Recent research in the area of urban and spatial patterns provides many candidates for exemplars. Regarding macrospatial distributions, this chapter now reviews a compilation of points developed in several recent analyses by Frey (1995), Glaeser and Shapiro (2003), Glaeser and Kohlhase (2003), and Kasarda (1995), all of whom draw on traditional demographic perspectives leavened with selected insights from new perspectives to explain and interpret trends in regional and intermetropolitan population shifts in recent decades. Kasarda (1995) and Frey (1995) draw on urban ecological theory; Glaeser and colleagues draw on the new economic geography (Anas, Arnott, and Small 1998; Fujita and Thisse 2002). These theoretical perspectives are highly compatible, so it is not surprising that they advance similar findings and interpretations. The impact of long-term changes in communication and transportation technologies receives the greatest attention, with secondary emphasis going to factors such as climate and amenities, government (although not always for the reasons suggested by critical theorists), and the implications of globalization.
While all four stress the impact of technological change, Glaeser and Kohlhase (2003) are especially forceful. "A revolution in transportation technologies unfolded in the twentieth century; it would have been surprising had a revolution in urban form and location not occurred as well'' (2003: 31). They argue that at the beginning of the 20th century transportation costs were high and created powerful economic incentives for cities to be located near natural resources and natural transport hubs. Transportation costs declined dramatically over the century and became less hub-centered. At the same time, cities shifted from being centers of goods production to centers of service production. This made it increasingly possible for population to redistribute, spatially guided by business and consumer preferences regarding climate, amenities, and the business environment. Cities still continued to be centers of economic activity because they brought people together in close proximity, creating significant agglomeration econ omies especially for educated and higher-skilled workers. But urban economies became more ''footloose'' in space, and cities were less tied to resources and break-in-transpor-tation locations. Consequently, population followed business and consumer preferences and shifted from the interior of the country and toward coastal regions, toward warmer and dryer regions, and ''favorable'' business environments. In addition, the population increasingly concentrated in cities which, due to the consequences of reduced transportation costs within cities, shifted from being organized around an urban core to being organized around sprawling urban regions. As dramatic as these trends are, Glaeser and colleagues speculate that they would have been even stronger if not for the massive inertia in the urban system associated with sunk costs in infrastructure and housing.
Glaeser and Shapiro (2003) and Glaeser and Kohlhase (2003) document the long-term declines in specific transportation costs and in the declining fraction that transportation costs represent in the total economy, both key factors in their overall arguments. They also find support for the hypotheses they advance by drawing on a variety of data sets to perform relevant empirical analyses. Analysis of urban growth patterns in recent decades shows, as they predict, that certain kinds of cities gained relative to others and certain kinds lost: cities with resource-based economies, such as extraction and agriculture, lost; cities with warm, dry climates gained; cities organized around auto and truck transport gained; cities with higher skill mixes gained; and cities specializing in goods production lost. Their arguments also generate predictions about urban form that figure into their overall analysis. Declining transport costs for goods freed manufacturing from rail and water transport hubs. With the reduction in economic incentives to locate near resources, customers, and suppliers, manufacturing moved from the center of the city to the fringe, and from larger to smaller metropolitan areas. Service industries requiring significant amounts of nonroutine face-to-face interaction continued to promote high densities in central cities, since time-costs of bringing people together over very long distances tend still to be high. However, service industries did not sustain the urban core in the same way as hub-and-spoke rail and water transport and mass transit systems. Thus, density was falling generally but was varying positively with services and level of education and negatively with manufacturing.
Finally, they present evidence on long-term population shifts consistent with the prediction that population is shifting toward large, sprawling urban regions. Specifically, they show that the fraction of national population found in counties in the bottom decile of density and in counties in the top 1% of density both declined substantially between 1920 and 2000, and the share of population found in the top decile (excluding the top 1%) increased substantially.11
Essentially these regions are large population and employment ''domes,'' which may or may not have single, multiple, or no dominant centers. Even when dominant centers are present, the majority of employment is spread diffusely, since the primary transport consideration is moving people and, in cities with car-oriented transport systems, movement within the suburban ring (now the most common kind of commute) has lower time-costs than movement from periphery to center due to congestion effects. The emergence of these large, sprawling urban regions is leading to important shifts in the national urban hierarchy. This is consistent with central place theory because the
11 Between 1920 and 2000 the population share for the bottom decile of counties by density fell from 19% to 9% and the share for the top 1% fell from 20% to 11%. The share for the top decile (excluding the top 1%) increased from 30% to 49%.
decline of cities at the lower tiers is expected under conditions of declining transportation costs for goods and information.12 Interestingly, Ogburn (1946) more than half a century ago forecast that this would occur based on his analysis of how maturation of auto and air transit would lead to consolidation of city-based trading areas into progressively smaller numbers of large regions.
Kasarda (1995) or Frey (1995) would take issue with little in this capsule summary. Both endorse the notion that cities that made the transition from producing goods to providing services fared better in intermetropolitan restructuring over recent decades. They also note the resumption of the long-term decline of nonmetro areas despite the ''blip'' of the 1970s; population movement toward the West and the South; population movement from the interior to coastal areas (also see Rappaport and Sachs 2003), and the vitality of population growth in metro areas with high education and skill mixes (especially after controlling for immigration). Kasarda gives greater attention to globalization and how it is implicated in some of the macrospatial changes observed in recent decades. One interpretation of the positive education-growth relationship and the negative manufacturing-growth relationship is that in a global economy, cities specializing in most forms of manufacturing are vulnerable to global competition due to the dynamics identified in the product-cycle theory of Vernon (1966) and the neo-Marxian profit cycle theory of Markusen (1985). Both predict that high-wage regions must continually innovate to maintain their relative standing in a spatially integrated economy, especially one in which transportation costs are declining, political barriers to trade are falling, and centers of high-skill labor are emerging in the developing world. Thus, Kasarda reports that cities specializing in new ''leading'' industries fare better than those specializing in ''lagging'' industries, which are vulnerable to ''filtering.''
While all these studies emphasize the growing importance of consumer preferences in location decisions, Kasarda (1995) gives more attention to business ''climate'' and quasipolitical economic concerns. His point is that in an economy where declining transport costs increasingly free businesses from specific locations, areas within the United States must increasingly compete with each other and with extranational locations on considerations such as tax incentives, regulatory burdens, and related factors to capture and retain business and industry. Labor costs and unionization factor into both this and product cycle theory. Closely related is the question of local government policies regarding growth and the provision of government-based services such as public education, crime, and low pollution. These considerations do not always line up nicely, but Kasarda notes that the match between local government policies and the class ideology of affluent business leaders and elite workers is likely to play an increasing role in location decisions. Ironically, this political-economic consideration is elevated in importance because conventional ecological dynamics give affluent decision makers greater discretion in location decisions. Finally, Kasarda notes that some have hypothesized that federal fiscal policies have influenced interregional restructuring, but the impact of this relative to other factors is not well established.13 In sum, the four studies highlighted here (plus the larger literature they draw on) document important long-term trends in interurban and intraurban population distribution and link them to changes in
12 The spacing between nodes in central place theory is an inverse function of the cost of communication and transportation. Along these lines Cottrell's (1951) study ''Death by Dieselization'' documented how a small layer of the central place system in rural areas disappeared as advances in rail transport technologies obviated the need for closely spaced stops.
13 For a discussion, see Markusen (1987).
factors that have long been central to traditional demographic perspectives on urban and spatial distribution of population.
Was this article helpful?