So a few days ago, I was anchoring the CNN Max show, State of the Race, basically a politics show. And I was talking to CNN's senior data reporter, Harry Enton, about who Trump might pick as his running mate. And Harry said this.

Betting markets and measure the conventional wisdom and try and give you an understanding, okay, where does it seem like the conventional wisdom is standing right now? And what we see here is- And I have to admit, I've heard I've heard him talk about betting markets before, but this time I just had to clarify.

And to be clear, in the slide from before, those are like Randos betting on this, right? The market for betting on elections is wild wooly and maybe meaningless?

It is Randos betting on it. But in fact, one of the more interesting things, I'll tell you, Audi, is that in studies that have been done on polling versus the betting markets, oftentimes the betting markets beat the polling.

So that interaction got me thinking, how popular is election betting? And what can we learn about the future of our elections from political prediction markets? I'm Audi Cornish. This is The Assignment. And joining me is Thomas Grukka. He's a marketing professor at the University of Iowa's Trippy College of Business. More importantly, he is the Director of the Iowa Electronic Markets, which is an online futures market that includes political betting. Thomas Grukka, Welcome to the assignment. First, there are all kinds of what are known as predictive markets, right? So before we get to the election part, what are the predictive markets?

So a prediction market is a futurist market that's organized to aggregate information that's embodied in a whole bunch of different places, usually different people. Different people have different information, and we want to put all this information together in one place.

Now, How is it different from the stock market, though?

Stock market, you actually own something. That's the difference, right? You own a piece of a company. But in a prediction market, the way we put this information together is we allow traders to buy and sell contracts, what we call futures contracts. The value of that contract is based on a future event. Usually, we set it up so that the price is, if the contract says, yes, this will happen, you get a dollar. If it does happen, you are holding the yes contract. If it doesn't happen and you're holding the no contract, you get zero. We're looking for events which have a lot of importance to people. This is really an important to Distinction between us and betting is gambling, betting is creating risk where none existed for the benefit of entertainment and the people who are running it make money. A prediction market is different, at least the way we run our prediction markets. We look at events that are happening in the future where there is a significant impact on other people. For example, who's going to win the next presidential election?

How far back does this go? Has there been election betting as long as there's been Wall Street? Or help us understand when these markets came to be.

Okay, so election betting itself goes back, way, way back into at least as early as the middle 1840s. The reason we know this is because they would publish the odds in newspapers. And the reason they did that was we really didn't have predictive polling then. We didn't really have ways to find out what the electorate was thinking. And so people who were gathering information from all over the country, many of whom happened to be on Wall Street, were gathering information on commodity prices, all these different things, which they need to understand what stocks they should trade, what bonds they should trade, and so forth. And at the same time, asking people about, Hey, what are you hearing about this candidate or that candidate. That actually was very, very popular, at least until probably about the 1916 era. After that, it fell off. There were more more and more limitations on state-level bans on gambling. We started to get polls and other ways to see what's going to happen in the future. It fell off.

That's interesting that even back then, it was affected by regulation on gambling. It's still a cousin to gambling.

Yeah, absolutely. That's one of the big issues is what separates out a prediction market from gambling. Now, the new The incarnation of this happened in 1988. That's when the Iowa electronic market started. It sounds like a joke, but three economists and a political scientist walk into a bar. They literally did. What had happened is Jesse Jackson was running but not running because he did that. He did extremely well in one of the primaries, and no one had him on the card saying that he's going to do well. All the polls were wrong. The economists were sitting around saying, Hey, is there some better way to aggregate information than a poll? The political scientist said, No. Do you have economists have any idea? The economist said, Well, we Is it there's markets to aggregate information to come up with one price. Is there some way we could do that with a political contest?

Got it. That's how they made it up. Everybody in the market decides, Okay, based on how much is available and what else is going on? This is how much we're all going to pay for corn. Exactly. Or whatever. But you're saying that they thought they could apply... This is what people mean when they talk about the wisdom of- The crowds. Crowds or markets.

