School 4Working Paper
Welfare Consequences of Upgrades: Evidence from the Airline Industry
Using upgrades - fees that customers pay to access a premium quality product after the purchase of a regular one - can significantly affect consumer welfare. On the one hand, regular consumers benefit from potentially accessing the monopolist's higher-quality goods at a discounted price. On the other hand, the monopolist will seek to capture surplus from these gains from trade by offering a different price menu for all goods. Whether consumer welfare will rise or fall when a firm introduces upgrades will depend on the relative magnitudes of these two effects. The aim of this research is to disentangle these two effects in the context of an international airline that gives economy class passengers the option to pay an additional fee to upgrade to business class. I develop and estimate a model of airline pricing in order to assess the effects of upgrades via counterfactual simulations. I show that the upgrade option improves the allocation of passengers among cabins over time and increases consumer and producer welfare by 1.4% and 2.2% respectively.
(Selected for) Presentations: UCLA proseminar, 2024 INFORMS (Revenue Management and Pricing Conference 2024), 2024 EEARIE 2024, PROSPER Seminars, Università degli Studi di Siena, 2024 European Winter Meeting of the Econometric Society, NHH, Liverpool School of Business, Università Ca' Foscari, Nova School of Business and Economics, EMAC, 2025 INFORMS (Marketing Science Conference)
Work in Progress
Impact of in-session price increases in the cart on consumer behavior: Evidence from an online marketplace (joint work with Serkan Mehder, Ovunc Yilmaz, Jiankun Sun, and Zifeng Zhao)
(Selected for) Presentations: 2025 INFORMS (Revenue Management and Pricing Conference 2025), ESADE Pricing Symposium, EWG Retail Operations Annual Meeting 2025
Flexibility and Revenue Management: Experimental evidence from an international airline (joint work with Sungtak Hong, and Jinglong Zhao)
The value of information: Evidence from the airline industry (joint work with William Burton, Daniel Fry, Adam Thorsteinson, Jason Wong)
This paper demonstrates that leveraging traveler-level data significantly enhances the prediction of cancellations and no-shows, and that incorporating these predictions into revenue management (RM) decisions leads to more efficient seat allocation. Using proprietary data from Alaska Airlines, we first show that booking-level characteristics improve the accuracy of cancellation and no-show forecasts compared to traditional airline methods based on historical flight-level data. Next, after estimating a structural model, we evaluate the impact of incorporating these improved forecasts into RM optimization - specifically in pricing and virtual capacity decisions - using counterfactual simulations. Our results show that RM decisions informed by more accurate predictions lead to increases in both revenue and consumer surplus by reducing excessive overbooking and seat spoilage.
(Selected for) Presentations: 2025 AGIFORS, 2025 SEA
Welfare Consequences of Upgrades: Evidence from the Airline Industry
Using upgrades - fees that customers pay to access a premium quality product after the purchase of a regular one - can significantly affect consumer welfare. On the one hand, regular consumers benefit from potentially accessing the monopolist's higher-quality goods at a discounted price. On the other hand, the monopolist will seek to capture surplus from these gains from trade by offering a different price menu for all goods. Whether consumer welfare will rise or fall when a firm introduces upgrades will depend on the relative magnitudes of these two effects. The aim of this research is to disentangle these two effects in the context of an international airline that gives economy class passengers the option to pay an additional fee to upgrade to business class. I develop and estimate a model of airline pricing in order to assess the effects of upgrades via counterfactual simulations. I show that the upgrade option improves the allocation of passengers among cabins over time and increases consumer and producer welfare by 1.4% and 2.2% respectively.
(Selected for) Presentations: UCLA proseminar, 2024 INFORMS (Revenue Management and Pricing Conference 2024), 2024 EEARIE 2024, PROSPER Seminars, Università degli Studi di Siena, 2024 European Winter Meeting of the Econometric Society, NHH, Liverpool School of Business, Università Ca' Foscari, Nova School of Business and Economics, EMAC, 2025 INFORMS (Marketing Science Conference)
Work in Progress
Impact of in-session price increases in the cart on consumer behavior: Evidence from an online marketplace (joint work with Serkan Mehder, Ovunc Yilmaz, Jiankun Sun, and Zifeng Zhao)
(Selected for) Presentations: 2025 INFORMS (Revenue Management and Pricing Conference 2025), ESADE Pricing Symposium, EWG Retail Operations Annual Meeting 2025
Flexibility and Revenue Management: Experimental evidence from an international airline (joint work with Sungtak Hong, and Jinglong Zhao)
The value of information: Evidence from the airline industry (joint work with William Burton, Daniel Fry, Adam Thorsteinson, Jason Wong)
This paper demonstrates that leveraging traveler-level data significantly enhances the prediction of cancellations and no-shows, and that incorporating these predictions into revenue management (RM) decisions leads to more efficient seat allocation. Using proprietary data from Alaska Airlines, we first show that booking-level characteristics improve the accuracy of cancellation and no-show forecasts compared to traditional airline methods based on historical flight-level data. Next, after estimating a structural model, we evaluate the impact of incorporating these improved forecasts into RM optimization - specifically in pricing and virtual capacity decisions - using counterfactual simulations. Our results show that RM decisions informed by more accurate predictions lead to increases in both revenue and consumer surplus by reducing excessive overbooking and seat spoilage.
(Selected for) Presentations: 2025 AGIFORS, 2025 SEA
