Ahead of November’s midterm elections, betting markets suggested a red wave was approaching, as they gave Republicans over a 70% chance of winning the U.S. Senate and around a 90% chance of taking control of the House of Representatives.
So where did betting markets go wrong?
Anthony Pickles, an anthropologist conducting research into political gambling and prediction markets, says several factors tripped up those wagering on American elections this year.
In a phone interview with MarketWatch, Pickles — a lecturer at the U.K.’s University of East Anglia who earned his doctorate at the University of St. Andrews — also talked about his views on betting on the 2024 presidential race, including whether to put money on Florida’s Republican governor, Ron DeSantis.
Pickles is both working on a book about political betting and engaging in it himself with small wagers of his own money, saying it’s “completely essential” as an anthropologist that he experiences it. He says he’s scored a profit of about 20% so far this year, not because of his own wisdom but thanks to tips from “really savvy gamblers” that he has met through his work.
The Q&A below has been edited for clarity and length.
MarketWatch: Betting markets didn’t get this year’s midterms right, and they also struggled in 2016 with the U.K.’s Brexit vote and that year’s U.S. presidential election. How good are they in general in predicting outcomes?
Pickles: There are arguments that they are generally better or the same as the polls once you factor in noise and you stretch it out over the period of an election. But when you speak to people who are in the community of political gamblers, they tend to be pretty skeptical about the markets as predictors. There are lots of cases where the markets appear kind of manipulated, or they are just following common wisdom or overreacting to recent events. They tend to be pretty good at following polls, but they’re also very likely to get caught up in a narrative, just as much as anything else is.
Betting markets are generally quite good on large elections, where there’s lots of polling data, but they’re very poor once the reliability of polling data comes into question, and that’s what’s happened over the last few cycles, especially in the U.S.
In the U.S., particularly, they’re also just quite right-leaning. They’ve got quite a right-leaning clientele, more so than they have in the U.K., and so they tend to favor Republicans.
MarketWatch: You covered the main reasons why betting markets got this year’s midterms wrong in a recent article for The Conversation. What are those reasons in a nutshell?
Pickles: I think it was the the trend towards the Republicans in the polls, responding to economic data. There was also one thing I didn’t mention in the article, but it is really important: The clientele of political gambling are overwhelmingly male, and therefore people weren’t as attuned to the importance of the overturning of Roe v. Wade, and how that was likely to result in larger turnout in certain areas than was necessarily being picked up in the polls.
Whereas political gamblers are often looking out for that reason for why the polls might be undersampling or underrepresenting particular groups, they don’t tend to do it that well. They tend to think about it more in economic terms. They are not likely to think about it in terms of bodily autonomy in the same way that they would if there were 50% women trading on gambling markets.
And the history told us that this is what happens in the midterms, and therefore you couldn’t trust that it would be close. You must assume that there’s something missing in the way that the polling is appearing. But the problem with that is that all the polling and all of the predictions, they all factor that in as well. So people kind of put in an extra 2% as a correction, but it doubles the assumption that the party in opposition is going to make gains in the midterms.
MarketWatch: What’s your advice to anyone feeling burned by betting markets given how the midterms turned out?
Pickles: The really long-term, successful political gamblers perform what they call value betting, which is basically where you’re looking for misprices, and then applying that to all kinds of outcomes. So if if you are going to make money on this stuff long term, you need to take a much more measured and rigorous approach.
I’ve been burned by betting for what I want to happen. I’ve been burned by trading news — buying high and selling low because you’re just reacting to events and getting caught up with the noise. It happens to everybody. We shouldn’t assume that political markets are in any way easy things to be successful in.
You’re often better betting from outside of a market than inside. The British political gamblers tended to come at the American election in a much more circumspect way than the ones in the American market did — a lot of whom got burned. Whereas we would often assume that Penny Mordaunt was the value bet for the Conservative leadership election, but we were too close to the action.
Quite often, being distant and just being able to look at the fundamentals without all the noise puts you in a better position.
MarketWatch: What’s your view on what betting markets are saying about the 2024 presidential race? Have you thought about that race?
Pickles: I think about little else. Well, that and the next general election in the U.K.
In the betting markets, there are a lot of people who question why Joe Biden is such long odds to win the presidency again.
He’s the incumbent president, with no obvious challengers inside his own party, and nobody out there who can actually tell him, “No, sorry, you can’t run. You’re too old.” And he has a track record of running many times. He’s a great bet. He should be much, much better odds.
He has the kind of belief that he’s the only person who can do things. I suspect that he will run, and I don’t know whether he will win, but he certainly is therefore an evens bet. His odds should be around evens.
On the other hand, Donald Trump — I don’t have the kind of stakes to short him as much as I would like, but I shorted him to the maximum of my capacity.
I have a little bit on on DeSantis, mostly because his star is rising, and it’s likely to continue to rise at least until we have a competitive primary.
I’m basically just shorting shorting Donald Trump and assuming that Biden’s star is going to rise. But do I know who’s gonna win? No.
MarketWatch: How did you get interested in political betting?
Pickles: I did my Ph.D. studying gambling in Papua New Guinea. Gambling is one of those things that we assume is universal, but actually it started at particular times. It didn’t cover the whole world. It came in during colonialism in various places.
And it obviously had a big effect on the way people think about things, because once you start thinking in terms of gambling, it pretty much colors your approach to risk in lots of ways. So that’s where my expertise started from, and then I was looking for a second project, now that I’m a fully fledged researcher and lecturer, so I came across political gambling, out of kind of personal interest. Then it just sort of slowly dawned on me that because political gambling is tied up with what we want to happen, and what we think is going to happen, at the most important scale, then it really floats between like culturally important understandings of how things work and mathematical, statistical approaches to what are likely outcomes — probabilistic thinking.
So I’m really interested in where probabilistic thinking and cultural understanding of what are the parameters of things that are possible, how those things come together. And how people make decisions based on that intersection.