The divides exist, but this map does virtually nothing to shed illumination on the matter. I think everyone here can figure out why the methodology and sampling size would lead to inaccurate results...and at the same time, I think everyone would also agree that this sampling would vastly undercount the actual amount of racism among our citizenry (as frequent Twitter users have a much different demographic than the voting demographic).

Is there a way to see what actual tweets were classified as racist? Or is it just ones that used words commonly accepted to be hate speech? If the latter, then my main issue with this is that this is nothing more than social porn...a way to set the bar for racism so blatantly high that we forget that the most damaging, pernicious kinds of racism are not going to be expressed through explicit language by a few scattered morons.