What we learn from Iowa caucuses

The weekend press has been full of predictions about tomorrow's Iowa caucuses, with new polls showing massive voter shifts. Gingrich now appears destined for third or even fourth place, with Mitt Romney having a late surge along with Ron Paul and Rick Santorum. High drama with the nomination hanging in the balance.

Or not. The only politicians for whom it can be said that Iowa launched them on their way to the nomination are Democrats — Jimmy Carter in 1976, Barack Obama in 2008 and John Kerry in 2004.

On the Republican side, Ronald Reagan lost Iowa to George H. W. Bush in 1980, Bush lost to Bob Dole and Pat Robertson in 1988 and John McCain ran behind Mike Huckabee. Romney and Fred Thompson in 2008. All three won the nomination regardless.

The reason we pay attention to the Iowa caucuses in spite of this track record is because they are the first opportunity for real voters to express themselves. Without them, we have nothing but polls and pundits to tell us what's going on. In sports terms, picking a nominee before the caucuses is like picking a Super Bowl winner before the first games are played. Predictions about the Super Bowl winner change after each week's results, as simple hunches at the beginning of the season become calculated analyses at its end. In Las Vegas, bookmakers publish numerical odds on each team's chances.

There is now a similar process going on with respect to elections. Intrade.com is an online market where one can buy a position either for or against a candidate's chances of winning. There are tens if not hundreds of thousands of people who participate, so the sample of opinion is very large, made up of folks who are so sure they have figured it out that they are willing to put their money on the line.

As I write this, Intrade says there is a 74.1 percent chance Romney will be the Republican nominee, with Ron Paul in second place at 7.3 percent. In the general election, Intrade puts Obama's chances at 52.1 percent, Romney's at 36.5 percent and Paul's at 3.8 percent.

Don't misunderstand: a 52.1 percent number for Obama's chances in November doesn't suggest that these people think he will get 52.1 percent of the vote. It means that 52.1 percent of the current participants in the market believe he will win.

Like the football odds, Intrade's predictions will change as the political season moves forward. As more information — polls, primary results, endorsements, debate performance, financial reports, etc. — becomes available, more participants will join the pool or change their minds. The farther away we are from an election, the less definitive Intrade's numbers will be.