Political Predictions

If you're right about that, you could make a lot of money betting on Intrade -- that's at the higher end of estimates for seat gains for Republicans.

Five-thirty-eight agrees precisely.

I was guessing about 53 a couple months ago. Now I am unsure... I am thinking anywhere from 51 to 65, independents have been breaking really hard towards republicans.

Nate Silver catching up to Intrade?

As of Oct 22, Five Thirty Eight is predicting 230 GOP seats in the House of Representatives, up 3 from Oct 18.

Intrade is holding steady at 234 seats, since Oct 18, based on midprices of the 55+ and 60+ contracts.

Posted at 8:20am this morning at http://caveatbettor.blogspot.com/2010/10/nate-silver-catching-up-to-intrade.html

P.S. This prediction is a swing of 57 seats, from the 177 GOP seats held on Nov 19, 2009. I've got some money on the line with my favorite Harvard physicist that Intrade is more predictive than Five Thirty Eight.

@caveat bettor: I assume your favorite Harvard physicist teaches economics at a law school. Regardless, which side of the bet are you taking?

[Comment removed for supplying false email address and for rudeness. Email the webmaster@econlib.org to request restoring your comment privileges. Your own email address is required to post comments on EconLog.--Econlib Ed.]

Start with Real Clear Politics latest polls and you get the GOP picking up 60 in the House.

Review the turnout models the polls are using and you may sense that the polls could be underestimating the GOP turnout enough to give them an additional 4% each race.

That would translate into 51-52 seats total in the Senate and plus 75-80 in the House for the GOP if turnout ratios are closer to 1994 than to 2006.

@fishsticks:

The joke's on you. No one is forcing you to read David's great posts. I for one enjoy them.

Real Clear Politics has the following:
House: Dems 180 toss up 33 Repubs 222
Senate: Dems 48 toss up 8 Repubs 44

Common assumption is midterm elections go 60 to 70% for the out of power party.

This election is an uncommon wave towards the out of power party. I assume the election will tilt +80% towards the Repubs.

House should be Dems 187 Repubs 248
Senate should be about 50/50 but I am assuming the Repubs will get 51% and Dems and Ind 49.

~50 house seats to GOP sounds right to me. My guess is the Dems holding the Senate 51-49 is the most likely scenario, with a strong possibility of a 50-50 tie.

@John Fast: I get paid for 229 GOP seats or more, and must pay if 228 seats or less. No, my favorite Harvard physicist is actually a recent hire who sits behind me. We work at a bank.

David, this is a strange post for someone who believes in markets. Intrade says it will be better than 55.

Actually, has there been a study on prediction markets lately?

I see one issue with political prediction markets: once they're no longer a secret to pundits, political junkies will be motivated to purchase for status reasons, diminishing knowledge aggregation about elections.

Hyena: That isn't a big problem because that draws in lots more smart money. They say someone artificially bid up the McCain wins item in 2008 but the price returned to normal in a short time as people swooped in to cash in on the skew.

Fructose,

Good to know that there are enough traders to fix that problem. Well, not so good to know, part of me hoped there wouldn't be and that I'd make a killing.