The Effects of Commute Trip Reduction Program on Employee Non-SOV Travel Frequency

摘要

Initiated in 1991, Washington State Commute Trip Reduction (CTR) program was one of the earliest employer-based transportation management program in the nation which requires employers to implement strategies to encourage alternative travel modes. This research investigates the effects of the CTR program on employees' non-SOV travel frequency, by controlling employee demographic features and worksite built environment characteristics. The database was assembled from 2015/2016 WA State employer and employee survey data. 84,878 employees in 379 worksites were selected for data analysis and model fit. Poisson model, negative binomial model, zero-inflated model and hurdle model were implemented individually. The hurdle model was eventually selected as the final model due to better goodness-of-fit, the accuracy of model prediction and the practical sense of model interpretation. The model is a combination of a binomial logistic model to predict the odds of being a non-SOV commuter versus an exclusive-SOV commuter and a negative binomial model to predict the number of non-SOV commuting day counts for the non-SOV commuters.;The results of this study corroborate the mainstream view that demographic features, built environment characteristics and CTR policies all have significant effects on employees' non-SOV model choice and travel behavior, yet more dynamic relationships were found among CTR policy components. Controlling for other variables, the parking fee management, transit subsidy, teleworking option and worksite amenity are expected to increase the odds for an employee of being a non-SOV commuter by 52.52%, 58.32%, 15.44%, 23.77%, respectively, while carpool/ vanpool subsidy and worksite service unexpectedly decrease the odds by 12.88% and 17.73%. Speaking of the non-SOV travel frequency, the parking fee management, transit subsidy, walking/biking subsidy, worksite amenity and service increase the rate of non-SOV commuting days by 4.72%, 2.19%,1.67%, 3.97% and 3.91%, whereas the carpool/ vanpool, compressed working hour and teleworking option decrease the rate by 5.57%, 4.91%, 5.47%. Overall, the CTR policy package increases the probability for a "typical" employee of being a non-SOV commuter by 22%, though its effect on non-SOV day counts is modest.