Yeah, That's how they predict the future, is you take all the information of all the people willing to share it. They trade with each other. Some are going to be right, some are going to be wrong. Through this trading mechanism, the people who end up being right get rewarded. The people who were not correct don't get rewarded. And so that interaction allows you to have a real-time price in the market.

That brings us to the Iowa Electronic Markets. And you guys are nonprofit. So fundamentally, this is about research, correct?

100 %.

Our traders would like us- So tell us about the people who are not about research. Okay. Because that has developed since.

There are two groups. Predict It, which is a very strange hybrid. The license to operate is held by a New Zealand University, but all of what you see online is run by a for-profit company in America.

We should say that because election betting is actually officially illegal- Correct. Yes. You have to be affiliated with a research project or a university, right? Have this academic purpose, hence- Well, yeah.

I'm not sure how that all works because the iElectronic market is owned, operated. Chief cook and bottle washer is me. I don't have a staff. I have a student, part-time programmer, and that's it. We don't have millions of dollars and tons of background support and all those other things. We are literally, we do all of it. We don't farm out any of it. And then there's another group called Calshi. They've been running what they call event markets or prediction markets for a while, and they are starting to ask the regulators, Can we start moving into political events? Because they know that there's a lot of interest. And here's what really separates us from the other two. Every time you trade in those others and you make money, those businesses, those organizations benefit. I'll use predictive for an example. Every time you sell something for a profit, they take 10 %. Like a casino, right? You win big because they don't really care who wins or loses. They're going to take 10 % of the winners, so they're fine no matter what happens. We don't. We have no transaction fees. We don't take any of the money.

All the money that goes in and that's from traders comes out to traders. We don't take anything. Our entire motivation is to create academic research.

The government actually tried to shut down one of these organizations, predicted, saying that they basically blew past their whole research focus mandate. They didn't really buy it. Then they got a court decision that said, Well, wait a second. You can keep the lights on until we figure this out. Right. But I had two questions going into this process, which is, number one, who are the people who do this?

There are information junkies of all types. There are people that just love politics, and they love to be right. And there's a lot of just keyboard warriors on online, right? Saying, I think this is going to happen. I think that's going to happen. And at some point people say, put up or shut up, and then the typing goes on and on. Here, If you are smarter, if you are indeed smarter, you can make money. But you have to remember that everyone you're trading with, whether it's in the IEM, Predictit, or Cal Richie, those people all think that they're much smarter than everyone else, too. So everyone's above average and overconfident, which is a good way to create a lot of trading and a lot of money going back and forth. It's just amazing how many people are just very into it because they get into politics, then they start looking at polling, and then they start thinking about, Is there something better? Our research It shows that our prediction markets, specifically, are more accurate than polls. Once you start doing a little research, it gets you into that prediction market world.

What's the difference between a person who approaches a prediction market or an election market and a person who answers a pollster?

When they ask you, Who are you going to vote for? Generally, we think about our aspirations patience. We think about who he hopes is going to win. We speak with our heart and we say, The person that I'm supporting or if the election were held today, I would go out and vote for this candidate. Now, if you ask them who they think is going to win, you might say to yourself, I don't know. I love my candidate, but I think that this person is going to get crushed on election day. But I'm going to still vote for him because I really like them. The difference is with a prediction market, you are basically voting with your head and not your heart.

Do you think that they're better than polls? There's a little bit of a chicken or egg, right? Correct. The coverage that comes out of polls versus... Yeah, go on.

There are two answers. One answer comes from our experience. We've done these not only here in the US, but we've organized We did prediction markets outside the US. I think we did like 30 of them outside the US. And not all of these places had polling that was available to us. There was probably private polling, but there wasn't public polling. Our prediction markets did extremely well. So even in the absence of polling information, prediction markets can do very well, which suggests that they don't follow the polls. And we've done a lot of research which shows that our markets almost always improve on whatever the polls are saying at a given time. And we've looked at this for multiple elections over multiple years. But one thing that we also know, and you'll watch people on TV, watch them on CNN, and they'll say, Well, this is a re-election. This is a perfect example. We're in a re-election. And there's a bit of folk wisdom called the incumbent rule. So when an incumbent runs that the undecided vote breaks towards the challenger if you look at the last poll and what happens on election day. And this is mainly for down ballot, so governor, senators, people like that.

