Richard Florida is the great proselytizer of the creative class:
[The creative class is] a fast-growing, highly educated, and well-paid segment of the workforce on whose efforts corporate profits and economic growth increasingly depend. Members of the creative class do a wide variety of work in a wide variety of industries—from technology to entertainment, journalism to finance, high-end manufacturing to the arts. They do not consciously think of themselves as a class. Yet they share a common ethos that values creativity, individuality, difference, and merit.
I’ve previously written about Florida’s work and how his insights help us understand US politics and elections. His work has value, but I think he and his like-minded theorists tend to look at correlations and attribute too much to their favored factors, ignoring others. This post from yesterday is a good example. He looks at population density, the mix of industry, racial composition and other factors that may answer his question: “what is it exactly that makes big cities vote Democratic?” But there are several problems with his question, and how he attempts to answer it. First, there isn’t an “exact” answer to such a question; most American cities have been voting Democratic since the late 1920’s, long before there was any significant “creative class,” and when there were few sizable cities west of St Louis or south of Washington DC. History can’t be distilled to an “exact” answer. Furthermore, he asks about cities, but the data he analyzes is metro areas; those aren’t the same thing. And there are many correlations and comparisons he doesn’t consider, including the obvious correlation that almost all of the biggest metro areas are in Democratic states.
Of the thirty largest metro areas in the United States–all but Las Vegas, at 1,969,675, have over two million residents–twenty-four are partially or entirely in states Obama won both times. Seventeen are in states that have gone Democratic in every election since 1992. And five–New York City, Boston, Minneapolis, Seattle and Portland–are in states that have been Democratic since 1988, when they went for Dukakis. But “metro area” is not the same as a city, and it isn’t a particularly useful measure of political geography. While Obama did lose the Houston and Dallas metro areas, he won the cities of Houston and Dallas. So in terms of cities themselves, almost every single one over 500,00 went for the Democrat. Partisanship of a metro area is most likely determined not by the votes in its core city, but by the votes in its suburbs. If Florida is going to make statements about cities, he should analyze cities. If he’s going to analyze metro areas, he should make clear his claims aren’t about cities.
As for metro areas, according to Florida’s numbers, Obama won 150, Romney 214. But is that number meaningful? My Chicago neighborhood is less than 2 square miles and holds less than 1% of the population of the Chicago metro area. It’s one of the most racially, ethnically, religiously and linguistically diverse places in the world. And my little neighborhood has almost exactly the same number of people as are in the 168 square mile Carson City NV metro area. Each of the 50 smallest metro areas has fewer than 120,000 people, and most are overwhelmingly white and homogeneous. The total population of those 50 metro areas–roughly 1.8 million–is about the same as the population of the 38th largest metro area, Columbus, OH, and less than one tenth the population of Greater NYC. Lumping Mankato, Minnesota and Pocatello, Idaho in with the Dallas and Boston metro areas doesn’t tell us anything that’s particularly useful about “cities.”
I love maps, so I spent some time looking at the map accompanying Florida’s post to see if there was something useful:
In fact, there is something useful, just not what I think Florida had in mind. Look at the map, and imagine you decided to color in the rest of the map, using a red pen for the counties won by Romney, and a blue pen for the counties won by Obama. You’d barely use your blue pen; the blue areas would continue to look like islands, but the red puddles would become a sea of red: But what about recent trends, you may ask? Well, trends are kind of important to look at if one is going to say economic change effects current voting. But Florida’s analysis is not a look through time, it’s a snapshot of the moment. We have no idea, if we look at only one point in time, what the trends are, if there are any. But if we look back at the last few elections, other than the the major shift from Democrats to Republicans in Appalachia and the Ozarks, the 2000 map looks pretty much the same (on this map, from the terrific web site run by David Liep, red is for Democrats and blue is for Republicans):
From election to election, the states contested for their electoral votes haven’t fluctuated much since 1996. Many big cities are in states that weren’t contested, while others–in particular those in Florida, Pennsylvania and Ohio–were bombed with ads and blanketed with canvassers and GOTV workers. One shouldn’t generalize too much about election results without considering that difference.
Rural America isn’t entirely white. The “black belt,” an arc from southeast Virginia through the Carolinas, Georgia and in to central Alabama is mostly rural and heavily African-American. It’s the same along the Mississippi River from Memphis and down through Arkansas, Mississippi and Louisiana. In the Rio Grande Valley of Texas and in much of the mountains of New Mexico and Colorado there are significant populations of Hispanics, some whose families have been there for 400 years. And Native Americans are clustered on reservations across the west. But most of rural America is overwhelmingly white. And other than New England and some heavily Scandinavian and German areas of the Upper Midwest, rural America has been voting overwhelmingly Republican for decades. Looking at his data, Florida ends up in a familiar place,
America is divided between cities of knowledge and skill and the rest…This divide is as economic and geographic as it is partisan. America’s polarized politics is a product of its deeply-etched geography of class.
Maybe. But probably not. Big cities, even in conservative states, vote Democratic. Metro regions generally vote consistent with their state and section of the country. There isn’t a single geographic divide in America, and national section–South, Plains and Great Basin vs Northeast, Great Lakes and Pacific Coast–is probably a far stronger correlation to how a metro area votes than does it’s creative class economics and demographics. To see that, all you have to do is look at Florida’s map…and use your imagination to color in what’s missing.