Saturday, April 30, 2011

What Can We Learn from Intrade? Part II

Continued from Part I.

As mentioned in the previous post, Intrade markets are not very good at predicting election outcomes (until right before the election), but they are a decent reflection of the conventional wisdom. More to the point, they reflect the conventional wisdom of Intrade investors.

Although anyone can place bets on Intrade, the people who do are not a representative sample of Americans. Judging by the previous behavior of Intrade markets, they are more likely to be from the "Washington establishment," and they tend to be more liberal. Until just before an election, Intrade markets consistently overestimate the performance of Democrats (presumably out of investors' optimism).

The Intrade market for the 2012 Republican primary is located here. As discussed earlier, there is some fairly high betting for even the least likely winners, because the potential payout for a long-shot bet is great. The candidates leading in the Intrade markets tell us something about the conventional wisdom of the Washington establishment.

At present, Mitt Romney leads the pack on Intrade with about a 25% chance of being elected the Republican nominee for president. This reflects the establishment consensus that he is the front-runner and the "next in line," as Republicans so often choose.

Second place is more interesting: Tim Pawlenty, who sits comfortably in second place with about 16%. Obviously this does not reflect Pawlenty's performance in the polls. But the Intrade investors recognize that his low poll numbers are probably due to low name recognition. Pawlenty's position here seems to reinforce the notion that he has successfully positioned himself as the leading "generic," unknown candidate.

Similarly, Intrade puts Mitch Daniels next at about 11%. Daniels is even less known than Pawlenty. But establishment Republicans see him as a potential consensus candidate. Recognizing Daniels shows some insight on the part of the Intrade investors.

However, you can also see the weaknesses of the Intrade investors' thinking. If Pawlenty can be selected because he's an inoffensive, "generic" candidate, why not Daniels just as easily? Pawlenty's higher numbers reflect the slightly greater name recognition that he has now, which will diminish once the race gets underway.

But Pawlenty is also higher on Intrade simply because he has formed an exploratory committee, while Daniels has not. The question of whether or not a candidate will run seems to weigh heavily on the investors' minds: Christie is at less than 2% on Intrade, and Huckabee at about 8%. If and when they form exploratory committees (or jump into the race), their numbers on Intrade will spike. Rapid swings like this undermine the idea that Intrade markets can predict future events; such a shift should only occur in response to a truly unexpected development, not something as ordinary as a contender officially joining the race.

Then again, neither Daniels nor Huckabee has formed an exploratory committee, but Daniels is ahead of Huckabee--despite Huckabee having a much clearer path to the nomination. This reflects the Republican establishment's preference of Daniels over Huckabee, whom they have never accepted as one of their own.

That's not to say Intrade is never impressed by name recognition. In 2010, Palin was consistently in the lead. Then her numbers cratered in January 2011, giving rise to Romney. Palin suffered from a drop in support around that time, and her visibility lessened in subsequent months. This could be a reflection of the (liberal) establishment view that the Tea Party lost influence after its 2010 victories.