Intrade Betting is Suspicious

FiveThirtyEight: Intrade Betting is Suspicious (HT: Greg Mankiw)

In the past few weeks, I've done posts about Intrade and the presidential election. Here is an interesting piece on alleged discrepancies.

There's something funny going on over at Intrade with respect to the pricing of the Obama and McCain contracts.

Right now, Obama is trading at 52.3 points. That is, Intrade implies that he has a 52.3 percent chance to become the next President.

Now, I happen to think that is a patently absurd price. But you don't have to take my word for it. Over at BetFair, another large UK-based gambling and futures site, you can also buy an Obama contract. But the price there is 1.62, which implies a 61.7 percent chance that Obama will become the next President.

That is a huge spread, 51.5 points versus 61.7 points. ...

... It does seem to be Intrade specifically that's out of line, rather than Betfair. At Iowa Electronic Markets, yet another political futures exchange, the probability of the Democrats winning the popular vote is about 61 percent. They don't have an electoral vote contract, but if they did, presumably that number would be a little higher because of the structural advantages Obama has in the electoral college this year that we've discussed here at length. ...

 

Comments

[Rob]

I would tend to think the problem is with the UK and Iowa markets, not with Intrade.

Most people closely following the election in the US see that it is neck-and-neck. If either candidate has had a lead since the conventions, it has switched twice.

In the US, at sites such as Real Clear Politics, there is a wealth of information from all "sides" available that shows how close and how volatile the election is.

On the other hand, as friends from the UK have told me, the perception there is that the election was decided months ago, and that Obama's landslide will be huge.

People following extremely closely will also note how close Rasmussen tracking has the Likely Voter spread. This is significant in that Rasmussen is a computer phone poll. If there is a "Bradley" or "Wilder" effect in the polling (i.e. Democrats being embarrassed admitting they will not be voting for Obama, presumably because of race), it's been posited that people are more likely to be "honest" about it to a computer. There were a couple of states during the Democratic primaries, where Obama significantly overpolled compared to the final vote.

Though I'm not a better, if I was I wouldn't take any odds much worse than even money for either candidate right now.

Lastly, if one looks at the internals of the polls, the ratio of Democrats to Republicans and Independents in some of the polls is a larger spread than has existed in any election in the last 50 years. Most of the other polls don't show that kind of discrepancy.

Certainly not a boring election, though.

[Michael W. Kruse]

Rob, I'm with you. I think Intrade probably has the better read. It just seems to match up better with other sources of information.