Publication Date


Degree Program

Department of Geography and Geology

Degree Type

Master of Science in Geoscience


In previous studies of urban commutes, little attention has been paid to commute patterns in smaller urban areas. In this study, the concept of "excess commute" (EC) is applied to the Bowling Green-Warren County Metropolitan Statistical Area (BGWCMSA) in Kentucky. EC quantifies the portion of commute distance explained by the overall spatial separation of jobs and households. Results in this thesis research show that approximately 65% of commute distance by persons driving alone in the study area can be explained by the physical locations of homes relative to job sites as well as the existing roadway network, leaving an EC of 35% attributable to other factors. This EC of 35% is less than those of larger metropolitan areas in previous studies, suggesting that EC does decline with the sizes of urban areas to a certain degree. I low ever, the analysis of "used commute potential" (UCP) reveals that workers in the study area on average use a higher percentage of its total potential in comparison to larger cities. A possible explanation is that BGWCMSA is the regional employment center for south central Kentucky. There is a relatively large percentage of commuters living in the rural areas and the surrounding counties, causing a significant number of commutes with long distances. In addition, the analysis of job distribution shows that BGWCMSA has developed a number of specialized employment subcenters. With some subcenters located in the outskirts of the urbanized area, cross-commuting between suburbs also accounts for a substantial portion of the overall commutes in the region, leading to trips with longer distances as well. Both EC and UCP are also applied to the data disaggregated by household income levels to determine if workers with lower household income are more likely to be spatially separated from their workplaces, necessitating longer commutes. In the disaggregate analysis, all workers in the study area are assigned to four household income groups; 1) those with less than $30,000 annually; 2) between $30,000 and $49,999 annually; 3) between $50,000 and $74,999; and 4) $75,000 or more. Results show that it is not the first income group but the second and third income groups of workers that, on average, travel the longest distances with the highest EC and UCP. Workers in the $75,000 or more income group are, on average, the most efficient commuters by both excess commute and commute potential measures. In summary, this work, by highlighting the presence of excess commuting methodology in the smallest metropolitan statistical area yet studied, provides an impetus for planning agencies in smaller urban areas to obviate the negative effects inherent in automobile use. As cities grow, there is a unique opportunity to develop policies and programs to reduce nonspatial factors that affect the amount of time and distance spent in the automobile in the journey to work (JTW). Nonspatial factors that may be impacted by policies include congestion, lack of transit, and parking availability, among many others. The prevailing trend of urban growth in recent decades is the emergence of employment subcenters on the urban fringe, with some being very specialized in employment type and others of a more mixed nature. Results from this stud} confirm the findings of previous work that smaller urban areas are more likely to use more of their commute capacity and are thus less efficient than larger ones, due to the lack of exurban centers with mixed land use types. Specifically, where there is already a regional jobshousing imbalance, the lack of such centers exacerbates the condition of longer commutes and higher UCP. This suggests that the placement and type of employment centers are critical to the commuting characteristics of a given area.


Earth Sciences | Geography