Among the dumbest of many dumb things passed off as intelligent commentary on Election Night 2010 was the constant emphasis on color changes on the big national Congressional district maps that all of the major networks kept on prominent display. Fox, CNN, and ABC (and probably others, but I can't watch everything at once) are in agreement, apparently, that viewers are 5 years old and thus best able to absorb the events of the evening through brightly-colored graphics. Look! Look at the blue turn into red. That means things are changing!
Anderson Cooper in particular was heavy on the "Just look at all of this red, Wolf!" commentary. Where there was once some blue there is now just a "sea of red", a metaphor he managed to use dozens of times without making a joke about the Red Sea. What AC and the gang were so breathlessly reporting underscores near total ignorance of a rapidly increasing trend in both presidential and midterm election years.
This is a county map of the Illinois Senate race. Say it with me, guys: Just look at all that red!!
Bearing in mind that the race was actually pretty close (48%-46%, with the final difference around 100,000 votes), note that the Democrat Giannoulias won three counties in Illinois. Three. Out of 102. One of them – St. Clair County, home of East St. Louis, he won by 0.1% (about 200 votes). Another, Alexander County, had all of 2400 votes cast. And then of course he won Cook County…by 450,000 votes. 450,000!
Basically Giannoulias did what Democratic candidates always do in Illinois, and increasingly what Democratic candidates do in every state – win the major urban areas by so much that losing the entire rest of the state won't matter. Alexi G lost, of course, but the race was damn close. He came within a hair of winning and he lost 99 of 102 counties in the state.
This pattern reminds me of 2004, when John Kerry won a number of states in a similar fashion. In fact, nationwide Kerry won 583 counties…to Bush's 2,530. That is a margin of nearly five to one, yet the popular vote was within 2%. Among the 5 counties with the largest margin of victory for Kerry were San Francisco County, Washington D.C., and Bronx Borough. Bush's were Glasscock County, TX, Madison County, ID and other similarly fly-blown places.
In short, gaping at the color balance on the map is ridiculous because Republicans have proven beyond any doubt in the past 30 years that they are absolutely dominant in areas where no one lives. Huge, sparsely populated districts/counties/states are their cup of tea. And they win a ton of them. Democrats, on the other hand, make their victories count by winning – and winning big – where humans actually live. The level of intrastate polarization between rural and urban areas in modern elections is hard to overstate. It really boils down to city folk vs. country folk to a degree that might be unequaled in our history.
Yes, American elections have always been divided by some more-or-less clearly defined cleavage, be it North vs. South, Coasts vs. The Middle, or industry vs. agriculture. But in recent years, Obama's near-blowout in 2008 excepted, both parties have struggled to win votes beyond their rural and urban comfort zones. Even in the reddest Red State, Democrats do far better in urban areas. Even in the bluest Blue State, the rural areas are highly conservative. Elections end up turning on turnout differentials and the "swing" areas in the suburbs – which lean Republican far more often than not.
This raises an interesting question for Democratic strategists, and it is the primary ideological debate within the party at present. Does it make more sense to try to alter the message enough to achieve some nonzero level of success in rural areas, or should they just go all in on urban-focused liberal policy to maximize turnout in the cities? In other words, should they even try in rural areas or is it rational to say "Ah, fuck it, let's just try to win Cook County by 550,000 next time" given that a well-oiled turnout machine could feasibly accomplish the latter? Perhaps just as interesting is the question of which elections are "normal" and which are anomalous: the 2000-2004 years in which Democratic success was limited strictly to urban areas or 2006-2008, when they had success elsewhere?
Unfortunately the results in 2010 suggest the former as the status quo, with the 2006-2008 results attributable mostly to anti-Bush sentiment and, in 2008, the spectacularly bad presidential ticket on the GOP side. This year snapped the Democrats back to their pattern of winning only in the cities and trying to make that hold up against the big ocean of rural red that so enthralls the crack professionals inside the Beltway.