« Probably gone for a week, a couple discussion topics | Main | "Libertarian" Economist Bloggers Rationally Racist Against Obama To Preserve White Male Status? »

September 06, 2008


Douglas Knight

It seems hard for me to reconcile an 80% chance of Obama winning with the claim that his VP choice was a mistake, or at least with the claim that he should incur the cost of changing the VP.

Can you make any predictions to test your claims about the value of Palin, in particular, that she appeals to Hillary supporters? There is a wide range of claims about what she is good and bad for. eg, fundraising shows she's good for the base; she
generated a lot of news coverage, which some said was good, but she's now fleeing back to AK, which suggests that the campaign disagrees.

There is a large chunk of Mankiw's graph where O & M are strongly positively correlated. So...the market isn't liquid. I guess the right thing to graph is the difference.

Hopefully Anonymous

I think I was pretty clear that Obama seems strongly favored to win even with Biden, but that switching to a more credentialed female veep than Palin in the immediate aftermath of McCain's reactive pick would improve his odds even further, in my estimation. By the way, I think the optimal window for that action has now passed. It doesn't seem hard to reconcile to me.

Douglas Knight

Maybe the disagreement about reconciliation is only from disagreement about the cost of switching.

Anyhow, what about predictions that would test the claim that she appeals to centrist women? (do you make that claim?)

(my expectation of her effect on the campaign and any particular demographic is pretty close to zero, with high variance. It's pretty hard to test the claim that this is the correct belief, though. It's based on seeing a lot of people have very strong divergent beliefs.)


Remember when we got into an argument over whether Obama was really, really smart or world-changing genius? An old classmate is now betting him a million dollars his GPA is higher.
Wayne Allen Root is a rather funny, laughable character. I'll also add that I think standardized test scores are better indicators of intelligence than GPA.

Hopefully Anonymous

TGGP, I'm not sure what the point is, unless maybe you have an emotional attachment with maintaining doubt about Obama's intelligence?

My theory is to explain why we have a 1/2 black guy doing so well politically without substantive precursors, and also why being african american is one of the greater stigmas in American society. Basically it's that african americans are one of the less intelligent trait cohorts in America (thus other trait cohorts are able to put stigma on that trait set rather than other trait sets, such as short people or east asians), but Obama is one of the smartest people in America that desires to be President. Thus we have the surprising result that Obama is likely to be President despite being part of a trait cohort that is one of the highest stigmatized.

Hopefully Anonymous

Wow, I was offline for a week and I see that Obama/McCain are almost at 50/50 parity on intrade.


I have to admit, I wasn't expecting that.

I'll have to look further to see if this is further headline chasing or if something else is going on.

Hopefully Anonymous



This seems silly, McCain is now trading higher than Obama on intrade. And yet Obama still leads handily in electoral college advantage. I suspect this intrade traders engaging in more headline chasing. I plan to have my assistant buy a substantial amount of obama stock on intrade today.


I think blacks do better than asians politically, considering the number with college degrees (basically a pre-requisite for entrance to the elite). In Freakonomics they had an analysis of the Weakest Link which found no bias against blacks, a bias against Hispanics at the beginning (so they are thought of as a drag rather than competition) and a constant bias against the elderly (because we just don't like being around them).

Douglas Knight

And yet Obama still leads handily in electoral college advantage.

maybe that was true when you wrote that post, but here's a mistake you might have made: to simply use a threshold of 50% in polls or intrade to assign states to candidates.

This page computes expected number of electoral votes using Intrade data. It also gives "leaning" numbers, favoring Obama, which I think means threshold. I'm not sure expected electoral votes is the right way to go, but it's a quick consistency check on intrade. The point is that there are a bunch of swing states that are all leaning to Obama, but he needs them all, and they're just barely blue.

A month ago, it wasn't clear which states mattered, but now it looks like it's as focused as in the last two elections.

Hopefully Anonymous

Last I checked Obama had the edge in the electoral college, and in the "needs them all" stats (McCain's "need them all" situation was even more dire). But that's not Obama's only advantage. He always wins in the House of Representatives and has the fiercer litigation for office/refuse to withdraw history (although McCain does have the advantage in the Supreme Court).

McCain's numbers seemed to shift on intrade with the popular vote, which I don't think means too much (more numerical support in already solid red states due to Palin?) And intrade traders have already showed themselves to be plenty dumb with the republican vice presidential selection debacle.

I'm leaning towards the McCain/Obama price inverse being due to dumb money entering intrade on the heel of the republican convention.


Balko shows McCain with an electoral advantage in both total and safe states:
It's pretty close though, and who knows what will happen by the election

Hopefully Anonymous

What's the general trend and consensus of people doing these type of electoral college projections, though?
If McCain has managed to turn it around at the electoral college level, that's pretty impressive. If he turns that into a win, then it'll be a long time before either party leaves a woman off the election ticket. The funny thing is that the polls were clear before this election began that women are a voting majority and all things equal claim to prefer to vote for a woman. I forgot earlier, that Obama also seems to have a turnout machine advantage, which I think would be a huge plus when the polling is close. I don't think religious conservatives can restart their turnout machine with anywhere near full effectiveness at this late date (once again, only relevant if the election is otherwise close).

Hopefully Anonymous

okay, a couple more thoughts:

1) you're right, it seems like the consensus is now that on the face of it McCain now has a slight electoral college edge:


2) the electoral college results seem very close, and another pro-Obama factor is that independents tend to break for change, though I've heard people also tend to be less likely to vote for a black candidate in the privacy of the voting booth so that would be a pro-McCain factor. It's tough for me to sense though how the counterdiscriminations stack up between a young health taller black candidate, and an old, weaker, shorter white candidate.

Hopefully Anonymous

The intrade presidential market seems to me to have proved itself to be smarter than me. Also perhaps Mankiw (although I intuit he got lucky while performing one of his standard McCain alignments -still can't argue with the outcomes).

About 1 week before Obama's electoral college advantage vanished intrade shifted from 60/40 Obama to about 50/50. McCain recently has a slight EC edge, and Intrade's been reflecting that too -it did IN ADVANCE.

This may be a powerful personal demonstration of the superior intelligence of markets that meet a certain threshhold standard of traits.

The performance is a marked contrast to the republican Veep market.

The comments to this entry are closed.