Urban Metabolism

It's not very often that I come across new research which I think is absolutely seminal to how we understand cities and regions.  But a new study of "urban scaling" by researches affiliated with the Santa Fe Institute is promises to change the way we think about cities. It's a must read!  Here is a quick distillation of two recent press summaries.

A new report in this week's Proceedings of the National Academy of Sciences USA confirms what many city dwellers, who account for the bulk of people on Earth, have claimed for years: Cities have an almost magical ability, spurred by increased human interaction, to stimulate innovation and increase wealth. The report also pooh-poohs the popular comparison of the growth of cities with biological organisms. An animal slows as it balloons in size ; in contrast, the researchers note, cities speed up as population and everything from crime to per capita income grow. Cities create a sort of "urban economic miracle," says study co-author Luis Bettencourt, a research scientist in Los Alamos National Laboratory's Theoretical Division. "When you integrate all these people and all these activities and the struggle to make a living, total productivity increases," he says.

The researchers sifted through an extensive amount of data on many urban systems—mostly big American cities, but also European (primarily German) and Chinese urban areas. The researchers mathematically modeled these factors according to population growth to see how each respond when more people move to a city. They found that human needs, such as employment, utility consumption and housing, correspond directly with the population ...At the other extreme, researchers found that increases in social activity and production outpace population growth. In other words, if the number of city denizens doubles, these factors—both negative (crime) and positive (wealth creation, total wages and gross domestic product)—will more than double.

“The one thing that we know about organisms, whether it be elephants or sharks or frogs, is that as they get large, they slow down,” Lobo says. “They use less energy, they don’t move as fast. That is a very important point for biological scaling. “In the case of cities, it is actually the opposite. As cities get larger, they create more wealth, and they are more innovative at a faster rate. There is no counterpart to that in biology.” What was surprising to the team was that the creative output (jobs, wealth generated and innovation), as cities grow, becomes faster and faster per capita. “It isn’t like if you double the size of a city, you double its creative output,” Lobo said. “But it does increase by about 10 percent to 30 percent." “Cities are really one of the most important innovations in human history,” Lobo says. “We need to think of them as being very human entities, and as engines of our collective creation."

The paper is here.

Comments

[Barkley Rosser]

I think it is important not to get too excited here or overstate what has been found in this study. It does perform a useful function in showing that scaling relations hold for a variety of variables related to urbanization. But the basic idea of scaling in relationd to urban populations has been around for a long time. Zipf's major book dates to the late 1940s, and Pareto was advocating scaling relations to explain income distribution more than a century ago.

Also, one needs to be a bit cautious about universalistic statements regarding unlimited internal economies of scale in cities. A well-established factoid of this literature is that the coefficient on the scaling can change over time within a nation, with a cliche story being that the scale tends to start out very flat in pre-industrial societies, becomes very steep during industrialization, with a major primate city likely to emerge, often driven by extreme economies of scale in heavy industry and strong agglomeration effects, with this flattening out at higher stages of development as transportation costs and logistical and information transmission costs decline, allowing for some decentralization of creative class activities and virtual economies through computerized links. The rise in population of LA relative to NYC is a sign of this, even if LA is now down in the dumps according to the latest "creative class hot spot" map, in comparison with such non-humongous metropoli as Cincinnati-Lexington-Louisville.

[Richard]

Barkley - Thank you for these insights. I am actually very interested in the Zipf distribution of cities, going back to my paper with Axtell on the subject. With Axtell and Tim Gulden we have been using the global mega-region data to try to get at this. My working hypothesis is that globalization stretches out the Zipf distribution (rank-size rule) from a series of national distributions to a global one. Thus, all national city-systems are coming under increased competition and strain, which hits particularly hard at mid-sized and smaller cities down the chain. My sense is that these national city systems will experience the same sorts of shake-outs that hit (national) industries when globalization hit. I am inclined to believe that we are moving gradually to a single global Zipf pattern for cities, though I am not sure of the time scale on this. I'd very much appreciate your thoughts on this.

[Barkley Rosser]

Well, I am sure that there is a Zipf distribution for the whole world that is itself undergoing some kind of evolution as globalization proceeds. However, I would not expect individual nations to converge to it. The problem is the varying size of nations. Thus, we have always tended to see flatter distributions in larger countries, think China, India, and the US, which are too large and decentralized to have really overwhelming primate cities, although New York was way ahead of the rest for a long time in the US, and Mumbai is moving out far ahead of the pack now in India.

But in much smaller countries, there simply is not the chance to have such decentralization with serious secondary cities. Vienna will always be way ahead of
Salzburg; London will always be way ahead of Birmingham (or Manchester or York or wherever); Paris will always be way ahead of Marseilles or Lyon, Copenhagen in Denmark, Stockholm in Sweden, Helsinki in Finland, and so on and so forth.