A factor-analytic generalized nested logit model for determining market position of airlines

• This study illustrates market positions and flexible substitution patterns of competing airlines.
• We develop a factor-analytic specification incorporated into the generalized nested logit model.
• The proposed model captures the effects of excluded variables and common error components.
• The results indicate that international air passengers are willing to pay more for better service quality.

This study presents a factor-analytic specification incorporated into the generalized nested logit model to gain insights into passenger service perceptions through positioning analysis of closely competing airlines. The model can illustrate market positions of competing airlines, while simultaneously capturing substitution patterns among airlines. The data used comes from a stated preference survey containing travelers’ airline preferences for international trips from Taiwan to Japan. Two latent dimensions comprised of service attributes that were not considered in the stated choice design are identified, namely “amenities and ground service” and “reputation and credibility.” The choice map identifies two highly competing groups based on similarities on latent attributes and shared error components. The factor-analytic generalized logit model statistically outperforms the factor-analytic multinomial logit and offers important behavioral and managerial suggestions for airlines and policy-makers.