Will Autonomous Vehicles Lead to Greater Sprawl or Greater Density? Yes.
Will self-driving cars achieve the urbanist dream of infill construction, parking lots and garages transformed into parks and affordable housing, and universal first/last mile transit? Yes. Will self-driving cars create greater sprawl? Yes.
How is this possible?
In the debate over the impact of autonomous vehicles on cities, we often forget about individual choice. The characteristics that define self-driving cars as a mode of transportation don’t lend themselves to a singular impact on land use in the same way that earlier modes such as ships, trains, electric streetcars, or cars did, respectively. Their owners don’t need to worry about having or finding a parking space in proximity to their final destination, but they also don’t need to worry about all of the things that can currently make driving unpleasant or even dangerous: falling asleep at the wheel, being drunk or otherwise not sober, or merely losing productive time while behind the wheel. So what are we left with?
Individual choice. We are fortunate enough to live at a time — perhaps the first in modern history, if not all of history — when both high and low density areas offer attractive residential options. Whereas in other epochs of American history whether you lived in a city was more or less determined by socioeconomic conditions — from the industrialization of the country until cars [1], if you wanted to participate in the global economy you lived in a city; after cars, if you were middle class and able to move out of the city you did so as quickly as you could — today individual choice is a real factor. Individuals can make a true choice between competing lifestyles and decide what suits them best — density, cosmopolitanism, and action; or space, homogeneity, and privacy — and whichever they choose, autonomous vehicles will make that lifestyle better [2].
The result is what I call the “Manhattanization” of urban areas: the urban core will become more dense, as would-be urbanites who were previously prevented from moving downtown are newly enabled — the two categories that come to mind are parents of small children who need direct, door-to-door transportation to carry their offspring and assorted paraphernalia (strollers, diaper bags, soccer uniforms, dance outfits, etc.) but found it too difficult and expensive to maintain a car in the city; and those for whom the city becomes newly affordable as new residential units replace parking lots and garages (admittedly the factors enabling this category are more speculative, but the case for NIMBYism over converting a parking garage to housing, or building over a parking lot, seems a lot more tenuous than that over replacing single-family homes with multi-family development. Does anyone really want empty parking structures dotting their neighborhood?) — while the suburbs will become more spread out as no longer having to drive the car yourself makes it possible to spend your commute getting work done or viewing TV/movies/VR/whatever people view for entertainment in the future.
In other words, urban areas will start to look more like the New York metro area, and less like the LA metro area. As illustrated in the figure above, New York City has an average density over 8x the average density of the rest of its metro, while the density of the City of LA is only 1.2x the average density across the entire rest of its region[3].
[1] By “until cars” I mean the period after WWII when the auto-dependent suburbs became the only real housing option for new families in the white middle class. Flight from the cities to the suburbs obviously did not occur all at once, but rather occurred proportionally to the affordability of the automobile as a mass market good and the availability of suburban housing developments to which to move. The earliest modern, car-oriented suburbs were already being built as far back as the late 1920s.
[2] We are fortunate that the advent of fully self-driving cars seems like it will occur at a time when the cultural zeitgeist is in favor the return to cities and the urban core. If autonomous vehicles had first become available in the 60s, when the National Interstate and Defense Highways Act had just been passed, the the nation was in the throes of its orgiastic suburbanization, and the Robert Moses’ and their traffic engineers around the country were running the Interstate Highway System through city centers, there would have been no thought to using AVs to increase density. The city center would continue to have been regarded as a destitute and unseemly place to live, let alone raise a family, and the result would have been an approach to AVs that focused only on their utility for the facilitation of sprawl (though not in those terms, of course).
[3] Thanks to Ashley Hand of @CityFi for pointing out this concept in Urban Mobility in a Digital Age, her transportation technology strategy for the City of Los Angeles.
