth Warren, I think, underperforms the generic Democrat by about 13 percentage points. All women and women of color, I will point out. Well, except for Elizabeth—well, I guess there's the debate about Elizabeth Warren if you really want to re-ass that one. I said women, comma, women, comma, and women of color. Just a point. Making me back to, what, November of 2020? But so to be clear, like, there's evidence backing this up. But like most, like, for example, if you look at AOC, she is a talented politician. She performs at about a generic Democrat level. You can be to the left and you can make up for it with your political talents.

44:26But there is just some evidence of candidates like this underperforming. And so I think that is where some of the maybe anxiety within the Democratic Party comes from. And I just want to add one thing, which is to put a number to what you said, Nate. The average error in a governor or Senate primary, so a statewide primary, is 13 percentage points. 20 points is more than that. So yes, it's still a larger than average error. But you should not go into these affairs thinking that the polls are going to get you within two or three points of the answer because they never have. Yeah, the margin of error is like literally, if you calculate it empirically, meaning based on how accurate polls actually are and not the sample size, it's like literally a 25-point margin of error.

45:10When you hear a primary poll, especially for a down-ballot race, a 25-point margin of error, you have to apply that in your head to get somewhere close to reality. Then why are, look, 25 point margin of error in races that are often decided by a couple of points. Like what, why not? Look, let's just take that money and I don't know, like let's do something fun with it. You know what I'm saying? Oh, Nate, I want to ask you one question because you wrote, you know, we've been talking about sort of which candidates overperform, which candidates underperform. Now, there is this problem, right? because even if candidates that are further to the left may do worse in a general election, the Democratic base is really dissatisfied with what centrist and center-left Democrats have been offering.

45:54There's a lot of anger and mistrust because if the bargain is you get less than you want, but we win, and then we don't win, it doesn't really work. And so I think there's like two big problems in there. And you wrote a piece about what your questions were for the sort of center-left as you defined it. Uh, one is, uh, uh, what does it take to persuade people inside of a primary? Right. But more broadly, uh, what is it that the center left is offering that is not just, uh, like Jerusalem Demsos talks about this at the argument, right? Like what is the center left that is not just incremental or cowardly leftist, right?

46:37Or like people that want part of the left agenda. Like, is there a way in which you could have a liberal politician who, in the way that Obama did, in the way that I think Bill Clinton did at the time, that can have a vision that is as big and clear and a worldview as defined as that of a Bernie Sanders, but that isn't from the left? Well, look, liberalism is not the same thing as leftism. And in fact, it's liberalism that was the dominant political philosophy of the late 20th century, pretty much. And liberalism, if you look at how it's defined in Europe, means you want more liberty, so you want free speech, you want free markets.

47:18You also are concerned, though, typically about institutions like rule of law. And liberalism is trying to solve the problem of like, okay, we want to give people freedom, basically, but we don't want to have total anarchy here. We're not libertarians, exactly. And so how do we mediate that? Liberal societies often tend to be very rich as well. So you have the luxury maybe of letting people kind of have their cake and eat it too, different strands of liberalism will redistribute more or less. But yeah, I wish American media discourse separated out those two things, right? To say, I care a lot about individual freedom.

47:57I care a lot about institutions. And I think as an empirical matter that market-based solutions work better than central planning. Those are things that are fairly coherent and consistent and across different strands of that. But like, look, the average Democratic voter or the average media member is not going to care about abstract political philosophy so much. I think the establishment, I think the center-left, right? The centrists or the center-left liberals have kind of let themselves be defined with the Democratic establishment when they shouldn't be. You know, the notion that like Kamala Harris was a centrist, I think is ridiculous.

48:34You know, she is a, the Democratic Party, the main line of the party is a progressive party. Maybe not leftists. I mean, kind of cross-theistic, progressive. You kind of need, I mean, I think what you sort of need is, I mean, Obama was sort of classic where you need like a service policy, right? Like, so he had, we're going to reform the healthcare system. Elizabeth Warren, I felt that that presidential primary was notable because Harris and Warren and Kirsten Gillibrand were all running and they were all talking about childcare as an important economic issue. That felt like a real sea change.

