Comparing the impacts of local land use and urban spatial structure on household VMT and GHG emissions.

To highlight the role of sustainable urban spatial structure in reducing household vehicle miles traveled (VMT) and CO 2 emissions, this empirical study of the 121 largest urban areas (UAs) in the U.S. compares the effects of local land use and UA scale spatial structure in a multilevel analysis framework. The results show that centralized population and mezzo scale jobs-housing balance as well as higher UA population density can significantly reduce VMT and CO 2 emissions. The combined effects of all UA level variables, including population-weighted density (PWD), are found to be on par with the elasticity of VMT with regard to a census tract level compactness index. Further, we find that urban spatial structure moderates local urban form effects on travel behavior. For example, while 10% more compact census tracts are associated with 5% fewer VMT in UAs with the sample average PWD, such as St. Louis and Pittsburgh, this estimated local effect increases to 7.5% and 10% in UAs where PWD is as high as in Chicago and New York, respectively. The findings of this study strongly support policy programs that aim to boost "articulated densities" in the urban region and call for stronger institutional frameworks for regional planning. • Urban area density, centrality and jobs-housing mix affect VMT and CO 2 emissions • Urban structure affects VMT and CO 2 emissions as much as local land use • Urban structure moderates local urban form effects on travel behavior • Elasticity of VMT w.r.t. local compactness increases with urban area density [ABSTRACT FROM AUTHOR]