Sprawl: a Compact History, Part VII -- Density/Los Angeles

Bruegmann has a lot of interesting things to say about density, by which he basically means population density. One of his basic points is that suburban density is increasing in many places by many measures, even as the anti-sprawl chorus gets louder.

(Now, I don't think that the increasing density and the increasing concern about sprawl are as contradictory as Bruegmann seems to think, because I don't think that low density development as such is what most people mean when they complain about sprawl.)

Perhaps the most provocative density-related statement Bruegmann makes is the following, concerning Los Angeles:

Los Angeles, for example, often taken to be the epitome of sprawl, has become so much denser over the past fifty years that it is now America's most densely populated urbanized area, as measured by the census bureau. It is considerably denser than the New York or Chicago urbanized areas, for example. Although this might seem preposterous since Los Angeles has no neighborhoods with densities anything like parts of Manhattan, Los Angeles has a relatively high density spread over an extremely large area. Los Angeles also has none of the very low-density exurban peripheral growth see in the New York region. In fact, quite unlike Eastern cities, Los Angeles has almost no exurban sprawl at all because the high cost of supplying water makes relatively compact development almost inevitable.

This idea is so counterintuitive that it was featured prominently in the first couple of mentions of Bruegmann's book that I saw on the net. I must say, it doesn't match my own preconceptions about Los Angeles, but, then again, I've never been there (like most other people, I guess).

Key to this statement about Los Angeles is the idea of the "urbanized area". I'll give an extended quote from Bruegmann about what this means. This is from Chapter 5, "Sprawl Since the 1970s".

The "urbanized area," by contrast, was devised by the [U.S.] census bureau to provide a functional definition of the [U.S.] city. It includes all the land with a strong connection back to the central population centers and more than 1,000 people per square mile. Although by no means a perfect measure for defining city and suburbs, the 1,000 people per suqare mile figure does approximately correspond to the threshold between the regularly developed suburban subdivisions and the exurban areas beyond because 1,000 people per square mile is today about the lower limit at which full city services like water supply and wastewater treatment can be provided in a way that most public and private agencies consider economical.

Bruegmann has collected and/or gone through quite a bit of data having to do with density of various urbanized areas worldwide, and has produced some interesting graphs, which are in the book.

One of the types of graphs he discusses he calls a "density gradient graph". This is a graph with miles from the city center on the x axis and residential population density on the y axis.

The mental picture that I (and probably most other people) carry around with me about the nature of cities includes a very densely populated city center with density gradually dropping as you get further and further from the center. This situation is represented on one of these density gradient graphs as a jagged diagonal line generally trending from the top left (city center/high density) to the bottom right (distant place/low density).

A variation of the density gradient graph shows the graph for a single city at various times, and Bruegmann gives an example of this for London in chapter 1, "Defining Sprawl". What this graph shows in a fairly intuitive manner is that the density gradient for London has been "flattening" over time -- the diagonal line on the gradient graph gradually gets less diagonal and more horizontal. In real world terms, this means that, as time goes on, the center gets less dense, outlying areas get more dense, and the total urbanized area increases in area covered. This fits in with my internal model of what has happened with cities over time, but it's interesting to see it graphically represented so clearly.

You could imagine a completely flat density gradient, which would basically mean that the entire urbanized region has rougly the same population density with no noticeable center of population at all. Bruegmann gives Phoenix as the American city that most closely approaches this flat gradient.

Another kind of graph that Bruegmann gives many examples of (in Chapter 5, "Sprawl Since the 1970s") is a graph of density of an urbanized area over time. (This is using the U.S. Census Bureau definition of "urbanized area", which I give above). As examples, the graph for Los Angeles shows the density continually increasing since 1950, which of course matches what Bruegmann has to say about Los Angeles in the above quote and elsewhere throughout the book. The graph for Cleveland (and many other rust belt cities) shows a big drop in density between 1950 and 1960, and a gradual decrease since then.

The interesting thing about the set of graphs as a set is that they're all noticeably different from one another (and he has graphs for 28 different U.S. cities). Some show general declines in density (Cleveland), some show general increases (Los Angeles) and some show substantial decreases up until some point (often the hinge is 1980) followed by gradual increases (Denver, Sacremento). There are some other patterns too.

As an Ohioan, I would love to the see the graph for Columbus, too. In 1950, Cleveland was a lot bigger and more important than Columbus, but the years since have mostly involved decline for Cleveland and growth for Columbus. It would be interesting to see this on the graph.

I like what Bruegmann has to say quite a lot when he's in his "let's see where the data takes us" historian mode. He has obviously analyzed a lot of data. I think the density gradient graphs and density over time graphs are great ways of discerning and understanding broad patterns in what's actually happening in these vast complicated urban areas.

Having said that, it sometimes seems that whole reason he's focused on density is so that he can accuse people who are complaining about sprawl in areas where the density is actually increasing of being incorrect about there being any sprawl at all. This only makes sense if you consider "sprawl" to be the same as "low density development". Bruegmann draws exactly this equivalence in several places, but I don't really agree with it myself, so the whole "See, the density is increasing" argument is kind of lost on me. 

¶ 1/28/2006 08:52:00 AM