Chapter 4: Trends in Sizes and Structures of Urban Areas

The Sizes of Urban Areas:

1. Although over 75 percent of the population live in urban areas, there is a tremendous diversity among the urbanized areas in terms of their population size.

2. The size of an urban area has an important effect on attitudes and life-styles of its inhabitants. But, there is less variation today than historically, partly because of the effects of the media and communication technologies.

3. There is perhaps more diversity in life-styles within urban areas (central city versus suburbs) than among suburban dwellers in different urban areas.

4. The two most important measures of the size of an urban area are its total population and its total land area. The two are combined to determine population density (people per square mile).

5. Population size is used more than other measures because of its availability. Comparisons of relative population size show considerable variance among urban areas. New York City CMSA is 300 times larger than Enid, Okla. MSA.

6. Table 4.1 shows population and ranking of the top 15 U.S. cities in 1990. Population decline occurred among many of the largest cities from the Northeast and Midwest, especially during the decade of the 70s. Chicago, Detroit, Philadelphia, Baltimore, Washington D.C., Cleveland, and Pittsburgh continued this decline during the 80s.

7. Table 4.2 shows the top ten rankings for 1950, 1970, 1980, & 1990. Ranks are relatively persistent among the largest cities, but Los Angeles CMSA passed New York City in 1990.

8. San Diego, Phoenix, and Miami are relative newcomers who have moved up considerably over the past three decades, but have not yet cracked the top ten. Dallas ranked number one in number of new jobs generated in 1994.

9. Population changes from central cities are due, in part, to massive suburbanization as well as interurban migration. The failure of annexation to keep up with population decentralization in some areas but not in others makes comparisons of relative population change difficult.



Estimates of MSA Size Distributions:

1. A systematic way of examining size distribution of MSAs is in terms of a frequency distribution based upon an array of urban sizes based upon their populations, beginning with the largest.

2. This frequency distribution, as predicted by central place theory, will be skewed to the right. A cumulative frequency distribution would follow a Pareto Distribution

G(x) = A x(^-a) where a = 1. Hence, x G(x) = A.

3. x G(x) = A is the rank size rule: The product of an urban area's rank and population is a constant equal to the size of the largest urban area. Table 4.3 shows the rank x population for a sample of U.S. metropolitan areas.

4. Note that a statistical estimate of the Pareto distribution results in the following:

log G(x) = 6.833 - 0.905 log x r-squared = .99 (both coefficients sign diff (0.025) (0.005) from zero)

5. Rank size rule is relatively accurate, but the estimate of a = .905. Hence, populations of lower ranked urbanized areas are smaller than indicated by the rank size rule.

6. In other countries, the estimate of a is even lower, so that there is more disparity in population size among cities in other countries than in the US.

7. The gradual relaxation of technological and economic constraints to the larger size of urban areas has been taken advantage of by a relatively small number of cities. But, with cheaper long-distance transport costs, it is even more feasible to concentrate higher ordered economic activities in a small number of widely dispersed urban areas.

Suburbanization

1. Suburbanization is measured by the growth in the urban ring around a central city. However, if central city boundaries are moved outward, it becomes difficult to measure the degree of suburbanization based upon jurisdictions.

2. In recent years, annexation of central cities has slowed relative to outlying incorporated areas, so that suburbanization is more easily measured.

3. Nevertheless, the following difficulties still exist:

a. Jurisdiction measures do not permit careful cross-sectional comparisons because of historical central city boundaries that have allowed for more expansion in some area than in others.

b. Even in time series some central city boundaries have been moved outward through annexation, especially in the South and Southwest.

4. Table 4.4 shows data on the central city versus suburban ring of population and employment in selected MSAs. Note the decentralization of both population and jobs.

5. Manufacturing and retailing are more suburbanized than and wholesale trade and services. Also, overall employment is less suburbanized than population, giving rise to journey-to-work trips from the suburbs to the central city.

6. However, the pace of suburbanization of jobs is slightly faster than the suburbanization of people. But, while and manufacturing and trade are fleeing the central cities, "information processing" jobs are racing to central cities. See Table 4.5.

A Model for Suburbanization

1. The income elasticity of the demand for "space" versus the income elasticity of the demand for transportation cost savings?

2. In the past the demand for "space" has outstripped the demand for "transportation cost savings.

3. What determines transport costs? Actual and opportunity cost of time spent in commuting. Will the information age change the relative importance of suburb versus central city living?

4. What about technological changes that influence the cost of commuting?