Although many studies explore built environment (BE) effects on commuting behavior, most overlook BE characteristics at workplace locations and their non-linear impacts. More importantly, limited effort is placed on the integrative effects of the BE and transportation policies. Using the data in Washington, D.C., this study applies a gradient boosting logit model to examine the influences of BE characteristics at both residential and workplace locations and commuting programs (transit/vanpooling subsidies and parking provision) on commute mode choice. We found that BE variables collectively contribute to 65% of the predicting power for mode choice. Although workplace BE variables are more important than residential BE elements, the difference is mainly due to distance to CBD (central business district). Furthermore, most variables show non-linear effects on car mode choice. There are also synergistic effects between BE variables and parking policy and between BE variables and transit/vanpooling subsidies. Therefore, land use policies will be more effective where supportive transportation policies exist.
|Original language||English (US)|
|Number of pages||15|
|Journal||Transportation Research Part A: Policy and Practice|
|State||Published - Dec 2018|
Bibliographical noteFunding Information:
This work is supported by the National Natural Science Foundation of China ( 71503018 , 61773040 , U1564212 , U1664262 , and U1764265 ), and Young Elite Scientist Sponsorship Program by the China Association for Science and Technology ( 2017QNRC001 ).
© 2018 Elsevier Ltd
- Built environment
- Land use and transportation integration
- Machine learning
- Travel behavior
- Work travel