Traffic congestion drives urban evolution

As a city grows, it evolves from a so-called monocentric organization, with activities concentrated in the same geographical area—typically a central business district—to a polycentric organization, in which a number of subcenters coexist. Researchers Rémi Louf and Marc Barthelemy, both at France’s Atomic Energy Commission in Saclay, have now developed a stochastic model that accounts for the emergence of that transition as an effect of rising traffic congestion. According to a 30-year-old economic model, the location at which individuals choose to work is one that maximizes income, taken as the difference between their wages and commuting costs. But cities are not static distributions of households and businesses; as populations increase, so do congestion and the time and thus cost of a commute. That trend, incorporated into the researchers’ new model, reduces the attractiveness of some areas in which individuals might otherwise opt to work and raises the attractiveness of others, which become new subcenters. The model predicts that there always exists a critical value of the population above which a city abruptly turns polycentric and that the total commuting distance and number of subcenters in a city scale sublinearly with its population. Those predictions are consistent with data—specifically, the number of employees per ZIP code—gathered from some 9000 US cities between 1994 and 2010. (R. Louf, M. Barthelemy, Phys. Rev. Lett. 111, 198702, 2013, doi:10.1103/PhysRevLett.111.198702 .)