My esteemed friend Jonathan Cohn has a piece today at TNR that makes the useful if well-known point that the constitutional set-up of the United States Senate discriminates against full representation for urban Americans, which in turn is bad for liberalism.
As illustration, he uses a chart from a political science paper rating the “liberalism” of American cities with population over 250,000, and unsurprisingly, most of them are left-of center.
While I’m sure the basic conclusion is right, and the point Jon is making is absolutely right, I’d warn a bit about comparisons of central cities without some recognition of their wildly variable relationship to the metropolitan areas they inhabit. The most obvious example of a misleading numbers based on artificial boundaries is Washington, DC, limited by the boundaries of the District of Columbia. As of the 2010 census, DC’s population was under 11% of the population of the Washington Metropolitan Statistical Area (MSA). But it’s not entirely an outlier: the city I used to live in before DC, Atlanta, represents less than 8% of the population of the Atlanta MSA. But now the big city I live closest to, San Jose, has over 51% of the population of its MSA. Why? Well, according to Wikipedia, during the 1950s and 1960s, San Jose executed 1389 annexations, expanding its land area from 17 to 149 square miles. By the 1960s, Atlanta suburbanites had permanently organized themselves to resist annexations by Atlanta (race became an obvious factor as Atlanta became a majority-minority city during the late 1960s). Georgia’s constitutionally-based system of very small counties was a factor, too (the Atlanta MSA includes 28 counties; the City of Atlanta is mostly in one and contains a small portion of another).
In any event, the term “city” means a lot of different things in different areas, and that’s worth keeping in mind when they are compared.