Researchers from the MIT Senseable City Lab have produced a series of maps visualizing commuting habits across Chinese cities. Titled Potato Project, the study used mobile phone location data from 50 million individuals across 234 cities to understand commuting patterns between a person’s home and work locations.

The research project was launched in response to the group’s observation that despite the significant increase in the size of cities over decades, commuting distances and times in larger cities have remained stable versus smaller cities. “The conserved commuting properties are quite counter-intuitive,” the team notes, “as the distance from the periphery to the urban center in larger cities is obviously greater than that in medium- or small-sized cities.”

With little existing large-scale modeling on the subject, the researchers analyzed the mobile phone location data to better understand the similarities in commuter behavior across different environments. In an accompanying paper, titled The universality in urban commuting across and within cities, the group argues that commuting behaviors across cities obey “remarkable regularities” which can be summarized as two laws.

Firstly, the team observed that “commuting distance at the city level is independent of city size.” Despite some subject cities containing populations 100 times greater than others in the study, the average commuting distance remained similar across cities, at approximately five miles. Similar results were found when ‘population size’ was replaced with ‘built-up area’ as a variable.

“This scale-invariant finding is surprising, as we might expect larger cities to have longer commuting distances,” the group notes in their paper. “However, even though Shanghai’s population is ten times larger than that of Daqing, their average commuting distances are very similar (8.82 km in Shanghai and 8.64 km in Daqing).”

The researchers, however, note that their data did not allow them to calculate specific commuting times due to a lack of information on transport modes such as walking, cars, or buses. “Considering that congestion is more severe in large cities, it is possible that commuting time is greater in large cities than in small and medium cities for a given commuting distance,” the group adds.

The study's second derived law, which they title the “attractiveness law,” is that “commuting distance within a city follows an inverted U-shaped function of the distance from the city, which implies that the city center has an upper bound on the employment attractiveness of residents.”

Data from the research found that commuting distance relative to a city’s urban center obeys a universal pattern across cities. From the city center to the suburbs, city commuting distances obey an inverted U-shaped curve, indicating that a city center exerts a limited attraction to urban dwellers as a place of work. People beyond this limit are more likely to find a job in a closer “sub-center” of the city rather than the main center. 

Across dozens of cities studied, the upper limit of the urban center’s “employment attractiveness” is between 6 and 9 miles from its center, including Shanghai at 7 miles and Chengdu at 6 miles.

“The existence of these upper bounds does not imply that a city’s outward expansion will stop after reaching a specified distance from the center,” the group, however, notes. “Instead, new development will spill into sub-centers when the commute to the main center from fringe areas exceeds certain constraints. In other words, cities are ‘forced’ to switch from a monocentric to a polycentric model, and such a process is driven by commuting constraints.”