“Social interaction potential,” or SIP, is Professor Steven Farber’s metric for the chance that “any random pair of a city’s residents can meet based on where they live, where they work, and, given those, how long they have to rendezvous”:
Drawing on census data, travel times, employment densities, and land-use patterns among other statistics, Farber calculated the SIP of 42 U.S. metropolitan regions with at least a million residents. Unsurprisingly, the largest cities–New York, Los Angeles, Chicago, and Washington–also boast the highest aggregate SIP scores. But once you control for size, all four cities underperform compared to smaller peers. For one thing, their vastness works against them–“super-commuters” to Manhattan, the District, or the Loop head straight home at the end of the day, sharply limiting their opportunities for happy hour.
But long commutes pale in comparison to two other factors: decentralization, which produced the edge cities of the 1980s and ‘90s, and the thin schmear of unchecked sprawl that flourished during the housing bubble. “The impact of decentralization is 20 times stronger than commute times,” says Farber. “It’s far more important to move people spatially to the same place than it is tagging on a few minutes to their commutes.” … “Our results clearly show that more sprawling regions make it harder and harder for people to have social interactions with each other,” Farber adds.
The Dish 