Clare Nelson | University of Vermont

Social Media Handle: 

Presentation Description

Smart growth is one strategy that’s widely promoted to reduce greenhouse gas emissions and vehicle miles travelled from transportation. While smart growth strategies like densifying and increasing transit access are likely to reduce vehicle dependence, the existing literature does not have clear estimates for the expected reduction in vehicle miles travelled from implementing such strategies. Prior studies have wide varying estimates for expected reductions in vehicle miles travelled, largely due to methodological and data limitations. Most studies use cross-sectional study designs that cannot look at causality and cannot account for residential self-selection biases. For example, if people who prefer to walk choose to live in denser neighborhoods, a cross-sectional study will find a strong association between density and walking. However, this does not indicate that densifying areas will increase walking. Longitudinal panel studies, on the other hand, can assess causality by looking at changes in travel behavior as a function of changes in the built environment across time, like increasing density and increasing access to transit. In this study, we respond to the call for more before and after studies on smart growth effectiveness using a longitudinal data set of vehicle miles travelled developed from the Vermont Department of Motor Vehicle registration and inspection records. This data set contains each registered vehicle’s address linked with an annual odometer reading from the inspection for all vehicles registered in the state between 2017 and 2023. By identifying households that relocated between 2017 and 2023, we examine the extent to which changes in the built environment explain the observed changes in vehicle miles travelled. We find that while the built environment has some effect on travel behavior, it is nuanced and inelastic. Our results suggest the expected percent reduction in vehicle miles travelled by relocating to an area developed with “smarter” growth strategies is at the lower end of previously reported elasticities. Our methodology presents a promising approach for states with similar data, and contributes to the limited, yet highly desired number of before and after studies looking at causal relationships between land use and reductions in vehicle miles travelled.

Speaker Biography

Clare Nelson is a research scientist at the University of Vermont Transportation Research Center. She recently graduated with her Masters of Science in Civil and Environmental Engineering from UVM and is now continuing her work studying transportation policy and travel behavior through a data science lens. Her current projects include using cross-cutting, large-scale data sets to create public dashboards with the aim of better advising policy makers about transportation planning.

Co-presenters

Gregory Rowangould
University of Vermont

Presentation File