Recent theoretical work has examined the spatial distribution of unemployment using
the efficiency wage model as the mechanism by which unemployment arises in the urban
economy. This paper extends the standard efficiency wage model in order to allow
for behavioral substitution between leisure time at home and effort at work. In equilibrium,
residing at a location with a long commute affects the time available for leisure
at home and therefore affects the trade-off between effort at work and risk of unemployment.
This model implies an empirical relationship between expected commutes and labor
market outcomes, which is tested using the Public Use Microdata sample of the 2000
U.S. Decennial Census. The empirical results suggest that efficiency wages operate
primarily for blue collar workers, i.e. workers who tend to be in occupations that
face higher levels of supervision. For this subset of workers, longer commutes imply
higher levels of unemployment and higher wages, which are both consistent with shirking
and leisure being substitutable.