Modelling air carrier choices with a Segment Specific Cross Nested Logit model

A number of customer segments may exist in any air travel market, differing in the willingness-to-pay, income, age, time of travel or airline preferences. Not accounting for this diversity may introduce a significant variance into any choice model. In this paper we present a Segment Specific Cross-Nested Logit with Brand Loyalty (SSCNL-BL) model that explicitly accounts for this heterogeneity. The model is estimated using data from a stated preference choice experiment conducted among Australians travelling to the United States. The resulting SSCNL-BL model performs better than the more complex Mixed Logit model without the additional computational burden.