98. Nate Silver: Can we predict who will win the 2024 US election?

24:24So you quickly see how utilitarianism becomes self-justifying and self - rationalizing. But also I think he's a degenerate gambler. You know, he really likes to make bets and really likes to have action. I spent a year as part of the book betting on NBA games, the basketball league in the U.S. And like you have a lot of adrenaline from like making trades and making bets. He had built a nice, not quite safe, but like a, you know, a crypto exchange. You don't have to actually make bets, right? You have to have a good user interface and customer service and whatnot. And you kind of take fees. and he couldn't abide by the idea that he had$10 billion sitting in a bank account.

25:02And all he had to do was change some code on the back end and wire it to his girlfriend's hedge fund. And then he can like make$10 billion worth of bets, which he thought he couldn't lose. That's not quite true. He actually did think he could lose, but he decided that like, it's so fun to make bets and the opportunity to become a trillionaire. I mean, he thought about becoming the world's first trillionaire. He gave himself a five to 10 % chance of becoming the American president at some point. And he was really, really, really willing to gamble on the upside. How do you think you'll be handling jail?

25:32Maybe he'll be a bit crafty and entrepreneurial. I suppose I talked to his co-founder at the original Alameda, who was named Tara McCauley. And she said that actually, Sam is kind of a person who doesn't enjoy, he self-diagnosis is having anhedonia, which means an inability to feel pleasure, right? So most of us might feel, okay, we work hard, but at the end of the day, we go and I don't know what, you know, we ride in our motorcycle or we, we have sexual relations or we eat and drink or do drugs or whatever other things we enjoy or travel and get the thrills from doing that. And that might motivate us.

26:05Sam seems to not feel those things as much to him, you know, being in like a nice prison might actually be in some ways just fine. There's order and routine. I mean, you know, again, his ex co-founders told me on the record that he thought maybe going to prism wasn't that much worse than his everyday life. Because people thought, oh, it's Bahamas, it's party central. I think Sam is not somebody who's a big partier, for example. He's about kind of work and I think he derives his thrills from trading and gambling, but not from pleasures of the flesh, so to speak. Tell us a little bit about gambling.

26:39I mean, that's another very big feature of every civilization, but also particularly of our culture. And you were a gambler. Tell us about gambling. Tell us about the difference between being a professional poker player and somebody who bets on politics? I think all these distinctions are fluid. And one of the more obvious, I suppose, conceits of the book is that the big successful people and the big failures in Silicon Valley and tech are fundamentally gambling. Now, they're gambling with bets that have a positive expected value in the long run, which you're not if you're playing roulette or something in a casino.

27:09But the personality types all across the river are more similar than I expected. You know, look, gambling is also interesting on its own as a cultural phenomenon. You know, almost every culture, probably 90 % of culture throughout the history of some form of gambling. It was related to divination originally of trying to understand fate and things like that. You know, the fact that different cultures have different forms of gambling. The UK kind of sanitizes a bit, right? You have like the bookmaker on the corner of the street, right? You have a lot of ads for gambling. But these nice kind of, you know, we kind of in the UK, they kind of civilize vice a little bit.

27:42Whereas in the US, we celebrate it. We have the biggest casinos in the world, outside of Macau, at least, in Las Vegas. But you couldn't bet on sport until recently. You couldn't until recently outside of New Jersey and Nevada. That's true. Yeah. So you don't have that embedded in the nationwide culture in the way that we do. But now that you can, of course, Americans go away overboard with it, right? And we've had multiple betting scandals. Shohei Otani, maybe the best baseball player in the world. His interpreter was a D-Gen gambler who bet$15 million of Shohei's money on non-baseball sports, or so he claims.

28:15We had a player suspended for life from the NBA for trying to win player prop bets, where you get two and a half rebounds and not three rebounds by faking injuries and things like that. And the proliferation of advertising is, so yeah, in the US, we tend to go all out or all in rather on things. But if I'm right, the two biggest betting markets ever in the world have been the last two American presidential elections. And it's getting bigger and bigger. And by the way, I should disclose, I consult for a prediction market called Polymarket. Right. Which is an online betting site for politics or?

28:47More or less, yeah. And, you know, there's serious money in this and people are like actually wagering a lot. But again, I would say like, you know, people on Wall Street, because everything is political risk and geopolitical risk, like, you know, Wall Street has been forced to analyze what's the risk of Trump winning again or of Kamala Harris winning and various combinations of Congress and things like that. You know, yeah, I mean, they were worried about the UK election in India and France and all these places, too. But like, if everything becomes politics, the financial markets need to price in political risk and betting markets are an efficient way to do that.