49:14That's like a sort of service policy that you can have. So you need almost like, A, the rhetoric and aesthetics of a big tent liberal party, right? Which is positivity, no sort of litmus tests. I mean, I think Bernie Sanders in 2017 went on this interesting tour where he was endorsing pro-life Democratic candidates. I can't remember what city it was, maybe in Nebraska, something like that. Yeah, and Bernie on guns, right? Yeah, exactly, right. So it's basically like we need positive rhetoric, a big tent, and like two big, splashy bread and butter policies that help people. And then kind of these tertiary things of like, well, here's the social issue that you might care about if you're on the far left and that's something you're voting on.

50:06Yeah, it's interesting. My take on your question, John, you mentioned that, well, Democrats get less of what they want, but they win. To me, I think the exchange is a little bit different, which is we were in a place in 2024 where actually, like, Democrats didn't really want, or, like, Democrats and left-leaners didn't really want like the immigration policies, for example, that Joe Biden was offering. And so what you really want to do is give them close to 100 % of what they really, really, really want. Say that's like healthcare, say it's a public option, something like that. And then you want to sort of break with your party on issues where either Democrats don't really care or Republicans are on the right side of public opinion.

50:58So you say, for example, and this is where the vision comes in, I'm so left on healthcare. You know, I want, you know, X, Y, and Z, whatever. Like I'm Bernie Sanders, but better on healthcare. And guess what? Joe Biden, you do the James Tallarico on immigration. Guess what? Joe Biden completely fucked up on immigration. We need a secure border and we don't think that people should, you know, the position of the Democratic Party from here on out will be that, you know, one illegal border crossing is one illegal border crossing too many and we take this issue seriously. And so the idea of like the set, like what Kamala Harris tried to do was like be in the middle on everything.

51:34That's what like Susan Collins tries to do is be in the middle on everything. And that is, I mean, that can be popular in some contexts. At this point, like, I feel like those years are a little bit behind us. Right now, if you're going to feel authentic, if you're going to feel like you have a vision, right? You want to go left on one thing, on authentically left on one thing, but go right on another. And I mean, we're going to go back to Barack Obama. he was like to the left of Hillary Clinton on the Iraq war, but then to the right of her on basically everything else. Bill Clinton, again, he was like to the left on healthcare, but then he was to the right on crime.

52:06And so these visionary Democrats that you've mentioned are not saying, hey, left, I'm going to give you everything you've ever wanted. They're saying, I'm going to give you the thing that you can't live without, but then I'm going to appeal to the rest of the country on the things where Democrats have shit the bed. Yeah, the thing I would add though, it's like, right, Like that is a kind of whatever mercenary way to look at the different sets of issues, right? But a great politician, both on the matter of being persuasive, but also I think great in the sense of being driven by a genuine sort of set of ideological precepts, right?

52:41Like there is a there is a you can make an ideologically coherent argument for believing in free markets, believing in Medicare for all because you believe in decoupling health care from the job market because of your understanding of the way markets work and believing in having immigration that is sort of generous, but also very well kind of controlled legally. Right. Like a generous legal regime. There is an ideological worldview that would be the kind of classic liberal worldview that could piece these things together with a story that is kind of exciting to hyper-engaged people, in part because it can be both optimistic, it can be sort of big and grand without assuming that the biggest and grandest answer will be coming from the far left.

53:33There just hasn't been a center-left or a kind of liberal figure since Obama who seems able to do that. Yeah, if you look at the polling going back to like the Iowa Democratic Caucus in 2008, Obama won both among the most liberal voters and the most moderate voters. Their moderates were convinced he was a postpartisan pragmatist. And because of Iraq and things, the liberals were convinced that he was a true, I should say, more left progressive, right? A real progressive. Kamala Harris is the polar opposite of that progressives think she's a centrist centrist think she's super woke and far left and to some extent that just gets to like to raw political talent I don't want to keep going back to Michigan you know I think El Saeed has a fair amount of like raw political talent I do well it's rhetorical when you say I don't want to keep bringing up this thing you're like Susan Collins and you don't want to talk about Troy Jackson's second cousin who cares who slept with who's second cousin not me Susan Collins Rob Sand might be among the more talented candidates of the cycle.

