On Cities, GHG Emissions, Apples & Oranges


Every day or so a new greenhouse gas emission report crosses my desk. Often these reports are very useful, other times they add little of value to the subject. The problem is separating the “wheat” from the “chaff.”

This dilemma is well illustrated by a paper called “Greenhouse Gas Emissions from Global Cities,” authored by 10 academics. I had received notification of the paper from Science Daily, a useful website that provides notification of new research on a wide range of scientific subjects.

The Science Daily article indicated that Denver produces the most greenhouse gases per capita annually, while Barcelona produces the least. I am always interested in reports that compare the performance of “cities,” both out of general interest and because of the gross errors that often are the result of invalid comparisons. So, immediately I ran down the report, and to my surprise the report dealt with only 10 “cities.” This seems rather a small number, since the smallest in the sample, Geneva, is not even among the top 700 urban areas in the world. This seems to be a rather incomplete sample: 10 out of more than 700.

That was just the beginning. There were serious problems of comparison between the 10 “cities.” Whenever someone starts talking about “cities,” it is best to ask what they mean. The word “cities” has so many meanings and is subject to such confusion that I generally avoid using it.

“Cities” might be municipalities, such as the city of New York or the ville de Paris.

Cities could be urban areas (urbanized areas or urban agglomerations), which are the urban footprints one observes from an airplane on a clear night.

“Cities” could be metropolitan areas, which are labor markets and are generally larger than urban areas, because people commute from rural areas (outside the urban footprint) to work in the urban area.

In nearly the entire world, with the exception of China, urban areas and metropolitan areas are larger than municipalities.

Or, “cities” could be used in the sense of Chinese prefectural, sub-provincial or provincial level cities, which tend to be far larger than any reasonable definition of a metropolitan area. Nearly all of China is divided into cities, in the same way that most of the United States is divided into counties.

These Chinese “cities” themselves often contain county level “cities” that are separate from the principal urban areas.

These differing definitions of municipalities make any international comparison of these entities difficult and often misleading. The ville de Paris represents barely 20 percent of the Paris region. The “city” of Atlanta represents barely 10 percent of its metropolitan area. The “city” of Melbourne represents only 5 percent of its metropolitan area. Yet, other “cities” are larger than their metropolitan areas, such as Chongqing, China, which has at least five times the population of its genuine metropolitan area (the “city” covers an area the size of Austria or Indiana). The city of San Antonio, with its vast stretches of suburbanization is surely not comparable to the city of Hartford, which is dominated by an urban core.

Any genuine comparison of “cities” must be at the metropolitan area or urban area level. These definitions both represent the city as the organism it is, rather than simply the happenstance of municipal boundaries. Of course, comparisons must be either between metropolitan areas or urban areas to be valid. It will not do to compare metropolitan areas with urban areas; they are as apples and oranges. Moreover, there are no international standards for delineation of metropolitan areas, which makes metropolitan comparisons more complex.

All of this raises the principal problem with the “Global Cities” paper. There is no consistency to the city definitions the paper uses and its results are thus meaningless (though “headline grabbing”). For example, “Global Cities” uses the geographic areas of the following barely comparable “cities”:

The municipality of Barcelona, which represents less than one half of the urban area and excludes the expansive suburbs that stretch in every direction but the Mediterranean.

The municipalities of Bangkok, Denver and New York, which are only parts of their respective metropolitan or urban areas.

The municipality of Cape Town, which could be considered a metropolitan area because of the large expanse of rural area under its jurisdiction.

The canton (province) of Geneva might probably qualify as a metropolitan area, except that it excludes the suburbs in France, from which virtually free movement of labor is permitted.

The Greater London Authority which is nearly co-existent with the London urban area, while Prague as the report defines it is somewhat larger than its urban area.

The Greater Toronto Area which meets none of the “city” definitions above and is larger than both the metropolitan area and the urban area as defined by Statistics Canada.

Los Angeles County, which meets none of the “city” definitions and is part of the larger Los Angeles-Orange County metropolitan area.

All in all, as charitably as it can be put, the “Global City” compares four municipalities, three metropolitan areas, two urban areas, one area larger than a metropolitan area and one that is part of a metropolitan area. Put another way, it tries to make comparisons between four apples, three oranges, two peaches, one banana and one sweet potato.

Granted, the paper indicates the geographical definitions it uses. That, does not, however, change the fact that treating apples and oranges as comparable is simply invalid.

There are other problems with the “Global Cities” paper, but one more is enough. In the obligatory fashion, the authors stress how important it is to adopt “smart growth” policies in North America. They cite a US Department of Transportation study to indicate that a doubling of density reduces vehicle miles traveled by 40 percent.

A bit closer reading would have indicated that the study says doubling density would reduce new vehicle miles by 40 percent, where population densities are already 6,000 to 7,000 per square mile. Only two large urban areas in the United States have densities that high, San Francisco and Los Angeles (which the authors characterize as having urban densities at least 40 percent below the US Bureau of the Census number for the Los Angeles urban area). A 40 percent reduction in “new” vehicle miles means that overall vehicle miles traveled increase 60 percent when the population is doubled, rather than 100 percent. Thus, even with the high density qualification in the US Department of Transportation study, vehicle miles would increase 60 percent as population densities double.

Maybe tomorrow will bring a better report. One can always hope.

Photograph: The “city” of Chongqing (part of its vast rural countryside)

Wendell Cox is a Visiting Professor, Conservatoire National des Arts et Metiers, Paris. He was born in Los Angeles and was appointed to three terms on the Los Angeles County Transportation Commission by Mayor Tom Bradley. He is the author of “War on the Dream: How Anti-Sprawl Policy Threatens the Quality of Life.