Andrew Gelman is one of a rare breed -- someone who's an honest-to-god empiricist about political subjects. He has
a new book out that engages in some non-partisan, data-rich mythbusting about the intersection of demography and politics in the USA, and if there's any justice every political pundit in the country ought to be forced to read it. He has
a website up to go along with the book, and on the front page you'll find a list of the main myths he punctures in the book. (It will not escape the astute reader that some of them are in direct conflict with eachother -- but hey, it isn't Gelman's fault the discourse just is incoherent.)
It actually grew out of
a paper he wrote carrying the same title, which in all fairness really contains most of the core content of the book. So if you're not up to the full-length treatment you can at least read the paper. I don't expect it to make much of a splash -- facts are so boring after all -- but it's a breath of fresh air anyhow.