54:34I think that's a great example because this is what he's doing. He's running against both parties. Rob Sand is, yeah. In an environment where the Democratic Party is not popular and the Republican Party is not popular, you want to put yourself above the partisan fray. And Rob Sand does that literally every campaign. He does. Rob Sand does something that I see. First of all, that great interview with Jonathan Martin where he's just like, nope, nope, nope, great. But I saw him talking about education in a town hall or just a small sort of whatever political meetup of some kind. And he basically was so straightforward.

55:09He said, look, I'll be dealing with a Republican legislature. These are the things I think can get done that I think would be helpful, right? There's a kind of like treating people as if they're smart and being pretty direct. that I think because it's a state race and not a national race, I think there's like some space for that, right? In part because you don't have to be as worried about kind of the traditional partisan dynamic and anything you say and kind of being dumped on your head from either the right or the left. I was there, like I talked to him a couple of weeks ago and went to a bunch of the town hall.

55:42Like I kind of did the journalist day with Rob Sand when I was out there for other stuff for my book. And it's interesting. There's a couple of things that I think Like, one, it's like he's been in the state for a long time doing these kind of, I guess wonky. Like being a state auditor is a little wonky, right? And he wrote like the book about the big lottery scandal in Iowa. He kind of does the Iowa thing where it's like, I'm going to show up. I go to all the town halls. And he is a little bit, he's kind of like when he's on the stump, he's kind of doing the like, aw shucks. but then there's like a little acid thrown onto it.

56:20Like he'll like twist the knife a little bit, which is interesting. He's also a former runway model. So people like looking at him, right? Like there's also like the thing. This is the first thing Claire showed me when she got back from the trip. She's like, have you seen his photos from Milan? I don't think the Milan photos are really being, are a huge help in Cedar Rapids. But what is true is that he did go to, he went to the East Coast for college, but then in a way that I think like, you know, people with political ambitions do, like, and also he grew up in Iowa and wanted to go back there.

56:52But like, he went to Iowa for law school, right? He like stayed in the state. But when you actually hear him talk about issues, he often does skirt. Like I went to a town hall where a candidate, a voter kind of like, kind of laid out this like question for him to take a big smack at like big ag. And he didn't write big agriculture business in Iowa. And he didn't, right? Which I was like, oh, that's an interesting thing that just happened here because he is dealing with a, you know, a state with vested business interests and also, again, a Republican legislature. But he was like, I texted a couple of people and was like, this is like a very notable political talent in the same way that like my former classmate, John Ossoff, is really good on camera and a very sort of poised presence.

57:41And those are, there are people who are, I don't know what John is like on the stump, right? John, we're going by first. We're going by John. No, but like, I don't know if he's good on the stump, but he's very good, like, in the way that AOC is very good, sort of being the sort of big communicator via video, which is incredibly important in this age and like Obama, you know, like Obama. And that we haven't had in quite some time.

58:14This episode is sponsored by BetterHelp. While we're big fans of BetterHelp, the real story is the feedback from people using it every single day. That is the real story. That's the real story. The media won't tell you about it. BetterHelp has an average live therapy session rating of 4.9 out of 5 based on over 1.7 million client session reviews. The feedback from real people who have actually used the service speaks for itself. More than 6 million people have used BetterHelp globally. BetterHelp therapists have at least three years and 1 ,000 hours of hands-on experience. They'll hold all required education, training, and certifications across a range of mental health specialties, meaning the perfect therapist for you could be a few clicks away.

58:46Better help makes it easy to get started. Answer a few questions, get matched with a therapist based on your needs and switch therapists at any time if it's not the right fit. Therapy, it's so important. So important. I got to get back into therapy. Oh, wow. Yeah. What are you doing? I'm doing Love It or Leave It. I was talking about the intrusive thoughts I have just from the walk from backstage onto the stage and it horrified the crowd. Yeah, that's a perfect reason to go to therapy. I don't know. That's the better place for that. John, do you ever - Maybe you can get better help. If I'm on the street, all right, and like, I will see like, let's say there'll be a commotion or people arguing about a traffic spot or something, or, you know, just see something.

