Mailing ListForum
City Places for City People
The Form of Cities: an Introduction to Carfree Cities

by J. H. Crawford


Parma, ItalyIntroduction
For some years now, I have advocated the development of entirely carfree cities. My position is based on a wide range of considerations. While the discussion here will center more on energy and the environment, I believe that the most compelling reasons to adopt carfree cities are in fact social and aesthetic. We are simply making a mess of our cities, and doing huge damage to our social systems, by our single-minded insistence on cars, cars, and more cars. The "need" to accommodate cars in our cities has led to the acceptance of an appalling degradation of our public spaces: the streets and plazas in which most of our lives could be conducted. It's no longer pleasant to walk through our cities or to linger in their squares, so we drive home and sit in front of the TV.

Despite the importance of the social and aesthetic issues, I suspect that energy and environmental considerations may become the more powerful early forces encouraging carfree cities. We could maintain "carburbia" in substantially its present form only if a sustainable energy infrastructure can be established with enough capacity to keep filling all those SUVs with cheap fuel. If, however, the ultimate capacity of our sustainable energy infrastructure is wholly inadequate to the demands of carburbia (or simply too expensive), then the nearly universal development of carfree cities becomes almost inevitable. Indeed, if energy limitations prove more serious than I anticipate, we might ultimately be forced to rely almost entirely on walking and bicycling for our transport.

I have proceeded from what I believe to be the most likely long-term circumstance: quite a lot of sustainable energy will be developed, enough to run efficient public transport systems (and supply other essential uses, albeit with much-improved energy efficiency), but insufficient to sustain carburbia. This assumption underlies the "reference design" for carfree cities that I have developed and which is presented in considerable detail in my book, Carfree Cities. (Such a "reference design" is an idealized point of departure; real-world conditions usually dictate major changes from a reference design, but its guiding principles are maintained.)

Roads to the Future
Reference design for a city of one million.  It is  about ten miles per side.When contemplating idealized city forms, transport and energy consumption are two of the most important design considerations and have profound effects on the resultant urban form. Indeed, the prevailing assumption that automobiles must be accommodated, no matter what the cost, dictates the form of carburbia. Roads and parking subsume most other considerations, with disastrous consequences for our cities and for the health of the planet.

It seems difficult to escape the conclusion that per-capita energy consumption in the industrialized nations is going to start declining within ten years, as a result of falling oil and natural gas production and limitations on the combustion of fossil fuels because of global warming. The eventual development of plentiful energy from nuclear fusion seems distant at best and unlikely at worst. If we assume that we must substantially reduce our energy consumption, at least over the medium term, then certain principles of development must be adopted, whatever the prospect for longer-term energy supplies may be. Several imperatives become evident:

Nancy, FranceThese basic measures can secure dramatic reductions in the energy required to run an advanced civilization and improve the quality of life in the bargain. We can best achieve these changes by simply returning to the way we shaped our cities before the use of fossil fuels became widespread. Until quite recently, cities were always built three to six stories high (no elevators!), streets were narrow (no cars!), one building always touched the next (no driveways!), and green space was kept to an absolute minimum.Reference design for a district, approximately half a mile across, housing about 12,000.If we make one adjustment to this paradigm--the addition of an underground metro to provide transport from one district to another--then we can considerably increase the amount of green space behind each building. We arrive then at the reference district for carfree cities, as shown here. It differs in one other respect from traditional medieval city patterns: a boulevard is included, which coincides with the route for the metro system.

City Topology
This leads directly to the matter of city topology. Topology is the branch of mathematics that deals with spatial relationships, so city topology is thus the study of the spatial arrangement of various urban elements and their relationships to one another. Deformation does not alter the topology, even though the form changes.

Same topology, different physical shape, accommodating design to geography.Virtually every aspect of city life is affected by the topology selected, in particular the transport arrangements: changing topology alters the route structure of the transport system, usually with large effects on its effectiveness and efficiency. The question thus becomes: how do we arrange our districts so that the transport system functions best?

