Are very large cities better?
Visitor Question: What are the consequences that we can expect if we continue to allow and encourage larger and larger cities to be built? I think cities in the Third World are much too large to be comfortable for most people. The huge cities in China look very uncomfortable and unhealthy to me also, but the local residents don’t seem unhappy, and that is what counts. I’m not sure about the best size for cities in Europe and America. What do you think?
Not so long ago anyone who tried to answer this question would be referring to his or her own experience base, fitting sample facts under a theory, or looking at the very limited historical research. In the last decade, the pace for collecting "big data" from new sources has picked up dramatically, so now there is actually some data to consult in talking about the efficiency of cities. These new data sources include cell phone usage and locating features, social media such as Twitter, sensors and public cameras, public transportation origin and destination information, better data-sharing arrangements by international entities such as the United Nations and within metropolitan areas, and simply the more creative use of time-honored data points as a result of more powerful computers.
Now "comfort" is subjective, so we won't even try to address that part of your question. Let's talk instead about economic efficiency, since improved economic efficiency, when shared with the population, presumably would allow consumers to purchase more comfort as they themselves define it.
So let's look at what research about city population size and economic efficiency is telling us. Let's add a third variable that has been researched, which is quality of transportation.
At the Santa Fe Institute, Geoffrey West and Luis Bettencourt have made a name for themselves through their finding that a city twice as large as its neighboring city is likely to be 15% richer, meaning that income, savings, and patents are all 15% larger than simple doubling would provide. They also show that when a city doubles in size, the length of electric lines, number of gas stations, and other infrastructure measures per inhabitant declines about 15%. In another article in the journal Nature
, they said that if population doubles, economic productivity increases by 130%. (Academic types have taken to calling this phenomenon super-linear scaling.)
Some other groups have delved into why population growth improves economic performance disproportionately. As an example, the MIT Media Laboratory's Human Dynamics Lab thinks that more urban population density gives people more opportunity for face-to-face interaction, and they compute a social-tie-density score for urban areas. Other research outfits such as the Cornell Social Dynamics Laboratory tend to agree.
Social scientists also have learned that all cities tend to exhibit the same growth rate in the long run, but they say younger cities grow faster. Here are Useful Community Development, we think that’s slim comfort to older Rust Belt cities that would love to have a growth rate period. Let's see how this plays out with more research.
OK, we said we were going to talk about how transportation quality interacts with population and economic growth. Various experts have pointed to the fact that this relationship between population growth and an even greater growth in economic productivity does not seem to work in poorer countries, as our site visitor who asked the question knows already. The working speculation about why this is true is that transportation within the urban area is much worse in poor countries, so in fact the population is not benefiting from the possibility of a greater number of face-to-face interactions in the same way that a transportation-rich city such as London allows.
Others say that larger cities invent more ways to provide cross-links to the transportation system. For example, light rail systems may criss-cross the road system, which tends to be built on a hub and spoke model radiating out from the center (or sometimes from several centers). If there is a separate tram, that too can cut across the other transportation routes to make even more efficient and effective ways to get from Point A to Point B.
A final point is that extreme congestion gets in the way of the face-to-face interaction that promotes a dynamic and efficient economy. Experts speculate but can't yet prove that cities that are super-congested, such as some of the largest Chinese cities, will tend to realize less than the expected uptick in economic efficiency as population doubles, just because the increase in face-to-face interaction will not occur as rapidly as in a city where that interaction is not so inconvenient. Time will tell.
In this and other examples, we think that interventions in the forms of plans and resulting policies can do quite a bit to alter the trends. Individual cities will make themselves exceptions to the rules. In addition, more sophisticated research and data collection methods may show that generalizations such as superlinear scaling actually hide other forms of important information.
So in sum, yes, we think very large cities are counterproductive in poor countries where people do not really move all the way across the city to interact face to face, and where most interactions in person are at a village scale anyway.
We don’t really know yet whether megacities such as those being built in China and elsewhere are more or less economically efficient than simply very large cities.
We ourselves haven't formed a conclusion about the "right" size of an American or European city, in part because many sizes seem to work well in societies that are basically wealthy. What we would say is that a city declining in size faces a downward spiral unless drastic measures are taken.
A final point is that when cities lie within a prosperous nation, many quality of life factors other than economic efficiency play into resident satisfaction. This means that the limits on our discussion based on economic efficiency must be recognized fully.
- Community Development ›
- Questions ›
- Are Very Large Cities Better?