59:25I not only imagine them fighting and then a crime and then a death and then a trial, I am standing on the street. I am testifying at that trial. Wow. Like that's fully the full story unfolds in a second of two people just being mad in a parking lot. Oh my gosh. Yeah, that's something I should get talked to somebody about. Yeah, yeah. I'm just thinking about the news. Yeah. For you, it's the news. Visit betterhelp.com slash PSA. That's betterhelp.com slash PSA. I'm Nate Silver. I'm Galen Druke. And I'm Claire Malone. We're back, this time with our friends at Crooked Media for a new show called Still Counting.

1:00:14Numbers are hard. Takes are easy. And frankly, numbers have only gotten harder. What does the data actually say? How do we reconcile that with the other things that we're seeing in American politics and around the world? And how do we use that data to understand or maybe even dare to predict the future of American politics? New episodes every Wednesday starting September 9th. Subscribe to Still Counting on youtube.com slash at Still Counting Podcast or wherever you get your podcasts. Hello, appliance fans. I'm here at Grand Appliance where Katie just bought a new Samsung laundry pair. Talk about your experience, Katie.

1:00:51My salesperson was amazing. He knew everything about appliances and was even able to get it on a delivery and install route for tomorrow. Wow. Next day delivery? Yes, my laundry is piling up, so this is a lifesaver. There you have it. Grand for the win. Check out grandappliance.com for full details.

1:01:15Now, I could keep talking to the three of you, but I do want to just close with some very quick, and as I promised, quite pedantic questions about modeling. I'm going to bring the model up on my laptop. These are, all right, I'm going to go. You can take these. I'm going to try to get through a few of these. All right. In your deluxe model, there are, let's say, three kinds of data. Polling, fundamentals, and ratings by expert. Presumably the fundamentals, incumbency, say economic conditions, fundraising, whatever you're incorporating, are in some way reflected in the polling. And then the ratings by experts are also reflective of the polling.

1:01:52So I just want to understand qualitatively, what does it mean to include data that is dependent on the other data? For example, does the fundraising data slowly recede because the fundraising advantage would predict a change in polling as the election gets closer? Or does the fundraising data continue to matter as you get closer to the election day, even though presumably that would have to be accounted for in the polls if it does matter? So we've lost. This is the part where Claire and I pull out. I'm so sorry. I just start doing shots at the table. You guys, I'm sorry. Is this okay? I really have.

1:02:25These are things that have annoyed me for a long time. Okay. Can I just, can I do that? Yeah, please do. Okay. I'm sorry. Answer number one. Polling gets more important as you get closer to the election. It's more accurate. You get more of it. So it becomes more of a polls mostly model by election day. Number two, the reason we look at fundraising is because it's a measure of grassroots support. and one of the few things you have that isn't a poll that is actively data collected within that election that shows people movement or people power. So we only look at money raised from individual contributions, not PAC money.

1:03:01We actually put a big multiplier within state contributions because those are much more likely to reflect how voters in that state feel. So like I think - And how do you decide what that multiplier is? Like how - Empirically. Yeah. Empirically, what could that possibly mean? what the fuck does it mean to empirically weigh the value of a grassroots donation come on John do you know that Nate went to the University of Chicago he does I feel like we've been lured here under false pretenses this is what this interview is really about I want you to know Nate I almost went to the University of Chicago Law School and I spent two days at that place and after that experience which felt like six years I did not enroll go away so but seriously how do you empirically know Draw a scatterplot.

1:03:49We have lots of data for Congress. That's one nice thing. We have data on every congressional race going back to 1990s, so thousands of races. Draw a scatterplot of fundraising versus the outcome using all fundraising data. Then draw one using only in-state fundraising data, and the in-state is a tighter fit because that reflects, like, how much money you have to spend in a Senate race? But that's not my question. But that's not my question. I agree that in-state fundraising is more valuable than out-of-state fundraising. I am asking, how can you possibly decide what the multiple is of the in-state fundraising?

1:04:20Because whatever you're doing, you are weighting the value of the fundraising in total. So the weight that you're applying to the in-state fundraising in some way determines the impact it has on the overall model. And I'm saying, how do you decide empirically how to weight that information? Through the magic of multivariate regression analysis. Yeah. Because we have a lot of data, you can take a bunch of factors that, you know, in Michigan, El Saeed's raised a lot of money, by the way. That's another factor. On the other hand, El Syed is very progressive. If you have thousands of races, we can compare those two things.

