Cities grow and evolve over time, forming patterns guided by an array of physical, economic, political, and technological influences. While no two cities are exactly alike, and urban form – the arrangement or layout of urban land uses; offices, houses, shops, roads, etc. – can vary significantly from one region to another, several urbanists have tried to identify and explain patterns of urban structure and urban growth at various points in time. While cities have evolved, so to has our understanding of their form and function. Here we take a brief look at a just a few of the models – the idealized and simplified representations of reality – that have been developed over the last 100 years to describe and explain the structures of humanity’s greatest artifiact: our cities.
Concentric Ring Model
Models of urban form developed in the early 20th century emphasized a monocentric (single-centered) form. In the land use model developed by Ernest Burgess in 1925, the city was thought of as a series of five concentric zones radiating out from the center of the city, or the central business district (CBD). The first and most central zone, which included the CBD, was considered the heart of the region both in terms of commerce and culture. This first zone was also where all major transportation lines converged. The second zone contained a mix of residential, commercial and industrial areas. According to the Burgess model, this zone was typically occupied by low-income residents and was likely to contain slums and other blighted areas. The three outer rings of the city were exclusively residential, with high-income residents most likely to live in the two outermost rings. In the Burgess model, all urban development occurs from the central city outward.
Expanding upon the Burgess model, Homer Hoyt in 1939 developed a sector model of urban form that considered direction as well as distance from the CBD. Hoyt observed that rent levels throughout 25 U.S. cities did not usually vary by concentric ring, but rather by sections radiating out from the central city. The general pattern suggested that rent levels (and, presumably, land values) within each section of the city, like slices of a pie, were more similar than those found among concentric circles.
Hoyt reasoned that urban growth must occur from the center outwards along particular corridors or wedges. High-income residential areas, for example, move successively outward from their initial location within the central city in a direction that maximizes their access to major transportation routes and recreational or scenic amenities such as water fronts, open countryside and commercial centers. Directional growth resulted in the sectoral partitioning of the city not only in terms of rents and land values but also land uses. Areas with high property values effectively excluded undesirable uses, such as heavy industry, from developing within that sector.
Multiple Nuclei Model
By the 1940′s, the automobile had begun to transform American cities, facilitating the broad dispersal of people into suburban enclaves ever further from the city center. Cities that once had a very clear monocentric urban form began to evolve into a more complex patchwork of suburban centers strung along major road and highways. Observing this transformation, Chauncy Harris and Edward Ullman in 1945 developed their multiple nuclei model, in which they envisioned a polycentric (multi-centered), or polynucleated, urban form.
According to the multi-nuclei model, the city is composed of a number of sections, each with its own functional specialization. Different land uses and economic activities tend to form around the city center in no universal concentric order or direction, but rather assemble according to the complex interaction of four primary variables.
First, many land uses require access to specialized facilities or locational amenities. Large retail shopping centers, for example, require access to major transportation routes. Second, some activities, such as those associated with the financial, insurance, and real estate industries (FIRE), benefit from agglomeration, and tend to cluster together in central locations. Third, the incompatibility of certain land uses, such as heavy industry and high-income residential housing, assures a certain degree of regional differentiation.
Lastly, high land values restrict the development of all but a few land uses in certain locations (e.g. office buildings in CBDs or large retail outlets at major highway intersections). These nodes of similar economic activity form either as the expanding city envelops surrounding settlements, or when new nodes develop in accordance with the variables discussed above. Thus, according to Harris and Ullman’s model, the city does not simply grow outward from the CBD, but rather evolves from the complex integration and development of separate functional nuclei.
Urban Realm Model
Expanding on the multi nuclei hypothesis, James Vance proposed the urban realms model in 1964. According to the model, the city is composed of autonomous nuclei (or urban realms) largely independent of the traditional CBD or central city. Vance argued that polynucleated belts of urban development formed over the previous century primarily due to the extrodinary growth in population and areal extent of cities, as well as the introduction and mass adoption of the automobile.
The size, character, and structure of urban realms depend on four primary factors. First is the topography of the landscape. Mountains, water bodies, and other natural features can both direct the spread of urbanization and influence the type of development within an urban realm. The second factor is the size of the metropolitan region, with larger urban areas tending to have larger, more numerous, and more differentiated urban realms. Third is the level and character of economic activity within each realm, and fourth is the layout of infrastructure and overall accessibility within and between realms.
Particularly important in the model is the presence of major circumferential highways and other transportation corridors that link the various urban realms. Airport connections allow urban realms to connect with other cities, providing additional economic independence from the central city. As urban realms become more powerful, the core-periphery relationship begins to weaken, resulting in what Vance described as a “sympolis” rather than a metropolis. Many large cities have distinct urban realms, including the greater Los Angeles area, whose western urban realm – the “Inland Empire” of Riverside and San Bernardino – is larger than all but a few U.S. cities with a population of over 4 million.
White’s Revised Concentric Model
Michael White (1987) provided a revised Burgess model to reflect new social, economic, and political forces influencing urban growth. White envisioned late-20th century urban form as a complex patchwork of concentric zones, corridors, epicenters (i.e. nuclei or nodes) and enclaves. The CBD, though having maintained its position as the economic and cultural heart of most large cities, had become more specialized in finance and management as retail and other activities migrated to the suburbs. Surrounding the CBD is the zone of stagnation, which White argued suffers not only from lack of investment, but also slum clearance, highway construction and relocation of industry to more peripheral locations. At one time it was expected that the CBD would expand and revitalize the zone of stagnation, but in most cities the CBD has expanded up rather than out.
Most of the remainder of the city from the zone of stagnation outward is composed of a patchwork of wealthy enclaves and (mostly poor) immigrant pockets held together by a spatially diffuse realm dominated by the middle class. Dotting the urban landscape are clusters of specialized activity, such as industrial parks, universities, and hospitals that can exert significant influence on local land use patterns. Finally, as standard among polycentric models, there are epicenters and corridors of economic activity along major transportation routes, especially where radial and circumferential highways intersect. These nuclei and corridors “form a latticework that extends over the entire urban region,” that increasingly “challenge the hegemony of downtown” (White 1987).
The Megapolitan Model
A new model of urban structure developed by Robert Lang and Paul Knox (2009) moves up in scale and beyond the single metropolis. They argue that the expansion of metropolitan areas, and the transportation linkages that bind them, have given rise to larger trans-metropolitan urban agglomerations, termed megapolitan regions. To be classified as a megapolitan region, neighboring metropolitan areas must share at least 15 percent of new commuters from 1995 onward. Using this criteria, Lang and Knox (2009) have identified nine megapolitan areas in the U.S. The concept of the megapolitan region borrows from Pickard’s (1970) urban regions, described as areas with “high concentrations of urban activities and [an] urbanized population,” and Lewis’ (1983) ‘galactic metropolis,’ which contains “varying sized urban centers, subcenters, and satellites [that are] fragmented and multimodal, with mixed densities” (Lang and Knox 2009).