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.
In a previous post, I pointed to the car as a key feature in the nation’s deepening economic and political fissures. It’s becoming clearer that how we get around our cities and towns is a significant aspect of these divides. That’s the big takeaway from my analysis, with my colleague Karen King at the University of Toronto School of Cities, of recently released 2017 commuting data from the American Community Survey. The dataset covers 270-plus metropolitan areas.
Drive alone to work: More than three-quarters (76.4 percent) of commuters drive to work alone. But in the New York metro area, the share is just about half. It’s 57 percent in San Francisco; about two-thirds in Boston, Washington, D.C., and Seattle; and about 70 percent in Chicago and Portland. And, as might be expected, a smaller-than-average share of workers drives to work alone in more compact college towns such as Boulder, Colorado; Corvallis and Eugene, Oregon; Ann Arbor, Michigan; and Ames and Iowa City, Iowa.
Transit: Five percent of U.S. commuters use transit to get to work. New York City, with its extensive subway and rail system, is the big outlier here—more than 30 percent of workers get to their jobs by transit in greater New York City. The only other metros where 10 percent or more of workers commute via transit are San Francisco (17.4 percent); Boston (13.4 percent); D.C. (12.8 percent); Chicago (12.3 percent); Seattle (10.1 percent); and Bridgeport-Stamford, Connecticut (10 percent).
Walk to work: Less than 3 percent (2.7 percent) of Americans walk to work. But more than 5 percent of workers do in New York City (5.9 percent), Honolulu (6.5 percent), and Boston (5.2 percent). An even larger share walks in smaller metros and college towns including Flagstaff, Arizona (9.7 percent); Iowa City, Iowa (8.7 percent); Jacksonville, North Carolina (8.6 percent); State College, Pennsylvania (8.5 percent); Corvallis, Oregon (7.6 percent); Ann Arbor, Michigan (7.5 percent); Ames, Iowa (7.1 percent); Lafayette, Indiana (6.4 percent); Burlington, Vermont (6.3 percent); and Bloomington, Indiana (6.1 percent).
Bike to work: Just half of a percent of Americans nationwide bike to work. But nearly 7 percent do in Corvallis, Oregon, and more than 4 percent in Boulder; Ames; and Santa Cruz, California. Two or three percent of commuters get to work by bike in other college towns: Gainesville, Florida; State College; Ann Arbor; and Madison, Wisconsin. Among large metros (with more than 1 million people), almost 2 percent of commuters get to work by bike in San Jose and San Francisco.
Carpool: Roughly 9 percent of workers carpool to work. Around 10 or 11 percent carpool in the tech hubs of San Jose, San Francisco, and Seattle, as well as in San Antonio, Houston, and Phoenix.
Work from home: Just over 5 percent of Americans work from home, which is a much larger share than those who walk or bike to work. Nearly 14 percent of workers in Boulder work from home, followed by 11 percent in Lawton, Oklahoma, and 10 percent in Asheville, North Carolina. Among large metros, roughly 7 or 8 percent of workers work from home in Austin, Denver, Portland, San Diego, San Francisco, and Phoenix.
To get at the specific factors associated with America’s geography of commuting, my colleague Charlotta Mellander ran a basic correlation analysis and a cluster analysis. As usual, I will point out that correlation does not in any way infer causation, but simply points to associations between variables. Still, some clear patterns stand out that are worth highlighting.
Size and density are positively and significantly associated with using transit (0.64), biking (0.36), walking (0.43), carpooling (0.68), and working from home (0.66). While density is negatively and significantly associated with driving to work alone (-0.36), population size is not significantly associated with it.
Education is another piece in the picture of how Americans get to work. People are less likely to drive to work alone and to use alternate modes in metros where more adults are college graduates. The share of adults with college degrees is negatively and significantly (-0.41) associated with driving to work alone, and positively and significantly associated with using transit (0.56), biking (0.62), walking (0.56), and working from home (0.50), although it is not statistically associated with carpooling.
The same basic pattern holds for class. Across metros, the share of workers who are members of the knowledge-based creative class is positively associated with using transit (0.56), biking (0.62), or walking (0.56) to get to work, as well as working from home (0.50), and it is negatively associated with driving alone to work (-0.44), and the same holds for the local concentration of high-tech industry jobs. But the reverse is true for the working class. Across metros, a higher concentration of working-class jobs is positively associated with driving alone to work (0.36) and negatively associated with using transit (-0.48), biking (-0.39), and walking to work (0.32).
Money matters, too. In metros with higher wages, a larger share of workers walks (0.43), bikes (0.50), or uses transit (0.66) to get to work, and a smaller share drives to work alone (-0.34).
Weather plays a role, but not necessarily in the way you’d think. People are more likely to drive to work where the weather is warm (0.32) and less likely to use transit (-0.19), bike (-0.23), and walk (-.34) to work. So the way we commute is more closely related to demographic and economic characteristics of metros than to their climate and weather.
The way we get to work is also related to our political cleavages. On the one hand, commuters in more progressive metros—those where Hillary Clinton got a bigger share of votes in the 2016 election—are more likely to walk (0.44), bike (0.44), or use transit (0.59), and less likely to drive to work alone (-0.36). Commuters in more conservative metros, where a larger share voted for Trump, are more likely to drive to work alone (0.44). This reflects the fact, though, that more liberal metros tend also to be denser, more affluent, and more educated.
Our commuting patterns are associated with key dimensions of what I dub the new urban crisis. Housing is less affordable, inequality greater, and economic segregation higher in places where commuters are less dependent on the car. Median housing costs are positively and significantly associated with transit (0.59), biking (0.48), carpooling (0.49), and walking (0.38) to work, and so are income inequality and economic segregation. These associations again reflect the fact that denser, more affluent, educated metros are more expensive, more unequal, and more segregated.
Our analysis shows a country and a people divided in how they get to work. Americans cleave into two distinct nations based on commuting: One, based in smaller, less advantaged, and more sprawling metros, depends on the car, while the other, based in large, denser, more advantaged, and more educated metros, uses a variety of alternative modes. Driving to work alone in a car is negatively and significantly associated with each and every alternative mode, especially so with biking (-0.47) or walking to work (-0.53).
This pattern also comes through in a statistical cluster analysis that Mellander did, which looks at how different types of commuting styles cluster together. One distinct cluster reflects metros where a larger share of commuters drive to work alone; a second reflects metros where larger shares of commuters walk and bike to work; and a third reflects metros where a larger share of commuters carpool and work from home.
Differences in how we commute are baked deeply into our economic geography. Larger, denser metros tend to have the most extensive transit networks, and tend to be where more affluent and educated people have opted to locate close to work in the city center or along transit lines. Smaller, more sprawling metros tend to be less educated and less affluent; they also often have less congested roads and are easier to navigate in a car.
We are cleaving into two nations—one where people’s daily lives revolve around the car, and the other where the car is receding in favor of alternative modes like walking, biking, and transit. Little wonder that bike lanes have emerged as a symbol of gentrification and “the war on cars” has become a way to call out the so-called urban elite.
CityLab editorial fellow Nicole Javorsky contributed research and editorial assistance to this article.