The British travel writer Jonathan Raban has a long piece on the Beeb’s website about political geography and the difference between urban and rural. Raban lives in Seattle, site of next year’s AAG.
His case is fairly familiar:
something close to a divorce has happened as the interests of the city and the countryside have diverged, and the acrimony and hard feelings that go with so many domestic splits have infected the politics of Washington state.
The “metropolitan area”, centred on Seattle and spread around the edges of Puget Sound, has more than 3.5m people living in it, and so can narrowly outvote the other 3m people in the state.
The metropolitans tend to be Democrats (our “leftie” Seattle congressman, Jim McDermott, was re-elected on 2 November with an 83% share of the vote), while most of the farmers, ranchers, loggers, miners, rural developers and contractors are staunch Republicans.
This much should be evident to any thoughtful or interested observer. Since the US elects presidents by state electoral colleges there is always a tendency to see a state as the unit of analysis. It is misleading however, as Raban argues, because there are within state differences. And usually these are along rural-urban lines. So starting about 10 years ago we have seen the rise of “purple states,” the concept that states are neither purely Democratic nor Republican.
However, that’s not the whole story. As the Democratic strategist Ruy Texeira has long argued, it’s not just urban/rural, but rather urban/inner suburbs/exurbs/rural.
Geographically, Democratic success in “mature” and “emerging” suburbs is more than offsetting Republican strength in exurban areas, while Republican majorities in high-growth states are being eroded by the very elements of their population that are growing most rapidly.
This applies perfectly to states like Virginia right now and possibly to states like Georgia in the future. Texeira is therefore very bullish about long-term trends favoring Democrats.
The reason for this is that urban growth is out-pacing exurban and rural growth. The counties around Atlanta, which is now over 5 million people are some of the fastest growing in the nation. Republicans might have a tricky time creating a new district in the suburbs. And Atlanta will likely be about half the state’s population when census numbers are released in a few months.
If Texeira is correct we would also need to factor in another layer of variables: ethnicity and income. If the idea of “tribalism” is correct, people vote less now for a party than with people of a similar nature (and for people who “shares my values” as the poll questions put it).
As the Republican pollster David Hill put it
Instead of party, it strikes me that candidates and their voters are becoming captives of affinity groups. They may reject party labels but don’t eschew being the Hispanic candidate, a suburban candidate, union-backed, rural, business-endorsed, regional favorite, gunner, greenie, God-fearer, etc., depending on where and how you live. It’s as if candidates know we are all looking for someone “like us,” something parties and partisans can’t be in a universal sense, that causes us to lunge instead for candidates who represent our “tribe.”
No doubt this is a bit overblown and not exactly a new trend (I remember coming to the US in the 1980s from the UK and being surprised how people didn’t just vote automatically for their party as in the UK).
Additionally, it’s probably more true at the state and local scale than at the presidential scale. And I think it works regionally as well. In the Senate elections on November 2, in a traditional blue collar state like Pennsylvania those identifying themselves as “liberal” voted 91% Democratic, and those identifying as “conservative” voted 88% Republican. But in next door Ohio liberals voted only 80% Democratic, and conservatives voted 86% Republican.
What we’d like to know (and some political scientist can crunch the numbers and we’ll map it) is what of these many factors is most important, where, and why. Is race more important in the south? Is percent urban in a county more important in the Midwest? Is class and income more important in industrialized states?
Cue the regression analyses.