29:22Okay, Nate, Rory, quick break.

29:52USA. Fixed Broadband Experience Report August 2024.

30:25One of the things we've been talking about a lot recently in conversations is where we think the world is going in 10, 15 years' time, and in particular, the balance between democracies on the one hand and populism or autocracy on the other. From your point of view, if somebody asked you to make that kind of prediction, sort of 10, 15-year big picture view on whether the world's going to get more democratic or more autocratic. Is that something you can even make sense of with numbers? What would be your response if somebody asked you to make that kind of prediction? I don't know you can make sense of that with numbers very much.

31:01I mean, you know, I suppose the book does a little bit of rumination on kind of where we stand as a society now. And I'm worried for liberal democracy, I suppose. I mean, the conceit of technology is that, which I agree with, I'm a techno-optimist, is that for many generations, technology has helped lift human beings out of poverty and extended human lifespans and improved quality of life. And we've even been generous enough to expand rights to people who aren't just straight white men, slowly a little bit at a time, not enough, obviously. And it's all kind of gone quite well, but you can make real cases for stagnation in the West.

31:38You look at, again, EU or UK GDP, and the pie is not growing particularly fast. You look at US lifespans, especially for men, and they have leveled off. You know, the pace of innovation, well, things are impressive. I mean, certainly, you know, I rode in a driverless car in San Francisco two weeks ago and I'm like, okay, we actually are making progress still. This is very futuristic and cool. But in other ways, like the recent round of like technologies, social media, is that from a utilitarian standpoint, net positive for society? I think probably not. Mobile phones, I think probably, yes, especially in the developing world, but maybe for like school children, I think the evidence is like, maybe not potentially.

32:17And, you know, so although there are things like Ozempic seems to be a very good pharmaceutical and mRNA vaccines did a lot of good. So we do see progress, but like, it's a little bit more even handed. And then also I worry about like, you know, even though I'm kind of like University of Chicago, which is kind of a neoliberal economic school, right? That was my undergrad degree. I'm an unapologetic neoliberal, I suppose. However, when you have the wealthiest 10 people in the world getting wealthier by 20 % per year. Well, I know how compounding interest works and that becomes very wealthy indeed, but they're kind of eating the world and functionally speaking, Elon Musk is almost treated like a quasi-governmental entity, right?

32:58The government of Brazil needs a policy that I disagree with, by the way, but a policy of an Elon Musk containment policy in essence. And the British Prime Minister, Rishi Sunak, sits down and it wasn't clear who was the more powerful figure as they sat on stage together. Absolutely, right? And that's probably genuinely debatable whether Rishi Sunak or Elon Musk actually has more power over the fate of the world. And that to me seems, I don't know. I mean, look, in some ways, liberal democracy has been remarkably resilient, and there are reasons for that. But it hasn't been that long, right? I mean, progress has only occurred.

33:30If you look at growth in GDP, for example, there was no progress for 95 % of human civilization. Then around the time of the Enlightenment, the Industrial Revolution, In the 1770s, 1780s, roughly, you begin to see a spike in progress. It's continuing ever since. But I don't know that we're necessarily guaranteed that exactly. And with AI in particular, most technologies are disruptive because fledgling startups are able to found them and disrupt these big bloated enterprises, right? With AI, because the amount of computing power that you need, and it's very much more a winner-take-all technology.

34:05So, you know, Meta and Google and Amazon and the Chinese government, OpenAI, Anthropic, right? Those six or seven basically are going to gain even more power to the extent AI is more powerful. And that's a change from the traditional paradigm of disruption in Silicon Valley. My final question, two questions in one. Who do you want to win the American election and who is going to win it? I will be voting in New York where my vote doesn't really matter for Kamala Harris. And for me, you know, I'm fairly centrist. To me, the decision is simple, which is that I thought that January 6th was horrifying and disqualifying.

34:45If Biden had run, I thought it was profoundly irresponsible to nominate someone to be president who was 86. I would probably have voted for some third party. But I have no disqualifying reason to vote against Kamala Harris. I do have disqualifying reason for Donald Trump. And so that's that's how I'll vote. And I'm transparent about that when I make our forecast. Look, if you had an election today, Harris would be favored narrowly. I think on November 5th, which is election day, closer to 50-50. If anything, her momentum has stalled out a little bit in the polls recently. For the purposes of this broadcast, I'm going to say 50-50.

