Doctor of Philosophy (PhD)


Geography and Anthropology

Document Type



Based on the CTPP data 1990-2010, this research analyzes the temporal variability of commuting patterns and efficiency (in both distance and time) in Baton Rouge, Louisiana. It proposes a simulation-based method to measure commuting by simulating individual resident workers, jobs, and trips between them, in order to mitigate the aggregation error and scale effect that are commonly encountered in existing studies. Specifically, the Monte Carlo simulation approach is adopted to simulate individual resident workers and jobs that were consistent with their spatial distributions across the areal unit (e.g., census tract), and then simulate individual trips that were proportional to the existing journey-to-work trip flows. The results indicate that average commute distance kept climbing between 1990 and 2010, whereas average commute time increased between 1990 and 2000 but then slightly dropped toward 2010. As commuting is a trip linking one’s residence to employment, this research follows the long tradition of using the urban land use pattern, namely the spatial separation between residential housing and job location, to explain the observed commuting pattern. Three land use measures are used: distance from the CBD, jobs to resident workers ratio, and a gravity-based job proximity index. The research finds that these land use measures remained a good predictor of commuting pattern in Baton Rouge over time, and the best model explained up to 90 percent of mean commute distance and about 30 percent of mean commute time. Furthermore, nonspatial factors such as a worker’s socioeconomic attributes also influence commuting. Foremost, income plays an important role in one’s residential choices and thus commuting. This research focuses on the role of wage rates of resident workers in commuting pattern. It is reported that commuting behaviors varied across areas of different wage rates, captured by a convex shape. Initially workers living in more affluent neighborhoods tended to commute more, but those in areas with the highest wage rates retreated for less commuting. This trend remained relatively stable over time. Wasteful (excess) commuting is also examined as the overall commuting efficiency metric for the study area. Wasteful commuting is measured as the proportion of actual commute that is over minimum (optimal) commute when assuming that people could freely swap their homes and jobs in a city. This research identifies two contributors resulting in the miscalculation of wasteful commuting: reporting errors and the use of aggregate zonal data. The former tended to overstate the actual commuting length and led to overestimate wasteful commuting; and the latter (especially the use of large areal unit) led to underestimate wasteful commuting. This research indicates that the percentage of wasteful commuting increased significantly between 1990 and 2000 and stabilized afterward.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Wang, Fahui