A GIS-based spatial statistical approach to modeling job accessibility by transportation mode: case study of Columbus, Ohio

Improving job accessibility based on transport connectivity helps to address equity issues. Spatial autocorrelation (SA) is also a focus of interest in transportation planning, but has been neglected in analyzing job accessibility in metropolitan areas. In this study, GIS-based job accessibilities by walking, transit, and car are computed for the metropolitan area of Columbus, Ohio, and three transport-based spatial autoregressive (SAR) models are estimated to account for the SA of job accessibility among neighboring block groups, while controlling for built-environment and socioeconomic factors. SA intensities and extents are compared in order to better understand local spatial clusters of job accessibility across the region. Direct and indirect spillover effects due to an investment change in transportation facilities are estimated and provide important transportation planning information. The results also show that walking-accessed jobs are primarily related to physical settings (e.g., land uses) at the local level. Locations with a higher share of zero-vehicle housing units have better job accessibility by transit. There is a spatial mismatch between Asian population clusters and transit-accessed jobs, possibly because of the car-oriented residential clusters around Honda of America Manufacturing in suburban areas. More importantly, locations with a higher share of single-parent households are at a disadvantage in overall job accessibility. Due to its complex transportation needs, a society friendly to single parents should spatially integrate accessible jobs with other needed activities via land-use and transportation planning. Alternatively, car-ownership programs and non-spatial social supports also might be effective to help secure job opportunities and perform daily life activities.