Political institutions don't often take risks. But the 2024 election is different, says Nate Silver

Listen to part one of this conversation here.

Nate Silver says that risk-takers are controlling society. Silver is an election forecaster, and admits that political institutions don’t take as many risks as sectors like technology or finance.

But this 2024 presidential race is different, Silver says. In his new book “On The Edge: The Art of Risking Everything,” he writes that the Democratic Party’s decision to oust President Biden and nominate Vice President Kamala Harris instead was a good risk management decision.

“I come from a gambling background. I played poker before I ever made an election forecast,” Silver says. “I understand full well that doesn't translate very well to most people's understanding of politics. They see the whole idea of betting on elections as being kind of dubious and morally cringe. They see each election as unique and not some part of some larger probability distribution.”

Silver, however, urges readers to rethink betting on elections and watching forecasts for real-time election coverage.

“Whenever people put money on the line, I tend to think that actually works,” he says. “That's kind of how the capitalist economy works. And that's kind of how Las Vegas is very hard to beat.”

Do people look at election models and forecasts with too much certainty?

“Our model is all about uncertainty. Hillary Clinton was ahead in the polls in the key states, but not by very much. And what our model is saying is that historically, empirically, the polls can be wrong. ‘Dewey Defeats Truman,’ for example. And so therefore we're modeling based on history, how wrong they might be.

“A lot of models are overconfident. I don't think ours was, but it is true that in general, people, when they build models, don't understand some of the real-world complications. There are technical properties, you can fit a model to pass data, but you don't necessarily guarantee that it will map the future very well.”

What does your model say about the 2024 presidential election?

“It's pretty close to 50/50. If you want to be more precise, it's Harris 54%, [former President Donald] Trump 46%. I mean, there are some other contingencies. If you're a poker player, then you take a 54-46 edge. But for all intents and purposes, the election’s a toss-up.

“I think there's a little bit of risk of Democrats becoming complacent. Harris has had momentum, and she'll probably, frankly, rise further in the polls after the convention, but when we get to after Labor Day and the first debate on Sept. 10, then Trump still has an advantage if it's close in the Electoral College, as he did in 2016.

“I think Trump has run in a really bad campaign recently and been hitting on Harris's race and other themes that are not very popular, but people still have trepidation about the economy, about the border. Harris ran quite a bit to her left in 2020. And so if the race is reversed and there are expectations that she's a front runner, then that can be a more difficult race to run.”

How was pushing President Biden out of the race a good risk management decision?

“They were able to make a rational calculation. In our model, they basically more than doubled — from 27% to 54% — their chances of winning by having Harris instead of Biden.

“The party was bitterly divided for a period of several weeks with [Rep.] Nancy Pelosi on the one side, maybe the [Democratic National Convention] and the Biden folks on the other side, because they thought it was worth fighting that fight because they think Donald Trump is an existential threat to the country. I think they played their hand pretty well as did Harris, by the way.

“She didn't wrap up the presumptive nomination within 24 hours on accident. She was very prepared. She was heady, in the moment, and willing to take a gamble and move kind of full speed ahead.”

Trump has owned casinos. Does that help him understand odds and risk in his campaign better?

“I think you have to give Trump some credit for the 2016 primary, where he realizes that, ‘I can actually run against the unpopular Republican establishment and have a more populist bent, and beat out 16 other candidates.

“Populist can be a euphemism for often things involving xenophobia, racism, et cetera, too.

“Trump, like Joe Biden is really up there in age. He's 78. And I think maybe he doesn't know his limitations either. I think the decision to pick JD Vance was short-sighted and kind of reckless where he thought, ‘I'm going to win because they're going to run Biden. So I can do whatever I want and get this guy who's in my legacy and become like the next MAGA president,’ when instead, they should have hedged and prepared for plan B if Harris were to take over the campaign instead.”

In your book, you point out that a lot of people in media, academia and government tend to lean left. What impact does that have?

“In the Trump era, people who have college degrees, especially advanced degrees, tend to be overwhelmingly Democratic, or at least moderates who are anti-Trump. And that's even more true in academic institutions like Harvard or Yale or Georgetown or whatnot.

“I freelance for the New York Times. I've worked in some of these institutions myself. I do not think that every journalist has a left-wing bias, but I think some do. And I think news consumers on issues like coverage of the lab leak, for example, were now the the official position of the Biden White House. Different agencies think different things, but the position is that we don't know.

“That was termed misinformation during an election campaign when, you know, that would be seen as giving too much credit to Trump.

“So that was an example, or censoring coverage of Hunter Biden's laptop, which I don't think is a terribly important story, but still involved labeling things as misinformation that in retrospect weren't.

“As media struggles, they're actually drawing from a narrower segment of society where you're getting mostly the Ivy's and the elite small private schools and Stanford and MIT and University of Chicago and places like that are disproportionately represented on the editorial mastheads of the New York Times, the Washington Post,  and those are people who tend to be liberal.”

Samantha Raphelson produced and edited this interview for broadcast with Michael Scotto. Grace Griffin adapted it for the web.