The Jobs-Housing Balance and Urban Commuting

This paper applies geographical information system (GIS) techniques and a piece-wise, non-linear model-spline functions—to analyse empirically the relationship between the jobs-housing ratio and urban commuting patterns in terms of vehicle miles travelled (VMT) and trip length. A dynamic buffering process in GIS is developed to measure the jobs-housing ratio within floating catchment areas of a 5-7 mile (8.05-11.27 km) radius as opposed to pre-defined and arbitrary jurisdictional boundaries. This study found a non-linear relationship between the jobs-housing ratio and VMT and trip length in the Portland, Oregon, metropolitan area. Only when the jobs-housing ratio is less than 1.2 or larger than 2.8 do VMT vary noticeably as the jobs-housing ratio changes. Marginal changes in VMT are small when the jobs-housing ratio is between 1.2 and 2.8. Since most areas in the study region have a job-household ratio between 1.2 and 2.8, any changes in the jobs-housing ratio will have little impact on VMT. Consequently, hardly any land use policy changes could affect the jobs-housing ratio enough to produce a significant change in VMT per capita, and jobs-housing policy will have limited impact on VMT at the regional level.