35:19Somebody's got to win, though. Your famous 538, that was called 538 because that's the electoral college. Yeah. Somebody's got to take it. You know when you make a bet, you go to Ladbrokes or Paddy Power or whatever, right? They take a 4 % cut, right? I don't think it's worth betting with the 4 % cut, basically. I don't think I have enough profit. I think it's close up to 50-50. Partly too, we just had Labor Day in the US. So there's kind of like a pause on polling. So we don't have a great measure on where the race stands now. A week before the first debate, two weeks after the convention, we're kind of looking back in time a little bit.

35:52That was good for Kamala, but she's fading a little bit. And so if you look at our model, it says 55-45 Trump, actually. So you can abide by that, but that might change in the span of a week and probably will change in two weeks time after the debate. Nate, what are the factors that we tend to miss? So let's say we get this predictions about this forthcoming American election completely wrong. What would be two or three things that we're likely to have missed that can explain how we screw these judgments or predictions up? Well, it depends on in which direction it's wrong, right? You know, if Trump wins by a wider margin than people predicted, I think you might look toward what's happening toward younger Black, Hispanic, and Asian American voters in the U.S.

36:36So for a long time, because of the civil rights era, that African American voters voted overwhelmingly Democratic in the United States. But those memories are fading. And there is some polling data showing that among younger Black voters, especially younger Black men, that they may be more open to voting for Republicans. And same for younger Hispanic men. Voting for Republicans or voting for Trump? Is it a Trump thing? Good distinction for Trump. For Trump, maybe more than other Republicans. And what do they see in him? Partly it was when Joe Biden was the candidate that, you know, Biden was very alienating, I think, to younger voters.

37:11I mean, he comes, he was 81 years old, going on 82, would have been 86 by the end of his second term and comes across that way. I mean, his references are dated. And remember, like younger voters in the U.S. have not gotten the candidate they want for quite a long time, right? They backed Bernie Sanders in 2012, got Hillary Clinton, backed Bernie Sanders in 2016, got Joe Biden, did not have a competitive primary in 2020. And so losing younger voters is a real concern, I think, for Democrats. And also maybe feeling as though like they were able to take for granted, like the black vote, for example.

37:42I mean, if you ask black voters, Hispanic voters, Asian voters, they often say, actually, I'm pretty conservative culturally, but I appreciate the welfare state and I maybe trust the government to do something, some redistribution, protect the working class. So as Democrats have become more college educated, focusing more on culture war issues that I think can be alienating to working class voters of color, potentially. Do you feel that Kamala Harris's speech at the convention was speaking to that? I thought she did something smart at the convention, which is that she was speaking mostly to men.

38:18If you look at the rhetoric, it was about the United States is competing with China and we've got to kick China's butt, right? And she said, our military force is really lethal. It isn't that cool, right? We're really strong and powerful military. An unusual speech for a Democrat, unusual speech for a female Democrat in particular. and she did not talk about like race or gender unlike hillary clinton who comes out in this white dress and is introduced by her daughter and you know glass ceiling shattering animation when she clinches a primary there's like a glass ceiling literally on the screen that's shattered right kamala harris did none of that did not mention gender or race very much at all and so i thought it was a smart play to go for for the center potentially why do you think it's faded a it since then?

39:01I think in part because she had a period of several weeks where Trump seemed very much on his heels and her press coverage was very good. I mean, people have short memories in the US and the media is impatient. Like ultimately people are correct that the media suffers from some degree of groupthink. And we have these really long elections. This is for us a short election, only five months, but it's not five weeks like you have here. And so the narratives change and the storylines change. I think they were maybe caught like a little bit flat footed as far as like they were going off vibes, right?

39:31They were going off momentum and not providing a lot of policy substance. They were very slow. The Harris campaign to do interviews. I think they got a little bit paranoid, actually, where the Biden campaign was very reluctant to have Biden do media hits because he screwed up half of them. Harris doesn't have that issue. I mean, she might not be the most natural politician in the world, but you're not running an existential risk anytime you go on television with Harris like you were with Biden. And so, you know, she retained a lot of the Biden advisors, which I think anyone who worked in the Biden campaign should never work in politics again, basically.

