IE 11 is not supported. For an optimal experience visit our site on another browser.

Bookies seen outbidding election polls

With just two weeks to go until the presidential election and most polls putting the two candidates neck and neck, you might be better off talking with a bookie if you want to know whether Sen. John Kerry or President Bush will win the Oval Office this November.

With just two weeks to go until the presidential election and most polls putting the two candidates neck and neck, you might be better off talking with a bookie if you want to know whether Sen. John Kerry or President Bush will win the Oval Office this November.

A number of online betting sites are offering punters around the world the opportunity to bet on the outcome of this November’s extremely tight presidential election race, as well as the outcomes of hotly contested Senate races in states like Colorado and South Dakota.

The business of online betting barely existed four years ago, but in recent years has blossomed into a multi-million dollar industry. And with millions of dollars being wagered on the U.S. elections this year, experts say the odds these sites offer are proving to be far more accurate real-time predictors of election outcomes than political pundits or over-quoted public opinion polls.

At the current odds offered at — a Dublin-based online trading exchange for politics, economics and sports — Bush has a 58 percent chance of wining the White House this November, while Kerry has a 42 percent chance.

“The odds on our site react to the latest news event immediately, and so in terms of the reliability of our election data we think we are way ahead of the opinion polls which only tend to survey a few thousand people,” said Mike Knesevitch, a spokesperson for

Liquidity is the key to the accuracy of betting sites like, according to Leighton Vaughan Williams, director of the Betting Research Unit at Britain's Nottingham Trent University who has studied how well efficient betting markets predict outcomes of events such as elections.

Unlike Web sites where bettors can buy and sell futures contracts based on the election, the sheer volume of wagering attracted to online betting sites — which are seeing millions of dollars change hands for bets on anything from who will win the election to the nation’s unemployment rate should George W. Bush be reelected — makes them more accurate, he notes.

The thinking among economists is voters don’t have any incentive to tell pollsters the truth about their voting plans, but when they are betting their own money they will tend to make informed choices based on the best information available to them, Vaughan Williams said. “The more money involved, the more efficient and accurate the market,” he added.

While election betting took in only a few thousand dollars four years ago, Vaughan Williams says recent figures show has raked in some $11 million from wagering on the 2004 presidential election, while London-based Betfair has taken in some $4 million. A very close presidential race and the intense media scrutiny surrounding this year’s White House race have made it a rich resource for bookmakers.

Vaughan Williams also asserts that when compared with poll results, betting odds offer a more consistent snapshot of where candidates stand in voters’ minds. Unlike past years, online bettors today can access real-time information about the campaign through the Internet and through 24-hour news channels, and so betting odds are reacting nearly instantly while polling is little more than a rear-view mirror picture of public opinion, he said.

In early October, for example, betting sites showed President Bush had a 61 percent chance of winning re-election, but that forecast dropped to a 58 percent chance as September’s disappointing jobs data were released. Similarly, betting odds flickered minute by minute between Bush and Kerry as each one was perceived to edge past the other in the recently televised presidential debates, notes Vaughan Williams.

Predicting elections is, of course, an inexact science, but betting sites have earned their chops. A case in point is the Australian federal election on Oct. 9. The campaign attracted intense interest in betting circles, and although news polls had predicted a tie between Prime Minister John Howard’s Liberal-National coalition and the Labor opposition, Centrebet — the nation’s largest sports bookmaker — accurately predicted that Howard would comfortably win a fourth term.

Similarly, academic studies have shown bookmakers’ odds outshined opinion polls in the 2001 Australian election, as they accurately predicted the outcome across a whole range of regional elections. “If you had looked at the public opinion polls you wouldn’t have been so sure who would win,” said Vaughan Williams. “They were all over the place.”

The notion that an open market can accurately predict the outcome of an election was pioneered by academics at the University of Iowa where the now oft-quoted Iowa Electronic Futures Market, a real-money market that allows punters to buy futures contracts based on the outcome of economic and political events, was established in 1988.

Run by the university's business school and regulated by the Commodity Futures Trading Commission, the Iowa market has spawned similar futures markets at and, and they have been remarkably prophetic. Compared with 596 national polls in four presidential election cycles between 1988 and 2000, the Iowa market was closest to the actual election result 76 percent of the time notes Thomas Rietz, a finance professor at the University of Iowa and director of the Iowa futures market.

Indeed, early on in the 2000 White House race, the polls put George W. Bush as the clear winner, but the Iowa futures market predicted a dead heat as early as May that year, Rietz said.

But Vaughan Williams says despite the Iowa market showing itself to be an accurate predictor of elections, traders on the exchange are limited to betting $500. When larger amounts of money are wagered in online betting sites, there is a tendency toward greater accuracy, he says.

Still, some analysts doubt that betting odds can always accurately predict an election’s outcome. Bill Thompson, a professor of public administration at the University of Nevada in Las Vegas and a gambling scholar, said online betting can sometimes skew toward the investing class, which likely vote Republican, or voters who don’t bet or use the Internet, and gamblers usually bet with their hearts and not their heads.

“There is some validity here, but you need to look at the issue from all the angles,” said Thompson. “Who is betting? Are they pooling their investments? And are these people betting on the candidates they want to see win and simply cheering for their side?”

Thompson also points out that while political wagering may be popular in very close elections, it is unlikely to garner as much interest in elections where there is a clear winner. “We are seeing the height of insanity in this year’s election because it is so close and people have so much invested in it emotionally,” he said.