Chapter 6: Discrete Choice Analysis of Shippers’ Preferences

Freight choice models should focus on two objectives: quality of service and total logistics costs, both of which are important to shippers. This chapter examines discrete choice models that apply these objectives to enhance existing freight choice models. A total logistics cost model based on revealed preferences (RP) data is examined first. The model is then extended to include service quality. However, because service perceptions are unobservable, we use a factor analytic formulation to relate the service perceptions to perceptual indicators. These indicators are extracted and the extended model is then jointly estimated using RP data along with these indicators. Estimation efficiency of the extended model is further improved through the joint use of two data sets — one measuring RP, and one measuring stated preferences (SP). We find that the simultaneous use of both sets of data produces superior estimation results in the extended model. This demonstrates the significance of perceptions of service quality in explaining shippers’ choices.