40:01It was one of the worst run campaigns, one of the worst ideas to try to run this man who was clearly too old, right? They should all be banished off to Crete or something as far as I'm concerned. So you've talked to us about some of the ways in which if Trump won by a bigger margin predicted, that might be because we've underestimated young, male, black, and Hispanic votes. If the Democrats win by more than we currently anticipate, what might we have missed? I think you might look at enthusiasm. I mean, clearly, it's an advantage to have, you know, so Democrats are the people, the political junkies are mostly Democrats now and political junkies turn out and vote.

40:37So we saw in midterm elections, for example, or we have various special elections, which I guess you call by elections here or something, you know, Democrats tend to do very well in those. They're raising lots and lots of money. They have booked more advertising in the closing weeks of the campaign. They'll probably have better data and turnout operations. And so, yeah, I mean, it would kind of be, hey, actually being like the educated party that has really motivated voters is actually quite an advantage potentially. Maybe you'd say the pollsters are so terrified to underrate Trump again that they wind up underrating Harris instead.

41:11That wouldn't be completely shocking, I suppose, from a psychological standpoint. My final one then, and then back to Alistair, you will be under pressure a lot from companies and individuals who want to think about the next 5, 10, 20 years because they need to often make investments in bets, which are 5, 10, 20-year investments. How do you give honest advice on that when presumably your mathematical models suggest that it's very difficult to make accurate predictions that far out? I mean, is it not a little bit like predicting the weather 5, 10 years out? Well, I mean, weather's a good analogy, right?

41:43Where we can, at high fidelity, forecast the weather about three days out, right? And somewhat high fidelity forecast the weather about 10 days out, but like long-term weather forecasts are not much better than random. No. So part of what you're trying to do is where do you have an edge? And often when you have an edge is in short-term forecasting, you know, so on election night, a lot of money will be made or lost based on who was able to read the initial data that comes in from certain states or precincts and extrapolate that on election night to who's going to actually win the electoral college when it's declared a day or two later.

42:14Right. So that's the kind of thing. Yeah. People don't have a good sense for the degree of difficulty involved in long-term versus short-term predictions. I said my final question earlier. I'm just going to have one last one. You mentioned J.D. Vance, who has sucked up a fair amount of attention since Trump appointed him. First of all, do you think that was a mistake to appoint him? And given we've talked earlier about Pennsylvania, do you think that Harris made a mistake in not appointing Josh Shapiro? Or have you, a bit like Rory, falling in love with Tim Walsh? I think they were both mistakes.

42:49Look, I think that Trump picked J.D. Vance at a time when he had just been shot, or well, I guess he was shot, grazed with a bullet, which could have been quite a bit worse if he had turned a different direction. You know, at that moment, I was actually in a poker tournament at Las Vegas, right? I just busted out of the poker tournament and texted my friend Maria, I'm out of the tournament. And by the way, someone shot Trump, you should probably turn on the news. But it's a moment he was riding high when it seemed like Democrats were not going to succeed in their effort to push Biden aside. Even I at that point was like, oh, forget the models, right?

43:17This is just divine intervention that Trump is going to win this election. So he picks Vance at a moment when he thinks he can't lose and that he'll therefore cement the MAGA, make America great again legacy for the future. Vance is young and clearly quite bright, right? But, you know, that doesn't sell very well to the average swing voter in Pennsylvania. I think Walls is closer to being a defensible, I think, On his own, he'd be a perfectly fine choice, right? I think they were a little bit short-term focused, right? They wanted to kind of keep the good vibes going. They had a lot of momentum, but it was, you know, but it's July and you have to continue the momentum until November.

43:55And there's going to be something that upsets the apple cart at some point. And so I would just have made the percentage play. Pennsylvania determines the election results about 35, 40 % of the time. You have a very popular governor in Pennsylvania who clearly wanted the job, very talented. and yeah, we just have said, let's look toward the long term and like in the long term, momentum, momentum, I want that extra point in Pennsylvania come November. Will you personally be putting a bet on the election? If they're both kind of, you basically say it's a toss up. If it's even, so you do back. Either person.

44:25Yeah. Then I'd just be wasting my money, right? I mean, or you could have a gamble and flip a coin or something. No, on election night, I'll be doing, as well as writing for many other subscribers, I'll be doing some consulting work. I will probably not be betting directly on the outcome, but the people I'm talking to might, and I have no moral problem with that. Well, thanks for all your time. Thank you, Alistair. Thank you, Brody. Thank you very much for coming.

