Fifteen years ago, Amazon's Pat Bajari and I started to work together on an urban economics paper studying household choice of housing type (i.e number or rooms, building year built) and neighborhood within major metropolitan areas. At first, we assumed that a household head already had a job and then searched for a place to live. This sequential approach would have greatly helped us because if people work in different locations, then the same neighborhood offers a different length commute for different people. This variation is very useful for allowing an economist to estimate the disutility from commuting. We ran into the challenge that our critics said that we couldn't assume that people found a job and then found a housing unit. While this makes sense for Professors and Investment bankers, it may not be true for high school graduates who may live in some neighborhood and then they find a job.
I tell this story because in today's NY Times sports section there is a great story about the New Jersey Nets and how these NBA players know that they have a job in Brooklyn and because they want a short commute and because Brooklyn is on the rise, they are searching for housing in Brooklyn. Since players are traded and the location of the 30 NBA teams are all well defined, the Bajari/Kahn original framework could have been used to study their preferences over commuting, structure and neighborhood.