Tuesday, February 12, 2008

Delegate counts and Intrade

As of writing, Senator Obama has moved ahead of Senator Clinton in the delegate race - including superdelegates! As the results from Maryland and DC come in, and the full results from Virginia come in, the Obama-Clinton gap will widen - possibly as high as 67 total delegates, including ~145 pledged delegates.

On to the main subject of this post - political futures trading markets like Intrade. ST sends me a link to a NYT article that looks at Intrade (thanks!) Apparently, Intrade did very well predicting the states W won back in 2004, as well as the states that went for the poor man's JFK - much unlike the early exit polls that leaked in the middle of the day. But this time, Intrade was badly off on the California Democratic primary, predicting an Obama win. If you were living outside the USA or under a rock stateside - Senator Clinton won that battle by 10%.

What happened? David Leonhardt, writing for the NYT, says there is a lack of trading volume - the market does not adjust to changing conditions rapidly enough. That is possible - I am not an economist, and admit to having to spend a little time trying to figure out what the numbers on Intrade's website meant.

But there is also the factor of misleading information. Most of the surveys released just prior to the California race showed either a tight race or - egads - a substantial Obama win (your bad, Zogby!) Only SurveyUSA got it right; ARG generally is seen - fairly or not - as a shill for Hill.

What happened in California? I think it is a combination of early voting and Latino support.
1. SurveyUSA's last poll showed 1/3 of voters had sent it in; these favored Senator Clinton 54-37 (7% to Edwards before he dropped out?) In-person voters were closer - 51-45. Question is, did other polls account for the early-voters?
2. Latinos were ~30% of the California voters based on exit polls, I think almost twice the Hispanic fraction in earlier elections; they broke for Clinton 70-30 or so.

Misleading information inevitably leads to misleading plays; and possibly no sane person could account for a historic turn of events. So I would not fault Intrade all that much. Maybe someone with a better handle on economics could find a better reasoning.

No comments: