Predicting outcomes
More than 60 countries – representing nearly half the world’s population – are holding democratic elections this year. As more voters than at any point in history head to the polls, error is an occupational hazard for those involved in predicting political outcomes.
And in today’s interconnected and data-driven world, opinion polls are not the only sources of truth for democratic parties and their representatives – betting markets also play a valuable role during elections.
Although notable studies have tracked similarities between punter predictions and polls, the key difference is that a poll will show which party or candidate voters would prefer to win, while price changes within political betting markets indicate who the public thinks will win.
Last week’s Super Tuesday is a perfect example. It has effectively confirmed that the forthcoming US presidential election will be a clear-cut, two-horse race, with President Joe Biden and Donald Trump set to become Presidential nominees for the Democratic Party and Republican Party, respectively.
However, what appears as certainty can be misleading. David Fleming, a Specialist Trader at Flutter, ruefully remembers both the US presidential election and UK’s EU referendum in 2016. At the time, the company was so convinced Hillary Clinton would be victorious that its UK brands – Paddy Power and Betfair – paid out £800,000 worth of bets ahead of the result. Of course, Donald Trump triumphed, an upset that cost Flutter £5m in total.
“2016 was a painful year. Trump and Brexit were very bad results for us,” says David as he reflects on what was both a pivotal year in modern history and the point when political betting entered the mainstream. In fact, the £250m matched in 2016 on the Betfair Exchange – the Flutter brand’s peer-to-peer betting platform and the biggest in the world – grew exponentially to a record £1.7bn in the 2020 election, making it one of the biggest betting events of all time.
While the bookmakers got Trump and Brexit wrong, so too did the opinion polls. Anthony Wells, Head of European Political and Social Research at YouGov, the opinion polling company, recounts how polls took a reputational hit when they predicted a Clinton win against Trump and, separately, a ‘Leave’ result in the EU Referendum vote.
Anthony suggests that these knocks aren’t entirely fair. Pollsters define likely margins of error, usually at around four percentage-points. While 4% may not sound like much, it is a difference which could see a 48% opinion poll turn into a 52% outcome – and vice versa.
In search of a true price
For Flutter, activity on the Betfair Exchange provides an important, unique source of information as the most liquid market in political betting. As with other markets like sports-betting, political probabilities analyse data from several sources, including historical data, polling stats, expert research, the news and information derived from trading on the Exchange itself.
Sam Rosbottom, Betfair Senior PR Manager who looks at political betting trends, sees the Exchange in the same light as a stock market, acting dynamically as money is wagered and matched in real time, with punters often turning to the platform to, as Sam says, “have their say”. “Often the odds on the Exchange are better as there is no bookmaker’s margin. When it comes to political betting this is a great way of getting a sense of what the nation really thinks, not just through ticking a box on a pollster’s survey but by people putting their money where their mouth is,” explains Sam.
“The accuracy,” he continues, “and sheer volume of bets is why it is so trusted. Odds on sports are generally most interesting after the fact, if there is an underdog win for example. But, as a product of pure opinion, the Exchange marries perfectly with politics and voting, where every voice matters.”
A symbiotic relationship
In short, neither political bets nor polls can predict an outcome with absolute certainty; but that’s not entirely what they’re there to do. In some ways, all bets are a type of opinion poll. If you take on a bet, or lay it, you are putting into action your belief that the performance of a particular horse, soccer team, golfer, or political party will match your opinion about a future outcome.
In sports, commentaries around a game and the betting odds are often quite separate. In politics, however, the odds often become part of the mainstream debate. Political bookmaking may even influence the progress of an election campaign. After all, politicians often change course if they see policies going down badly with an electorate. Betting markets, as well as opinion polls, simply highlight the popularity, or otherwise, of policies and the personalities behind them.
Subtle differences between betting markets and opinion polls also mean that they complement each other, rather than compete. YouGov’s Anthony Wells highlights this distinction succinctly: “My job isn’t to predict, it is to measure.”
As Anthony explains, pollsters are looking at opinions expressed at any point in time, essentially asking the same question frequently and regularly – how would you vote if there was an election tomorrow? Betting markets are expressions of opinion about the outcome of an election result in the future, which could be days, weeks, months or even years before the event.
Given those who are placing and taking bets are backing their opinions with money, you could say betting markets are a more powerful source of truth. However, bettors are also a self-selecting audience, while opinion pollsters canvas carefully selected – hopefully, representative – groups of people. In reality, there’s a symbiotic relationship between the two.
So, what do we think is going to happen this November? Coming back to the Betfair Exchange, despite President Biden being slightly ahead in most polls, the Exchange gives Trump a near 50% chance of winning the US Election, with Biden trailing at around 28%. Both voters and bettors alike will have to wait until November to see who is right!