Planet of Cities is for a specialized audience, but it has one very big point that I didn’t realize:
New empirical evidence on the average population density of cities across space and time confirms that these densities have been in decline almost everywhere for a century or more. The new evidence is counterintuitive, since numerous academic researchers believe that urban densities have been on the increase. Were that true, it would lend encouragement and support to those favoring densification. However, urban density decline has been persistent and global in scope, and it predated the automobile. It is not restricted to the United States or other industrialized countries, but is pervasive in developing countries as well. [. . .]
The forces driving density decline—rising per capita incomes, cheap agricultural lands, efficient transport, and income inequality—are quite formidable. Accordingly, absent a highly effective policy intervention or a steep increase in travel costs in the future, there is little reason for the global decline in densities to slow down anytime soon.
I like living in cities (if New York were less expensive—as it was until the ’90s—I’d happily live here forever) and attack legal rules that prohibit higher buildings. But the research Angel has both conducted and cited indicates that, even in the absence of such rules, cities would still probably be getting less dense, or at most evening out. That is news to me; he also says, “No matter how we choose to act, however, we should remain aware that conscious and conscientious efforts to increase the density of our cities require the reversal of a powerful and sustained global tendency for urban densities to decline.” Newer cities, like Phoenix or Dallas, are even less dense than older cities, which is an innovation problem for the reasons described in Steven Berlin Johnson’s Where Good Ideas Come From and Edward Glaeser’s The Triumph of the City.
Cities like New York and Boston remain so important because creating dense neighborhoods like those found in both areas is effectively impossible in today’s climate of urban land-use controls. Building such areas doesn’t have to be impossible, but we, collectively, make it so. Take Seattle, a city I’m very familiar with. In downtown Seattle, rents have effectively increased by 40 – 50% from 2002 – 2012, even though some construction has been permitted. Rapid increase indicates that much more could be commercially built, and that many people want to live there and will pay to do so.
Some cities have attempted to limit horizontal urban sprawl, and presumably increase vertical urban height. Portland is a case in point, because it created an “urban growth boundary” (UGB) and thus has been a test case for some city planning. By density, which is arguably the most important measure of whether an UGB succeeds, the UGB failed:
The chief aim of the UGB was to contain urban sprawl and preserve the natural beauty of the surrounding countryside within reach of city residents. Sprawl was not defined precisely, but presumably it included both low-density development and fragmentation. After examining the change in built-up area density within the UGB between 1973 and 2005, we found that densities decreased rather than increased.
Even cities that have explicitly attempted to contain sprawl by increasing density have not fully succeeded. This points to the obvious need to plan for sprawl, and to plan for less centralization:
The possibility that cities worldwide are now in a process of transformation from a monocentric to a polycentric spatial structure poses an interesting challenge. It suggests that if public transport is to be a viable option in areas of expansion to economize on the energy expended and to limit greenhouse gas emissions, then it cannot be limited to continued reliance on radial routes to the city center. The transport network must be two-dimensional, providing frequent and reliable service among suburban destinations over the entire metropolitan area, rather than a one-dimensional network of radial routes into the city center. Some public transport systems that already provide such service are the bus lines of Edmonton and Toronto in Canada.
To provide reliable point-to-point service throughout metropolitan areas and to function effectively, bus lines or new transportation technologies will need to operate on a grid of arterial roads.
Again, this is certainly true in Seattle: I can’t find a citation, but for a time there was more Class-A office space in downtown Bellevue than downtown Seattle. In the greater Seattle area, there are major population and employment centers in Seattle itself, Bellevue, downtown Kirkland, Redmond (where Microsoft is headquartered) and Renton (where Boeing has or had many facilities). Many people I went to high school with live on the Eastside and seldom cross Lake Washington, and they can live a relatively urban experience in downtown Bellevue or Kirkland. Seattle as a city still imposes severe height limits in areas adjacent to downtown, and, as a result, some of the development and population infill that might otherwise take place, say, east of 12th Avenue or North of Olive / John on Capitol Hill, instead moves to Kirkland and Bellevue.
The style in Planet of Cities is unimpressive, verging on non-existent, as the quotes above demonstrate. But its insights impress, as does its re-evaluation of the conventional wisdom about rising densities in cities. As I wrote in the first paragraph, it’s a specialist book for specialist interests. Still, most people hold naive views on city planning, even if they don’t realize they hold such views. Planet of Cities replaces naiveté with knowledge. It could use panache, but that’s true of almost all textbooks and monographs.
EDIT: Anthony Flint of the Lincoln Institute of Land Policy wrote to ask me to note that the Lincoln Institute of Land Policy published this book. I’m not sure why this is important, but here it is.