The Industry Website

Web's Largest listing of crowdsourcing and crowdfunding events


Web's Largest Directory of Sites

2,968 crowdsourcing and crowdfunding sites

Polls and the Crowd Pick Winners Between Obama, Romney on U.S. Election Eve
© Image: / Rob Boudon

Polls and the Crowd Pick Winners Between Obama, Romney on U.S. Election Eve

In his 2006 book Infotopia, Cass Sunstein, a Harvard Law professor and the previous Administrator of the Office of Information and Regulatory Affairs, makes the case for prediction markets. Whereas groups of people can be swayed into answering a question one way, he argues, creating a marketplace out of a question tends to reward those with more knowledge of the issue at hand.

In recent years, the market and crowdsourcing dynamics have been applied to a number of events, including the U.S. presidential elections. That’s not to say people didn’t bet on the elections before – betting, in some shape or form, has probably been around since the most ancient of marketplaces. But digital media have made it simple to crowdsource people’s proverbial money where their mouths are.

This election cycle, pundits are looking closer than ever at prediction market results, perhaps encouraged by their accuracy the last time around.

Intrade, an Ireland-based website that collects bets from across the world (including America), was only one electoral college vote off in 2008, for example. If the site can deliver similarly accurate results this year, President Barack Obama has little to worry about (the Intrade election map has him winning 303 electoral college votes; the amount needed to win is 270). Overall, Intrade’s betting crowd thinks incumbent Obama has about a 67 percent chance of winning over challenger Mitt Romney. Britain-based Betfair (which doesn’t accept Americans’ money) gives an even weightier advantage to Obama – 77 percent. 

Since the aggregate Real Clear Politics poll-of-polls shows the presidential campaign to be a virtual tie in terms of the popular vote, such bullish behavior on prediction markets is puzzling. Several factors may offer an explanation.

Consider, for one, the American presidential election model, which counts not the popular vote, but rather operates under the electoral college system. By most accounts, Obama has a relatively large lead in that regard, and needs to capture only several toss-up states to secure the 270 electoral college votes needed for re-election. Two large swing states are Wisconsin and Ohio, and Obama seems to be leading in both; if he can also secure Nevada, as predicted, the election is his.

From a crowdsourcing perspective, however, the more interesting reason for Obama’s significant lead in prediction markets is the question posed to participants. Whereas polls generally ask individuals who they will be voting for, prediction markets ask them to bet on who they think will win. Here, the oft-mentioned crowd wisdom kicks in. Because individuals have access to a large host of information, and because information spreads quickly among individuals in a crowd, they take a more educated guess than just blindly picking a candidate. In the same way, crowd wisdom may prevent individuals from crowdfunding fraudulent or unappealing companies. Since many people are likely to know more than one individual (and because one individual's knowledge will spread throughout the crowd), investments will be made with a more complete understanding of the risks involved.

Obama's supporters also tend to be younger and more social media-savvy, meaning they see more of their friends endorsing the president on Facebook and Twitter. These endorsements may matter more than the ones given out by celebrities or political pundits. This can also go against Sunstein's hypothesis, however, and sway more people to bet on Obama just because others appear to be doing so. It would be interesting to examine how individuals would bet on Obama's chances of reelection before and after seeing Intrade's numbers. 

Prediction markets aren't perfect. The New York Times FiveThirtyEight blog’s Nate Silver explained that due to the favorite-long shot bias, people are more willing to support the underdog (because, for example, they enjoy the risk or are allured by the prospect of a higher payout if he can pull off the win). Indeed, professional bookmakers put Obama’s chances of winning as high as 81 percent. On top of that, the prediction markets can be skewed by large inputs of cash, as recently happened when someone (or a few individuals) invested almost $18,000 into Romney’s chances on Intrade, pushing his victory odds from around 41 percent to 49 in a matter of minutes, before quickly collapsing back.

So should Obama feel confident just because the crowd is confident in him? By no means – Romney’s chances are still fair; an upset is well within the realm of possibility. And that’s even more reason for all Americans to exercise their Constitutional right to vote.

Flag This



  • Nov 07, 2012 05:19 am GMT

    The crowd was right: Obama wins U.S. 2012 presidential election with more than 285 electoral votes.

  • Eric Blattberg Eric Blattberg Nov 07, 2012 05:30 pm GMT

    The final tally, as it stands this afternoon: 303 electoral votes for Obama and 206 for Romney. Popular vote: 50% Obama, 48% Romney.

  • Anton Root Anton Root Nov 07, 2012 07:37 pm GMT

    The crowd off by one state - Florida, which looks like it's going to Obama in a pleasant election-day surprise for the President. Other than that, the crowd predicted every state correctly.

 Join or Login