Are Betting Odds More Accurate At Predicting Electoral Outcomes Than Polling Data?

2574 Words Nov 13th, 2016 11 Pages
Who Will be First Past the Post in the Political Grand National?
Are betting odds more accurate at predicting electoral outcomes than polling data? If so, Why?
Background
The 2015 General Election was notable for polls predicting a hung Parliament , much of the commentary after the election was about how the polls got it so wrong. Additionally, 2015 was one of the first General Election in the UK to see large sums of money involved in betting on the multiple outcomes. The favourite was for the Conservatives to have the most seats, but it also predicted there would be a hung parliament. In this scenario, while both the polls and betting odds were wrong, the odds had more confidence in the Conservatives getting more seats than Labour. Following the election Polling companies conducted a review, to work out where they got it wrong . Betting companies have more to lose from predicting the wrong winner, as both the company and the better have a financial interest in the result. The polls getting it wrong, was not a story that is unique for 2015, or unique to the UK, with 2014 Mid-terms and the 2016 Presidential elections also failing to predict the winner. Is there evidence to suggest that the polls are becoming less accurate?
Historically, in the US, bets were placed on elections before polling data was available and these were consistently accurate at predicting the outcome, this could suggest that betting markets are able to function in isolation to the polls and…

Related Documents