It doesn't necessarily hold for presidents because it's such a big deal and there aren't as many. What we found was exactly that. There is an inherent bias in the polling for state-level offices. And how the IEM became more accurate was our traders understood that the support for the incumbent in the polls, it's pretty close. It's within 2 or 3% percentage points. But for the challenger, it's off, but it's almost always off in one direction. It's almost always too low. So if you see a polling result, the incumbent leading 52 to 46 six or whatever, and then you have a bunch of undecideds, they're not going to split those undecideds. Most of them are probably going to go to the challenger. Because if you think about what people are saying, it's a week before the election. This person in office has had two years, four years, six years, maybe even decades to earn your trust, earn your vote. You're saying, I still don't know. I'm undecided. A lot those people appear to choose the challenger because they're like, You know what? They didn't win my vote by now. I'm going to pull the other lever.

Okay, stay with us. I'm talking with Thomas Grukka of the University of Iowa. We'll be back in a moment.

Hey, John Favreau here. There's no shortage of political takes in 2024, but quantity doesn't cut it. We need a better conversation about the latest, biggest election of our lives. On Pod Save America, me and my co-host cut through the noise to help you figure out what matters and how you can help. Every Tuesday, Wednesday, and Friday, Pod Save America is breaking down the political news that makes us laugh, cry, and snap our laptops in half. Expensive year for laptops. Make sure to check out new episodes of Pod Save America on your favorite podcast platform or our YouTube channel now.

We're back with Thomas Skruca. He's Director of the Iowa Electronic Markets, which is basically a research-driven online prediction market, but we're also talking about election betting. We've talked on this program about the growth of online gambling just as an industry. As this is happening, you have the Commodity Futures Trading Commission basically coming up with a new proposal this past spring saying, Look, we want to permanently ban betting on elections, and a bunch of other things in commercial prediction markets. The reason why the chairman said that this was the idea, Rostin Benham, is about election integrity in the democratic process. What are the concerns here?

One of the things that appears What it means to me that they're worried about is if you allow these to become unbounded, that there could be positions taken which if a certain election goes one way or another, one person stands to gain or lose million dollars, tens of million dollars, whatever. And if it's allowed that any event could be part of an event market, then you could have a lot of these that may be subject to manipulation. They don't want to spend the resources to scrutinize every single application for, let's say you're going to run a prediction or what they would call an event market on a county level seating Johnson County here in Iowa. Someone could conceivably take a position if they can find someone on the other side. They would pay $10 million if a certain candidate wins. Well, there may be some incentives for that other candidate to throw the election or someone play around with the results. So a certain person gets paid that money.

What I hear you saying is that they're afraid that if this gets so big, the incentives to actually try and affect these real-world events will just be stronger, just like in gambling, where people get worried about players or coaches or other participants betting and then therefore, throw in the game, so to speak.

My wife's nephews are big traders, but they're like baseball stat junkies. They also do a lot of fantasy football and things like It's just that same analytics thing.

Just somehow it feels different because baseball is all stats in so many ways. But then when someone takes that same verve, so to speak, and applies it to who might be the Senate majority leader, that just feels crazy to me.

Yes. That's why what we focus on Because we don't have anywhere near as many traders as predicted, because we only focus on the most important things that really will impact people, control of Congress, President. All that other stuff is fun and interesting, but the big shifts are going to be based on these other contests.

Then the other thing I thought is that one of the things that is causing stress in the world of of online sports gambling are these very tiny bets on tiny events, meaning, is this person going to make this many free throws? Is this person going to... And the players hate it. Oh, yeah.

