Wednesday, August 13, 2008
Watching the Race According to a Futures Prediction Market
US Public Opinion Polls using standard random sampling techniques have been notoriously unreliable during the primary season this year at predicting the outcome of the various nomination contests, even those that have been taken just before actual balloting. I think it has something to do with the "lazy" telephone-based sampling techniques. But there does seem to be a new alternative to them in the form of real-money futures trading markets in which the point is not to find out whom respondents WANT to win and would vote for, but who they think WILL win and must put their money where their mouth is. One of the most successful of these is the Iowa Electronic Market whose head-to-head data on Barack Obama versus John McCain since 2006 has never predicted a victory by the GOP candidate, at least by the participants of this particular market. It is typical of the genre and may depress some while cheering others. The data is from their "Winner Take All" sub-market. It remains to be seens if this "wisdom of the crowds" approach will eventually displace the traditional random sampling method of predicting future events. It has one disadvantage: there is no quantitative prediction of how big the winning margin might be, though in this particular case the prediction trend is amazingly consistent among the market's participants who've bet consistently since 2006 that Barack Obama will be the next US President. The most significant data are those since May 2008 when it became clear these two candidates would be the GOP and Democrtic party nominees. The race may narrow in these last hundred days or so before the historic 2008 US presidential election.