From LEE SCHIPPER
Few dispute the relationship between city density and travel or car
use, but figures cited in Mick Hamer’s ‘City planners against global warming’
(24 July) hide ruses that hopefully readers have already spotted.
In the graph of urban density versus petrol consumption (from Peter
Newman), readers get suspicious finding that Toronto has a higher density
than New York, as does even Los Angeles. The problem is that no one knows
how the lines are drawn around ‘cities’, so we don’t know what population
(or driving) is included.
Newman and colleagues have not measured ‘petrol consumption’ per capita,
but only inferred it from very sketchy data about cars and traffic around
cities. As we found in our own studies, virtually no one knows the petrol
consumption of citizens in the American or European cities shown in the
graph; the best that is known is sales within a certain boundary, as evidenced
by tax receipts. These are only distantly related to how much those who
live there travel, ignoring traffic through the city, particularly for the
low-density towns in the US. Houston is full of filling stations, but try
to find one in downtown New York, Stockholm, or Paris. Hence the vertical
axis of this graph does not measure car fuel consumption related to the
mobility of people living in the city in question.
Surveys show that people who live in the city centre own fewer cars
and travel somewhat less than those living elsewhere, a demographic and
income effect (higher car insurance and maintenance costs) not a density
effect alone. Those that do own cars drive them less and walk more, to be
sure. Those living in suburbs (which may or may not be counted fairly in
the Newman data) drive more. The lower fuel consumption apparent from comparing
Europe and the US results more from the low auto ownership in actual cities
in Europe (as opposed to the suburbs), the lack of American-style gas guzzlers,
and lower driving per car in Europe, than from some magic of density alone.
Advertisement
In the final analysis, population density does influence car use, but
it also has a large effect on virtually every other feature of everyday
living, both positive and negative. While it would be nice to control sprawl
of our American cities, one should look beyond the distortions of the simplistic
figures of the Newman study to a better representation of the costs and
benefits of living differently.
Lee Schipper Lawrence Berkeley Laboratory California