The topology of the reference design flows from a variety of needs, including the creation of a clearly defined downtown area and the presence of rural areas adjacent to each district. The most important single concern, however, was the optimum arrangement of the transport system. The reference topology permits the use of a modified hub-and-spoke route system, which provides the fastest service for most trips. (This is similar to the routings used by most major air carriers nowadays.) No trip requires more than one transfer, even though there are three transfer stations downtown.

A few trips would be better served by a system with tangential routes in addition to the routes radiating from downtown, but this would greatly increase the length of the transit system that must be constructed, maintained, and operated. Although only a single transfer is all that's necessary, three transfer stations reduce congestion.The tangential routes only slightly increase the number of passengers who enjoy direct service. Even with tangential routes, most trips are actually faster through the hub, and the concentration of ridership on the radial lines results in higher utilization, better efficiency, and more frequent service, with its attendant shorter waiting times. (In larger cities with greater distances to downtown and more riders, the tangential lines start to make sense and also increase the capacity of the system.)

I believe that the reference topology is ideal for cities between about 500,000 and 2,000,000 inhabitants. Outside this range, a simpler or more complex topology will serve better.

Topology for a large city.Topology for a very large city. High-speed rail connects the centers of the six sub-cities, allowing rapid movement between any two parts of the city. This topology works for populations up to about 12,000,000.
Rationale for the Reference Topology
My work draws heavily on that of Christopher Alexander and especially his magnificent work, A Pattern Language. While the application of Alexander's patterns is not discussed explicitly in this article, most of the design is largely or entirely in accord with most of the patterns he has defined. This matter is taken up explicitly in Part III of my book, Carfree Cities.

While developing the reference design, I might have chosen almost any set of design parameters, but I thought it most useful to choose attributes for a medium-sized city, as follows:

These attributes are achieved in the reference design. The attributes are discussed in the following sections.

Six-Lobe Design
The city is divided into 99 districts, which are in turn arranged into six "lobes," each of which is similar in form and function. The six lobes plus the center form the complete city. This arrangement conveys several advantages:

  • There are only three metro routes, each beginning in one lobe, running into the center, and back out to an adjacent lobe.
  • The metro routes never cross one another, but each can be brought adjacent to the others for convenient transfers. More than three routes cannot be arranged in such a pattern.
  • The closed-loop metro arrangements provide redundant service in the event of an outage in one direction; some journeys can only be made by way of a longer route, but all destinations remain accessible.
  • Each district is surrounded by at least a narrow greenbelt.
  • The transport system is efficient, densely utilized, economical to construct, and holds travel time to an almost irreducible minimum.
  • All areas of the city are within easy reach of the center.

100 Districts
About 100 districts would be constructed, each 2500 feet in diameter, with streets radiating from the central transport halt. Of the 100 districts, the 18 farthest from the city center are "utility areas." These non-residential areas are reserved for various infrastructure requirements, heavy industry, and parking. (These are shown in gray on the illustrations.) The remaining 81 districts are mixed-use developments. While faster public transport could be provided with fewer, larger districts (requiring fewer stops and permitting higher top speeds), walking distances become excessive and door-to-door times for the longest journeys actually increase. The optimum number of inhabited districts for a city of 1,000,000 is about 80.

The 18 nonresidential utility areas are reserved for various infra-structure requirements, heavy industry, and parking for residents and visitors alike. Utility areas are located at the extremities of the city because they require direct access to the external rail and road networks. Siting the utility areas at the periphery keeps the heavy transport infrastructure away from inhabited areas and mitigates noise problems. It also helps conserve valuable land near the city center. The actual number of utility areas constructed will depend on the needs of a particular city and might be considerably more or less than 18.

20% Urbanized
Maximum density, minimum green space.In order to provide plenty of open space, only 20% of the area of the full site would be developed. The remaining 80% of the site is rural, excepting only a small amount of land occupied by the connections to the external rail and road systems. The rural areas can and should be used for a wide variety of purposes, including parks, garden allotments, lakes and streams, playing fields, golf courses, farms, forests, nature reserves, and wilderness.

