A Future in Politics?Regardless of your political persuasion, November’s presidential election will always be remembered as a remarkable one, not least because of the record turnout. Two million more votes were cast than in the 2000 election, a phenomenon ascibed, in part, to the intensive ‘get-out-the-vote’ campaigns by the two parties, and, perhaps more significantly, by the sheer intensity created by the narrow margin between the two. Nothing captures the public imagination like a closely fought battle, and this election was the hardest to call in history. We may as well have suspended the democratic process and flipped a coin for the presidency.
When the dust settled on those Ohio ballot boxes, Bush had received the largest number of votes of any Presidential candidate in US history, but, amazingly, Kerry also received more votes than any candidate in any previous U.S. election. Bush totaled 2.5% more than Kerry; the closest margin ever, slimming Woodrow Wilson’s narrow 3.2% victory in 1916.
Bookmakers, too, reported overwhelming interest in the presidential race, with hundreds of millions staked, not just in the US, but all over the world. It’s clear that political wagering is a rapidly expanding obsession. Whether this reflects a more politicised populace or a nation of gamblers who will bet on anything, it’s hard to tell. What is interesting, however, is the unfailing ability of betting markets to get the right result.
Imagine if you have a jar filled with jellybeans and you ask one person to guess how many jellybeans he or she thinks it contains. Well, the likelihood is he or she will be way out. Ask a million people, however, and you’ll find that the mean number of a million guesses will, allowing for a small margin of error, be just about right.
That’s the (simplified) theory anyway. The Bush vs Kerry battle was interesting because the markets had them practically neck and neck, and so accounting for the tiniest margin of error could swing it the either way.
To complicate matters, in the months leading up to the election, the markets themselves, though in general slightly favoring Bush, swung back and forth between the two camps – a reaction to information available to the public on each candidate at that particular time. Two borrow a phrase from the Republicans, there was a lot of ‘flipp-flopping’ going on: one week Kerry was favorite; the next it was Dubya – it was clear that nothing would be apparent until the bets were off.
By the time the ballot boxes and the betting markets had closed, however, the writing was on the wall. 'Based on the betting action we received on the election, we accurately predicted the overall winner even before the official pronouncement that Bush won,' said Alex Czajkowski, marketing director, Sportingbet Plc Americas Region. 53 per cent of Sportingbet customers had bet on Bush; 47 percent Kerry, that’s 0.5 percent out. The punters had got it spot on, right down to who would win the individual ‘battleground’ states, Pennsylvania (Kerry) and Ohio (Bush).
Betting markets are clearly more accurate than opinion polls in predicting future political results. Unlike the betting markets, opinion polls conducted by different companies often produce results that vary widely from each other, and generally prove ineffective in predicting the actual result. They are inaccurate because they offer no incentives to answer honestly; they impose themselves on people who may not have made up their minds, haranguing them into making an on the spot decision. Opinion polls take up time and people find them irritating. So if pollsters place dozens of phone calls to get just one answer, then they are not reaching a true cross section of the public.
Bettors, on the other hand, are willing participants and are ‘putting their money where their mouths are’. People, en masse, are likely to bet the way they vote – why bet against yourself? It’s not logical. And those who are sufficiently interested in politics to place money on an election will be more than likely to cast their vote on the day.
It’s no surprise therefore, that academics and statisticians have begun to adapt the fundamental principals of the ‘book’ for the serious purpose of political analysis. Future markets, such as the Iowa Electronic Markets, are kind of university funded, non-profit bookmakers, devoted to analysing the future. This must be the only university project that allows you to bet, win and lose money! Iowa Electronic Markets insist that money must be staked, as the only way to ensure accuracy is for participants to have a vested interest in the outcome.
Operated by the University of Iowa the scheme is part of the Henry B. Tippie College of Business’ research and teaching mission. Trading accounts can be opened for $5 to $500 online. Participants then use their funds to buy and sell contracts and therefore have the opportunity to win or lose their money.
The markets were first used in the 1988 U.S. presidential election and, since then, have predicted election results with an eerie prescience, with an average error of only 1.37 percent. And the more people on board, the more accurate the results. So, next time there’s an election looming, why not, in the name of academic progress of course, sign up and try your luck. Until then, there are plenty of websites out their offering topical politics prop bets. Okay, so putting your money on the next Justice appointed to US Chief Justice Seat might not have quite the thrill of sportsbeting, but, then again, some people will bet on just about anything…