With Madelyn Glickfeld of the UCLA Extension Public Policy Program, we have been examining infrastructure investment as a function of spatial location in California metropolitan areas. One of the underlying issues is the dynamics of urban growth processes. Some analysts feel that suburbanization leads to excessive sprawl and high expenditure levels by local governments, not to mention permanent damage to the environment.

On the other hand, our analysis of fiscal expenditures in 465 cities and counties between 1972 and 1992 don't support this position. Instead, we find that infrastructure investments are generally less the further out on the periphery. However, there are trade-offs. Roadway and sewer infrastructure expenditures are generally cheaper, whereas water distribution and quality infrastructure expenditures are generally more expensive. On the other hand, the most expensive local infrastructure investments are in the old suburbs, those cities which were 'bedroom' communities 40 or 50 years ago, but which have since become mixed use, higher density with sizeable employment.

What has to be realized is that this process occurs over time and that a 'snapshot' view at a single point in time may lead to wrong conclusions. For example, even though the Los Angeles metropolitan area was developed as a series of dispersed communities, connected primarily through an extensive surface rail system (in the 19th century) and a modern freeway system (in the 20th century), the metropolitan area has grown together to form a compact region holding about 16 million persons.

Above is a graph showing 1990 population density (by census tracts) as a function of distance from downtown Los Angeles (city hall). The graph has been smoothed with a moving average to see the general trend. As seen, the density drops off almost logarithmically. However, out to about 35 miles away, there are more than 5000 persons per square mile on average. Even out to about 60 miles away, the density is greater than 2000 persons per square mile.

Now contrast this with the second graph below showing the 1990 density-distance function (by block groups) for the Washington, DC metropolitan area, a supposedly compact city. At the core, the density is almost the same as in the Los Angeles region. However, the density drops off very rapidly. By 10 miles out, it is below 5000 persons per square mile. At around 30 miles out, the densities become extremely low, characteristic of sprawling communities. The basic reason is that huge tracts of open space have been earmarked for preservation, presumably under the guise of discouraging sprawl and encouraging compact development in Maryland and Virginia. However, these policies failed to consider population growth so that new developments have simply leapfrogged beyond these areas, just the opposite of what was intended by the plans. Population growth can debilitate even the most well-thought-out plans!

Publications

• Ned Levine, "Credit distributed, new points raised". Journal of the American Planning Association. 1997, 63(2): 279-82.