35 Minutes to Any Doorstep
Assuming that an underground metro is selected to provide passenger transport, the longest transit time between any two points in the reference design averages only 35 minutes. This occurs when the origin of the trip is at the outside edge of the outermost utility area of one lobe, and the destination is at the edge of the outermost utility area located on a different metro line. Many trips are within the district (since basic shopping is located at the center of every district), and the walking time to the district center does not exceed 5 minutes from any doorstep.

In many cases, bicycling will be the fastest means of transport, especially if the origin or destination is near the edge of the district or if the journey is to an adjacent lobe. It would be possible to bicycle anywhere in the city in under an hour.

1,000,000 Residents
A million residents is a reasonable balance between small and large cities. A city of a million is big enough to require the solution of difficult transport issues. A city of this size can support major cultural institutions such as orchestras, museums, and sports teams. It is not so big that it creates extreme concentrations of pollution, nor does the supply of water and other necessities become a terribly difficult issue. The reference topology is useful for cities ranging in population between about 300,000 and 3,000,000. While the size and number of districts will change, the basic 6-lobe topology can be adapted to cities within this range.

100 Square Mile Land Area
The reference design occupies a roughly square parcel of land about 10 miles on a side. A parcel of this size allows the city to be separated from adjoining uses by a green-belt at least a mile wide (more is better!).

Metro Freight
Because the burden of truck traffic on city residents is even greater, on a per-vehicle basis, than car traffic, the reference design is also intended to be entirely truck-free. A heavy-duty freight distribution system is included in the design. It is based on the use of standardized shipping containers, which already move by the millions in global commerce. Adapting standard metro technology to handle shipping containers appears to be a relatively simple task. This would permit the delivery of shipping containers directly onto loading docks of business located along the dedicated metro-freight line (which parallels the regular metro, separated from it by commercial buildings). This accounts for the bulk of freight deliveries; the remainder would be distributed from freight depots located in each district, using a combination of freight bicycles and small, slow, battery-powered freighters. The streets of the city would thus be free from all forms of conventional motorized transport excepting only emergency vehicles (and possibly ready-mix concrete trucks, for which a useful substitute is difficult to arrange).

Alternative Topologies
For a variety of reasons, the reference topology is not suitable to every case, and even within the reference design, a wide range of changes may be made.

Adjustments for Size
The number of districts could be easily adjusted to suit the desired size of the city. Fewer than about 50 districts indicates either that the topology could be simplified.

Here is a simplified topology for a city of 400,000; it can readily be expanded into the reference topology:

Simplified topology for a city of 400,000.

Above a population of about 2,000,000, it becomes necessary to adopt a more complex topology. "Necklaces" can be added to serve a population of several million.

Topology for a population of over 2,000,000.The outer necklaces greatly increase the urban area and population without fundamentally altering the basic topology. At some point, the capacity of the metro system starts to become inadequate. The downtown sections of the metro are all four-track, so a population of several million could be supported by this arrangement.

A grid topology for cities of over 5,000,000.With still larger cities, it is possible to adopt a grid system, with metros running in one direction and either metros or trams running in the other direction. The capacity of the transport system is greatly increased, as is necessary. By deforming the grid, it is possible to increase the amount of green space in the outer areas.

Graduated Density
One obvious variation is to graduate the size and density of the districts, with those districts farthest from the city center being the largest and least densely populated. As one moves away from downtown, the size of the districts increases and the density declines, permitting a form of development characterized by the streetcar suburbs of the 1920s. Travel time from the edge of the outer districts increases by about 12 minutes. The density remains considerably above that for contemporary US suburban development, because the building lots are fairly small and relatively little land is devoted to streets.

Construction density declines as the distance from the city center increases. It is also possible to achieve a reduction in density by adding green areas between the districts. Because the extra distance is covered at top speed, the increase in metro riding times is actually fairly small.

Geographic Constraints
With few exceptions, cities are greatly affected by the nature of the site upon which they are built: rivers and hills have shaped almost every city. A few cities in the Netherlands developed almost entirely free from topographic constraints, but in these cases, soil conditions have exercised a considerable influence. The reference design will almost certainly have to be adjusted to local geography, and in many cases substantial changes will be necessary. In particular, the closed-loop topology may have to be abandoned if a particular site does not lend itself to this arrangement. The lobes would then be open at the ends, which only slightly degrades the design and somewhat complicates the operation of the transport system.

