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.
Despite its merits, in the U.S., density peaked in the 1950s and has declined since then.
The biggest issue shaping the future of our cities, and our nation, is the question of how we grow. Do we continue to try to sprawl our way to the American Dream, or do we add the density that powers innovation and economic growth?
A new study by the urban economist Issi Romem of BuildZoom sheds light on both America’s historical pattern of growth and development as well as how these patterns shape the trade-offs and choices we have about our future. Using detailed data from the American Community Survey, Romem tracks the connection between housing development, sprawl and density since the 1940s.
In his previous research, which I wrote about here, Romem made a basic distinction between so-called “expansive” and “expensive” cities or metro areas. “Expansive cities” as their name implies, have sprawled outwardly and in doing so remained largely affordable. “Expensive cities” are more hemmed in by geography and have not been able to expand nearly as much, and have seen their housing prices increase as a result. The developed footprints of “expansive” Las Vegas and Atlanta, for example, have expanded by more than 200 percent between 1980 and 2010, compared to just 30 percent for expensive New York and San Francisco.
But looking at the developed footprints of metro areas is just one way to measure sprawl. A metro can still sprawl without adding to its footprint, by developing the majority of its housing in low-density areas.
The Bias Toward Sprawl
In fact, in his latest research Romem finds a more all-encompassing bias for sprawl that extends across the entire United States and virtually all of its metro areas.
Roughly 60 to 70 percent of America’s existing homes are located in low-density areas, defined as those with fewer than four homes per acre. The overwhelming majority of new homes built across the country since the 1940s (90 percent of them) have been developed in low-density areas.
In the 2000s, nearly a quarter (23.3 percent) of all new homes built were built in undeveloped areas, a third (33.2 percent) were built in areas with a prior density below one home per acre, and another third (31.9 percent) were in areas with a prior density between just one and four homes per acre.
This bias toward sprawl comes through clearly in the chart below, which compares the share of new homes built in low-density areas to the share of existing homes in these same low-density places. Metros that lie above the 45-degree line are those where the development of new homes has been more concentrated in low-density areas. You’ll notice that not a single U.S. metro lies below the line. In each and every one—older versus younger, Sunbelt or Frostbelt, expensive or expansive—new housing development has been highly skewed toward low-density sprawl.
Share of Low-Density Housing (Existing vs. New)
The great majority of all new U.S. housing, including housing built in dense and expensive metros, has come from suburban expansion. For the lion’s share of the top 40 most populated metros, more that 90 percent of all new housing built between 1980 and 2010 has come from low-density development.
This is true not just of Sunbelt metros such as Las Vegas, Phoenix, Atlanta, and Charlotte, but also tech hubs such as Seattle and Austin, and Rustbelt metros such as Detroit and Cleveland. New York is the biggest outlier of the bunch, but still almost two-thirds (63.8 percent) of all new housing built in Greater New York between 1980 and 2010 has come from low-density development. In high-priced San Francisco, nearly 80 percent of new housing developed over the same period has come from low-density development.
This bias toward sprawl turns on the simple fact that it is far cheaper and far easier to add new housing at the periphery of metros than it is to build in already built-up areas. Undeveloped areas have by definition little or nothing on them. They are essentially virgin territory, so there is much less regulation of or opposition to new development. It is harder to assemble property in already built-up areas and such development faces greater land-use restrictions and more resistance from existing residents. This is also why a great deal of new urban development take places in older industrial districts, such as New York’s Hudson Yards, where land is easier to assemble and there is little, if any, residential development.
America’s Long-Running Density Decline
Urbanists like to think that the back-to-the-city movement has increased density and mitigated sprawl, but Romem shows that America has seen a long-run decline in density. As the next two charts show, increasing sprawl has contributed to a long-term decline in densification. According to both measures, density peaked in the 1950s and has declined since then.
Share of developed U.S. land density (threshold and percentage)
The first one traces the share of all developed land that has crossed different levels of density—one house, four houses, and ten houses per acre—each decade between 1950 and today. The share of developed land that crossed the four or ten homes per acre threshold, still a fairly suburban pattern of development, has remained strikingly small over this entire period. In fact, it has been declining rapidly since the 1950s.
The same density decline can be seen in the second chart as well, which traces the share of developed land that has been built out by adding 0.5 houses, one house, or two houses per acre. Again, note the dramatic decrease in densification since the 1950s.
The Land Use Trilemma
Romem frames the issue in terms of a complex tradeoff that he dubs the “land use trilemma.” He writes,
Cities confronting growth pressure face a tradeoff between accommodating growth through outward expansion, or accepting the social implications of failing to build enough new housing. Sprawl is not something to be welcomed. But people must understand that with neither outward expansion nor meaningful densification, U.S. cities cannot provide enough housing to prevent equally unwelcome changes to their social character.
While most urbanists—including Romem—think sprawl is bad, it does offer communities a way to deliver both affordable housing and preserve their urban neighborhoods. The downside here, of course, is increased congestion and environmental degradation. Alternatively, metros that choose to limit sprawl and densify can make housing more affordable, but that can also come at the expense of character and “charm” of their existing urban neighborhoods. For metros that do nothing—that neither sprawl nor densify—housing prices will go through the roof. Existing homeowners and the rich gain, while the poor and working classes are priced out.
This long era of cheap, suburban growth has its limits. Many of our most expensive metros have reached the limits of outward expansion, and many others are headed toward it. Not to mention, reurbanization and densification are required both for future innovation and growth, and to address the mounting inequality and spatial segregation of America’s expensive cities. Our cities and the nation as a whole face some very tough choices ahead.