The prop bets. Yeah.

The prop bets. In a way, I feel like what you're saying is that trying to do those specifics in elections is the equivalent of trying to do prop bets.

Yeah. I'll give you a quick example. I don't think Predict It did this, but let's say that they did every congressional race. Now, that would generate a lot of trading and a lot of revenue for Predict It, but there's about a 95% chance of an incumbent being reelected since 2000. Do we really need to predict those? Because I can tell you right now, the answer is yes. And I'd be right 95% of the time. Why are you messing around running these markets? Well, if you're Predict It, it's you make money that way.

Someone somewhere might.

Yeah. Someone might think that, Oh, yeah, that's my neighbor. He's going to unseat this 15-turn congressman. Yeah, no, he's not. But the pool of traders that are smart is limited, so you don't want to waste them on unimportant things.

Right this week, as we're talking about this, the Conservatives Party in the UK, They're dealing with a news story about this parliamentary assistant or aid who placed a bet before a July poll.

Right. Yeah. And they're fairly unregulated over there for their betting. The only thing that would be of concern to us, and we've had people talk about this as one of the reasons they were not thrilled with the IEM, is they say, You guys predict so well. Maybe people look at your numbers and they don't go out to the polls and vote because you say that this candidate is going to win the popular vote, so my vote doesn't count. We used to get that a lot back in the early to late '90s. We don't hear that criticism anymore because I think most people have realized that the way to ensure that your candidate wins is to get out there and get all your friends out there, too.

What can you tell Tell us from the people who are participating in the Iowa elections market right now.

Okay. I looked up some of the prices this morning. Our traders are very convinced There are about 75 % chance that Biden will win, and actually the Democratic, because we haven't put the names on it yet, because we don't put the names on until they're nominated. But the Democratic candidate will win the popular vote. And in our vote share market, they're looking at quite a... At least the prices I saw today, a landslide, a huge victory on the popular vote side. Now, if you don't believe that's true, Everyone is free, every adult is free to come to the IEM, apply, get an account. It only costs you $5 to get an account. You can put in up to $500 of your own money and go buy the other contract that you think will pay more than the contracts that we're looking at right now. I haven't looked at the other ones, but for people who get excited about this stuff, we are running markets on who will control the house, who will control the Senate. And then we also have a Congressional control combination market where we have four outcomes of Republican House, Democratic Senate, vice versa, and so forth.

How should we, as viewers, listeners, readers, interpret information that is reported about what the markets are saying about elections? When someone gets on air or in the paper and says, Well, right now people are saying the odds makers, that's the phrase I usually hear, the odds makers are saying that so and so will be the vice presidential candidate, or that this will happen with the House, this will happen with the Senate. I know you don't do specifics, but how should we at home interpret that information when we hear it?

The thing is that even though people will put their money on a particular outcome and drive that price of that outcome up higher than it should be. On election day, we figure out who's swimming without their trunks pretty quickly. And a lot of times for certain types of decisions, those predictions are not necessarily very accurate, especially this far out. We are not on a news cycle basis, but from a people paying attention basis. We are really far out. Those numbers can change dramatically between now and when we get to election day. The party who's trailing based on whatever you're seeing from those odds makers, that party is doing everything it can to make sure that that prediction is wrong.

Tom Kruka, thank you so much explaining this all to me. My pleasure.

It's one of my most favorite topics in the whole world.

Thomas Kruka is Director of the Iowa Electronic Markets. And that's all for this episode of The Assignment: A Production of CNN Audio. This episode was produced by Asoke Samuel. Our senior producer is Matt Martinez. Dan DeZula is our Technical Director, and Steve Ligtai is the executive producer of CNN Audio. We got support from Haley Thomas, Alex Manisari, Robert Mather's, John Dianora, Lenny Steinhart, Jamis Andrést, Nicole Pessereau, and Lisa Namerou. Thanks, as always, to Katie Hinman. I'm Audi Cornish, and thank you for listening.