Do Highways Frustrate Mobility?

Traditionally, the most common argument for new and widened roads seems (to me) to be that they will reduce congestion. In response, environmentalists and urbanists argue that induced demand may frustrate this alleged benefit—that is, that roads may actually increase congestion by making long car commutes easier and thus increasing the demand for driving. 

Pundit Matt Yglesias recently responded to this “induced demand” argument by stating that even if roads lead to more driving, “People got to go more places and do more stuff.” So if a road cuts the commute from downtown to (for example) a suburban WalMart from 20 minutes to 15 minutes, this is a great thing because more people get to go to the WalMart.

But this argument overlooks the longer-term effects of highways. City-to-suburb highways open up suburbs for development, thus causing people and jobs to migrate to those suburbs. Once people spread out into suburbs, commutes become longer. 

For example, suppose that in 1950, I lived in a neighborhood at the eastern edge of Cleveland, such as Glenville. Glenville is six miles from downtown Cleveland; if I was taking surface streets, that commute might have taken me 15 or 20 minutes. Because of the highway, I can now move to a suburb eight or ten miles out and still get to downtown in 15 or 20 minutes. So now I live further out, and my commute is the same. I haven't really saved any time at all, so I don't get to go more places or do more stuff at all. 

And if my suburb (like most suburbs) is more thinly populated and less walkable than Glenville, every destination I might wish to visit in my suburb (such as neighborhood restaurants, groceries or dry cleaners) is further away than it would be in Glenville. So even if I work from home, I still might be less mobile because I have to travel farther to reach anything other than my job. If this is true, I still don't get to go more places or do more stuff. 

Admittedly, there are some winners from this process. Some people really do enjoy larger yards, and may think that even if they are less mobile than they would be in a city, this loss is outweighed by the benefits of suburbia. But I'm not sure that this generalization applies to all suburbanites. Some people who move to suburbs might actually value bigger yards and low density. But others move because Glenville is no longer a desirable area. 

And why isn't Glenville desirable anymore? Because of the highways. First of all, highways make urban neighborhoods near them (such as Glenville) noisier and more polluted, as they become flooded with suburbanites' cars. Moreover, as the highway-induced growth of suburbia led to middle-class flight, Glenville and many other Cleveland neighborhoods became drained of their middle class, causing their schools to become poverty-packed and thus less attractive. (And as judges moved to suburbs, they were more willing to support desegregation orders that targeted central cities while allowing suburban schools to remain all-white). And as neighborhoods like Glenville declined, even more prosperous urban neighborhoods became less desirable, as rising poverty gutted Cleveland's tax base and political talent pool.  (A side note: Republicans often complain about Democrat-led cities, but overlook the fact that without highway-generated suburban sprawl, cities like Cleveland would have more Republican inhabitants and would thus be more politically diverse).

But even if I'm just as mobile in suburbia, the growth of suburbia makes other people less mobile.  When cities decline, all sorts of things start to move to suburbia—not just jobs, but even movie theaters and churches and other civic amenities. For example, many Cleveland neighborhoods once had movie theatres; now, only a couple of theatres are left in the entire city of Cleveland. As a result, if I stay in Glenville and don't drive, I have to change buses to get to a theatre, and may have to change buses to get to my job if it is not downtown (assuming that my job is even on a bus line at all). Thus, I might be less mobile than I was in 1950. And even if I do drive, the movement of jobs and amenities to suburbia might mean that I have to drive to reach some destinations that I could have walked to in 1950, or drive more miles than I would have driven in 1950, so I am still worse off.

One might argue that all of this is speculative: for example, how do I know that highways didn't reduce commuting times? Isn't it possible that suburbanites saved more time from faster speeds than they lost from longer distances? I have found no data for the first few post-suburbanization decades after World War II. However, since 1980 average commute times have steadily increased: from 21 minutes in 1980 to 22 in 1990 to 25 in 2000 to almost 26 in 2013. This trend isn't limited to states with lots of transit users; in Mississippi (where the largest city, Jackson, has no bus service after 8 pm or on Sundays), the average commute increased from 20 minutes in 1990 to 23 minutes in 2010. 

Another way of examining commute times is to see how many people have a relatively easy commute (say, under 20 minutes). Between 2010 and 2020, the number of commuters nationwide increased by 9 percent (from 131 million to 142 million), while the number of under-20-minute commuters increased by about 1 percent (from roughly 57 million to 58 million).* Similarly, among drivers the number of people with short commutes barely budged (from 52.2 million to just over 53 million) while the overall number of commuters increased from by about 9 percent, from 118.1 million to 128.7 million. Thus, it appears that even drivers are spending more time commuting, rather than “going more places and doing more stuff.”**

In sum, highways don't make it easier for suburbanites to reach destinations in the long run, because even though suburbanites win time by being able to drive faster, they might lose just as much time by having to drive more miles. And as commutes become longer, people without cars are worse off, because longer distances mean longer bus rides. 

*For 2010-20 commuting time data is available at Table B8134 of the American Community Survey.  

**Of course, it is too early to tell how the growth of telecommuting has affected this.