Enough already with the words. Think of this as The Big Sort scorecard for the election, several different ways of seeing how the geographic clustering of like-minded citizens plays out in presidential elections.
First, the sort itself . Here we compare the "landslide counties" in the 1976 and 2004 elections. (Landslide counties are those in which one candidate won by 20 percentage points or more, counting only Republican and Democratic votes.) Both '76 and '04 were close contests, but the distribution of the vote changed dramatically over those 28 years.
In 1976, 26.8 percent of voters lived in a landslide county. (Democratic landslide counties are in black; Republican landslides are gray.) In close elections, the percentage of voters living in landslide counties rose steadily. By 2004, 48.3 percent of voters lived in a county where the contest between George W. Bush and John Kerry wasn't close at all. About six of every 10 counties were won by landslide margins in '04.
In 1976, Democrats won the vote in rural America. Bill Clinton broke even in rural America in 1996, and then rural counties went solidly Republican. So did the exurbs. Cities—particularly cities that produced loads of technology and patents—swerved Democratic.
Basically, the more dense the population, the greater the Democratic vote.
One way to track the election is by this rural/exurban/urban breakdown in each state. Thanks to Tim Murphy and the Daily Yonder , we do that in the chart below for the 2004 election. You can see that rural and exurban votes were essential to President Bush's victory.
(For those who wonder about such things, "rural" here are what the OMB and the census classify as "non-metro." "Urban" counties are called "metro" by the feds. Geographer Tim Murphy created an "exurban" category from "metro" counties where 40 percent to 50 percent of the residents live in rural settings. On the bottom line, you can see that rural counties had 17.4 percent of the vote, exurban counties had 9.2 percent of the vote, and urban counties had 73.4 percent of the vote.)
Given The New Yorker 's recent interest in the political significance of ancestry, I wondered why we shouldn't go straight to the source. This map comes from our friends at the U.S. Census. It shows the ancestry group with the largest population in each county. (Ancestry is self-reported.) It's sort of fun to switch back and forth between the 2004 landslide map and the ancestry map. You can see that Democrats didn't do well where the dominant ancestry group was "American." Maybe these are the folks Gov. Sarah Palin had in mind .