National utility-based framework for evaluating U.S. work trip accessibility

This study has a dual purpose, first to define and operationalize a national-scale multimodal consumer surplus work travel accessibility metric, filling critical gaps in the literature through several interdisciplinary innovations including the development of a machine-learning derived interconnected set of geographic typologies leveraged to estimate the first set of nested logit travel demand models for the entire U.S. population. Employing these travel demand models harnesses travel behavior methodologies in the development of a utility-based job accessibility metric defined for every census tract in the U.S., from very rural to very urban. The second purpose is to demonstrate the value of this metric by evaluating its relationship to economic vitality across the urban to rural spectrum. We quantify larger variation in work travel accessibility between very urban and very rural region types compared to within them. Additionally, within the two most urban region types we demonstrate evidence of higher job accessibility in the urban center neighborhoods compared to suburban or ex-urban subregions. In the most rural region types, road network topology drives variation in intra-regional accessibility, with the most circuitous network type locations exhibiting the lowest accessibility. We find evidence that higher job accessibility is correlated with higher median household income, particularly outside the densest urban center areas. The scale of this accessibility framework enables novel insights at the national level.