New York Times to the Rescue- this is how well Obama did. He appears to have overperformed on everybody, even in Kerry's home state, save in the Redneck areas wherein people report their ethnicity as "American" (538 did a study on it, showing that percentage of "American" ancestry links one to rural white social conservative voters who are pretty much the dregs in the appalachians and that's it. It's quite hard to prove these people are racists, but are they almost assuredly the highest percentage of racists per capita? You betcha.)
- Map of people who report their ancestry as "American".
No Blacks.pdf - A map of counties in the US where there are fewer than 25 blacks reported on the census.
Percent Blacks.pdf - A map of counties showing the percentage black.
Unfortunately we haven't a spare parralel universe to run a White Obama so we can't directly compare, but generally we can say there needs to be proven one thing if Obama's race were to have negatively impacted him
1) It must be proven that his gain in Blacks provided less votes than his loss among Whites.
2) It must be proven that a white politician could have gained those white voters
3) It must be proven that a white politician would not have lost those black voters.
Now, #1: this seems relatively easy even to me- since blacks make up about 13% of the country, he must gain 8 times as many black voters as he loses whites.
The problem is with #2. Obama does better among whites than any democratic politician since the Dixiecrats save Lyndon Johnson. Obama could have lost the vote of 30%, maybe even 50% of the blacks in the country to McCain or anyone else and still won easily.
Further, looking at the map Obama doesn't appear to have gained much in those black regions of the south where they actually make up massive portions. He has gained in the midwest in counties where there are no black people just as much.
Further, it's hard to tell with #3 since every democratic presidential candidate save Obama has been white so we have little basis to compare black voters compared to the race of the president. No, Jessie Jackson or black senators don't count because they come from majority black districts and black caucuses- you can't gauge their white support easily, and they are far shorter in national appeal unlike Obama who would have won every single state save Arkansas, Tennessee, Oklahoma, and perhaps Kentucky if it were a matter of Δvote and not actual vote.
I'm not claiming his Race wasn't an issue- I believe a White Obama would have overperformed so massively that it would have made McCain look like McGovern.
But I can't think of a way to show it using what sparse data we have. Anyone have some clues on a good rule of thumb we can use?
Generally I'm somewhat skeptical of the idea that a huge amount of black support helps, because if only 1 out of 10 americans is a racist that causes him to get a net loss, even if almost every black in the country votes it would barely make it up. (And, Blacks are statistically the poorest, most urban voters in the country. Poor, Urban voters have a very high chance to not vote compared to others, in historical elections as of 2000.)
However, it's hard to prove it was a net loss simply because of the data paucity.
I think we can at minimum state that Obama's race had almost no chance of helping him inasmuch as that black people were not what carried Obama over the top- it was icing on top of his massive performance among whites, especially traditionally conservative or moderate whites. It could have helped in things such as "Message" or "Brand" that we can't measure reliably without more samplings, but that is rather meaningless for a discussion of statistics.
However, that doesn't mean race necessarily hurts him- Obama, after all, overperformed on all demographics racially compared to Kerry, and it's hard to prove he had a stunted overperformance among whites (that is, instead of +20% he would have gained, say, +40% if he were white) to the satisfaction demanded by my conscience.