44:48So, Nate Silver, Mr. 538. I mean, obviously, I see him as an extraordinary example of all these different strange strands in American culture, which we kind of touch on glancingly, but right at the heart of his world. It touches with, as I said, your interest in sport, political betting, politics, Elon Musk, effective altruism, Silicon Valley, quality tax. I do think he's a really interesting route through. You can see his personality is obviously very, in a sense, mathematical, risk-based, quantitative. He shares a lot in common with, I guess, the mindset of the sort to Teals or... But he's not of them.

45:35But he's not of them. No, he's not of them ideologically at all. I think I got it wrong, didn't I? Because it was 2008, he got 49 out of 50. 2012, 50 out of 50, the States. 2020, 48 out of 50. But 2016, I'm not having that where he said, no, I wasn't wrong. No, I wasn't wrong. What he means by that is that... He was less wrong than other people. No, no, he won on the betting, I guess. Yeah, yeah. But no, if you're saying 71 % chance, if I read that rightly, he's basically saying Hillary's winning. Yeah. Right. Just for the historical record, Hillary, Alaskan, did not win. And it is quite amazing if you think, I can remember at the time, this hype that there was around him because he was the guy, the one guy who successfully predicted the election.

46:15But most elections, you know, he admitted, didn't he, that, you know, Alabama and California, it's not a difficult call. And I thought he was quite, you know, are there really 10 states in play? Well, as he said, something pretty marginal. Yeah. You were good on pushing him on Pennsylvania. I remember though, I mean, I don't know, like me, you sat up through that night of 2016 and there was a New York Times tracker. Given that Nate was central to the New York Times team, I don't know to what extent he helped design that tracker, but there was this awful thing where they had the probability of Hillary Clinton winning.

46:46Oh, they kept going. Yeah, yeah. It started the night, I remember 95 % chance. Hello, just through the lights, it was going down and Trump was going up. I remember that now. Presumably, this is one of the reasons why you're a bit skeptical about polling. Obviously, see a lot of pollsters listen to this, get very angry with us saying this. But again, the truth of the matter is people basically got 2016 wrong. They got the Brexit vote wrong. They got the 2017 British election wrong. Well, it's not that they got them wrong necessarily because they presumably were recording accurately what people were saying.

47:18But I think the other thing in American, I talked to another pollster out in Chicago. He was saying that a lot of their polling, they don't even check whether people are registered. So you can ask people how they're going to vote, who aren't going to vote because they don't have a vote. So yesterday, Fiona sort of said, oh God, have you seen this poll in America? And it was like Harris's was down a bit. But it's like a nationwide poll. It's utterly pointless. One, it's a vast country. Two, that does not speak to the outcome. What it does say is, you know, you'd rather, if you're the Democrat, you want her to have a reasonably sized lead because you're going to need that to carry the Electoral College.

47:56I'd far rather that there was every week a decent poll in Michigan, a decent poll in Wisconsin, a decent poll in Pennsylvania, et cetera. Well, let's maybe finish on what I thought the kind of killer point that you made at the end was, which is this question of whether if it goes wrong, people are going to come back to, you should have gone with Josh Shapiro. And the Wolfe decision is a weird luxury. A governor of a state that doesn't really matter in this election, that's voted against the Republicans every election since Nixon in the early 70s, my whole lifetime, as voted Democrat. I mean, I don't know.

48:29I'm jury out on whether it makes that much difference who the vice president candidate is. I think you may be right, actually, because I have a sort of in the back of my mind when Gore was vice president, he didn't manage to carry his own state. I don't think he carried Tennessee. I think there's so much focus on it. I think you might be right that Trump didn't really care about who was his running mate. And he just, all those Peter Thiel people saying, you must take Vance, you must take Vance, you must take Vance, and he took Vance. I think he probably is regretting it. Anthony Scaramucci is convinced that he's regretting it.

48:57Whereas I think that, you know, we were both in Chicago. If you're thinking about the people that have to be won over, we both saw Shapiro's speech and Tim Walsh's speech. And everything you saw about Tim Walsh at that convention, there's been a couple of hiccups since, but everything you saw was like, you just said he was the person who was going to appeal more to more people. communication. Yeah. So anyway, maybe he's right. The thing we know is whatever the result, it will be more complicated than one factor. Also, it's interesting that he said that Trump, January 6th, disqualifies him from ever even thinking that he could vote for him.

49:35I mean, why aren't there more people like that? I never thought there'd be millions of people like that, but you don't meet many of them who say it as clearly as he did. No, it's weird that that isn't. And I suspect he was going to vote Democrat anyway. Good. Thank you. Onwards and upwards. Thank you.