Abstract
This study investigated the effectiveness of the Smart Commute program, a well-established travel demand management (TDM) program in the greater Toronto and Hamilton area of Canada. The study exploited a data fusion technique to combine data collected through cross-sectional ex ante and ex post surveys at the workplace of each Smart Commute member. Two types of approaches were used: aggregate statistical analysis and disaggregate choice modeling with the fused-combined data set. The results clarify that aggregate investigation may not always uncover many behavioral details. Aggregate comparisons of the survey data showed that the performance of the Smart Commute program varied by sociodemographic attributes of the employees, including age, employment status, employment shift, and regional municipality of employment. The aggregate investigation also showed that the stated willingness to consider bike, walk, and telework options was not reliable in evaluating the effectiveness of any TDM policy. Complementary to the aggregate investigation, the study used an advanced mixed logit model framework to explore the effects of Smart Commute on commuters’ perceptions about travel attributes. The results of the empirical model revealed more variation in the perceptions of commuters about travel attributes after implementation of TDM interventions. In addition, the perceptions of travel times became more negative after TDM implementation.