Carfree cities offer us a way to have our cake and eat it too. We can retain the benefits of easy mobility without making our cities into hideous places. We can do this while dramatically reducing our energy consumption. We can restore the public realm to its rightful place as the public living room. We can even keep our cars for travel to distant cities or into the countryside, assuming the fuel is available. All of this can be achieved with existing, proven technology at a cost that is surely considerably less than the cost of sprawl development, with its terribly expensive infrastructure. In return, we get quiet cities with clean air, beautiful streets, and safe places for our kids to play. I think it's a pretty good deal.

©J. H. Crawford 2001

Further Study Carfree Cities was published by International Books in 2000 and includes a foreword by James Howard Kunstler. For more information about the book, or about carfree cities in general, visit:

Suggested Reading
Alexander, Christopher et al. A Pattern Language: Towns, Buildings, Construction (New York: Oxford University Press, 1977). A ground-breaking work magnificently prepared by Oxford. In my opinion the most important book on architecture and urban design in the 20th century.

Appleyard, Donald et al. Livable Streets (Berkeley: University of California Press, 1981). This is the definitive work describing the effects of traffic on community life. Essential.

Goddard, Stephen B. Getting There: The Epic Struggle between Road and Rail in the American Century (New York: Basic Books, 1994). An excellent scholarly work, clearly written, de-scribing the death of US rail transit at the hands of the Road Gang. Goddard is a lawyer, so there is some emphasis on legal aspects. Includes an interesting discussion of how the road interests, led by GM, broke the law in their efforts to destroy rail passenger transport and how they in essence got away with it.

Hale, Jonathan. The Old Way of Seeing: How Architecture Lost Its Magic (And How to Get It Back) (Boston: Houghton Mifflin Company, 1994). Brilliant if tentative look at how modern civilization lost a gift that was once nearly universal: the ability to design beautiful things. Desperately needs further followup.

Jacobs, Allan B. Great Streets (Cambridge: MIT Press, 1993). A simply marvelous compendium of great streets from around the world. Beautifully illustrated, with dimensions.

Kay, Jane Holtz. Asphalt Nation: How the Automobile Took Over America and How We Can Take It Back (New York: Crown Publishers, 1997). Takes a comprehensive look at all the problems caused by automobility in the USA and considers ways to reduce car dependency.

Krier, Léon. Architecture: Choice or Fate (Windsor, UK: Andreas Papadakis Publisher, 1998). Wonderful examination of the impact of good urban design and architecture on the public realm, by a great contemporary architect. Krier designed the new town of Poundbury for Prince Charles, which project is also treated.

Kunstler, James Howard. Home From Nowhere: Remaking Our Everyday World for the Twenty-First Century (New York: Simon & Schuster, 1996). Continues the themes of The Geography of Nowhere, this time with an emphasis on remedies, particularly the restoration of traditional civic design. Includes entertaining cultural criticism and an assault on the US automobile mania.

Newman, Peter and Jeffrey Kenworthy. Sustainability and Cities: Overcoming Automobile Dependence (Washington: Island Press, 1999). Dense source book, full of tables and graphs. Essential for serious students.

Sitte, Camillo. City Planning According to Artistic Principles (New York: Random House, 1965; first edition [German]: Vi-enna, 1889. Translated by George R. Collins and Christiane Crasemann Collins). Hard to find and expensive, but a gold mine. Full of drawings and fascinating explanations of why some urban areas are attractive and others are not. Most of the places that Sitte criticized little more than a century ago are today regarded as architectural treasures. It's not that these old areas have improved, it's that the newer areas are so much worse. Essential.

Van der Ryn, Sim and Peter Calthorpe. Sustainable Communities: A New Design Synthesis for Cities, Suburbs, and Towns (San Francisco: Sierra Club Books, 1986). An early work in the sustainable development effort.

Zuckermann, Wolfgang. End of the Road (Post Mills, Vermont: Chelsea Green Publishing Company, 1991). Thorough review of techniques for taming cars; considers unintended side effects.