2004 Correction

Relative to the discussion below it’s been pointed out to be that Andrew Gelman’s data that I was citing uses the average vote within congressional districts, not the nationwide turnout which is different thanks to differential turnout. Gelman’s methodology is probably the right one to use for some circumstances, but not for the issue I was discussing. The relevant correction is that in 2004, 49 percent of the total House vote went to the GOP and only 46 percent to the Democrats.