Is LA really denser than NYC?

by Sandy Ikeda
Thu, 7 Feb 2008 at 7:53 PM

Send

RSS

After my post on density and the ensuing discussion, I e-mailed Wendell Cox, who runs the website Demographia, asking how he reckons that Los Angeles has a higher population density than New York City, something he says has been true since 1980. He responded immediately and generously gave of his time over the phone.

To begin with, his concept of a city, as I understand it, is based on the physical contiguity of what urbanists call "the built environment." A city is thus an agglomeration of houses and other buildings that stand close together.

Next, Mr. Cox finds that the area of highest density in New York has almost three times the density as the area of highest density in LA. But as you go from higher- to lower-density areas in each city, there are more people per acre in LA than in NYC. That is, LA's density profile is like a "doughnut," typical of urban development since WWII, while NYC's looks like a "fried egg." You can view these data here. Mr. Cox also provides some striking aerial photos of this pattern. The images on page 8 of this pdf document dramatically illustrate much higher density of the suburbs 15 and 35 miles from LA and comparable suburbs near NYC.

Finally, defining metro LA in terms of contiguous development includes more of the surrounding dense suburbs (the thick outer doughnut) than does the Los Angeles "metropolitan statistical area" (MSA), a Census Bureau measure used to calculate densities, which raises metro LA's density. On the other hand, including more of the contiguously built area around NYC captures relatively fewer people per square mile (the fringe of the fried egg), and so his metro NYC's density falls in comparison with the Census Bureau's.

Is this right, and does it matter? Well, it matters because one of the principal goals of contemporary urban planners has been, for reasons I won't try to explain now (but you can read about on Cox's website), to fight "urban sprawl" (i.e., low-density development). So if it turns out that the classic example of sprawl, LA, is actually denser than the classic example desirable development, NYC, then it undermines the notion that "density is desirable" and weakens the case for government intervention (e.g., in the form of land-use restrictions and open-space policies) to combat perceived sprawl.

But I'm afraid I can't buy into this way of defining a city and of calculating density. For one thing, how close do buildings have to be to be considered part of the same urban area? It's not clear what principle is being used to make this decision. For example, Cox doesn't merge LA and San Diego or NYC and Philadelphia, even though there is fairly continuous development in between. Where he draws the line seems arbitrary.

A more rigorous and non-arbitrary approach that I like is to define a city in such a way that it becomes a basic unit of economic analysis. Jane Jacobs does this in her book "The Economy of Cities," where she defines a city as "a settlement that consistently generates its economic growth from its own local economy." The problem with her definition is it's hard to apply. (I think I've come up with a clever way to do this that uses something called "Zipf's Law," which I may post in the future.)

Incidentally, the Census Bureau, like Jacobs, uses an economic approach in some of its definitions. For example, its Consolidated Metropolitan Statistical Area (CMSA) is supposed to be a region of economically integrated settlements. Unfortunately, it views metropolitan areas too much as labor markets than as incubators of entrepreneurship. And it's also constructed by adding counties together (because their political borders are pretty stable over time), which is like trying to make a snowman out of cinder blocks.

So while Mr. Cox's result is startling, it doesn't shed light on things that I'm interested in, such as the relation between density and entrepreneurial development, and Jacobs's approach does. Nevertheless, his website is an extremely useful source of information, and I still highly recommend it.

Culture of Congestion Homepage