Tuesday, September 27, 2011

Congestion data, part 3: density and congestion

One common argument for anti-density regulation is that dense population leads to traffic congestion caused by packing lots of cars into a small area. A common counter-argument is that if population is compact enough, people will walk and take transit more and drive less. Does the TTI congestion data support either side?

To test the proposition in a fairly primitive way, I examine the central city* density of the 15 largest urbanized areas, and compare the congestion levels of these cities. In particular, I divide urbanized areas into three categories:

1.High-density regions where central cities have over 10,000 people per square mile- Chicago, New York, Boston, Philadelphia, San Francisco, and Miami. In all but one of these cities (Miami), over 20% of commuters use public transit to get to work.

2. Medium-density regions where central cities have 5000-10,000 persons per square mile- Washington, Seattle, Los Angeles, Detroit. These regions have varying levels of transit ridership. Washington has almost 10,000 persons per square mile and its ridership is comparable to that of the cities in category 1. In the other cities, between 7 and 20% of commuters use transit.

3. Lower-density regions where even central cities have less than 5000 people per square mile- Dallas, Houston, Atlanta, San Diego, Phoenix. In all but one (Atlanta) of these regions, even in the central city less than 5% of commuters used transit.

What do we find? In the high-density regions, hours lost to congestion as of 2010 were as follows: New York (54), Chicago (71), Philadelphia (42), Boston (47), San Francisco (50), Miami (38).

In the medium-density regions, hours lost to traffic were as follows: Washington (74), Los Angeles (64), Seattle (44), Detroit (33).

In the low-density regions, hours lost to traffic were as follows: Atlanta (43), San Diego (38), Phoenix (35), Dallas (45), Houston (57).

The average low-density region lost fewer hours to traffic (43.6) than the medium (53.8) or high (52) density regions.

It nevertheless seems to me that there is no clear pattern here. Although the low-density regions, on the average, suffered from traffic congestion less than regions with compact cities, residents of Houston lost more hours to traffic than residents of New York, Philadelphia or Boston. Moreover, there was no real gap between medium- and high-density cities.

*I use central city density because I think central city density is a better predictor of transit use and automobile dependency than overall regional density. For example, Los Angeles has a fairly high level of regional density but a medium level of central city density, because its population is spread more evenly throughout the metropolitan area than that of other cities. Its transit ridership tends to be on the low side, as compared with New York City which has a very compact central city and very low-density suburbs.

Michael Lewyn


Density, Transportation | Permalink

TrackBack URL for this entry:


Listed below are links to weblogs that reference Congestion data, part 3: density and congestion :