Advertisement
Want better election forecasts? Allow people to bet on the races | Column
Betting on the outcomes of political races could provide more accurate forecasts than traditional polling.
Published Dec. 18, 2020

Filmmaker Michael Moore predicted that Donald Trump would prevail in 2016, while polls showed that Hillary Clinton would win handily. Prior to the 2020 election, Moore warned that the Trump vote was again being underestimated because his supporters were suspicious of the so-called “deep state” pollsters quizzing them about their vote. So, were they hanging up on the pollsters or lying about who they were voting for? We do not know, but what we do know is that citizens are losing confidence in pollsters’ forecasts.

Allowing the public to bet on elections — currently illegal in all states — could provide more accurate forecasts. Not only should it be allowed, it should be encouraged by permitting all the brokerage houses to serve as conduits for betting which can be done via the Chicago Mercantile Exchange where futures contracts are traded. You should be able to sign into your brokerage account and place a bet. As easy as that. Ease of trading will encourage people to participate which could lead to more accurate forecasts. A basic tenet of finance is that low cost and ease of trading leads to more trading and better pricing.

Why more accurate forecasts? Because people do their homework when they put their own money at risk. All of this speculation would be subject to the oversight of the Commodities Futures Trading Commission which regulates trading on futures contracts for commodities and financial instruments.

Here is an example of how it could work. You can place a bet for a price between 1 cent and 99 cents with a $1 payoff to the winner. So, suppose that you are willing to bet that Jack is going to become president, and I’m willing to bet on Jill. You pay the going price of 48 cents for a Jack bet that goes into an escrow account. Believing that Jill will win, I take the other side of the bet and would have to place 52 cents in that escrow account. If Jack wins you get to keep the entire $1 in the escrow account, and if Jill wins, I get the $1. As simple as that. A small commission would be paid by both parties.

What determines the 48 cent price for a Jack bet? The forces of demand and supply. If there are a lot of buyers willing to bet on Jack, the price will rise from 48 cents to say, 52 cents. And if the buyers keep on coming, the price would continue to rise. Suppose that the price settles at 65 cents for a bet on Jack, while polls are forecasting that Jill is going to win. This information would be useful to Jill who might decide to change strategy.

Depending upon the level of interest, you could have markets for individual states and for members of Congress. So even if Moore and others like him voted for Clinton, they might have bet on Trump in Michigan and Pennsylvania, because they believed that they had a better feel for the sentiment of voters than the pollsters.

The only major concern is that if trading is thin — if not a lot of people make bets — political operatives may make substantial bets intended to distort the price to spread disinformation about who is favored to win.

But according to Fortune.com, overseas betting markets called the recent presidential election more accurately than polls. And if betting is allowed to grow, it could prove an even more accurate reflection of public sentiment. A few election cycles will determine if pollsters or the futures markets are more accurate at forecasting elections.

Murad Antia teaches finance at the Muma College of Business, University of South Florida in Tampa. In a former life he was a quantitative equity manager at a major bank in Florida.