A Scobit-Based Car-sharing Participation Model

A Scobit-based model, which considers individual heterogeneity, is proposed to analyze the residents' willingness to participate in car-sharing programs. In the Scobit model, skewness parameters are used to reflect the heterogeneity of individual attributes of different travelers. During the parameter calibration and elasticity analysis progress, the model is mutually validated with the results of Binary Logit model. It shows that the Binary Logit model and Scobit model are consistent with the calibration results and Scobit model performs better than Binary Logit model. Based on the results, the factors of age, occupation, household income, car hailing and car rental experience, vehicle cost, vehicle mileage, parking service, vehicle type and vehicle return method significantly affect people's choices. Besides, elastic value analysis results show that vehicle costs, vehicle's mileage, parking services have greater impact on the choice of residents than others. Finally, some advice is given to encourage the car-sharing participation willingness.