If you look at Intrade today, you see that Obama has 2/3 chance of winning. With polls very close, and Romney leading in many national polls, the question is why?
I think there are two reasons. The first reason is that it has gotten significantly harder to participate in In-trade. In 2008, all you needed to do was use a credit card. With the passing of Dodd-Frank, you would now need to either send a cashiers check to Ireland or pay for a bank transfer. Then you also need to send substantial proof of identification to Ireland. Both of these are barriers to participation. With less participation, you will have a less fluid market, reducing its accuracy.
The second reason is all of the bets are based on the same polling averages that everyone else is looking at. The overall chance of a Obama victory is keyed off the summation of the electoral college results.
The electoral college map is is largely based on the RCP no-tossup map, which includes all state polls, no matter how accurate they have been in the past. State polls don't happen as often as national polls, so there is more of a lag to get the latest sentiment into the polls. RCP averages uses state polls that are as old as 2-3 weeks.
Given the inclusion of polls with low accuracy, in the past two elections, the RCP average has overweighted the result of the Democratic candidate. The RCP average shows a tie. In 2004, the RCP average on election day showed a 1.5% Bush lead. He won by 2.4%. In 2008, the RCP average on election day showed a 7.6% Obama lead. He won by 7.2%.
By contrast, the Rasmussen no-tossup map show Romney with a victory. This is based on the following tossup states:
Colorado: 50% Romeny, 46% Obama
Virginia: 50% Romney, 47% Obama
Florida: 50% Romney, 48% Obama
Ohio: 50% Romney, 48% Obama
New Hampshire: 50% Romney, 48% Obama
Nevada: 50% Obama, 48% Romney
Iowa: 48% Romney, 48% Obama
Wisconsin: 49% Romney, 49% Obama
No comments:
Post a Comment