'Our responses to distance are quite predictable. Most of us will walk to a corner store rather than climb in and out of the car if it’s less than a five-minute walk—about a quarter mile—away.  Kids move by a similar calculus. Frank found that if there is a park or some kind of store within a half mile of their home, school-age youth are more than twice as likely to walk. If destinations are farther, they wait for a parental chauffeur.

Think of the implications: a community with one central mega-sports complex with several baseball diamonds and soccer fields can actually be bad for children’s health if it replaces small parks scattered every few blocks. In the finer-grained community, instead of begging Mom for a ride to a league game, a teenager might find it easier to organize her own game at the local park.

We won’t walk more than five minutes to a bus stop, but we will walk 10 to a light-rail or subway station, partly because most of us perceive rail service to be faster, more predictable, and more comfortable. This is the geometry perfected by streetcar city developers a century ago. It’s now being rediscovered by planners who find that simply introducing regular high-quality light-rail service can alter the habits—and the health—of people nearby. Less than a year after the LYNX commuter light-rail line was installed in Charlotte, N.C., people living near the line had started walking an extra 1.2 miles every day because the system changed their daily calculus. People who switched to the LYNX for their commute lost an average of 6½ pounds during that time.

“The way we organize most cities actually encourages individuals to make choices that make everyone’s life harder,” Frank told me. “The system fails because it promises rewards for irrational behavior.”