I don't much like Steve Sailor's politics or policy arguments, but I surely do respect his research.
He makes a pretty good case for a "Dirt Gap" - that the availability of land for suburban development drives political trends via housing costs. It makes me wonder if Houston and New Orleans are trending Democratic, if his theory is right. Doesn't *look* like it - Austin was more blue than Houston in the last election, although New Orleans is definitely an island of deep purple. In Austin/Houston's case, the difference is driven culturally - Austin is a countercultural exception to the Texas case, even though it's totally and comprehensively land-locked. Pittsburgh's another counter-case - the three rivers hardly take up enough land to affect land availability, although the steepness of Pittsburgh's many, many canyon-walls does significantly reduce the available real estate stock.
But the "Black belt" of Mississippi and Alabama cuts right across the area of cheapest housing in the country. A lot of Sailor's formulations carefully use the term "non-Hispanic white" to exclude distorting factors from his argument.
Link via Mickey Kaus.