Disentangling the influence of neighborhood type and self-selection on driving behavior: an application of sample selection model

• Published in: Transportation (Dordrecht) Vol. 36; no. 2; pp. 207 - 222
• Main Author: Cao, Xinyu (Jason)
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01.03.2009; Springer; Springer Nature B.V
• Series: Transportation
• Subjects: Applied sciences; Attitudes; Automobiles; Built environment; California; Causality; Economic Geography; Economic growth; Economics; Economics and Finance; Engineering Economics; Exact sciences and technology; Ground, air and sea transportation, marine construction; Innovation/Technology Management; Land use; Logistics; Marketing; Neighborhoods; Neighbourhoods; Organization; Regional/Spatial Science; Smart growth; Studies; Transport; Transportation; Treatment effect; U.S.A; Urban development; Urban environments; Variables; California; United States; North America; America; United States > US; Causality; Smart growth; Transportation; Land use; Treatment effect; Data analysis; Vehicle driving; Selection; Residential neighborhoods; Built environment; Modeling; Travel; Models; Behavior; Application; Comparative study; Land use; Smart growth; Transportation
• ISSN: 0049-4488; 1572-9435
• DOI: 10.1007/s11116-009-9189-9

The causality issue has become one of the key questions in the debate over the relationships between the built environment and travel behavior. Although previous studies have tested statistical and/or practical significance of the built environment on travel behavior, few have quantified the relative roles of the built environment and residential self-selection in influencing travel behavior. Using 1,479 residents living in four traditional and four suburban neighborhoods in Northern California, this study explores the causal effect of neighborhood type on driving behavior and its relative contribution to the total influence of neighborhood type. Specifically, this study applied Heckman’s sample selection model to separate the effect of the built environment itself and the effect of self-selection. The results showed that, on average, the effect of neighborhood type itself on driving distance was 25.8 miles per week, which accounted for more than three quarters of the total influence of neighborhood type and 16% of individuals’ overall vehicle miles driven. These results suggest that the effect of the built environment on driving behavior outweighs that of self-selection. This paper also discussed the advantages and weaknesses of applying the Heckman’s model to address the self-selection issue.