Railway shippers’ heterogeneous preferences with random parameters latent class model

Abstract

Abstract Railway shippers’ preferences heterogeneity is caused by their attitudes and perception towards attributes of various rail freight services, and is differentiated amongst shippers. Random Parameter Latent Class (RPLC) model is applied to elicit Chinese railway shippers’ preferences heterogeneity in both continuous and discrete way. With a Discrete Choice Experiment (DCE) conducted in southwest area of China, 83 respondents are counted to calculate their utilities towards five rail service attributes: cost, time, frequency, reliability and safety. Results by RPLC model are compared with MNL, ML and LC model, and shows that quality attributes are more preferred by railway shippers than price attributes. Further market segmentation with RPLC clarifies that heterogeneous valuation of railway freight service quality dimension does exist among shippers, and segmentation according to shippers latent variables can be more suitable than the traditional exogenous market segmentation way. For railway companies in China, multidimensional freight services could be provided by the combination of different attributes and levels, to attract more and more shippers making railway as their first choice.