1:04:53You can disaggregate and say which factor is more important over the long run. Just like with, it's kind of the same thing as like the money ball type analysis where you can say we have hundreds of baseball players. We can say, what contributes to winning? Is it batting average or OBP or home runs or pitching or whatever else? It's the magic of big data. You look skeptical. No, I'm not skeptical. I, I, this, I'm not, I'm so, some of this is skepticism. Some of this is genuine curiosity. I'm not really sure what the balance is. Well, I would have to, we'll have to run a bunch of simulations to figure out what the answer is there.

1:05:24But, but I guess because part of it, right. Is then, so you use that scatterplot to determine the influence of the, of the fundraising. Right. But a lot of that data is then also going to be in the model in other ways as well. Right. Like you're going to take that outcome data that has determined the value of the fundraising. It's going to go in the model in another place. And I guess when you're running 40 ,000 simulations, okay, how smart are those simulations? Do those simulations inside of each simulation, is there an understanding that the quote-unquote random, I'm not using that word, I'm saying the randomized sample that is taking place inside of those 40 ,000, is that a smart randomization that understands the ways in which various data is dependent?

1:06:05It's a super fucking smart randomization, John. I'll tell you that much. No, there were a lot of models in 2016. Famously, the Princeton election consortium model had Trump with only a 1 % chance of winning. The New York Times model had him at 10 % or something. Our model at 538 had him at 30 % instead of 29%. And the reason why is because our simulation architecture was better. Because errors in the polls are correlated. Okay. So the correlation of the errors is accounted for in the random sampling that takes place when you're running the 40 ,000 samples. And we can do fancy stuff, right? We can say one simulation will test what happens if Democrats really overperform among the Hispanic vote, right?

1:06:53That has a big impact in Texas and Florida, maybe not so much in Michigan, for example. Or what happens if Republicans have low turnout, but Democrats win the persuasion battle and things like that? So yeah, we're actually, we have micro data on every congressional district and how they're related. And then how do you make sure that in your random sampling, in which inside of those random samplings, there is an understanding of the relationship between the data, that you are then randomly sampling what the correlation could be when you're running those random samples? Do you know what I'm saying?

1:07:30Because I'm really good at this and I've done this kind of thing for a long time. No, I'm not. It's hard. It's hard, but this is what I'm trying to understand. Like, I'm sorry. I'm like putting on a—I very rarely get to just sort of like, you know, math out to this degree. You have to have a lot of knowledge of the system you're trying— So one very important thing about politics, most important thing in some ways, since the 1990s, we've had more and more polarization. Oh, no. I'm sorry. I'm sorry. Yeah, this is my fault. Is what's happening here basically like, you know, when you go to your doctor and like they're really smart.

1:08:05But at the end of the day, the doctor, being a doctor is an art as much as a science. Is that what you're trying to say here? That it's as much of an art as it is a science? Or am I— Well, first off, when you make predictions every year, you get a lot of feedback, right? You can say, where'd my model fuck up? And you're being accountable. We're putting this out there for the low price of— Do you ever take the model and go back and run it on a previous election in real time to see how it would have done? I mean, inherently, you're using past data to guide the parametrization of the model. But you have to be careful because we have these things that are called overfitting in statistics where you can be overly literal.

1:08:41Oh, a candidate whose name starts with C and ends with N and is a woman, they always lose the general election, right? So you don't want to nominate anybody with a last name C if they're a woman kind of thing, right? You have to look at – this is why you actually have to know something about politics. We can't use the same model that we would have when we first had a congressional forecast in 2010 because the climate is much more partisan. Things like candidate quality, things like incumbency matter much less than they once did. And you can observe that in the data, but also you can observe it by being a smart person who understands American politics.

1:09:19Right. But this, that's where I sort of, this is where I struggle to, I like have like a philosophical, I struggle. I'm not, I'm not saying in a pejorative way. I genuinely struggle to understand what qualitatively a model is given that every election is unique and only run once. And that the polling on which you base it is both, is both kind of, you're trying in the data to account for both the unknowable unknowables of what will happen and the unknowable knowables of what people think at this exact moment. And it's hard to, and I don't know if the model knows the difference. What I think people miss is that unpredictability is predictable.

