Hi Jake,

thoughts from afar:

>As I'm absorbing all this, it seems to make sense. Then what I think is THE number jumped out. 6 million! As in there were an estimated 6 million fewer ballots cast by whites in 2012 compared to 2008. I was going to spend my lunch today doing some spreadsheet math to see how this affects the "shares" but thankfully someone at RCP has done the analysis.
>http://www.realclearpolitics.com/articles/2012/11/08/the_case_of_the_missing_white_voters_116106.html
>
....
>While the Republican party clearly has an issue with reaching minority voters and needs to address their issues far better in the future, that's not why Romney lost. He lost because fewer whites voted for him than voted for McCain. 6 million fewer! They simply didn't show up. Why? I do not know. This was billed as the latest "biggest election of our lifetime", yet 5 million fewer voters total showed up than 2008. Why?

Equating those missing white votes as 6.5 mio fewer Mitt votes is wrong IMO.
Going from the assumption that 60/40 for Romney in whites holds roughly true for those abstaining as well,
having them vote would have added 1.3 mio surplus for Romney - still trailing in popular vote,
but might have tipped some swing states where the difference was really small.
No data analyzed here where those whites were missing,
but the article hints at no clearly visible lopsided distribution.

One other eyebrow raised on other unspoken assumption in the article :
Can the 55% turnout be said to be equal among the different ethnicities "normally" ?
And questioned at the end and important: Higher turnout in minorities a trend or a Obama-effect ?

regards

thomas