On Monday I got an email from the University of Iowa telling me that Barack Obama had a 75% chance of winning the Democratic nomination; Hillary Clinton was at 21%. I’m guessing I’m not the only business reporter out there who is most comfortable processing the horse race elements of national politics through Iowa’s prediction markets—run by professors at the Henry B. Tippie College of Business—which let people buy and sell futures contracts pegged to each candidate’s chances of winning. Ah, market forces. Hello, comfort zone.
I checked back in this morning to see where things stand post-Pennsylvania. As of 10 a.m. Central time, an Obama contract was trading at 78.9 cents (indicating a 78.9% chance of winning the nom), and Clinton was at 18.2 cents. Now, Clinton did win Pennsylvania, so I could make some snide comment about the extent to which PA votes matter, but being Pittsburghian by birth, I’ll instead take this as evidence that the commentators singing the song about how Clinton is still in it but with a hard road ahead are getting it right.
I should also say that even though the Iowa markets have a terrific track record—a recent study showed them besting polls 74% of the time—they are still markets. The thousand or so people actively trading can make money by being right, or by buying low and selling high irrespective of the underlying value of the contracts. Hence the little
arbitrage game that keeps John Edwards contracts selling at 1/10th of 1 cent.
Here’s a look at how things have unfolded over the past year:
DROF_NOM means “Democratic Rest of Field.” Those are the people hoping for an Al Gore write-in campaign or somesuch—that contract shot up to 6 cents the day after our very own Joe Klein suggested that Gore might be the answer.
The other interesting chart is the one that shows predictions on whether the Democratic or Republican nominee will win the general election. The never-ending Democratic nomination battle has helped John McCain, but blue is still in the lead. (You have to link to this one, by clicking on the chart):