InTrade.com

I like the markets...and InTrade is a pretty good predictor of politics given the 2004 election. Obama has soared in the polls on Intrade (essentially gambling website on politics and other things). The state-by-state electoral map looked close to even before the first Obama-McCain debate -- 290-260, or close to that. Now the "leaning" polls (which show electoral votes if the probabilities of today were the votes of 11/4) shows 338-200. Even the "average" votes which defy the electoral college and divide votes proportionally to the 'odds' on the state show Obama leading 314-224. Either way, the gambling markets show a Obama landslide.

In the age of Bush elections, this is a landslide. Bush "beat" Gore 271-266 in 2000 and Kerry 286-251 in 2004. There has not been a competitive election since the Bush fraud in 2000 (1996, 1992, 1988, 1984, 1980 both had winners with more than 350 electoral college votes out of 538 (# of reps + # of senators + 3 to Washington DC) or 65%). In this polarized stalemate, it takes 300 electoral votes to create a 'rout' (along with the popular vote) which is what the InTrade markets suggest is happening. It's not predicting a rout in the pre-2000 way, but in the 'new reality' that Bush created of 50%+1, 300+ electoral college votes is significant.

I will be happy if Obama takes the presidency regardless, but I will be sadly disappointed to see him win the electoral college without capturing the popular vote. The electoral college (the largest impediment to third paries, a shame) ranks people a different types of "fractional voters". In a state with 100,000 people, each person is 0.00001% of the population. In a state with 10,000,000 people, each person is 0.0000001% of the population. The electoral college system tries to iron this out using the proportionately represented votes plus the equally represented votes to make up the electoral college (plus 3 good will votes for DC, which has equal taxation but no representation). However, this creates issues for both parties: Using 2003 data on population except for 2000 data for DC, each electoral college vote represents 167,081 people in Wyoming, while each represents 650,545 in Texas and 645,172 in California.

This means that electorally, one vote in Illinois is worth 28% of a Wyoming vote while a Delaware vote is worth 61% of a Wyoming vote and an Arizona vote 30% of a Wyoming vote and an Alaskan vote is worth 77% of a Wyoming vote. Not that those statistics mean anything; voting is majority rules in each state, but it is still worth considering alternatives to the electoral system that are more reasonable.

The discrepancies described above are probably built into the markets if these 'efficient market' people are to believed so it's a wonder they're still so cheerily pro-McCain.