Joint Model for Last-Mile Delivery Service Selection in China: Evidence From a Cross-Nested Logit Study

This study analyses consumer last-mile delivery service choice behaviour by developing a cross-nested logit model (CNL); we then compare the analytical results with three nested logit models (NL). The model parameters are estimated using the data from a questionnaire collected from consumers residing in Beijing, Shanghai, Tianjin, Guangdong, Zhejiang, Jiangsu, and Shandong. The direct elasticities and cross-elasticities are then calculated to assess the change in probability of each alternative caused by utility variables. Parameter estimation results demonstrate that the CNL model outperforms the three NL models. Consumers are usually reluctant to change the way they are served when utility variables are altered. Moreover, elasticity analysis results suggest that service factors have the strongest effect on choice probability, followed by socioeconomic factors and delivery activity factors. Thus, enterprises should first strive to promote the service experience of consumers in corresponding delivery services, then account for the effect of socioeconomic factors, and finally consider changing delivery service fees if they want to induce consumers to select a specified delivery service.