1:09:56That people should know. Wow. Hey, man. People should know that in a gubernatorial primary in Wisconsin, where literally three candidates dropped out within the final month of the race, you don't have any fucking clue where the race is at based on a bunch of shitty internet polling by college kids, right? That's predictable. We don't forecast congressional primaries because the model would have these margins of errors of 25 or 30 points. It's predictable that in midterms, the error used to be a little bit higher than in presidential elections, although it's actually reversed lately to some degree.

1:10:33If you don't have a lot of polling in a race, if you have contradictory data, you should be less confident about those forecasts. But look, a model is, in a literal sense, it's a computer program. It's a computer program designed by me. Yeah. Did you get that, John? Yeah. It's a structured way of thinking, right? It's saying, okay, here's all the shit I know about American politics. And here's all the data I have. And we have lots of data in politics. If I'm looking at the model, am I really looking at you? Is the model but merely a mirror? Claire and Galen, I'm so sorry. No, it's fine. He's not telling you some things, I'm sure, because it's IP protected.

1:11:11Oh, that's what... Because I would love to, I really am like. We have these 10 ,000 word methodology sections. I want to get it. So Claire and Galen. Now I had pitched a name for this show and the name was Alpha Dogs. And nobody but me liked this name. And I just want to understand. It's too male. It's too male. It's too male. It's too male. Yeah, it's too male. It reads male. Even though I do believe I am an alpha. It is too male. I didn't even realize that this was one of your pitches. The main pitch that I don't even know that it had your name by it, but that I assumed was from you was Showpole.

1:11:46No, in fact, it wasn't. But thank you so much. I think Showpole would have been a great name for this. I put my initials by that. I was like, I can live with that. But perhaps, Claire, too male? Too male. Too male. Thank you to all three of you for taking the time to talk today. We're so excited about Still Counting. Everybody, please go subscribe. Nate, thank you for indulging me on this. I really am interested in like the... Sometimes I actually think my problem is just really that I worry that... And this is what I worried about certainly in 2016, was that people, they're not as responsible with what the model means as what you're responsible about.

1:12:30They don't take what you're saying about it. They just take the number. But I am just sort of interested in the sort of what it means to be modeling an election and the difference between like sort of prediction and sort of the whether a model is predicting the future or trying to predict the present, you know. But for another day, I think, for another day. And so everybody subscribe to Still Counting right now. Right now. Galen, Claire, Nate. We agree. Come join us. The pool, the weather, the water is warm. We have lots more data and hot takes than folks are today. That's for sure. We didn't even get to Calci.

1:13:16I wanted to talk to you about Calci. I'm afraid to start making bets because I think it's going to be a problem for me because I, you know, I got to - Don't get George Santos. I've got the devil. I've got the devil. I've got the devil in me. Should we start a crooked in-house fantasy league? How often will Nate swear on the - Where we bet fake money on Calci? on politics, because obviously we don't bet on politics for like a sense of propriety or whatever. Because of NCAA rules. Because of NCAA rules. But we could start a fantasy league where we do bet on politics. Rejected! We could see who wins.

1:13:45Thank you all so much, and we'll see you soon. Thank you. Thank you to Gail and Nate and Claire for joining. You can subscribe to Still Counting. The first new episode is going to be up on September 9th, and we at Politics of America will be back in your feeds after the holiday, and see you then.

1:14:22I'm Nate Silver. I'm Galen Druke. And I'm Claire Malone. We're back, this time with our friends at Crooked Media for a new show called Still Counting. Yeah. Numbers are hard. Takes are easy. And frankly, numbers have only gotten harder. What does the data actually say? How do we reconcile that with the other things that we're seeing in American politics and around the world? And how do we use that data to understand or maybe even dare to predict the future of American politics? New episodes every Wednesday starting September 9th. Subscribe to Still Counting on YouTube.com slash at Still Counting Podcast or wherever you get your podcasts.

1:15:01Hello, appliance fans. I'm here at Grand Appliance, where Bill just bought a premium Bosch Benchmark dishwasher. Talk about your experience, Bill. My salesperson was an appliance genius. She told me about their free install offer on Benchmark Appliances, and it was a done deal for me. Wow, free install from the pros at Grand? That's a no-brainer. Agreed. They make it so easy. There you have it. Grand for the win. Check out GrandAppliance.com for full details. Thank you.