On the big issue of who wins the presidential election, I’ve been as constant as the north star.
But for state-by-state estimates, I’ve been flipping back and forth like a corrupt politician (pardon my redundancy) trying to decide between two interest groups.
- Four months ago, I predicted that Obama would win reelection with 297 electoral votes, 27 more than needed.
- Back in July, I shifted Virginia to Romney’s column and predicted Obama would still win, but with 284 votes.
- Then in August, I predicted things were moving even farther in the GOP direction. By moving Colorado to the Republican side, I guessed the outcome would be 275-263 for Obama.
- Last month, I shook things up a bit and gave Florida to Obama and Wisconsin and Iowa to Romney, resulting in a 288-250 margin in Obama’s favor.
This month, I’m reversing everything from last month. I give Florida back to Romney, largely on the basis of his performance in the debate. Moreover, I was thinking of giving Virginia and Colorado back to Obama, thus changing what I did in July and August, but decided to leave those states in the GOP camp because of what happened on the stage in Denver.
But I decided I was wrong about Iowa and Wisconsin. The polls from those two states are simply too unfriendly and I’m guessing the Obama turnout operation will be stronger.
However, I’ve decided to shift New Hampshire to Romney, again because of the debate, so the net effect is a very close election. But Obama still prevails.
For what it’s worth, the folks at Real Clear Politics show Obama winning 303 electoral votes. The difference in our projections is that they give Nevada, Colorado, Virginia, and New Hampshire to Obama.
Are they right? Well, their estimates are based on polling data, so you have to ask yourself if the polls are accurate and/or if the polls today reflect what will happen on November 6.
Intrade says Obama is a 2-1 favorite, so the people putting money on the table certainly think the election isn’t that close. Then again, Intrade had Obama as a 3-1 favorite before the debate, so that number also can move a lot.