Richard Florida is a co-founder and editor at large of CityLab and a senior editor at The Atlantic. He is a university professor in the University of Toronto’s School of Cities and Rotman School of Management, and a distinguished fellow at New York University’s Schack Institute of Real Estate.
The biggest U.S. cities are still Democratic strongholds, but new research sheds light on why some of them aren't.
We’re all familiar with the map of America’s “red” and “blue” states. But what’s less familiar and certainly less well understood is how political allegiances break down inside smaller jurisdictions within those states.
That’s why a new study of voting patterns inside counties in and around the country's largest metro areas is so important. The study, by University of Dayton political scientist Joshua Ambrosius and published in the Journal of Urban Affairs this month, examines the past four presidential elections (2000, 2004, 2008 and 2012) to divine the factors that shape voting patterns and political leanings within America’s urban areas.
To do this, Ambrosius assembled a detailed data set spanning 92 of the nation’s “core counties”—that is, the counties that contain the center city for metro areas with more than 250,000 people (see the map below). These counties accounted for between one-quarter and one-third of all votes cast between 2000 and 2012, the period studied. While many of these places reliably vote Democratic, Ambrosius is particularly interested in better understanding the core counties that vote Republican.
In sync with the conventional wisdom, Ambrosius finds that core urban counties trend blue. Between 2000 and 2012, as the graph below shows, Democrats won between 57 and 62 percent of these core county votes depending on the election year. “A greater proportion of the presidential popular vote in the core counties supports the Democrats than in the nation as a whole,” Ambrosius notes. “Indeed, these urban counties are considerably bluer than the population of all counties.”
But (and this is a big but) the picture gets more complicated and considerably more interesting when Ambrosius looks at the number of core counties won by specific candidates. Not all core urban counties are blue—not by a long shot. The graphic below, from the study, shows the shift in red and blue counties throughout the 12-year period.
Notice the number and wide dispersal of the core counties in purple, indicating counties won by George W. Bush in 2004, but by Barack Obama in 2008. Bush won 39 core counties in 2000 and 2004, 42.4 percent of the total. By contrast, 2008 GOP candidate John McCain only won 18 core counties, 19.6 percent, and 2012 candidate Mitt Romney won 21, or 22.8 percent.
Next, Ambrosius turns to the factors that shape this red/blue divide. To do so, he conducts a detailed statistical analysis of three sets of factors that shape voting patterns: social and demographic characteristics like race, age, education, density and the city share of county population; cultural factors like church membership and the share of evangelicals, veterans and same-sex households; and economic factors such as the unemployment rate, the share of the labor force employed in manufacturing industries and the housing foreclosure rate.
Again, many of his key findings conform to the conventional wisdom. As of 2012, blue urban counties were notably more educated and economically advantaged, denser, more culturally progressive and had greater shares of African-Americans. Red urban counties had more evangelicals and military veterans, fewer same-sex households and were more culturally conservative. As Ambrosius notes, these differences have become more pronounced over time. (His statistical models explain a whopping 73 to 81 percent of the variation in support for Republican presidential candidates).
The number of factors that distinguish red from blue urban counties has grown over time as well. In 2000, the red/blue divide turned on four key factors. Red core counties had more veterans and evangelicals, while blue core counties had higher population densities and greater shares of same-sex households. By 2012, several more factors entered into the mix. In addition to the same four, blue places had more college educated adults, higher incomes and more African Americans.
While social and demographic factors play the biggest role in distinguishing red from blue urban core counties, Ambrosius’ analysis finds cultural factors are becoming more influential. Economic factors—like the foreclosure rate and the unemployment rate—have actually declined in importance, having played their biggest role in the 2004 election (which may well seem surprising in the wake of the economic crisis).
Interestingly, Ambrosius finds that though being located inside a red state was associated with increased Republican support in a core counties, this location factor declined in importance over the course of the four elections. In 2008 and 2012, when Barack Obama turned a number of classic red states blue, there was no significant correlation between location within a red state and status as a red core county.
Ambrosius' findings shed additional light on the political role of density. One study of the 2012 election identified a tipping point of 800 people per square mile, after which many places turned from red to blue. Between 2000 and 2012, Ambrosius finds the average cutoff point for core counties is similar: roughly 810 people per square mile. Above that, two-third of core counties voted for Democrats; below it, two-thirds voted Republican. Ambrosius finds this crossover point trended downward over time, from 840 people per square mile in 2000 to 790 in 2012. As he notes: “the Democrats are lowering this crossover point in urban areas by capturing more lower-density urban market share.”
Ambrosius reminds us that while these trends favor the Democrats in such places, more than four in ten urban counties went red during the Bush years. This reinforces the point made by USC political scientist Jefferey M. Sellers that the new defining element of America’s political geography is not the swing state, but the growing ranks of older, economically troubled suburbs or “distress burbs,” upon which tight elections increasingly turn.
America’s real electoral map is not the red and blue one we see on election night. It’s a much more complicated one that in many ways we are still just beginning to understand.